title
stringlengths 1
827
⌀ | uuid
stringlengths 36
36
| pmc_id
stringlengths 5
8
| search_term
stringclasses 44
values | text
stringlengths 8
8.58M
|
|---|---|---|---|---|
An Overview on Antimalarial Peptides: Natural Sources, Synthetic Methodology and Biological Properties | 754f76df-087e-42c3-b77e-2049e131d169 | 10708299 | Pharmacology[mh] | Malaria is a common disease in Africa and certain Asian countries and is one of the leading causes of death worldwide. Indeed, the World Health Organization (WHO) reported that the number of deaths increased by 10% in 2020 compared with 2019, but declined to an estimated 619,000 deaths in 2021. Furthermore, nearly half of the world’s population is at risk of malaria . Malaria is caused by a tiny protozoon belonging to the Plasmodium species that is spread to humans through the bite of an infected female Anopheles mosquito. Of the 172 Plasmodium species, five can infect humans, including Plasmodium malariae , P. falciparum , P. vivax , P. ovale , and P. knowlesi . Many drugs have been produced to treat and prevent this disease. For example, several peptides have been identified as possessing antimalaria properties against P. falciparum . These antimalarial peptides were initially isolated from plants or animals and the low yields have stimulated great advances in the development of the synthetic methodologies and synthetic analogs. This review will discuss the relevant literature published between 2015 and 2023 regarding the isolation techniques, chemical synthesis, and biological activity of the antimalarial peptides. Nature has become a source of antimalarial peptides. Various types of peptides with antimalarial activity have been isolated with cyclic and linear structures . Peptides with antimalarial activity are commonly found as cyclopeptides; for example, ribifolin is a cyclopeptide containing eight amino acid residues (ILGSIILG) that has antimalarial activity and was first isolated by Pinto et al. in 2015 from the aerial parts of Jatropha ribifolia . The peptide was extracted using Strata C18-E solid-phase extraction catridge using H 2 O and CH 3 CN as mobile phases (0% 30%, 70%, and 100% CH 3 CN in H 2 O), then purified by semipreparative HPLC using a linear gradient of 40−80% of CH 3 CN, containing 0.036% TFA in H 2 O containing 0.045% TFA for 120 min . Antimalarial cyclopeptides can also be isolated from alkaloid-producing plants, such as the cyclopeptide alkaloids hymenocardine ( 2 ), hymenocardinol ( 3 ), hymenocardine N -oxide ( 4 ), and hymenocardine-H ( 5 ) extracted from the root bark of Hymenocardia acida with 80% methanol by Tuenter et al. in 2016 . The crude extract was fractioned by liquid–liquid partitioning followed by flash chromatography, and then the compounds were purified by semipreparative HPLC with DAD and ESIMS detection. Tuenter et al. (2016) also isolated nine cyclopeptide alkaloids from two different batches of the roots of Ziziphus oxyphylla . The first batch was subjected to a general liquid−liquid partition and fractionation scheme, whereas a more alkaloid-specific fractionation procedure was used for the second batch to increase the number of alkaloids detected. The single compounds were purified by semipreparative HPLC with DAD and ESIMS detection with LC-DAD-SPE-NMR and identified as nummularine-R ( 6 ), O -desmethylnummularine-R ( 7 ), O-desmethylnummularine-R N -oxide ( 8 ), hemsine-A ( 9 ), hemsine-A N -oxide ( 10 ), ramosine-A ( 11 ), oxyphylline-C ( 12 ), oxyphylline-E ( 13 ), and oxyphylline-F ( 14 ) . Maluf et al. (2016) isolated a 42-amino acid polypeptide (YKCHKKGGHCFPKEKICLPPSSDFGKMDCRWRWKCCKKGSG, MW = 4889.81 Da) named crotamine ( 15 ) from rattlesnake venom . Six hundred milligrams of crude dried venom was dissolved in 5 mL of 0.25 M ammonium formate buffer (pH 3.5) and the major component crotoxin was eliminated by slow-speed centrifugation before crotamine was recovered in a narrow protein peak by raising the NaCl concentration and acid hydrolysis. The purity of crotamine was further confirmed by analytical RP-HPLC with a C18-column, LCMS, and ESI-MS. Two new cyclic octadepsipeptides octaminomycin A ( 16 ) and octaminomycin B ( 17 ) discovered by Jang et al. (2017) exhibited activity against P. falciparum . The octadepsipeptides were isolated from a microbial metabolite fraction library of Streptomyces sp. RK85-270 based on Natural Products Plot (NPPlot) screening using liquid chromatography for extraction and purification . Sweeney–Jones et al. (2020) isolated an antimalarial cyclopeptide named kakeromamide B ( 18 ) from Fijian marine cynobacterium by VLC using 80 mL of silica gel in a glass vacuum funnel . The fractions were collected using a stepped gradient of hexane, EtOAc, and MeOH (100% hexane, 10% EtOAc/90% hexane, 20% EtOAc/80% hexane, 40% EtOAc/60% hexane, 60% EtOAc/40% hexane, 80% EtOAc/20% hexane, 100% EtOAc, 25% MeOH/75% EtOAc, and 100% MeOH), then separated by reverse SPE phase (gradient decreasing polarity gradually starting with 15% MeOH in H 2 O to 100% MeOH). Fractions 3, 4, and 5 were separated again by HPLC using a C18-silica column to obtain kakeromamide B (tR = 17.5 min, 0.18 mg). Another cyclic peptide with antimalarial activity called pipecolisporin ( 19 ) was isolated in 2021 by Fernández-Pastor et al. from the endophytic fungus Nigrospora oryzae CF-298113 . A scaled-up culture of Nigrospora oryzae CF-298113 was extracted using methyl ethyl ketone (MEK) in BRFT medium and then purified using reversed-phase C18 medium pressure chromatography followed by semi-preparative reversed-phase HPLC on a phenyl column applying isocratic elution of H 2 O-CH 3 CN as the mobile phase (3.6 mL/min; 32% CH 3 CN for 40 min; UV detection at 210 nm). Pipecolisporin ( 19 ) was obtained as an amorphous white powder. Subsequently, the antiplasmodial peptides kosidachin A ( 20 ) and kosidachin B ( 21 ) were discovered by Watanabe et al. (2022) . The compounds are cyclic tetrapeptides isolated from a culture broth of Pochonia binensis FKR-0564 by HPLC using Capcell Pak C18 column with an isocratic solvent system. Two new compounds were acquired from two fractions with retention times between 12.5–14.0 and 15.0–16.5 min. Another antimalarial peptide georatusin ( 22 ) was isolated by Shi et al. (2018) from Geomyces auratus, which is a mesophilic fungus . The ethyl extract was fractionated using a flash purification system with a gradient elution of n-hexane:acetone and purified by semipreparative HPLC on a RP-HPLC with C-18 to obtain georatusin. Some linear peptides have also been reported to be active against malaria parasite. N -methylated linear peptides friomaramide B and shagamides A-F were purified from the marine sponge Inflatella coelosphaeroide by Bracegirdle et al. in 2022 . A collection of I. coelosphaeroides sponge samples was extracted, mounted onto silica gel and then partitioned using NP flash MPLC before fractionation by semipreparative C18 HPLC using a linear gradient. The individual fractions were then purified again by C18 HPLC to obtain friomaramide B ( 23 ), shagamide A ( 24 ), shagamide B ( 25 ), shagamide C ( 26 ), shagamide D ( 27 ), shagamide E ( 28 ), and shagamide F ( 29 ). Recently, Cabrera-Muñoz et al. (2023) reported that the carboxypeptidase NpCI ( 30 ) inhibited P. falciparum growth . The compound was isolated from the mollusk Nerita versicolor and purified using a CPA-ehtyloxy-Sepharose CL-6B matrix. shows the sequence of NpCI and modelled folded of NpCI displaying a central compact region, a long N -tail, and a short C-tail. Cyclic and linear peptides have been synthesized by various methods as outlined in the following section. 3.1. Cyclopeptides 3.1.1. Mahafacyclin B The antimalarial mahafacyclin B ( 32 ) is a cyclic heptapeptide that was synthesized by Fujita et al. (2013) using the soluble-tag assisted liquid–phase method . This method allowed peptide head-to-tail cyclization leading to the total synthesis of mahafacyclin B assisted by hydrophobic benzyl alcohol. The synthesis was initiated by reductive amination of hydrophobic benzaldehydes using amino acid methyl esters to obtain hydrophobically tagged Gly-Ome from its reaction with coupling reagent Fmoc-Phe-OH, and a combination of HATU, HOAt, DIPEA, and THF. Basic deprotection of the hydrophobically tagged dipeptide produces diketopiperazine when the C-terminal residue is proline. The hydrophobically tagged tripeptide was deprotected to obtain the cyclized compound . A less polar solvent was employed during cyclization to overcome the issue of the highly hydrophobic residues in the linear precursor. The amide nitrogen was used as the marking site allowing the rapid reaction checks and product isolation. 3.1.2. Aerucylamide B Aerucyclamide B ( 33 ) was synthesized by Peña et al. (2013) using two heterocycles and a dipeptide as building blocks . Dipeptide 34 and thiazole 35 were synthesized in the first step and then HBTU was used as coupling reagent to produce the thiazole precursor 37 through cyclodehydration of β-hydroxythioamides and further oxidation methodology . In the next step, two intermediate compounds 41 and 44 were prepared to obtain macrocycle 34 . Compound 41 was synthesized via routes I and II . Route I involved a coupling between the N -deprotected dipeptide 40 and intermediate 39 , while route II involved coupling between compound 42 and 43 . Macrocycle 34 was obtained from intermediates 41 and 44 using HBTU as a coupling agent , and then aerucyclamide B was synthesized from 37 via intermediate 44 , and from 36 via intermediate 41 . 3.1.3. Cyclomarin Cyclomarin A ( 46 ) is a marine cycloheptapeptide with antimalarial activity that can be synthesized in an enantiomerically pure form . Key steps in the synthesis of cyclomarin are asymmetric chelate−enolate Claisen rearrangement, asymmetric hydrogenation, and highly diastereoselective additions of organozinc and titanium reagents . The peptide synthesis developed to obtain the required building blocks is based on long-term experience in amino acid synthesis. Amino acid chelate enolate esters are used to obtain β-hydroxy through an aldol reaction, whereas γ,δ-unsaturated amino acids are obtained by transition metal-catalyzed allylic alkylation or Claisen enolate chelate rearrangement and syn-isomers are obtained by Claisen rearrangement. Thus, this approach was used to produce the desired N -protected amino acid (2 S ,3 R )- 50 . The protected 5-hydroxyleucine 56 was synthesized with the commercially available Roche ester 52 . The compound was reduced to the aldehyde 53 , and used in an olefination using Schimdt’s phosphonoglycinate 57 . The α,β-unsaturated amino acid derivative obtained was stereoselectively hydrogenated with ( R )-monophos as a ligand. Following N -methylation and saponification, amino acid 56 was produced with good yield selectivity . β-methoxyphenylalanine 61 is the third unusual building block that was prepared through several attempts shown in . The first attempt using phenylmagnesium bromide (3 equiv.) and DIBAL, which gave an acceptable yield of the addition products (60%), but only moderate diastereoselective products (7:3 ratio). Consequently, the titanium reagent was added as the single diastereomer. Subsequent O-Methylation, cleavage of the silyl ether, and oxidation then gave rise to an excellent yield of the target acid 61 . Another method was also reported for regioselectively prenylated from electron-withdrawing indole. The Pd-catalyzed protocol to 3-indolecarboxylate 62 was applied resulting in indole 63 . The reaction temperature was maintained at 0 °C to suppress N -prenylation, the ester was then saponified directly before the peptide coupling to avoid decomposition of the labile acid. The next building block was obtained by the ester saponification of compound 62 . The prenylated indole 64 was converted to the 3-iodo derivative 65 and the terminal double bond was changed to epoxide 66 . After several steps, niodosuccinimide (NIS) in MeOH produced methyl ester 72 . The O-silylated cyclomarin A was obtained by removing the two protective silyl groups using a two-step protocol in which the primary OH group was deprotected first using NH 4 F. After obtaining the necessary building blocks, cyclomarin A was then synthesized. Ring closure was planned between aminohexenoic acid and N -methylated leucine. BEP (2-bromo-1-ethylpyridinium tetrafluoroborate) 80 was used for coupling with protected valine. HCl was applied to suppress DKP formation during synthesis. The dipeptide salt was coupled to the next peptide with the activated compound 47 . The same protocol was then applied in the next coupling step. The HOBT reagent was used to activate N -methyl-4-hydroxyleucine 56 to avoid epimerization during the coupling reaction. This step obtained enantiomerically pure pentapeptide 77 . The methyl ester 72 was saponified and directly coupled using the BEP protocol to obtain compound 78 . Under Pd catalysis, the alloc protecting group was removed and combined with amino acid 50 , then the deprotected heptapetide was slowly added to an aqueous solution of PyBOP and base in CH 2 Cl 2 to obtain O-sylilated cyclomarin A 46 . The synthesis of another cyclomarin was also reported by Kiefer et al. (2019). Seventeen desoxycyclomarin-inspired derivatives ( 96 – 100 , 106 – 110 , 119 – 125 ) were synthesized using a straightforward solution-phase approach . First, an analog library was prepared stepwise by varying the N ′-side chain of the tryptophan moiety followed by simplifying the aminohexenoic acid unit. A Negishi reaction served as key step for the generation of N ′-isopropyltryptophan, with final saponification yielding the N ′-isopropyltryptophan derivative. d -hydroxyleucine units were replaced by d -hydroxynorvaline for additional structural changes in the peptide backbone. The synthesized fragment with protected glutamic acid ( 81 ) was activated as mixed anhydride and then reduced with NaBH 4 . The alcohol produced as TBS ether gives compound 82 , which is saponified and N -methylated to obtain N -methyl- d -hydroxynorvaline 84 . The first step of synthesis is focused on linear precursor assembly with varying tryptophan motifs and hexenoic amino acids . The intermediate compounds 86 – 90 were reacted with compounds 58 / 59 using the HOBt coupling reagent to produce compounds 91 – 95 , which were then cyclized with PyBOP and DIPEA coupling reagents to produce compound 96 – 100 . Then, compounds 106 – 110 were produced using the same synthesis steps. Intermediates 101 – 105 were cyclized with the same coupling reagent to produce the target compound. The second set of cyclomarin analogs was synthesized based on the key intermediate 106 . Several stages of coupling and deprotection were conducted to produce linear peptide compounds 112 – 118 . The linear compounds were then cyclized using PyBOP and DIPEA to produce compounds 119 – 125 . 3.1.4. Chaiyaphumine Gholap and Ugale (2017) reported the synthesis of chaiyaphumine-A ( 130 ) by convergent-based solution phase synthesis involving protection, deprotection, coupling and cyclization reactions . This novel class of peptide will the subject of much research to develop new antimalarial drugs. Five amino acids, L-threonine, D-phenyl alanine, D-alanine, L-proline and L-tryptophan, were assembled by coupling reactions involving HATU. The macrocylization of the side chain (-OH) of L-threonine and C-terminus of carboxylic acid from L-tryptophan using T 3 P yields chaiyaphumine-A . The synthesis of chaiyapumines was also undertaken by Lu and Batey (2022) . Fmoc-based solid-phase synthesis approaches were used for the synthesis of the requisite precursors. The solution-phase macrolactamization approach using EDC-HCl/HOBt was achieved without any detectable epimerization. This strategy was successfully applied for the total synthesis of the antimalarial compound chaiyaphumine A, as well as the natural products chaiyaphumine B, C, and D ( 136 – 139 ) using a late-stage acylation of the threonine amino group . Cyclized products were not produced by the macrolactonization method using MNBA/DMAP/iPr 2 Net, while the use of the additive Dy(OTf) 3 resulted in epimerization macrocycles. The slow rate of macrolactonization is influenced by the weak nucleophilicity of oxygen and the steric hindrance of threonine in the seco acid precursor, which allows competitive epimerization/cyclization to occur. Successful macrocyclization can be achieved via a macrolactamization approach that allows the total synthesis of all chaiyaphumine compunds. 3.1.5. Cyclopeptides Anolgs Prepared by Macrocyclization Fagundez et al. (2018) investigated the synthesis of a cyclopeptide using solid-phase peptide synthesis and on-resin macrocyclization, with various analytical techniques used to monitor reactions and purify compounds . The investigation compared solution- and solid-phase ring-closing reactions following the synthesis of linear precursor on a 2-chrotrytil chloride resin . Method A applied a combination of the solid-phase (linear precursor) and the solution-phase methods (macrocylization), whereas method B involved all reactions on-resin. A combination of DIC and Cl-HOBt as a coupling agent was used when the next amino acid added was Fmoc-L-Cys(Trt)-OH because it does not require the addition of a base and can decrease potent racemization. The solution-phase macrocyclization involved HBTU or HATU in the presence of DIPEA in DCM. The final products were obtained in moderate to good yields 26–80%. Eleven cyclic peptides ( 140 – 150 ) were synthesized (23 to 63% yield) using method A . Method B constructed Fmoc-L-Glu-OH and all amino acids on 2-CTC resin and the capping step took advantage of DIPEA and MeOH . Reagent [Pd(PPh 3 ) 4 ] in a solution of 10% piperidine in THF was used to remove the Fmoc-protected amino acid at the N -terminal group and produce a free carboxyl group at the C-terminal. This method yielded four antimalarial cyclopeptides ( 151 – 154 ) . A comparison of these methods is provided in . A year later after synthesizing compounds 140 – 154 cyclopeptides, Fagundez et al. (2019) developed cyclopeptides containing N -methyl amino acids with promising antiplasmodial activity against both erythrocytic and liver stages of malaria in vitro. These compounds could potentially be used as new and safe drugs to combat malaria . Compounds 156 – 161 were synthesized by employing solid-phase linear peptide synthesis and solution macrocyclization . The 2-chlorotrityl resin (2-CTC) was used to decrease the diketopiperazine formation, and HBTU and DIPEA were employed as coupling reagents in most cases. HCTU and DIPEA reagents were used to combine N -Me-Gly with the next amino acids because they are more effective. DIC and Cl-HOBt were then used as reagents to activate the Fmoc-L-Cys(Trt)-OH and coupling to resin-attached peptides to minimize racemization of Cys residues. Compounds 162 , 163 , and 164 were synthesized employing on-resin macrocyclization . First, Fmoc-L-Glu-O was anchored to 2-CTC resin with DIPEA, and then the Fmoc SPPS protocol was followed. After the removal of the allyl ester and Fmoc group with [Pd(PPh 3 ) 4 ] in a solution of 10% piperidine in THF, the ring closure was performed using DIC and Cl-HOBt. shows the structure of compounds 156 – 165 . 3.1.6. Hirsutellide A and Its Analogues Sahile et al. (2020) reported the synthesis and evaluation of hirsutellide A and its analogs for their antimycobacterial and antiplasmodial activities . Optimizing the structures of lipophilic antimicrobial cyclic peptides resulted in more membrane-permeable bioactive peptides, such as the lipophilic antimicrobial depsipeptide (hirsutellide A). The first report of the synthesis hirsutellide A was by Xu et al. in 2005; compound 167 had Ile residues instead of allo-Ile resulting in stereoisomers that do not have antimycobacterial potential . The compound was synthesized by a solution-phase method and initiated by the preparation of protected tridepsipeptide . D-Phe 168 was hydrolyzed into α-hydroxyl carboxyl acid 169 . After several steps, dipepsipeptide 172 coupling with N-Boc-allo-Ile-OH using HATU/DIPEA to generate tridepsipeptide 173 . The desired cyclic depsipeptide 177 was obtained via macrocyclization of the precursor 176 using HATU/HOBt/DIPEA in DMF . Furthermore, depsipeptide ( 178 – 181 ) analogs and peptide analogs ( 182 – 188 ) of hirsutellide A were designed and synthesized for preliminary SAR studies with a combination of solution and solid phase method . 3.1.7. Macrocyclization Strategies Recently, Zhang et al. (2022) synthesized macrocyclic peptides as species-selective antimalarial proteasome inhibitors . The compounds were prepared through several macrocyclization strategies such as Suzuki coupling, intermolecular amidation, intermolecular alkylation, intramolecular alkylation, and ring-closing metathesis (RCM) reaction. Synthesis of cyclic peptides 202 , 203 , and 204 took advantage of Suzuki coupling as a macrocyclization strategy . Benzylation and borylation reactions of N-Boc-3-bromophenylalanine 189 produced boronate. Deprotection of Boc in compound 191 was followed by combining with Boc-L-homophenylalanine to produce dipeptide intermediate 193 , which then underwent Boc deprotection and condensation with carboxylic acid 194 to produce compound 195 . Oxidative hydrolysis of boronic ester to boronic acid in 195 by NaIO 4 was followed by an intramolecular Chan–Lam coupling reaction to produce macrocycle 205 . Removal of the benzyl protective group by catalytic hydrogenation, followed by a coupling reaction of urea, produces 198 – 199 , which undergoes Boc deprotection to yield amines 209 and CP1. Furthermore, compounds 200 and CP1 undergo acetylation of free amines to produce macrocycles 202 and 203 . Macrocycle 204 is produced from amine 210 through subsequent alkylation with 4-bromobutanoate, Boc-deprotection, and lactamization using EDCI and HOBt. Intramolecular amidation as a macrocyclization strategy was used to synthesize compound 214 from compound 212 by hydrolysis of methyl ester and a coupling reaction using trifluoroethylamine. The synthesis began with deprotecting the tritile group in aziridine 205 , then protecting it again with the Boc group present in Boc-aziridine 206 . Compound 207 was then produced by an SN 2- type ring opening reaction mediated by BF 3 ·Et 2 O on aziridine 206 with 2-benzyloxyethanol. Boc-deprotection in compound 207 was followed by a coupling reaction with Boc-L-homophenylalanine to produce dipeptide 208 , then benzyl deprotection and Mitsunobu reaction of fragment 209 yielded the macrocyclization precursor 210 . Debenzylation and Boc deprotection of compound 210 was followed by intramolecular mediation to produce macrocycle 212 . The synthesis of compounds 229 – 237 were involved in intramolecular alkylation as a macrocyclization strategy . Dipeptide 225 was prepared from amino acid 224 by methyl esterification and HATU-mediated coupling reaction with Boc-L-homophenylalanine. Boc deprotection of dipeptide 225 and a coupling reaction with fragment 226 gave tripeptide 227 , which was then removed by two benzyl groups. The primary alcohol was chemoselectively protected with Tos groups, and Cs 2 CO 3 -mediated intramolecular. It underwent alkylation to give macrocycle 228 . After treatment of the methyl ester with sodium hydroxide, coupling of the free acid with various primary amines yielded macrocycles 229 , 230 , 231 , 232 , 236 , and 237 . Reduction of the methyl ester 228 with lithium borohydride formed alcohol 238 , and then Dess–Martin oxidation, followed by reductive amination with trifluoroethylamine, yielded macrocycle 233 . Condensation of the methyl ester 228 with hydrazine, followed by coupling with methyl trifluoroacetimidate gave the hydrazide 239 , which was thermally cyclized to give the 1,2,4-triazole 235 . Subsequent hydrolysis and dehydration of hydrazide 239 in the presence of the Burgess reagent gave 1,3,4-oxadiazole 234 . Macrocyclization with the intramolecular alkylation strategy was used to synthesize compounds 247 , 248 , 249 , and 254 . Fragments 231 – 233 underwent an amide coupling reaction with fragments 234 – 235 and subsequent deprotection, yielding dipeptides 236 – 239 , which were coupled to fragments 240 – 241 and 217 to yield tripeptides 242 – 246 . The tripeptides 242 – 246 were sulfated with a hydroxyl group, debenzylation of aryl benzyl ether, and intramolecularly substituted between the phenol group and the sulfonate ester, giving macrocycles 250 , 251 , 247 , 248 and 249 . Removal of the tert-butyl from compounds 250 – 251 using TFA was followed by an amide coupling reaction with a primary amine, which produced macrocycles 254 and TDI8. Synthesis of compounds 270 , 271 , 272 , and 273 was achieved by the RCM reaction . Fragments 231 and 255 underwent subsequent amide coupling reaction with fragment 256 ; Boc-deprotection and an amide coupling reaction with acids 261 – 264 afforded diolefins 265 – 268 , which were subjected to RCM in the presence of Grubbs Catalyst second Generation and reducing of the C=C double bond, yielding macrocycles 270 , 271 , 272 , and 269 . Removal of the Boc protection of 269 provided macrocycle 273 . 3.2. Linear Peptides 3.2.1. Falcitidin Somanadhan et al. (2013) succeeded in synthesizing linear antimalarial peptides. Falcitidin 284 is the first member of a new class of falcipain-2 inhibitors, which is a cysteine protease used by P. falciparum to degrade hemoglobin during the trophozoite stage of infection . Falcitidin contains isovaleric acid-D-His-L-Ile-L-Val-L-Pro-NH 2 and was synthesized by the solution-phase peptide synthesis method . The synthesis begins with deprotecting the Boc group from N-Boc-proline carboxamide using TFA. The amino-proline TFA salt was coupled with freshly prepared L-Ile-L-Val using HATU/HOAt to give L-Ile-L-Val-L-Pro-NH 2 . The dipeptide was obtained by coupling L-Ile-NHBoc with the in situ prepared bis-trimethylsilane (TMS) ether of L-Val under mixed anhydride conditions. The Boc group of the tripeptide was then cleaved using TFA and coupled to Na-Fmoc-N(im)-trityl-D-histidine using HATU/HOAt to give the Fmoc–tetrapeptide. The Fmoc group of was removed using piperidine and HATU/HOAt. The final synthetic target falcitidin was obtained after trityl deprotection of Trt-falcitidin using TFA in the presence of triisopropylsilane. Kotturi et al. (2014) synthesized falcitidin and its analogs ( 296 – 303 ) using the same solution phase peptide synthesis . First, N im -trityl-D-histidine 288 was converted into bis-TMS ether with TMSCl/Et 3 N 7 and coupled immediately with the mixed anhydride of isovaleric acid 287 using ethyl chloroformate/NMM to form the trityl protected N-acyl-D-histidine. In parallel, the N-Boc proline amide was deprotected with TFA, and the crude amine salt was coupled immediately with the known dipeptide N-Boc-L-Ile-L-Val by using HATU/HOAt to yield the key tripeptide intermediate N-Boc-L-Ile-L-Val-L-Pro-NH 2 . Lastly, the Boc group of tripeptide was cleaved using TFA and coupled to N-acyl-D-histidine using HATU/HOAt to give the N im -trityl protected tetrapeptide 292 . Falcitidin acylatide and its analogs produce eight analogs compounds ( 296 – 303 ) by diversifying the synthesis of the N-acyl tetrapeptide analog of falcitidin through the common tripeptide N-Boc-L-Ile-L-Val-L-Pro-NH 2 . 3.2.2. Gallinamide Gallinamide A is a linear peptide that has antimalarial activity. Conroy et al. (2014) designed and synthesized 18 gallinamide A analogs by solid-phase synthesis and substitution of amino acid residues . Four gallinamide A analogs ( 309 – 312 ) were originally designed with varying levels of saturation, including compound 309 , which is structurally identical to gallinamide A. Another analog was compound 310 with a reduced methoxy-enol moiety in the pyrolinone ring 309 . Part of the olefinic component of the 4( S )-amino-2-( E )-pentenoic acid unit is reduced in compound 311 and 312 has reduced olefin groups. Analog synthesis was initiated by synthesizing the N -terminal fragment 304 by solid-phase peptide synthesis using the Fmoc strategy. The 2-CTC resin is filled with Fmoc-Leu-OH followed by the subsequent incorporation of the amino acid. Reductive amination of the resin, followed by cleavage of the resin using hexafluoroisopropanol (HFIP), produced an excellent yield of tripeptide 304 which was coupled with imide fragments 306 and 307 prepared using a similar protocol. The coupling reaction involved HATU at low temperature to avoid epimerization to produce good yields of analogs 309 and 310 . At this stage, 309 and 310 underwent hydrogenation to give 311 and 312 . Analogs 313 – 318 were synthesized from 2-CTC resin filled with Fmoc-Ala . The coupling of Fmoc-protected α,β-unsaturated amino acids was followed by elongation through the Fmoc solid-phase peptide synthesis strategy. With reductive methylation of the N-terminus using formaldehyde and sodium cyanoborohydride, the peptide was cleaved from the resin using HFIP to produce the C-terminal peptide acid. C-terminal functionality was achieved through benzylamine coupling using PyBOP at low temperatures, and compounds 313 , 315 , 317 , and 318 were purified by RP-HPLC and demonstrated no significant epimerization. Compound 314 was produced by attaching 4-( R )-hydroxy L-proline methyl ester to the C-terminus of the two peptides with the addition of NMM as a hindered base. Aminothiazole was coupled to the C-terminus of one peptide acid using PyBOP at low temperature to yield compound 316 as a mixed diastereoisomer. A further six analogs ( 346 – 351 ) were designed, possessing the identical peptide backbones to 309 and 310 but with a variation in the side chain on the pyrolinone unit and the enol substitution of pyrolinone . The synthesis of compounds 346 – 351 was initiated by preparation of the requisite pyrolinones 333 – 338 from Fmoc-protected amino acids 319 – 322 . Then, Meldrum’s acid in ethyl acetate was coupled with amino acids 319 – 322 using EDC and DMAP followed by reflux of the Meldrum’s adduct in ethyl acetate. This step affected the condensate cyclization to give Fmoc-protected pyrolinones 323 – 326 , which were then reacted with methanol under Mitsunobu conditions using DIAD and triphenylphosphine, respectively, to produce O-methylated pyrolinones 327 – 330 . Compounds 323 – 324 were reacted with benzyl alcohol under the same conditions to produce 331 – 332 . Compounds 333 – 338 were produced by treating piperidine in acetonitrile. After each pyrrolinone building block was obtained, amino acid 339 was activated as the corresponding pentafluorophenyl ester. Separately, pyrolinones 333 – 338 were deprotonated with n-butyllithium at low temperatures before the addition of pentafluorophenyl ester to yield 340 – 345 (36–59%). Due to the unfavorable results, acidolysis of the Boc groups from 340 – 345 was followed by coupling to the N-terminal tripeptide 304 using the coupling reagent HATU and NMM as the base to produce the desired gallinamide A analogs 346 – 351 with excellent yields. The natural product gallinamide A was also synthesized by Stoye et al. (2019) and possesses potent inhibitory activity against P. falciparum cysteine proteases, namely falcipain, and therefore shows promise as a potential malaria treatment to breakdown hemoglobin in the parasitic food vacuole . Gallinamide A was synthesized using NMM in a solution of the imide fragments 353 – 354 (1.0 equiv, as the trifluoroacetate salt), tripeptide 334 – 336 (1.5 equiv, as the trifluoroacetate salt), HATU and HOAt in DMF/CH 2 Cl 2 (1:1 v / v ). After consumption of the starting material (as evidenced by LC-MS), the solvent was subsequently removed by an N 2 stream and the residue was purified by preparative RP-HPLC . This synthesis produced different yields of gallinamide A analogs 371 – 387 ( and ). 3.2.3. PfSERA5 Analogs PfSERA5 is an abundant asexual antigen that can inhibit parasitic growth in vitro and is a candidate malaria falciparum vaccine. Kanodia et al. (2014) reported the design and synthesis of peptides with similar sequences to the SERA5 protein, an inhibitor of malaria parasite development . The solid-phase synthesis of all peptides was initiated on Rink amide resin. The appropriate Fmoc-amino acid was dissolved in a solution of TBTU/HOBT and DIPEA before being added to the resin. Fmoc was deprotected by piperidine in DMF and then acetylated by a mixture of acetic anhydride and DIPEA. The nine SE5 P1-P9 ( 388 – 396 ) were cleaved from the resin using TFA/H 2 O/TIS/Thioanisole/Phenol and their sequences are shown in . 3.2.4. Angiotensin Silva et al. (2015) investigated angiotensin II, a peptide that has antiplasmodial activity, as an antimalarial drugs . They synthesized 10 peptides ( 397 – 406 ) via manual solid-phase synthesis . The Fmoc strategies were applied using Wang resins and deprotection was performed by treatment with 4-MePip in DMF. Couplings were conducted using DIC/HOBt in DCM/DMF (1:1, v / v ) and were monitored using the Kaiser ninhydrin test. Dry-protected peptidyl resin was exposed to TFA/H 2 O/anisole (95:2.5:2.5, v / v / v ) for 2 h at room temperature to produce the crude linear peptides. All crude peptides were precipitated with anhydrous diethyl ether, separated from the ether–soluble reaction components by filtration, extracted from the resin with 60% ACN in water, and lyophilized. They were purified by preparative RP-HPLC in 0.1% TFA/60% ACN in water on a Waters Associates system (Delta Prep 600). The peptides were loaded onto a Phenomenex C18 (21.2 × 250 mm, 15 µm particle size, 300 Å pore size) column at a flow rate of 10.0 mL/min and eluted using a linear gradient (slope 0.33% B/min) of TFA/ACN with detection at 220 nm. Selected fractions containing the purified peptides were pooled and lyophilized. In 2015, Silva et al. also synthesized nine octapeptides ( 407 – 415 ) of the renin-angiotensinsystem (RAS), which have been reported to have anti-plasmodium activity towards P. gallinaceum (88% sporozoite inactivation) . The angiotensin II analogs were synthesized by Fmoc/tert-butyloxycarbonyl (t-Boc) strategy on a solid phase, purified by liquid chromatography, and characterized by mass spectrometry. The amino acid Nα-terminal protecting group was removed with TFA in DCM in the presence of 2% anisole for 20 min. Coupling and deprotection reactions were carried out using the same reagent as before. Repeated couplings were performed for one hour using TBTU with DIPEA in DCM/NMP. Dry protected peptidylresin was exposed to 70% TFA/20% TFMSA in 10% anisole for 12 h. All crude peptides were precipitated with anhydrous diethyl ether, separated from the ether–soluble reaction components by filtration, and the fractions were lyophilized using 60% ACN (acetonitrile) in water. Ten angiotensin analogs ( 416 – 425 ) were also synthesized by Torres et al. (2015) . These lactams and sulfide bridge-containing peptides were synthesized manually by the solid-phase method using a t-Bu strategy with chloromethylated resin and a Fmoc strategy with Wang resin. Deprotection was conducted with TFA and DCM and the addition of 2% anisole. The resin was washed with anisol in isopropyl alcohol and TEA in DCM and MeOH, and the reaction was monitored with the Kaiser test. Recoupling and deprotecting the protective groups on each amino acid used the same reagents as before. Furthermore, the cyclization reaction was conducted with excess Castro reagent and DIPEA dissolved in DMSO/NMP, after which the resin was washed with TEA in DCM to obtain a dry peptidyl-resin. Deprotection of the sulfide bridge analogs was conducted by adding 4-MePip in DCM. The amino acid coupling reaction involved treating the protected amino acid acyl-resin with a molar excess of the Boc/Fmoc protected amino acid using the DIC/HOBt reagent. The disulfide bridges were formed by first dissolving the peptide in an acetic acid solution containing iodine. The peptides were then lyophilized after being extracted with water and diethyl ether and purified by RP-HPLC. Silva et al. (2017) synthesized eight angiotensin II hormones ( 426 – 433 ) . They optimized the synthesis of these peptides for their antiplasmodial activity and to reduce their vasoconstriction and rapid degradation characteristic. The peptides were synthesized by manual solid-phase synthesis and characterized by circular dichroism spectroscopy. The synthesis was similar to previously reported and involved manual phase peptide synthesis using the Fmoc strategy and Wang resin. The amino acid residues were coupled using the DIC/HOBt reagent in DCM. Each step was followed by alternating washing with DMF, methanol, and DCM to change the degree of resin swelling and promote the removal of excess reagents. Dry peptidyl resin was added to TFA/water/anisole to obtain unprotected peptides. The S-S bonds were formed by first dissolving the crude peptide in an iodine-containing acetic acid solution. The peptides were extracted with water and diethyl ether, evaporated, and then lyophilized. 3.2.5. Decoralin Torres et al. (2018) reported the re-engineering of a wasp venom peptide, decoralin, into a synthetic anti-malaria agent through modifications that removed its hemolytic activity . Dec-NH 2 and eight analogs ( 434 – 441 ) were synthesized and designed to preserve specific physicochemical structures. Certain amino acid substitutions significantly improved the antiplasmodial activity, giving new sequence principles for creating potent anti-malaria drugs such as replacing the original sequence by Arg, Phe and Trp. The synthesis involved a solid-phase method and purification by chromatography and characterization using MS. The antiplasmodial activities were assessed by fluorescence microscopy. All residue substitutions resulted in increased anti-Plasmodium activity , with [Arg] 1 -Dec-NH 2 , [Pro] 4 -Dec-NH 2 , and [Phe] 2 -Dec- NH 2 being the most active peptides tested. 3.2.6. Carmabin and Dragomabin Ye et al. (2018) synthesized carmabin A and dragomabin, which demonstrated antimalarial activity . Carmabin A and dragomabin synthesis was initiated by retrosyntethic analysis . The C-terminal amide of carmabin A ( 442 ) and dragomabin ( 443 ) was prepared via amidation of the C-terminal methyl esters of compounds 444 and 445 . Compounds 448 were further disconnected into two parts: the Mdya/Moya fragment and protected tetrapeptide 454 ( and ). Tetrapeptide 44 8 was prepared by repeated condensation of amino acids . The methyl group of Mdya/Moya 446 / 447 could be stereoselective. Carboxylic acid 454 was converted to the corresponding alcohol 453 in four steps involving acyl chlorination, amidation with benzyl-2-oxazolidinone, diastereoselective-methylation, and reduction. The addition of carboxylic 454 with N HCl led to the hydrolysis of the TMS group and the amide bond to yield compound 446 . shows the construction of tetrapeptide 447 , the treatment of compound 457 , and TFA resulting in Boc deprotection. The condensation of compound 458 with the coupling reagent HATU/DIPEA produced compound 459 and repeated condensation under the same conditions yielded tetrapeptide 447. Carmabin A was synthesized from the building blocks 446 , ent- 446 , and 447 ( and ). The coupling reaction between Boc-deprotected 447 and 446 and ent- 446 was followed by the addition of ammonia to produce compounds 442 and 442a . The configuration of carmabin A was further investigated using NMR, showing that only the 442 NMR data matched that of natural carmabin A. Dragomabin was synthesized using the building blocks Moya 447 and ent -447 . First, Moya 447 is generated from ent -452 by deprotecting with TMS and TFA, then the chiral auxiliary is removed (Schema 38). Dragomabin is obtained by deprotecting 448 with TMF under DCM to provide 445 , which is amidated with NH 3 to form compound 443 . The correct structure for dragomabin was revised as shown in 443a . As dragomabin and dragonamide differ in the stereochemistry on the Moya fragment, the stereochemistry of the alkyne fragment in these lipopeptides is varied, and correlation with other natural products is not reliable. The absolute stereochemistry at C35 and C37 of carmabin A was assigned as 35R, 37S and the absolute stereochemistry at C35 of dragomabin has been revised as 35R. 3.3. Biology Activity of Isolated and Synthesized Compounds Ribifolin ( 1 ) is moderately effective against the Plasmodium falciparum 3D7 strain with an IC 50 of 42 µM , while its linear analogue 1a had an IC 50 of 519 µM, demonstrating the importance of cyclization to enhance the biological activity . None of the tested compounds showed cytotoxic potential against human cells (HEK293aT), although normal growth was observed in the concentration range of 0.001–100 μM. The isolated cyclopeptide alkaloids ( 2 – 5 ) from the root bark of Hymenocardia acida demonstrated antiplasmodic activity against P. falciparum K1, with IC 50 values ranging from 12.2 to 27.9 μM. Only hymenocardine ( 2 ) exhibited cytotoxic properties against MRC-5 cells with an IC 50 of 51.1 ± 17 μM . The antiplasmodic activity of nummularine-R ( 6 ), O-desmethylnummularine-R ( 7 ), hemsine-A ( 9 ), ramosine-A ( 11 ), and oxyphylline-F ( 14 ) was evaluated against the P. falciparum strain KI, while cytotoxicity was tested on MRC-5 cells (human fetal lung fibroblast cells). Based on the structural features and IC 50 values for compounds 6 , 7 , 9 , and 11 , it can be assumed that the tryptophan moiety in the side chain is important for antiplasmodic activity. Compound 14 does not contain a tryptophan unit and also showed antiplasmodic activity, indicating that tryptophan could mediate, but is not vital for the antiplasmodic activity. The most promising compound was O-desmethylnummularine-R ( 7 ), which exhibited an IC 50 of >64.0 µM against MRC-5 cells . Maluf et al. (2016) reported that crotamine ( 15 ) selectively enters infected erythrocytes ( A) and has potent anti-plasmodial activity with an IC 50 value of 1.87 µM ( B) . The instability of H + homeostasis by cr otamine was also confirmed ( B,C). The authors suggest that crotamine alters the internal pH of the vesicle due to its abundant Lys and Arg residues and the resulting high net charge (8+). Internal pH regulation is important for parasite survival, as it regulates the activity of certain intracellular enzymes required for parasite growth. Thus, this polypeptide is a promising lead molecule for the development of potential new peptidomimetics that have selectivity for infected erythrocytes and the ability to inhibit malaria infection by their natural affinity for acid vesicles . The cyclic octadepsipeptides, octaminomycins A ( 16 ) and B ( 17 ) showed activity against P. falciparum . According to Jang et al. (2017), compounds 16 and 17 were not significantly cytotoxic at a concentration of 30 μM against human cervical cancer cells (HeLa), human promyelocytic leukemia cells (HL-60), mouse temperature-sensitive cdc2 mutant cells (tsFT210), and rat kidney cells, which were infected with ts25 (srcts-NRK). They were also evaluated for antimicrobial activity against Staphylococcus aureus 209, Escherichia coli HO141, Aspergillus fumigatus Af293, Pyricularia oryzae kita-1, and Candida albicans JCM1542, as well as antimalarial activity against the P. falciparum 3D7, Dd2, and K1 strains. Chloroquine was less effective against strains resistant to Dd2 and K1, whereas compounds 16 and 17 showed the same in vitro antimalarial activities against chloroquine-sensitive 3D7 and chloroquine-resistant Dd2 and K1 strains, with no antimicrobial activity up to 30 μM . The antimalarial activity of the isolated kakeromamide B peptide ( 18 ) was evaluated against asexual blood-stage and liver-stage P. falciparum . Compound 18 exhibited moderate activity against the blood stage of P. falciparum with an EC 50 value of 8.9 µM as well as moderate liver-stage antimalarial activity against P. berghei liver schizonts with EC 50 values of 11 µM. Although 18 displayed only moderate antimalarial activity, its ability to inhibit both the blood and liver life stages of Plasmodium, coupled with its low cytotoxicity in human cell lines, make it a promising lead compound for drug discovery . The novel cyclic antimalarial and antitrypanosomal hexapeptide, pipecolisporin ( 19 ), was isolated from cultures of Nigrospora oryzae CF-298113 and exhibited interesting activity against P. falciparum and Trypanosoma cruzi Tulahuen C4 parasites . The activity against the T. cruzi Tulahuen C4 parasites was the most remarkable, with an IC 50 of 8.46 µM, comparable to that of the standard drug benznidazole, currently used in the treatment of Chagas disease (IC 50 in the same assay of 2.21 µM) ( and ). The activity against P. falciparum was also in the micromolar range, with an IC 50 of 3.21 µM . It was not cytotoxic to the human cancer cell lines A549 (lung carcinoma), A2058 (metastatic melanoma), MCF7 (breast adenocarcinoma), MIA PaCa-2 (pancreatic carcinoma), and HepG2 (hepatocyte carcinoma) at the highest concentration tested of 50 µM . Koshidacins A ( 20 ) and B ( 21 ) demonstrated in vitro antiplasmodial activity and cytotoxicity against human MRC-5 cells . Compounds 20 and 21 demonstrated antiplasmodial activity against a chloroquine-sensitive P. falciparum FCR3 strain with IC 50 values of 17.1 and 0.89 μM, respectively. Similarly, compounds 20 and 21 also exhibited antiplasmodial activity against chloroquine-resistant P. falciparum K1 strain with IC 50 values of 12.5 and 0.83 μM, respectively. Compounds 20 and 21 showed cytotoxicity against human MRC-5 cells, with IC 50 values of 6.8 and 14.7 μM, respectively, suggesting selectivity indices ranging from 0.4 to 18.4 . In addition, when given intraperitoneally at a dose of 30 mg/kg/day for 4 days, compound 21 inhibited 41% of malaria parasites in vivo . Shi et al. (2018) evaluated georatusin ( 22 ) produced by a soil fungus Geomyces auratus . It had no obvious cytotoxicity, but displayed antiparasitic activities against Leishmania donovani (IC 50 = 9.1 μM) and P. falciparum (IC 50 = 1.6 μM) . This discovery offers new insight into the metabolic potential and ecological importance of Geomyces and may encourage further exploration of this genus. New highly N-methylated linear peptides, friomamaride B ( 23 ) and shagamides A-F ( 24 – 29 ), exhibited activity against three strains of blood-stage P. falciparum . All compounds were tested for their effectiveness against three blood-stage of P. falciparum using the blood-stage antiplasmodial and cytotoxicity assay. Friomamaride B ( 23 ), shagamides C ( 26 ) and D ( 27 ) with values less than 10 µg/mL all possess potential activity. The N-terminal of phenylalanine residue is essential for this activity. None of the isolated compounds demonstrated cytotoxic activity. The carboxypeptidase inhibitor NpCI peptide ( 30 ), which is related to the model enzymes of bovine carboxypeptidase A (bCPA) and porcine carboxypeptidase B (pCPB), was discovered by Cabrera-Muñoz et al. (2023) . The kinetic characterization of NpCI revealed that it was a slow inhibitor of bCPA and pCPB. While pCPB inhibition was not significant, the evaluation of NpCI inhibition was also performed by comparing the decrease in bCPA inhibitory activity caused by the substrate with the increase in substrate concentration. The Dd2 strain was used for in vitro antiplasmodial activity against P. falciparum , showing that the cycle was significantly delayed with an IC 50 of 5.5 µmol/L. Parasite growth can be slowed down by increasing inhibitor concentration. The growth inhibition by NpCI occurs during parasite development. The 3D7 strain displayed comparable inhibition with a delayed maturation mechanism. NpCI was not cytotoxic to human cells (IC 50 < 25%). Kiefer et al. (2019) synthesized 17 cyclomarin analogues ( 96 – 100 , 106 – 110 , 119 – 125 ) and evaluated their biological activities against chloroquine-sensitive Pfalcp strain 3D7 and multi-resistant strain Dd2, as well as against Mtb wild-type strain Erdma . The antitubercular activity was also evaluated, and the resazurin reduction microtiter assay (REMA) was used to measure the growth inhibition of Mtb. Three compounds ( 96 , 97 , 99 ) demonstrated excellent parasitic growth suppression . Compound 99 consists of a simplified γ,δ-unsaturated side chain, and the N’-methyl tryptophan unit. Desoxycyclomarin C, which was inspired by a natural product compound, has a similar range of bioactivity, but it is significantly shortened. The anti-mycobacterial activity was demonstrated by the five derivatives ( 96 , 98 , 99 , 108 and 110 ) when applied to Erdman wild-type strain. Due to a remarkable simplification by replacing two of the four non-canonical amino acids by L-valine and L-tryptophan, compound 110 represents a highly appealing natural product-derived lead structure for battling Mtb. Fagundez et al. (2018) revealed that most of the synthesized peptides have the potential to be antimalarials . The corresponding cyclopeptide effectiveness was evaluated against the chloroquine-resistant K1 strain of P. falciparum . Cyclo-Cys(Trt)-Gly-Thr(tBu)-Gly-Cys(Trt)-Gly (compound 140 ) showed potent in vitro and selective activity against this parasite with an EC 50 = 28 µM. The inclusion of a carboxylic group from Glu ( 151 , 152 ) can increase solubilization, and the substitution of hydrophobic amino acids can boost biological activity . Fagundez et al. (2018) also developed cyclopeptides containing N-methyl amino acids that demonstrated promising antiplasmodial activity, compounds 156 – 161 and 163 – 165 . In addition, a new class of antimalarial cyclopeptides that contain N-methyl Gly has been developed that exhibits enhanced antiplasmodial activity. The in vitro evaluation of the compounds against P. falciparum revealed that N-Me-Gly is required to maintain the activity in the presence of a Glu with a free carboxyl group. Moreover, none of the active compounds are toxic against HepG2 cells. Compounds 158 and 160 were assessed for their antimalarial activity in comparison to other antimalarial drugs . These results point to an antimalarial mode of action that does not immediately affect parasite viability. The prophylactic potential of compounds 158 and 160 is demonstrated by their low and submicromolar EC 50 values (0.018 and 0.355 µM) in the liver stage of the parasite. Compounds 157 and 165 are effective against P. berghei parasites, reducing parasitemia by 71 and 66% on day 5. To evaluate the oral bioavailability, the plasma pharmacokinetic of compound 157 in male Swiss Albino mice, following a single oral dosage, was examined, and shows a considerable half-life of 4.93 h. The hirsutellide A analog, compound 177 , exhibited moderate antiplasmodial activity (IC 50 = 2.3 μM) similar to that reported for hirsutellide A (IC 50 = 4.2 μM). Compound 177 is not cytotoxic (IC 50 > 100 μM) to Hep2G cells. ADME profiling for compound 177 displayed moderate stability in humans, but low stability in mouse microsomes . Additionally, the analogs have little to no activity against Mtb H37Rv. Peptide analogs generally have higher antiplasmodial activity (IC 50 = 1.8−7.7 μM) than the depsipeptide analogs (IC 50 = 7.5−20.1 μM), exhibiting a higher aqueous solubility, a high plasma stability and mouse plasma stabilities. Thus, the ester-to-amide substitution and the membrane permeability of hirsutellide A analogs appear to be dependent on the nature of the amino acid substituents. This study provides insight into the structural features relevant to the cyclic peptide-related drugs. Some compounds exhibited promising antimycobacterial activity and low cytotoxicity, making them potential candidates for further research and development as anti-TB agents. The developed macrocyclic peptides as proteasome inhibitors has potential for antimalarial drugs (Zhang et al., 2022) . Cyclic peptide 201 is a noncovalent inhibitor with strong antimalarial activity and high species selectivity, but has poor pharmacokinetics, therefore a docking model was developed. Compound 201 was acetylated using the N-Terminal amino group to boost its antiparasitic and inhibition activities. Five compounds were created, and P1, P3, P5, and P2–P4 linkers were examined for their contribution to the potency and drug-like properties of macrocyclic peptides ( , , and ) . Compound 220 demonstrated good potency against parasite and pharmacokinetics and made optimal use of P1 amide moiety modification . Compound 223 showed moderate inhibition against hu-β2c (IC 50 = 97.4 μM). Compounds 272 and 270 exhibited moderate inhibition of β1c and β1i. All other compounds showed <50% inhibition against all the four subunits, even at 100 μM. Similar to the observation of the simultaneous inhibition of β5 and β2 in tumors, co-inhibition of β5 and β2 is synergistic and the inhibition of Pf20s β5 is sufficient to kill Plasmodium. The compounds were optimized for aqueous solubility, passive membrane permeability, metabolic stability, and a clean off-target profile against CYP450s and the hERG channel. The proteasome function appears to be critical for parasites to survive. Falcitidin and its analogs ( 296 – 303 ) were synthesized by Kotturi et al. (2014) and showed moderate activity against P. falciparum 3D7, but only when N-tritylated on its histidine residue . The IC 50 activity of the new compounds, which ranged from 1 to 5 µM, was typically modest in whole cells. Compound 302 exhibited the highest activity (IC 50 0.14 µM). These new compounds represent an important new peptide chemotype that may be elaborated into improved antimalarial leads . Conroy et al. (2014) developed a new class of antimalarial drugs based on the natural product gallinamide A, which inhibits the falcipain cysteine proteases essential for malaria parasite survival . These gallinamide A analogs were as effective as chloroquine against the P. falciparum parasite. Gallinamide A analog 309 has strong inhibitory activity against FP-2, FP-3, and P. falciparum in vitro . Reducing the enol moiety in the acyl-pyrrolinone unit in compound 310 resulted in a two-fold reduction in antiplasmodial activity (IC 50 = 210 nM), but a slight improvement in activity against FP-2 and FP-3. Analog 311 showed no measurable inhibition of FP-3 and a 3 orders of magnitude drop in inhibitory effectiveness against FP-2 (IC 50 = 3710 nM), as well as significantly decreased activity against P.falciparum. Analog 312 lost the measurable inhibitory activity against the FPS and the parasite after losing both of its olefinic moieties. The activities of analogs 313 , 315 , and 317 were similar: they inhibited FP-2 at low micromolar concentrations (IC 50 = 3.4–11.5 µM), did not significantly inhibit FP-3 at 25 µM, and inhibited P. falciparum at nanomolar concentrations (IC 50 = 320–5400 nM) . Interestingly, peptide 318 demonstrated less potent antiparasitic activity (IC 50 = 6.6 µM) but inhibited FP-2 and FP-3 by adding an N-methylproline functionality at the N-terminus of the peptide while keeping the C terminal benzylamide. The lack of activity was especially noticeable for 315 because it has the same structure as analog 309 . The addition of a more flexible and highly functionalized hydroxyproline methyl ester to the C-terminus in 314 produced inhibitory activity against FP-2 and P. falciparum , which was comparable to the benzylamide-derived compounds. Low micromolar antiparasitic activity (IC 50 = 1.1 µM) and moderate inhibitory activity against FP-2 and FP-3 were achieved by C-terminal functionalization as a thiazole amide in compound 316 . Like analog 309 , the C-terminal N -acyl-pyrrolinone in 346 – 351 restored strong inhibitory activity against the FPs and P. falciparum . Compound 346 displayed similar inhibitory activity to 309 against FP-2, FP-3, and P. falciparum despite having no modification on the pyrrolinone molecule. The pyrrolinone ring of 347 was modified to include a more hydrophobic substituent, which improved its inhibitory efficacy against FP-3 and P. falciparum . As with compound 347 , the addition of aromatic side chains to the pyrrolinone ring did not modify the inhibitory activity against FP-2 but significantly increased activity against FP-3 and P. falciparum . The most effective inhibitor of FP-3 and P. falciparum was compound 349 , which had an indole side chain attached to the pyrrolinone ring . The compounds were screened against the P. falciparum , namely AP M1, AP M17, and AP M18 aminopeptidase (AP) enzymes. The effective N-acyl pyrrolinone analogs 309 and 346 – 349 , as well as the C-terminal amide derivatives 309 , 315 , and 317 , had IC 50 values of less than 600 nM against the 3D7 strain of P. falciparum . The CQ-resistant, Dd2 strain of P. falciparum was powerfully inhibited by every tested compounds (IC 50 = 29.0–421 nM). The compounds were selective inhibitors of P. falciparum over HEK298 cells, with 309 , 315 , and 317 showing no measurable inhibition of this cell line at a concentration of 50 µM, but having strong inhibitory activity against human cathepsin . The gallinamide A analogs, synthesized by Stoye et al. (2019), were also tested for activity against FP2, FP3 also CQ-sensitive 3D7 and CQ-resistant W2 strains of P. falciparum in vitro . Most substitutions in the compound structure were well tolerated, except for analog 387 which had a more potent activity parasite (IC 50 3D7 = 1 nM; W2 = 4 nM) and substituents on R 1 which markedly enhanced the stability in plasma and blood. Five gallinamide A analogs were assessed in vitro for their stability in mouse blood and plasma . The half-life in plasma and blood was significantly prolonged due to the substituents at R 1 and R 3 . Additional cytotoxicity assays in the HEK293 cell line of the compounds displayed no measurable cytotoxicity at 25 µm. The last three analogs 372 , 373 and 387 were assessed in vivo in a mouse model of cerebral malaria (CM), P. berghei ANKA (PbA) infection. Severe signs of disease as well as the rise in parasitemia were both significantly delayed by analog 387 . The peptides derived from PfSERA5 by Kanodia et al. (2014) inhibited the enzymatic activity of PfSERA5P50 protein, which in turn blocked the development of the parasites in an in vitro culture . Evaluation of PfSERA5 derivative against P. falciparum development and growth revealed that SE5 P1 ( 388 ) and SE5 P2 ( 389 ) have the highest parasite growth inhibition and invasion values of 60–70% . Analysis of the synthetic substrate Suc-LLVY-AMC revealed that PfSERA550′s proteolytic activity was greatly reduced by the two C-terminal peptides SE P1 ( 388 ) and SE5 P2 ( 389 ). Kanodia et al. also describe how they used computational modeling to examine molecular docking studies with the known crystal structure of PfSERA5 to explore the effects of SE5 P1 ( 388 ) and SE5 P2 ( 389 ) peptides. SE5 P1 ( 388 ) and SE5 P2 ( 389 ) occupied 50.1 and 57.5% larger than the substrate when the Suc-LLVY-AMC was applied to see the bond , suggesting that the peptides must interact with Glu638 and Ser640/Ser641 to be inhibitory . P. falciparum -infected RBCs uptake of labeled SE5 P1 and SE5 P2 peptides when the localization of biotinylated peptides therein inhibited the protease activity similarly to the non-biotinylated peptides. This indicates that the peptides have access to the intracellular parasites and are co-localized with the PfSERA5 protein. PfSERA5 plays an important role in parasite development and the final proteolytic cleavage, which can be produced as a new drug design and offers information on the potential use of these peptides as antimalarial therapeutics . Angiotensin II analogs synthesized by Silva et al. (2015) and tested in vitro to identify a short bioactive peptide as well as to verify the hydrophobic cluster’s influence on parasite-membrane interaction on both P. gallinaceum and P. falciparum . Fluoresence microscopy was utilized to examine the effects of the peptides on P. gallinaceum sporozoites produced by the salivary glands and the therapeutic index (MHC/MIC ratio). Higher values in the therapeutic index indicate more antimicrobial specificity. It is a parameter that measures the specificity of an antimicrobial agent and is calculated by the ratio MHC and MIC. The MHC/MIC ratio, which was higher in peptide 401 , indicates that the peptides have varied specificities . New peptides related to Ang II were designed, including the most hydrophobic amino acid residues (Val, Ile, Pro, and Phe), aromatic residues (Tyr, His, Pro, and Phe) and residues from the Ang II hydrophobic cluster (Tyr, Ile and His) in an attempt to verify the peptide–parasite interactions. Due to the influence of hydrophobic clusters, side chain aromatic rings, and hydrophobic residues, these peptides showed antiplasmodial activity in P. gallinaceum sporozoite (64–94%) and activity between 89 and 94% . The three peptides with the highest antiplasmodial activity were 1 ( 397 ), 5 ( 401 ), and 6 ( 402 ) with 94, 89, and 94%, respectively . The effect of the peptides in the P. falciparum erythrocytic cycle was assessed in vitro. All peptides reduced new ring formation at 10 −8 mol L −1 , which was studied by Saraiva et al. as the ideal concentration for these inhibition assays. Four analogs reduced the ring formed in the blood stage, but only analogs 5 ( 401 ) and 6 ( 402 ) showed inhibition that was higher by 50% than control . showed that after 24 h of incubation with 2–3% schizont-infected erythrocyte cultures in the absence (control) or presence of 10 − 8 mol L − 1 peptides, the percentage age of rings was evaluated (* p < 0.05 compared to the control, *** p < 0.001). The result is statistically significant compared to the (mean ± standard deviation, n = 2), as indicated by the dark grey shading . Another technique for determining the hemolytic effect of peptides demonstrated that peptides 401 and 402 had an effect of the Ang II to define the hemolysis. These peptides did not exhibit hemolytic effects . The hydrophobic portion and the Arg, Tyr, Pro, and Phe residues increased the antiplasmodial activity when they were present in the primary sequence. Furthermore, these peptides did not display hemolysis or contractile response activities, as shown in (*** p < 0.05 compared to control, n = 2) . The IC 50 values are more promising than the Ang II results, which were demonstrated by evaluating seven concentrations giving 7–65% inhibition . The results imply that intramolecular interactions cause conformational tendencies that are crucial for antiplasmodial activity. When present on the peptide primary sequence, the hydrophobic portion and the residues of Arg, Tyr, Pro, and Phe increased the antiplasmodial activity. Following in vivo model tests of this class of peptides, this type of research aids the development of new chemotherapeutics, which can be explored as antimalarial drugs . The synthesized RAS octapeptides were tested for their effectiveness against P. falciparum and P. gallinaceum , revealing that only [Ala 5 ]-Ang II showed equipotent Ang II activity with 45% of biological activity . At 10 −8 M concentration, P. gallinaceum erythrocytic cycle [Ala 6 ]-Ang II decreased parasite invasion by 49%. The efficiency of anti-plasmodial activity was significantly impacted by the presence or absence of amino acid substitutions. The anti-plasmodial activity of the Ang II molecule depends on specific amino acid side chains. The biological activities and binding affinities are affected when the biologically active peptide is replaced by an alanine residue. From , the percentage of rings was determined after 24 h of incubation of erythrocyte culture infected with 2–3% schizonts in the absence (control) or in the presence of 10 − 8 M analogs. Dark grey shading denotes that the result is statistically significant compared to the control. Regarding the contractile response, [Ala 5 ]-Ang II and [Ala 6 ]-Ang II did not promote contractile activity compared to Ang II, and carbachol [Ala 5 ]-Ang II and [Ala 6 ]-Ang II reduced parasite invasion in red blood cells in the erythrocyte hemolysis assays, and presented no effect on cells integrity at 10 −8 M concentration. This model directs new possibilities for peptide design that can act more effectively in preventing the erythrocytic cycle of the parasite and other phases of the human–malaria cycle. demonstrates that after erythrocyte cultures were infected for 24 h with 3% parasitemia, the percentage of total parasites was determined in the absence (control) or presence of 10 − 8 M analogs. Dark grey shading denotes that the result is not statistically significant compared with control . Other synthesized angiotensin analogs presented high antiplasmodial activity in the P. gallinaceum test . They have similar hydrophobic interactions between the isolated guanidyl (Arg 2 ) and hydroxyl (Tyr 4 ) groups and the side chains of alkyl and aromatic amino acid. The guanidyl (Arg 2 ) and carboxyl (Asp 1 ) groups seem to be less significant in this case. Lactam bridges have an important role in constricting the conformation of bioactive peptides. When evaluated against P. falciparum , all lactam bridge analogs lacked significant activity against P. gallinaceum and perform better against P. falciparum because of the rigidity of the peptide bridge and possibe increased resistance to degradation under this kind of restriction. Analog 416 and 418 presented significant antiplasmodial activity, however analog 425 , a disulfide bridge analog, was the most active peptide against P. gallinaceum . It was also active against P. falciparum because it has the same restriction size and implications when considering hydrophobic and hydrophilic interactions between side chains. All analogs frequently adopt a β-turn conformation, which is consistent with that of the analog that is believed to be most active against P. gallinaceum . Angiotensin II hormones were also synthesized to investigate erythrocytic cycle invasion . Peptides 426 and 427 exhibited >80% activity on P. gallinaceum and >40% activity on P. falciparum . The hemolytic effect, contractile response, and stability in human serum were determined , demonstrating that the peptides did not present hemolytic effects as they were inactive when compared to positive controls and were resistant to degradation in human serum after a prolonged exposure (6 h), especially peptides 426 and 427 , which displayed excellent stability. The study of synthesized Dec-NH 2 and eight analog peptides reported that certain amino acid substitutions significantly improved the peptide’s antiplasmodial activity, providing new sequence principles for creating potent anti-malaria drugs, such as replacing the original sequence by Arg, Phe, and Trp . The most active peptides examined were [Arg] 1 -Dec-NH 2 ( 438 ), [Pro] 4 -Dec-NH 2 ( 437 ), and [Phe] 2 -Dec-NH 2 ( 439 ); all compounds substitutions improved the effectiveness, with the gihest antiplasmodial activities achieved with mutations to the N-terminus of Dec-NH 2 . A higher antiplasmodial activity was achieved by the introduction of a positive charge, increased hydrophobicity, and introduction of a restrictor residue . 3.1.1. Mahafacyclin B The antimalarial mahafacyclin B ( 32 ) is a cyclic heptapeptide that was synthesized by Fujita et al. (2013) using the soluble-tag assisted liquid–phase method . This method allowed peptide head-to-tail cyclization leading to the total synthesis of mahafacyclin B assisted by hydrophobic benzyl alcohol. The synthesis was initiated by reductive amination of hydrophobic benzaldehydes using amino acid methyl esters to obtain hydrophobically tagged Gly-Ome from its reaction with coupling reagent Fmoc-Phe-OH, and a combination of HATU, HOAt, DIPEA, and THF. Basic deprotection of the hydrophobically tagged dipeptide produces diketopiperazine when the C-terminal residue is proline. The hydrophobically tagged tripeptide was deprotected to obtain the cyclized compound . A less polar solvent was employed during cyclization to overcome the issue of the highly hydrophobic residues in the linear precursor. The amide nitrogen was used as the marking site allowing the rapid reaction checks and product isolation. 3.1.2. Aerucylamide B Aerucyclamide B ( 33 ) was synthesized by Peña et al. (2013) using two heterocycles and a dipeptide as building blocks . Dipeptide 34 and thiazole 35 were synthesized in the first step and then HBTU was used as coupling reagent to produce the thiazole precursor 37 through cyclodehydration of β-hydroxythioamides and further oxidation methodology . In the next step, two intermediate compounds 41 and 44 were prepared to obtain macrocycle 34 . Compound 41 was synthesized via routes I and II . Route I involved a coupling between the N -deprotected dipeptide 40 and intermediate 39 , while route II involved coupling between compound 42 and 43 . Macrocycle 34 was obtained from intermediates 41 and 44 using HBTU as a coupling agent , and then aerucyclamide B was synthesized from 37 via intermediate 44 , and from 36 via intermediate 41 . 3.1.3. Cyclomarin Cyclomarin A ( 46 ) is a marine cycloheptapeptide with antimalarial activity that can be synthesized in an enantiomerically pure form . Key steps in the synthesis of cyclomarin are asymmetric chelate−enolate Claisen rearrangement, asymmetric hydrogenation, and highly diastereoselective additions of organozinc and titanium reagents . The peptide synthesis developed to obtain the required building blocks is based on long-term experience in amino acid synthesis. Amino acid chelate enolate esters are used to obtain β-hydroxy through an aldol reaction, whereas γ,δ-unsaturated amino acids are obtained by transition metal-catalyzed allylic alkylation or Claisen enolate chelate rearrangement and syn-isomers are obtained by Claisen rearrangement. Thus, this approach was used to produce the desired N -protected amino acid (2 S ,3 R )- 50 . The protected 5-hydroxyleucine 56 was synthesized with the commercially available Roche ester 52 . The compound was reduced to the aldehyde 53 , and used in an olefination using Schimdt’s phosphonoglycinate 57 . The α,β-unsaturated amino acid derivative obtained was stereoselectively hydrogenated with ( R )-monophos as a ligand. Following N -methylation and saponification, amino acid 56 was produced with good yield selectivity . β-methoxyphenylalanine 61 is the third unusual building block that was prepared through several attempts shown in . The first attempt using phenylmagnesium bromide (3 equiv.) and DIBAL, which gave an acceptable yield of the addition products (60%), but only moderate diastereoselective products (7:3 ratio). Consequently, the titanium reagent was added as the single diastereomer. Subsequent O-Methylation, cleavage of the silyl ether, and oxidation then gave rise to an excellent yield of the target acid 61 . Another method was also reported for regioselectively prenylated from electron-withdrawing indole. The Pd-catalyzed protocol to 3-indolecarboxylate 62 was applied resulting in indole 63 . The reaction temperature was maintained at 0 °C to suppress N -prenylation, the ester was then saponified directly before the peptide coupling to avoid decomposition of the labile acid. The next building block was obtained by the ester saponification of compound 62 . The prenylated indole 64 was converted to the 3-iodo derivative 65 and the terminal double bond was changed to epoxide 66 . After several steps, niodosuccinimide (NIS) in MeOH produced methyl ester 72 . The O-silylated cyclomarin A was obtained by removing the two protective silyl groups using a two-step protocol in which the primary OH group was deprotected first using NH 4 F. After obtaining the necessary building blocks, cyclomarin A was then synthesized. Ring closure was planned between aminohexenoic acid and N -methylated leucine. BEP (2-bromo-1-ethylpyridinium tetrafluoroborate) 80 was used for coupling with protected valine. HCl was applied to suppress DKP formation during synthesis. The dipeptide salt was coupled to the next peptide with the activated compound 47 . The same protocol was then applied in the next coupling step. The HOBT reagent was used to activate N -methyl-4-hydroxyleucine 56 to avoid epimerization during the coupling reaction. This step obtained enantiomerically pure pentapeptide 77 . The methyl ester 72 was saponified and directly coupled using the BEP protocol to obtain compound 78 . Under Pd catalysis, the alloc protecting group was removed and combined with amino acid 50 , then the deprotected heptapetide was slowly added to an aqueous solution of PyBOP and base in CH 2 Cl 2 to obtain O-sylilated cyclomarin A 46 . The synthesis of another cyclomarin was also reported by Kiefer et al. (2019). Seventeen desoxycyclomarin-inspired derivatives ( 96 – 100 , 106 – 110 , 119 – 125 ) were synthesized using a straightforward solution-phase approach . First, an analog library was prepared stepwise by varying the N ′-side chain of the tryptophan moiety followed by simplifying the aminohexenoic acid unit. A Negishi reaction served as key step for the generation of N ′-isopropyltryptophan, with final saponification yielding the N ′-isopropyltryptophan derivative. d -hydroxyleucine units were replaced by d -hydroxynorvaline for additional structural changes in the peptide backbone. The synthesized fragment with protected glutamic acid ( 81 ) was activated as mixed anhydride and then reduced with NaBH 4 . The alcohol produced as TBS ether gives compound 82 , which is saponified and N -methylated to obtain N -methyl- d -hydroxynorvaline 84 . The first step of synthesis is focused on linear precursor assembly with varying tryptophan motifs and hexenoic amino acids . The intermediate compounds 86 – 90 were reacted with compounds 58 / 59 using the HOBt coupling reagent to produce compounds 91 – 95 , which were then cyclized with PyBOP and DIPEA coupling reagents to produce compound 96 – 100 . Then, compounds 106 – 110 were produced using the same synthesis steps. Intermediates 101 – 105 were cyclized with the same coupling reagent to produce the target compound. The second set of cyclomarin analogs was synthesized based on the key intermediate 106 . Several stages of coupling and deprotection were conducted to produce linear peptide compounds 112 – 118 . The linear compounds were then cyclized using PyBOP and DIPEA to produce compounds 119 – 125 . 3.1.4. Chaiyaphumine Gholap and Ugale (2017) reported the synthesis of chaiyaphumine-A ( 130 ) by convergent-based solution phase synthesis involving protection, deprotection, coupling and cyclization reactions . This novel class of peptide will the subject of much research to develop new antimalarial drugs. Five amino acids, L-threonine, D-phenyl alanine, D-alanine, L-proline and L-tryptophan, were assembled by coupling reactions involving HATU. The macrocylization of the side chain (-OH) of L-threonine and C-terminus of carboxylic acid from L-tryptophan using T 3 P yields chaiyaphumine-A . The synthesis of chaiyapumines was also undertaken by Lu and Batey (2022) . Fmoc-based solid-phase synthesis approaches were used for the synthesis of the requisite precursors. The solution-phase macrolactamization approach using EDC-HCl/HOBt was achieved without any detectable epimerization. This strategy was successfully applied for the total synthesis of the antimalarial compound chaiyaphumine A, as well as the natural products chaiyaphumine B, C, and D ( 136 – 139 ) using a late-stage acylation of the threonine amino group . Cyclized products were not produced by the macrolactonization method using MNBA/DMAP/iPr 2 Net, while the use of the additive Dy(OTf) 3 resulted in epimerization macrocycles. The slow rate of macrolactonization is influenced by the weak nucleophilicity of oxygen and the steric hindrance of threonine in the seco acid precursor, which allows competitive epimerization/cyclization to occur. Successful macrocyclization can be achieved via a macrolactamization approach that allows the total synthesis of all chaiyaphumine compunds. 3.1.5. Cyclopeptides Anolgs Prepared by Macrocyclization Fagundez et al. (2018) investigated the synthesis of a cyclopeptide using solid-phase peptide synthesis and on-resin macrocyclization, with various analytical techniques used to monitor reactions and purify compounds . The investigation compared solution- and solid-phase ring-closing reactions following the synthesis of linear precursor on a 2-chrotrytil chloride resin . Method A applied a combination of the solid-phase (linear precursor) and the solution-phase methods (macrocylization), whereas method B involved all reactions on-resin. A combination of DIC and Cl-HOBt as a coupling agent was used when the next amino acid added was Fmoc-L-Cys(Trt)-OH because it does not require the addition of a base and can decrease potent racemization. The solution-phase macrocyclization involved HBTU or HATU in the presence of DIPEA in DCM. The final products were obtained in moderate to good yields 26–80%. Eleven cyclic peptides ( 140 – 150 ) were synthesized (23 to 63% yield) using method A . Method B constructed Fmoc-L-Glu-OH and all amino acids on 2-CTC resin and the capping step took advantage of DIPEA and MeOH . Reagent [Pd(PPh 3 ) 4 ] in a solution of 10% piperidine in THF was used to remove the Fmoc-protected amino acid at the N -terminal group and produce a free carboxyl group at the C-terminal. This method yielded four antimalarial cyclopeptides ( 151 – 154 ) . A comparison of these methods is provided in . A year later after synthesizing compounds 140 – 154 cyclopeptides, Fagundez et al. (2019) developed cyclopeptides containing N -methyl amino acids with promising antiplasmodial activity against both erythrocytic and liver stages of malaria in vitro. These compounds could potentially be used as new and safe drugs to combat malaria . Compounds 156 – 161 were synthesized by employing solid-phase linear peptide synthesis and solution macrocyclization . The 2-chlorotrityl resin (2-CTC) was used to decrease the diketopiperazine formation, and HBTU and DIPEA were employed as coupling reagents in most cases. HCTU and DIPEA reagents were used to combine N -Me-Gly with the next amino acids because they are more effective. DIC and Cl-HOBt were then used as reagents to activate the Fmoc-L-Cys(Trt)-OH and coupling to resin-attached peptides to minimize racemization of Cys residues. Compounds 162 , 163 , and 164 were synthesized employing on-resin macrocyclization . First, Fmoc-L-Glu-O was anchored to 2-CTC resin with DIPEA, and then the Fmoc SPPS protocol was followed. After the removal of the allyl ester and Fmoc group with [Pd(PPh 3 ) 4 ] in a solution of 10% piperidine in THF, the ring closure was performed using DIC and Cl-HOBt. shows the structure of compounds 156 – 165 . 3.1.6. Hirsutellide A and Its Analogues Sahile et al. (2020) reported the synthesis and evaluation of hirsutellide A and its analogs for their antimycobacterial and antiplasmodial activities . Optimizing the structures of lipophilic antimicrobial cyclic peptides resulted in more membrane-permeable bioactive peptides, such as the lipophilic antimicrobial depsipeptide (hirsutellide A). The first report of the synthesis hirsutellide A was by Xu et al. in 2005; compound 167 had Ile residues instead of allo-Ile resulting in stereoisomers that do not have antimycobacterial potential . The compound was synthesized by a solution-phase method and initiated by the preparation of protected tridepsipeptide . D-Phe 168 was hydrolyzed into α-hydroxyl carboxyl acid 169 . After several steps, dipepsipeptide 172 coupling with N-Boc-allo-Ile-OH using HATU/DIPEA to generate tridepsipeptide 173 . The desired cyclic depsipeptide 177 was obtained via macrocyclization of the precursor 176 using HATU/HOBt/DIPEA in DMF . Furthermore, depsipeptide ( 178 – 181 ) analogs and peptide analogs ( 182 – 188 ) of hirsutellide A were designed and synthesized for preliminary SAR studies with a combination of solution and solid phase method . 3.1.7. Macrocyclization Strategies Recently, Zhang et al. (2022) synthesized macrocyclic peptides as species-selective antimalarial proteasome inhibitors . The compounds were prepared through several macrocyclization strategies such as Suzuki coupling, intermolecular amidation, intermolecular alkylation, intramolecular alkylation, and ring-closing metathesis (RCM) reaction. Synthesis of cyclic peptides 202 , 203 , and 204 took advantage of Suzuki coupling as a macrocyclization strategy . Benzylation and borylation reactions of N-Boc-3-bromophenylalanine 189 produced boronate. Deprotection of Boc in compound 191 was followed by combining with Boc-L-homophenylalanine to produce dipeptide intermediate 193 , which then underwent Boc deprotection and condensation with carboxylic acid 194 to produce compound 195 . Oxidative hydrolysis of boronic ester to boronic acid in 195 by NaIO 4 was followed by an intramolecular Chan–Lam coupling reaction to produce macrocycle 205 . Removal of the benzyl protective group by catalytic hydrogenation, followed by a coupling reaction of urea, produces 198 – 199 , which undergoes Boc deprotection to yield amines 209 and CP1. Furthermore, compounds 200 and CP1 undergo acetylation of free amines to produce macrocycles 202 and 203 . Macrocycle 204 is produced from amine 210 through subsequent alkylation with 4-bromobutanoate, Boc-deprotection, and lactamization using EDCI and HOBt. Intramolecular amidation as a macrocyclization strategy was used to synthesize compound 214 from compound 212 by hydrolysis of methyl ester and a coupling reaction using trifluoroethylamine. The synthesis began with deprotecting the tritile group in aziridine 205 , then protecting it again with the Boc group present in Boc-aziridine 206 . Compound 207 was then produced by an SN 2- type ring opening reaction mediated by BF 3 ·Et 2 O on aziridine 206 with 2-benzyloxyethanol. Boc-deprotection in compound 207 was followed by a coupling reaction with Boc-L-homophenylalanine to produce dipeptide 208 , then benzyl deprotection and Mitsunobu reaction of fragment 209 yielded the macrocyclization precursor 210 . Debenzylation and Boc deprotection of compound 210 was followed by intramolecular mediation to produce macrocycle 212 . The synthesis of compounds 229 – 237 were involved in intramolecular alkylation as a macrocyclization strategy . Dipeptide 225 was prepared from amino acid 224 by methyl esterification and HATU-mediated coupling reaction with Boc-L-homophenylalanine. Boc deprotection of dipeptide 225 and a coupling reaction with fragment 226 gave tripeptide 227 , which was then removed by two benzyl groups. The primary alcohol was chemoselectively protected with Tos groups, and Cs 2 CO 3 -mediated intramolecular. It underwent alkylation to give macrocycle 228 . After treatment of the methyl ester with sodium hydroxide, coupling of the free acid with various primary amines yielded macrocycles 229 , 230 , 231 , 232 , 236 , and 237 . Reduction of the methyl ester 228 with lithium borohydride formed alcohol 238 , and then Dess–Martin oxidation, followed by reductive amination with trifluoroethylamine, yielded macrocycle 233 . Condensation of the methyl ester 228 with hydrazine, followed by coupling with methyl trifluoroacetimidate gave the hydrazide 239 , which was thermally cyclized to give the 1,2,4-triazole 235 . Subsequent hydrolysis and dehydration of hydrazide 239 in the presence of the Burgess reagent gave 1,3,4-oxadiazole 234 . Macrocyclization with the intramolecular alkylation strategy was used to synthesize compounds 247 , 248 , 249 , and 254 . Fragments 231 – 233 underwent an amide coupling reaction with fragments 234 – 235 and subsequent deprotection, yielding dipeptides 236 – 239 , which were coupled to fragments 240 – 241 and 217 to yield tripeptides 242 – 246 . The tripeptides 242 – 246 were sulfated with a hydroxyl group, debenzylation of aryl benzyl ether, and intramolecularly substituted between the phenol group and the sulfonate ester, giving macrocycles 250 , 251 , 247 , 248 and 249 . Removal of the tert-butyl from compounds 250 – 251 using TFA was followed by an amide coupling reaction with a primary amine, which produced macrocycles 254 and TDI8. Synthesis of compounds 270 , 271 , 272 , and 273 was achieved by the RCM reaction . Fragments 231 and 255 underwent subsequent amide coupling reaction with fragment 256 ; Boc-deprotection and an amide coupling reaction with acids 261 – 264 afforded diolefins 265 – 268 , which were subjected to RCM in the presence of Grubbs Catalyst second Generation and reducing of the C=C double bond, yielding macrocycles 270 , 271 , 272 , and 269 . Removal of the Boc protection of 269 provided macrocycle 273 . The antimalarial mahafacyclin B ( 32 ) is a cyclic heptapeptide that was synthesized by Fujita et al. (2013) using the soluble-tag assisted liquid–phase method . This method allowed peptide head-to-tail cyclization leading to the total synthesis of mahafacyclin B assisted by hydrophobic benzyl alcohol. The synthesis was initiated by reductive amination of hydrophobic benzaldehydes using amino acid methyl esters to obtain hydrophobically tagged Gly-Ome from its reaction with coupling reagent Fmoc-Phe-OH, and a combination of HATU, HOAt, DIPEA, and THF. Basic deprotection of the hydrophobically tagged dipeptide produces diketopiperazine when the C-terminal residue is proline. The hydrophobically tagged tripeptide was deprotected to obtain the cyclized compound . A less polar solvent was employed during cyclization to overcome the issue of the highly hydrophobic residues in the linear precursor. The amide nitrogen was used as the marking site allowing the rapid reaction checks and product isolation. Aerucyclamide B ( 33 ) was synthesized by Peña et al. (2013) using two heterocycles and a dipeptide as building blocks . Dipeptide 34 and thiazole 35 were synthesized in the first step and then HBTU was used as coupling reagent to produce the thiazole precursor 37 through cyclodehydration of β-hydroxythioamides and further oxidation methodology . In the next step, two intermediate compounds 41 and 44 were prepared to obtain macrocycle 34 . Compound 41 was synthesized via routes I and II . Route I involved a coupling between the N -deprotected dipeptide 40 and intermediate 39 , while route II involved coupling between compound 42 and 43 . Macrocycle 34 was obtained from intermediates 41 and 44 using HBTU as a coupling agent , and then aerucyclamide B was synthesized from 37 via intermediate 44 , and from 36 via intermediate 41 . Cyclomarin A ( 46 ) is a marine cycloheptapeptide with antimalarial activity that can be synthesized in an enantiomerically pure form . Key steps in the synthesis of cyclomarin are asymmetric chelate−enolate Claisen rearrangement, asymmetric hydrogenation, and highly diastereoselective additions of organozinc and titanium reagents . The peptide synthesis developed to obtain the required building blocks is based on long-term experience in amino acid synthesis. Amino acid chelate enolate esters are used to obtain β-hydroxy through an aldol reaction, whereas γ,δ-unsaturated amino acids are obtained by transition metal-catalyzed allylic alkylation or Claisen enolate chelate rearrangement and syn-isomers are obtained by Claisen rearrangement. Thus, this approach was used to produce the desired N -protected amino acid (2 S ,3 R )- 50 . The protected 5-hydroxyleucine 56 was synthesized with the commercially available Roche ester 52 . The compound was reduced to the aldehyde 53 , and used in an olefination using Schimdt’s phosphonoglycinate 57 . The α,β-unsaturated amino acid derivative obtained was stereoselectively hydrogenated with ( R )-monophos as a ligand. Following N -methylation and saponification, amino acid 56 was produced with good yield selectivity . β-methoxyphenylalanine 61 is the third unusual building block that was prepared through several attempts shown in . The first attempt using phenylmagnesium bromide (3 equiv.) and DIBAL, which gave an acceptable yield of the addition products (60%), but only moderate diastereoselective products (7:3 ratio). Consequently, the titanium reagent was added as the single diastereomer. Subsequent O-Methylation, cleavage of the silyl ether, and oxidation then gave rise to an excellent yield of the target acid 61 . Another method was also reported for regioselectively prenylated from electron-withdrawing indole. The Pd-catalyzed protocol to 3-indolecarboxylate 62 was applied resulting in indole 63 . The reaction temperature was maintained at 0 °C to suppress N -prenylation, the ester was then saponified directly before the peptide coupling to avoid decomposition of the labile acid. The next building block was obtained by the ester saponification of compound 62 . The prenylated indole 64 was converted to the 3-iodo derivative 65 and the terminal double bond was changed to epoxide 66 . After several steps, niodosuccinimide (NIS) in MeOH produced methyl ester 72 . The O-silylated cyclomarin A was obtained by removing the two protective silyl groups using a two-step protocol in which the primary OH group was deprotected first using NH 4 F. After obtaining the necessary building blocks, cyclomarin A was then synthesized. Ring closure was planned between aminohexenoic acid and N -methylated leucine. BEP (2-bromo-1-ethylpyridinium tetrafluoroborate) 80 was used for coupling with protected valine. HCl was applied to suppress DKP formation during synthesis. The dipeptide salt was coupled to the next peptide with the activated compound 47 . The same protocol was then applied in the next coupling step. The HOBT reagent was used to activate N -methyl-4-hydroxyleucine 56 to avoid epimerization during the coupling reaction. This step obtained enantiomerically pure pentapeptide 77 . The methyl ester 72 was saponified and directly coupled using the BEP protocol to obtain compound 78 . Under Pd catalysis, the alloc protecting group was removed and combined with amino acid 50 , then the deprotected heptapetide was slowly added to an aqueous solution of PyBOP and base in CH 2 Cl 2 to obtain O-sylilated cyclomarin A 46 . The synthesis of another cyclomarin was also reported by Kiefer et al. (2019). Seventeen desoxycyclomarin-inspired derivatives ( 96 – 100 , 106 – 110 , 119 – 125 ) were synthesized using a straightforward solution-phase approach . First, an analog library was prepared stepwise by varying the N ′-side chain of the tryptophan moiety followed by simplifying the aminohexenoic acid unit. A Negishi reaction served as key step for the generation of N ′-isopropyltryptophan, with final saponification yielding the N ′-isopropyltryptophan derivative. d -hydroxyleucine units were replaced by d -hydroxynorvaline for additional structural changes in the peptide backbone. The synthesized fragment with protected glutamic acid ( 81 ) was activated as mixed anhydride and then reduced with NaBH 4 . The alcohol produced as TBS ether gives compound 82 , which is saponified and N -methylated to obtain N -methyl- d -hydroxynorvaline 84 . The first step of synthesis is focused on linear precursor assembly with varying tryptophan motifs and hexenoic amino acids . The intermediate compounds 86 – 90 were reacted with compounds 58 / 59 using the HOBt coupling reagent to produce compounds 91 – 95 , which were then cyclized with PyBOP and DIPEA coupling reagents to produce compound 96 – 100 . Then, compounds 106 – 110 were produced using the same synthesis steps. Intermediates 101 – 105 were cyclized with the same coupling reagent to produce the target compound. The second set of cyclomarin analogs was synthesized based on the key intermediate 106 . Several stages of coupling and deprotection were conducted to produce linear peptide compounds 112 – 118 . The linear compounds were then cyclized using PyBOP and DIPEA to produce compounds 119 – 125 . Gholap and Ugale (2017) reported the synthesis of chaiyaphumine-A ( 130 ) by convergent-based solution phase synthesis involving protection, deprotection, coupling and cyclization reactions . This novel class of peptide will the subject of much research to develop new antimalarial drugs. Five amino acids, L-threonine, D-phenyl alanine, D-alanine, L-proline and L-tryptophan, were assembled by coupling reactions involving HATU. The macrocylization of the side chain (-OH) of L-threonine and C-terminus of carboxylic acid from L-tryptophan using T 3 P yields chaiyaphumine-A . The synthesis of chaiyapumines was also undertaken by Lu and Batey (2022) . Fmoc-based solid-phase synthesis approaches were used for the synthesis of the requisite precursors. The solution-phase macrolactamization approach using EDC-HCl/HOBt was achieved without any detectable epimerization. This strategy was successfully applied for the total synthesis of the antimalarial compound chaiyaphumine A, as well as the natural products chaiyaphumine B, C, and D ( 136 – 139 ) using a late-stage acylation of the threonine amino group . Cyclized products were not produced by the macrolactonization method using MNBA/DMAP/iPr 2 Net, while the use of the additive Dy(OTf) 3 resulted in epimerization macrocycles. The slow rate of macrolactonization is influenced by the weak nucleophilicity of oxygen and the steric hindrance of threonine in the seco acid precursor, which allows competitive epimerization/cyclization to occur. Successful macrocyclization can be achieved via a macrolactamization approach that allows the total synthesis of all chaiyaphumine compunds. Fagundez et al. (2018) investigated the synthesis of a cyclopeptide using solid-phase peptide synthesis and on-resin macrocyclization, with various analytical techniques used to monitor reactions and purify compounds . The investigation compared solution- and solid-phase ring-closing reactions following the synthesis of linear precursor on a 2-chrotrytil chloride resin . Method A applied a combination of the solid-phase (linear precursor) and the solution-phase methods (macrocylization), whereas method B involved all reactions on-resin. A combination of DIC and Cl-HOBt as a coupling agent was used when the next amino acid added was Fmoc-L-Cys(Trt)-OH because it does not require the addition of a base and can decrease potent racemization. The solution-phase macrocyclization involved HBTU or HATU in the presence of DIPEA in DCM. The final products were obtained in moderate to good yields 26–80%. Eleven cyclic peptides ( 140 – 150 ) were synthesized (23 to 63% yield) using method A . Method B constructed Fmoc-L-Glu-OH and all amino acids on 2-CTC resin and the capping step took advantage of DIPEA and MeOH . Reagent [Pd(PPh 3 ) 4 ] in a solution of 10% piperidine in THF was used to remove the Fmoc-protected amino acid at the N -terminal group and produce a free carboxyl group at the C-terminal. This method yielded four antimalarial cyclopeptides ( 151 – 154 ) . A comparison of these methods is provided in . A year later after synthesizing compounds 140 – 154 cyclopeptides, Fagundez et al. (2019) developed cyclopeptides containing N -methyl amino acids with promising antiplasmodial activity against both erythrocytic and liver stages of malaria in vitro. These compounds could potentially be used as new and safe drugs to combat malaria . Compounds 156 – 161 were synthesized by employing solid-phase linear peptide synthesis and solution macrocyclization . The 2-chlorotrityl resin (2-CTC) was used to decrease the diketopiperazine formation, and HBTU and DIPEA were employed as coupling reagents in most cases. HCTU and DIPEA reagents were used to combine N -Me-Gly with the next amino acids because they are more effective. DIC and Cl-HOBt were then used as reagents to activate the Fmoc-L-Cys(Trt)-OH and coupling to resin-attached peptides to minimize racemization of Cys residues. Compounds 162 , 163 , and 164 were synthesized employing on-resin macrocyclization . First, Fmoc-L-Glu-O was anchored to 2-CTC resin with DIPEA, and then the Fmoc SPPS protocol was followed. After the removal of the allyl ester and Fmoc group with [Pd(PPh 3 ) 4 ] in a solution of 10% piperidine in THF, the ring closure was performed using DIC and Cl-HOBt. shows the structure of compounds 156 – 165 . Sahile et al. (2020) reported the synthesis and evaluation of hirsutellide A and its analogs for their antimycobacterial and antiplasmodial activities . Optimizing the structures of lipophilic antimicrobial cyclic peptides resulted in more membrane-permeable bioactive peptides, such as the lipophilic antimicrobial depsipeptide (hirsutellide A). The first report of the synthesis hirsutellide A was by Xu et al. in 2005; compound 167 had Ile residues instead of allo-Ile resulting in stereoisomers that do not have antimycobacterial potential . The compound was synthesized by a solution-phase method and initiated by the preparation of protected tridepsipeptide . D-Phe 168 was hydrolyzed into α-hydroxyl carboxyl acid 169 . After several steps, dipepsipeptide 172 coupling with N-Boc-allo-Ile-OH using HATU/DIPEA to generate tridepsipeptide 173 . The desired cyclic depsipeptide 177 was obtained via macrocyclization of the precursor 176 using HATU/HOBt/DIPEA in DMF . Furthermore, depsipeptide ( 178 – 181 ) analogs and peptide analogs ( 182 – 188 ) of hirsutellide A were designed and synthesized for preliminary SAR studies with a combination of solution and solid phase method . Recently, Zhang et al. (2022) synthesized macrocyclic peptides as species-selective antimalarial proteasome inhibitors . The compounds were prepared through several macrocyclization strategies such as Suzuki coupling, intermolecular amidation, intermolecular alkylation, intramolecular alkylation, and ring-closing metathesis (RCM) reaction. Synthesis of cyclic peptides 202 , 203 , and 204 took advantage of Suzuki coupling as a macrocyclization strategy . Benzylation and borylation reactions of N-Boc-3-bromophenylalanine 189 produced boronate. Deprotection of Boc in compound 191 was followed by combining with Boc-L-homophenylalanine to produce dipeptide intermediate 193 , which then underwent Boc deprotection and condensation with carboxylic acid 194 to produce compound 195 . Oxidative hydrolysis of boronic ester to boronic acid in 195 by NaIO 4 was followed by an intramolecular Chan–Lam coupling reaction to produce macrocycle 205 . Removal of the benzyl protective group by catalytic hydrogenation, followed by a coupling reaction of urea, produces 198 – 199 , which undergoes Boc deprotection to yield amines 209 and CP1. Furthermore, compounds 200 and CP1 undergo acetylation of free amines to produce macrocycles 202 and 203 . Macrocycle 204 is produced from amine 210 through subsequent alkylation with 4-bromobutanoate, Boc-deprotection, and lactamization using EDCI and HOBt. Intramolecular amidation as a macrocyclization strategy was used to synthesize compound 214 from compound 212 by hydrolysis of methyl ester and a coupling reaction using trifluoroethylamine. The synthesis began with deprotecting the tritile group in aziridine 205 , then protecting it again with the Boc group present in Boc-aziridine 206 . Compound 207 was then produced by an SN 2- type ring opening reaction mediated by BF 3 ·Et 2 O on aziridine 206 with 2-benzyloxyethanol. Boc-deprotection in compound 207 was followed by a coupling reaction with Boc-L-homophenylalanine to produce dipeptide 208 , then benzyl deprotection and Mitsunobu reaction of fragment 209 yielded the macrocyclization precursor 210 . Debenzylation and Boc deprotection of compound 210 was followed by intramolecular mediation to produce macrocycle 212 . The synthesis of compounds 229 – 237 were involved in intramolecular alkylation as a macrocyclization strategy . Dipeptide 225 was prepared from amino acid 224 by methyl esterification and HATU-mediated coupling reaction with Boc-L-homophenylalanine. Boc deprotection of dipeptide 225 and a coupling reaction with fragment 226 gave tripeptide 227 , which was then removed by two benzyl groups. The primary alcohol was chemoselectively protected with Tos groups, and Cs 2 CO 3 -mediated intramolecular. It underwent alkylation to give macrocycle 228 . After treatment of the methyl ester with sodium hydroxide, coupling of the free acid with various primary amines yielded macrocycles 229 , 230 , 231 , 232 , 236 , and 237 . Reduction of the methyl ester 228 with lithium borohydride formed alcohol 238 , and then Dess–Martin oxidation, followed by reductive amination with trifluoroethylamine, yielded macrocycle 233 . Condensation of the methyl ester 228 with hydrazine, followed by coupling with methyl trifluoroacetimidate gave the hydrazide 239 , which was thermally cyclized to give the 1,2,4-triazole 235 . Subsequent hydrolysis and dehydration of hydrazide 239 in the presence of the Burgess reagent gave 1,3,4-oxadiazole 234 . Macrocyclization with the intramolecular alkylation strategy was used to synthesize compounds 247 , 248 , 249 , and 254 . Fragments 231 – 233 underwent an amide coupling reaction with fragments 234 – 235 and subsequent deprotection, yielding dipeptides 236 – 239 , which were coupled to fragments 240 – 241 and 217 to yield tripeptides 242 – 246 . The tripeptides 242 – 246 were sulfated with a hydroxyl group, debenzylation of aryl benzyl ether, and intramolecularly substituted between the phenol group and the sulfonate ester, giving macrocycles 250 , 251 , 247 , 248 and 249 . Removal of the tert-butyl from compounds 250 – 251 using TFA was followed by an amide coupling reaction with a primary amine, which produced macrocycles 254 and TDI8. Synthesis of compounds 270 , 271 , 272 , and 273 was achieved by the RCM reaction . Fragments 231 and 255 underwent subsequent amide coupling reaction with fragment 256 ; Boc-deprotection and an amide coupling reaction with acids 261 – 264 afforded diolefins 265 – 268 , which were subjected to RCM in the presence of Grubbs Catalyst second Generation and reducing of the C=C double bond, yielding macrocycles 270 , 271 , 272 , and 269 . Removal of the Boc protection of 269 provided macrocycle 273 . 3.2.1. Falcitidin Somanadhan et al. (2013) succeeded in synthesizing linear antimalarial peptides. Falcitidin 284 is the first member of a new class of falcipain-2 inhibitors, which is a cysteine protease used by P. falciparum to degrade hemoglobin during the trophozoite stage of infection . Falcitidin contains isovaleric acid-D-His-L-Ile-L-Val-L-Pro-NH 2 and was synthesized by the solution-phase peptide synthesis method . The synthesis begins with deprotecting the Boc group from N-Boc-proline carboxamide using TFA. The amino-proline TFA salt was coupled with freshly prepared L-Ile-L-Val using HATU/HOAt to give L-Ile-L-Val-L-Pro-NH 2 . The dipeptide was obtained by coupling L-Ile-NHBoc with the in situ prepared bis-trimethylsilane (TMS) ether of L-Val under mixed anhydride conditions. The Boc group of the tripeptide was then cleaved using TFA and coupled to Na-Fmoc-N(im)-trityl-D-histidine using HATU/HOAt to give the Fmoc–tetrapeptide. The Fmoc group of was removed using piperidine and HATU/HOAt. The final synthetic target falcitidin was obtained after trityl deprotection of Trt-falcitidin using TFA in the presence of triisopropylsilane. Kotturi et al. (2014) synthesized falcitidin and its analogs ( 296 – 303 ) using the same solution phase peptide synthesis . First, N im -trityl-D-histidine 288 was converted into bis-TMS ether with TMSCl/Et 3 N 7 and coupled immediately with the mixed anhydride of isovaleric acid 287 using ethyl chloroformate/NMM to form the trityl protected N-acyl-D-histidine. In parallel, the N-Boc proline amide was deprotected with TFA, and the crude amine salt was coupled immediately with the known dipeptide N-Boc-L-Ile-L-Val by using HATU/HOAt to yield the key tripeptide intermediate N-Boc-L-Ile-L-Val-L-Pro-NH 2 . Lastly, the Boc group of tripeptide was cleaved using TFA and coupled to N-acyl-D-histidine using HATU/HOAt to give the N im -trityl protected tetrapeptide 292 . Falcitidin acylatide and its analogs produce eight analogs compounds ( 296 – 303 ) by diversifying the synthesis of the N-acyl tetrapeptide analog of falcitidin through the common tripeptide N-Boc-L-Ile-L-Val-L-Pro-NH 2 . 3.2.2. Gallinamide Gallinamide A is a linear peptide that has antimalarial activity. Conroy et al. (2014) designed and synthesized 18 gallinamide A analogs by solid-phase synthesis and substitution of amino acid residues . Four gallinamide A analogs ( 309 – 312 ) were originally designed with varying levels of saturation, including compound 309 , which is structurally identical to gallinamide A. Another analog was compound 310 with a reduced methoxy-enol moiety in the pyrolinone ring 309 . Part of the olefinic component of the 4( S )-amino-2-( E )-pentenoic acid unit is reduced in compound 311 and 312 has reduced olefin groups. Analog synthesis was initiated by synthesizing the N -terminal fragment 304 by solid-phase peptide synthesis using the Fmoc strategy. The 2-CTC resin is filled with Fmoc-Leu-OH followed by the subsequent incorporation of the amino acid. Reductive amination of the resin, followed by cleavage of the resin using hexafluoroisopropanol (HFIP), produced an excellent yield of tripeptide 304 which was coupled with imide fragments 306 and 307 prepared using a similar protocol. The coupling reaction involved HATU at low temperature to avoid epimerization to produce good yields of analogs 309 and 310 . At this stage, 309 and 310 underwent hydrogenation to give 311 and 312 . Analogs 313 – 318 were synthesized from 2-CTC resin filled with Fmoc-Ala . The coupling of Fmoc-protected α,β-unsaturated amino acids was followed by elongation through the Fmoc solid-phase peptide synthesis strategy. With reductive methylation of the N-terminus using formaldehyde and sodium cyanoborohydride, the peptide was cleaved from the resin using HFIP to produce the C-terminal peptide acid. C-terminal functionality was achieved through benzylamine coupling using PyBOP at low temperatures, and compounds 313 , 315 , 317 , and 318 were purified by RP-HPLC and demonstrated no significant epimerization. Compound 314 was produced by attaching 4-( R )-hydroxy L-proline methyl ester to the C-terminus of the two peptides with the addition of NMM as a hindered base. Aminothiazole was coupled to the C-terminus of one peptide acid using PyBOP at low temperature to yield compound 316 as a mixed diastereoisomer. A further six analogs ( 346 – 351 ) were designed, possessing the identical peptide backbones to 309 and 310 but with a variation in the side chain on the pyrolinone unit and the enol substitution of pyrolinone . The synthesis of compounds 346 – 351 was initiated by preparation of the requisite pyrolinones 333 – 338 from Fmoc-protected amino acids 319 – 322 . Then, Meldrum’s acid in ethyl acetate was coupled with amino acids 319 – 322 using EDC and DMAP followed by reflux of the Meldrum’s adduct in ethyl acetate. This step affected the condensate cyclization to give Fmoc-protected pyrolinones 323 – 326 , which were then reacted with methanol under Mitsunobu conditions using DIAD and triphenylphosphine, respectively, to produce O-methylated pyrolinones 327 – 330 . Compounds 323 – 324 were reacted with benzyl alcohol under the same conditions to produce 331 – 332 . Compounds 333 – 338 were produced by treating piperidine in acetonitrile. After each pyrrolinone building block was obtained, amino acid 339 was activated as the corresponding pentafluorophenyl ester. Separately, pyrolinones 333 – 338 were deprotonated with n-butyllithium at low temperatures before the addition of pentafluorophenyl ester to yield 340 – 345 (36–59%). Due to the unfavorable results, acidolysis of the Boc groups from 340 – 345 was followed by coupling to the N-terminal tripeptide 304 using the coupling reagent HATU and NMM as the base to produce the desired gallinamide A analogs 346 – 351 with excellent yields. The natural product gallinamide A was also synthesized by Stoye et al. (2019) and possesses potent inhibitory activity against P. falciparum cysteine proteases, namely falcipain, and therefore shows promise as a potential malaria treatment to breakdown hemoglobin in the parasitic food vacuole . Gallinamide A was synthesized using NMM in a solution of the imide fragments 353 – 354 (1.0 equiv, as the trifluoroacetate salt), tripeptide 334 – 336 (1.5 equiv, as the trifluoroacetate salt), HATU and HOAt in DMF/CH 2 Cl 2 (1:1 v / v ). After consumption of the starting material (as evidenced by LC-MS), the solvent was subsequently removed by an N 2 stream and the residue was purified by preparative RP-HPLC . This synthesis produced different yields of gallinamide A analogs 371 – 387 ( and ). 3.2.3. PfSERA5 Analogs PfSERA5 is an abundant asexual antigen that can inhibit parasitic growth in vitro and is a candidate malaria falciparum vaccine. Kanodia et al. (2014) reported the design and synthesis of peptides with similar sequences to the SERA5 protein, an inhibitor of malaria parasite development . The solid-phase synthesis of all peptides was initiated on Rink amide resin. The appropriate Fmoc-amino acid was dissolved in a solution of TBTU/HOBT and DIPEA before being added to the resin. Fmoc was deprotected by piperidine in DMF and then acetylated by a mixture of acetic anhydride and DIPEA. The nine SE5 P1-P9 ( 388 – 396 ) were cleaved from the resin using TFA/H 2 O/TIS/Thioanisole/Phenol and their sequences are shown in . 3.2.4. Angiotensin Silva et al. (2015) investigated angiotensin II, a peptide that has antiplasmodial activity, as an antimalarial drugs . They synthesized 10 peptides ( 397 – 406 ) via manual solid-phase synthesis . The Fmoc strategies were applied using Wang resins and deprotection was performed by treatment with 4-MePip in DMF. Couplings were conducted using DIC/HOBt in DCM/DMF (1:1, v / v ) and were monitored using the Kaiser ninhydrin test. Dry-protected peptidyl resin was exposed to TFA/H 2 O/anisole (95:2.5:2.5, v / v / v ) for 2 h at room temperature to produce the crude linear peptides. All crude peptides were precipitated with anhydrous diethyl ether, separated from the ether–soluble reaction components by filtration, extracted from the resin with 60% ACN in water, and lyophilized. They were purified by preparative RP-HPLC in 0.1% TFA/60% ACN in water on a Waters Associates system (Delta Prep 600). The peptides were loaded onto a Phenomenex C18 (21.2 × 250 mm, 15 µm particle size, 300 Å pore size) column at a flow rate of 10.0 mL/min and eluted using a linear gradient (slope 0.33% B/min) of TFA/ACN with detection at 220 nm. Selected fractions containing the purified peptides were pooled and lyophilized. In 2015, Silva et al. also synthesized nine octapeptides ( 407 – 415 ) of the renin-angiotensinsystem (RAS), which have been reported to have anti-plasmodium activity towards P. gallinaceum (88% sporozoite inactivation) . The angiotensin II analogs were synthesized by Fmoc/tert-butyloxycarbonyl (t-Boc) strategy on a solid phase, purified by liquid chromatography, and characterized by mass spectrometry. The amino acid Nα-terminal protecting group was removed with TFA in DCM in the presence of 2% anisole for 20 min. Coupling and deprotection reactions were carried out using the same reagent as before. Repeated couplings were performed for one hour using TBTU with DIPEA in DCM/NMP. Dry protected peptidylresin was exposed to 70% TFA/20% TFMSA in 10% anisole for 12 h. All crude peptides were precipitated with anhydrous diethyl ether, separated from the ether–soluble reaction components by filtration, and the fractions were lyophilized using 60% ACN (acetonitrile) in water. Ten angiotensin analogs ( 416 – 425 ) were also synthesized by Torres et al. (2015) . These lactams and sulfide bridge-containing peptides were synthesized manually by the solid-phase method using a t-Bu strategy with chloromethylated resin and a Fmoc strategy with Wang resin. Deprotection was conducted with TFA and DCM and the addition of 2% anisole. The resin was washed with anisol in isopropyl alcohol and TEA in DCM and MeOH, and the reaction was monitored with the Kaiser test. Recoupling and deprotecting the protective groups on each amino acid used the same reagents as before. Furthermore, the cyclization reaction was conducted with excess Castro reagent and DIPEA dissolved in DMSO/NMP, after which the resin was washed with TEA in DCM to obtain a dry peptidyl-resin. Deprotection of the sulfide bridge analogs was conducted by adding 4-MePip in DCM. The amino acid coupling reaction involved treating the protected amino acid acyl-resin with a molar excess of the Boc/Fmoc protected amino acid using the DIC/HOBt reagent. The disulfide bridges were formed by first dissolving the peptide in an acetic acid solution containing iodine. The peptides were then lyophilized after being extracted with water and diethyl ether and purified by RP-HPLC. Silva et al. (2017) synthesized eight angiotensin II hormones ( 426 – 433 ) . They optimized the synthesis of these peptides for their antiplasmodial activity and to reduce their vasoconstriction and rapid degradation characteristic. The peptides were synthesized by manual solid-phase synthesis and characterized by circular dichroism spectroscopy. The synthesis was similar to previously reported and involved manual phase peptide synthesis using the Fmoc strategy and Wang resin. The amino acid residues were coupled using the DIC/HOBt reagent in DCM. Each step was followed by alternating washing with DMF, methanol, and DCM to change the degree of resin swelling and promote the removal of excess reagents. Dry peptidyl resin was added to TFA/water/anisole to obtain unprotected peptides. The S-S bonds were formed by first dissolving the crude peptide in an iodine-containing acetic acid solution. The peptides were extracted with water and diethyl ether, evaporated, and then lyophilized. 3.2.5. Decoralin Torres et al. (2018) reported the re-engineering of a wasp venom peptide, decoralin, into a synthetic anti-malaria agent through modifications that removed its hemolytic activity . Dec-NH 2 and eight analogs ( 434 – 441 ) were synthesized and designed to preserve specific physicochemical structures. Certain amino acid substitutions significantly improved the antiplasmodial activity, giving new sequence principles for creating potent anti-malaria drugs such as replacing the original sequence by Arg, Phe and Trp. The synthesis involved a solid-phase method and purification by chromatography and characterization using MS. The antiplasmodial activities were assessed by fluorescence microscopy. All residue substitutions resulted in increased anti-Plasmodium activity , with [Arg] 1 -Dec-NH 2 , [Pro] 4 -Dec-NH 2 , and [Phe] 2 -Dec- NH 2 being the most active peptides tested. 3.2.6. Carmabin and Dragomabin Ye et al. (2018) synthesized carmabin A and dragomabin, which demonstrated antimalarial activity . Carmabin A and dragomabin synthesis was initiated by retrosyntethic analysis . The C-terminal amide of carmabin A ( 442 ) and dragomabin ( 443 ) was prepared via amidation of the C-terminal methyl esters of compounds 444 and 445 . Compounds 448 were further disconnected into two parts: the Mdya/Moya fragment and protected tetrapeptide 454 ( and ). Tetrapeptide 44 8 was prepared by repeated condensation of amino acids . The methyl group of Mdya/Moya 446 / 447 could be stereoselective. Carboxylic acid 454 was converted to the corresponding alcohol 453 in four steps involving acyl chlorination, amidation with benzyl-2-oxazolidinone, diastereoselective-methylation, and reduction. The addition of carboxylic 454 with N HCl led to the hydrolysis of the TMS group and the amide bond to yield compound 446 . shows the construction of tetrapeptide 447 , the treatment of compound 457 , and TFA resulting in Boc deprotection. The condensation of compound 458 with the coupling reagent HATU/DIPEA produced compound 459 and repeated condensation under the same conditions yielded tetrapeptide 447. Carmabin A was synthesized from the building blocks 446 , ent- 446 , and 447 ( and ). The coupling reaction between Boc-deprotected 447 and 446 and ent- 446 was followed by the addition of ammonia to produce compounds 442 and 442a . The configuration of carmabin A was further investigated using NMR, showing that only the 442 NMR data matched that of natural carmabin A. Dragomabin was synthesized using the building blocks Moya 447 and ent -447 . First, Moya 447 is generated from ent -452 by deprotecting with TMS and TFA, then the chiral auxiliary is removed (Schema 38). Dragomabin is obtained by deprotecting 448 with TMF under DCM to provide 445 , which is amidated with NH 3 to form compound 443 . The correct structure for dragomabin was revised as shown in 443a . As dragomabin and dragonamide differ in the stereochemistry on the Moya fragment, the stereochemistry of the alkyne fragment in these lipopeptides is varied, and correlation with other natural products is not reliable. The absolute stereochemistry at C35 and C37 of carmabin A was assigned as 35R, 37S and the absolute stereochemistry at C35 of dragomabin has been revised as 35R. Somanadhan et al. (2013) succeeded in synthesizing linear antimalarial peptides. Falcitidin 284 is the first member of a new class of falcipain-2 inhibitors, which is a cysteine protease used by P. falciparum to degrade hemoglobin during the trophozoite stage of infection . Falcitidin contains isovaleric acid-D-His-L-Ile-L-Val-L-Pro-NH 2 and was synthesized by the solution-phase peptide synthesis method . The synthesis begins with deprotecting the Boc group from N-Boc-proline carboxamide using TFA. The amino-proline TFA salt was coupled with freshly prepared L-Ile-L-Val using HATU/HOAt to give L-Ile-L-Val-L-Pro-NH 2 . The dipeptide was obtained by coupling L-Ile-NHBoc with the in situ prepared bis-trimethylsilane (TMS) ether of L-Val under mixed anhydride conditions. The Boc group of the tripeptide was then cleaved using TFA and coupled to Na-Fmoc-N(im)-trityl-D-histidine using HATU/HOAt to give the Fmoc–tetrapeptide. The Fmoc group of was removed using piperidine and HATU/HOAt. The final synthetic target falcitidin was obtained after trityl deprotection of Trt-falcitidin using TFA in the presence of triisopropylsilane. Kotturi et al. (2014) synthesized falcitidin and its analogs ( 296 – 303 ) using the same solution phase peptide synthesis . First, N im -trityl-D-histidine 288 was converted into bis-TMS ether with TMSCl/Et 3 N 7 and coupled immediately with the mixed anhydride of isovaleric acid 287 using ethyl chloroformate/NMM to form the trityl protected N-acyl-D-histidine. In parallel, the N-Boc proline amide was deprotected with TFA, and the crude amine salt was coupled immediately with the known dipeptide N-Boc-L-Ile-L-Val by using HATU/HOAt to yield the key tripeptide intermediate N-Boc-L-Ile-L-Val-L-Pro-NH 2 . Lastly, the Boc group of tripeptide was cleaved using TFA and coupled to N-acyl-D-histidine using HATU/HOAt to give the N im -trityl protected tetrapeptide 292 . Falcitidin acylatide and its analogs produce eight analogs compounds ( 296 – 303 ) by diversifying the synthesis of the N-acyl tetrapeptide analog of falcitidin through the common tripeptide N-Boc-L-Ile-L-Val-L-Pro-NH 2 . Gallinamide A is a linear peptide that has antimalarial activity. Conroy et al. (2014) designed and synthesized 18 gallinamide A analogs by solid-phase synthesis and substitution of amino acid residues . Four gallinamide A analogs ( 309 – 312 ) were originally designed with varying levels of saturation, including compound 309 , which is structurally identical to gallinamide A. Another analog was compound 310 with a reduced methoxy-enol moiety in the pyrolinone ring 309 . Part of the olefinic component of the 4( S )-amino-2-( E )-pentenoic acid unit is reduced in compound 311 and 312 has reduced olefin groups. Analog synthesis was initiated by synthesizing the N -terminal fragment 304 by solid-phase peptide synthesis using the Fmoc strategy. The 2-CTC resin is filled with Fmoc-Leu-OH followed by the subsequent incorporation of the amino acid. Reductive amination of the resin, followed by cleavage of the resin using hexafluoroisopropanol (HFIP), produced an excellent yield of tripeptide 304 which was coupled with imide fragments 306 and 307 prepared using a similar protocol. The coupling reaction involved HATU at low temperature to avoid epimerization to produce good yields of analogs 309 and 310 . At this stage, 309 and 310 underwent hydrogenation to give 311 and 312 . Analogs 313 – 318 were synthesized from 2-CTC resin filled with Fmoc-Ala . The coupling of Fmoc-protected α,β-unsaturated amino acids was followed by elongation through the Fmoc solid-phase peptide synthesis strategy. With reductive methylation of the N-terminus using formaldehyde and sodium cyanoborohydride, the peptide was cleaved from the resin using HFIP to produce the C-terminal peptide acid. C-terminal functionality was achieved through benzylamine coupling using PyBOP at low temperatures, and compounds 313 , 315 , 317 , and 318 were purified by RP-HPLC and demonstrated no significant epimerization. Compound 314 was produced by attaching 4-( R )-hydroxy L-proline methyl ester to the C-terminus of the two peptides with the addition of NMM as a hindered base. Aminothiazole was coupled to the C-terminus of one peptide acid using PyBOP at low temperature to yield compound 316 as a mixed diastereoisomer. A further six analogs ( 346 – 351 ) were designed, possessing the identical peptide backbones to 309 and 310 but with a variation in the side chain on the pyrolinone unit and the enol substitution of pyrolinone . The synthesis of compounds 346 – 351 was initiated by preparation of the requisite pyrolinones 333 – 338 from Fmoc-protected amino acids 319 – 322 . Then, Meldrum’s acid in ethyl acetate was coupled with amino acids 319 – 322 using EDC and DMAP followed by reflux of the Meldrum’s adduct in ethyl acetate. This step affected the condensate cyclization to give Fmoc-protected pyrolinones 323 – 326 , which were then reacted with methanol under Mitsunobu conditions using DIAD and triphenylphosphine, respectively, to produce O-methylated pyrolinones 327 – 330 . Compounds 323 – 324 were reacted with benzyl alcohol under the same conditions to produce 331 – 332 . Compounds 333 – 338 were produced by treating piperidine in acetonitrile. After each pyrrolinone building block was obtained, amino acid 339 was activated as the corresponding pentafluorophenyl ester. Separately, pyrolinones 333 – 338 were deprotonated with n-butyllithium at low temperatures before the addition of pentafluorophenyl ester to yield 340 – 345 (36–59%). Due to the unfavorable results, acidolysis of the Boc groups from 340 – 345 was followed by coupling to the N-terminal tripeptide 304 using the coupling reagent HATU and NMM as the base to produce the desired gallinamide A analogs 346 – 351 with excellent yields. The natural product gallinamide A was also synthesized by Stoye et al. (2019) and possesses potent inhibitory activity against P. falciparum cysteine proteases, namely falcipain, and therefore shows promise as a potential malaria treatment to breakdown hemoglobin in the parasitic food vacuole . Gallinamide A was synthesized using NMM in a solution of the imide fragments 353 – 354 (1.0 equiv, as the trifluoroacetate salt), tripeptide 334 – 336 (1.5 equiv, as the trifluoroacetate salt), HATU and HOAt in DMF/CH 2 Cl 2 (1:1 v / v ). After consumption of the starting material (as evidenced by LC-MS), the solvent was subsequently removed by an N 2 stream and the residue was purified by preparative RP-HPLC . This synthesis produced different yields of gallinamide A analogs 371 – 387 ( and ). PfSERA5 is an abundant asexual antigen that can inhibit parasitic growth in vitro and is a candidate malaria falciparum vaccine. Kanodia et al. (2014) reported the design and synthesis of peptides with similar sequences to the SERA5 protein, an inhibitor of malaria parasite development . The solid-phase synthesis of all peptides was initiated on Rink amide resin. The appropriate Fmoc-amino acid was dissolved in a solution of TBTU/HOBT and DIPEA before being added to the resin. Fmoc was deprotected by piperidine in DMF and then acetylated by a mixture of acetic anhydride and DIPEA. The nine SE5 P1-P9 ( 388 – 396 ) were cleaved from the resin using TFA/H 2 O/TIS/Thioanisole/Phenol and their sequences are shown in . Silva et al. (2015) investigated angiotensin II, a peptide that has antiplasmodial activity, as an antimalarial drugs . They synthesized 10 peptides ( 397 – 406 ) via manual solid-phase synthesis . The Fmoc strategies were applied using Wang resins and deprotection was performed by treatment with 4-MePip in DMF. Couplings were conducted using DIC/HOBt in DCM/DMF (1:1, v / v ) and were monitored using the Kaiser ninhydrin test. Dry-protected peptidyl resin was exposed to TFA/H 2 O/anisole (95:2.5:2.5, v / v / v ) for 2 h at room temperature to produce the crude linear peptides. All crude peptides were precipitated with anhydrous diethyl ether, separated from the ether–soluble reaction components by filtration, extracted from the resin with 60% ACN in water, and lyophilized. They were purified by preparative RP-HPLC in 0.1% TFA/60% ACN in water on a Waters Associates system (Delta Prep 600). The peptides were loaded onto a Phenomenex C18 (21.2 × 250 mm, 15 µm particle size, 300 Å pore size) column at a flow rate of 10.0 mL/min and eluted using a linear gradient (slope 0.33% B/min) of TFA/ACN with detection at 220 nm. Selected fractions containing the purified peptides were pooled and lyophilized. In 2015, Silva et al. also synthesized nine octapeptides ( 407 – 415 ) of the renin-angiotensinsystem (RAS), which have been reported to have anti-plasmodium activity towards P. gallinaceum (88% sporozoite inactivation) . The angiotensin II analogs were synthesized by Fmoc/tert-butyloxycarbonyl (t-Boc) strategy on a solid phase, purified by liquid chromatography, and characterized by mass spectrometry. The amino acid Nα-terminal protecting group was removed with TFA in DCM in the presence of 2% anisole for 20 min. Coupling and deprotection reactions were carried out using the same reagent as before. Repeated couplings were performed for one hour using TBTU with DIPEA in DCM/NMP. Dry protected peptidylresin was exposed to 70% TFA/20% TFMSA in 10% anisole for 12 h. All crude peptides were precipitated with anhydrous diethyl ether, separated from the ether–soluble reaction components by filtration, and the fractions were lyophilized using 60% ACN (acetonitrile) in water. Ten angiotensin analogs ( 416 – 425 ) were also synthesized by Torres et al. (2015) . These lactams and sulfide bridge-containing peptides were synthesized manually by the solid-phase method using a t-Bu strategy with chloromethylated resin and a Fmoc strategy with Wang resin. Deprotection was conducted with TFA and DCM and the addition of 2% anisole. The resin was washed with anisol in isopropyl alcohol and TEA in DCM and MeOH, and the reaction was monitored with the Kaiser test. Recoupling and deprotecting the protective groups on each amino acid used the same reagents as before. Furthermore, the cyclization reaction was conducted with excess Castro reagent and DIPEA dissolved in DMSO/NMP, after which the resin was washed with TEA in DCM to obtain a dry peptidyl-resin. Deprotection of the sulfide bridge analogs was conducted by adding 4-MePip in DCM. The amino acid coupling reaction involved treating the protected amino acid acyl-resin with a molar excess of the Boc/Fmoc protected amino acid using the DIC/HOBt reagent. The disulfide bridges were formed by first dissolving the peptide in an acetic acid solution containing iodine. The peptides were then lyophilized after being extracted with water and diethyl ether and purified by RP-HPLC. Silva et al. (2017) synthesized eight angiotensin II hormones ( 426 – 433 ) . They optimized the synthesis of these peptides for their antiplasmodial activity and to reduce their vasoconstriction and rapid degradation characteristic. The peptides were synthesized by manual solid-phase synthesis and characterized by circular dichroism spectroscopy. The synthesis was similar to previously reported and involved manual phase peptide synthesis using the Fmoc strategy and Wang resin. The amino acid residues were coupled using the DIC/HOBt reagent in DCM. Each step was followed by alternating washing with DMF, methanol, and DCM to change the degree of resin swelling and promote the removal of excess reagents. Dry peptidyl resin was added to TFA/water/anisole to obtain unprotected peptides. The S-S bonds were formed by first dissolving the crude peptide in an iodine-containing acetic acid solution. The peptides were extracted with water and diethyl ether, evaporated, and then lyophilized. Torres et al. (2018) reported the re-engineering of a wasp venom peptide, decoralin, into a synthetic anti-malaria agent through modifications that removed its hemolytic activity . Dec-NH 2 and eight analogs ( 434 – 441 ) were synthesized and designed to preserve specific physicochemical structures. Certain amino acid substitutions significantly improved the antiplasmodial activity, giving new sequence principles for creating potent anti-malaria drugs such as replacing the original sequence by Arg, Phe and Trp. The synthesis involved a solid-phase method and purification by chromatography and characterization using MS. The antiplasmodial activities were assessed by fluorescence microscopy. All residue substitutions resulted in increased anti-Plasmodium activity , with [Arg] 1 -Dec-NH 2 , [Pro] 4 -Dec-NH 2 , and [Phe] 2 -Dec- NH 2 being the most active peptides tested. Ye et al. (2018) synthesized carmabin A and dragomabin, which demonstrated antimalarial activity . Carmabin A and dragomabin synthesis was initiated by retrosyntethic analysis . The C-terminal amide of carmabin A ( 442 ) and dragomabin ( 443 ) was prepared via amidation of the C-terminal methyl esters of compounds 444 and 445 . Compounds 448 were further disconnected into two parts: the Mdya/Moya fragment and protected tetrapeptide 454 ( and ). Tetrapeptide 44 8 was prepared by repeated condensation of amino acids . The methyl group of Mdya/Moya 446 / 447 could be stereoselective. Carboxylic acid 454 was converted to the corresponding alcohol 453 in four steps involving acyl chlorination, amidation with benzyl-2-oxazolidinone, diastereoselective-methylation, and reduction. The addition of carboxylic 454 with N HCl led to the hydrolysis of the TMS group and the amide bond to yield compound 446 . shows the construction of tetrapeptide 447 , the treatment of compound 457 , and TFA resulting in Boc deprotection. The condensation of compound 458 with the coupling reagent HATU/DIPEA produced compound 459 and repeated condensation under the same conditions yielded tetrapeptide 447. Carmabin A was synthesized from the building blocks 446 , ent- 446 , and 447 ( and ). The coupling reaction between Boc-deprotected 447 and 446 and ent- 446 was followed by the addition of ammonia to produce compounds 442 and 442a . The configuration of carmabin A was further investigated using NMR, showing that only the 442 NMR data matched that of natural carmabin A. Dragomabin was synthesized using the building blocks Moya 447 and ent -447 . First, Moya 447 is generated from ent -452 by deprotecting with TMS and TFA, then the chiral auxiliary is removed (Schema 38). Dragomabin is obtained by deprotecting 448 with TMF under DCM to provide 445 , which is amidated with NH 3 to form compound 443 . The correct structure for dragomabin was revised as shown in 443a . As dragomabin and dragonamide differ in the stereochemistry on the Moya fragment, the stereochemistry of the alkyne fragment in these lipopeptides is varied, and correlation with other natural products is not reliable. The absolute stereochemistry at C35 and C37 of carmabin A was assigned as 35R, 37S and the absolute stereochemistry at C35 of dragomabin has been revised as 35R. Ribifolin ( 1 ) is moderately effective against the Plasmodium falciparum 3D7 strain with an IC 50 of 42 µM , while its linear analogue 1a had an IC 50 of 519 µM, demonstrating the importance of cyclization to enhance the biological activity . None of the tested compounds showed cytotoxic potential against human cells (HEK293aT), although normal growth was observed in the concentration range of 0.001–100 μM. The isolated cyclopeptide alkaloids ( 2 – 5 ) from the root bark of Hymenocardia acida demonstrated antiplasmodic activity against P. falciparum K1, with IC 50 values ranging from 12.2 to 27.9 μM. Only hymenocardine ( 2 ) exhibited cytotoxic properties against MRC-5 cells with an IC 50 of 51.1 ± 17 μM . The antiplasmodic activity of nummularine-R ( 6 ), O-desmethylnummularine-R ( 7 ), hemsine-A ( 9 ), ramosine-A ( 11 ), and oxyphylline-F ( 14 ) was evaluated against the P. falciparum strain KI, while cytotoxicity was tested on MRC-5 cells (human fetal lung fibroblast cells). Based on the structural features and IC 50 values for compounds 6 , 7 , 9 , and 11 , it can be assumed that the tryptophan moiety in the side chain is important for antiplasmodic activity. Compound 14 does not contain a tryptophan unit and also showed antiplasmodic activity, indicating that tryptophan could mediate, but is not vital for the antiplasmodic activity. The most promising compound was O-desmethylnummularine-R ( 7 ), which exhibited an IC 50 of >64.0 µM against MRC-5 cells . Maluf et al. (2016) reported that crotamine ( 15 ) selectively enters infected erythrocytes ( A) and has potent anti-plasmodial activity with an IC 50 value of 1.87 µM ( B) . The instability of H + homeostasis by cr otamine was also confirmed ( B,C). The authors suggest that crotamine alters the internal pH of the vesicle due to its abundant Lys and Arg residues and the resulting high net charge (8+). Internal pH regulation is important for parasite survival, as it regulates the activity of certain intracellular enzymes required for parasite growth. Thus, this polypeptide is a promising lead molecule for the development of potential new peptidomimetics that have selectivity for infected erythrocytes and the ability to inhibit malaria infection by their natural affinity for acid vesicles . The cyclic octadepsipeptides, octaminomycins A ( 16 ) and B ( 17 ) showed activity against P. falciparum . According to Jang et al. (2017), compounds 16 and 17 were not significantly cytotoxic at a concentration of 30 μM against human cervical cancer cells (HeLa), human promyelocytic leukemia cells (HL-60), mouse temperature-sensitive cdc2 mutant cells (tsFT210), and rat kidney cells, which were infected with ts25 (srcts-NRK). They were also evaluated for antimicrobial activity against Staphylococcus aureus 209, Escherichia coli HO141, Aspergillus fumigatus Af293, Pyricularia oryzae kita-1, and Candida albicans JCM1542, as well as antimalarial activity against the P. falciparum 3D7, Dd2, and K1 strains. Chloroquine was less effective against strains resistant to Dd2 and K1, whereas compounds 16 and 17 showed the same in vitro antimalarial activities against chloroquine-sensitive 3D7 and chloroquine-resistant Dd2 and K1 strains, with no antimicrobial activity up to 30 μM . The antimalarial activity of the isolated kakeromamide B peptide ( 18 ) was evaluated against asexual blood-stage and liver-stage P. falciparum . Compound 18 exhibited moderate activity against the blood stage of P. falciparum with an EC 50 value of 8.9 µM as well as moderate liver-stage antimalarial activity against P. berghei liver schizonts with EC 50 values of 11 µM. Although 18 displayed only moderate antimalarial activity, its ability to inhibit both the blood and liver life stages of Plasmodium, coupled with its low cytotoxicity in human cell lines, make it a promising lead compound for drug discovery . The novel cyclic antimalarial and antitrypanosomal hexapeptide, pipecolisporin ( 19 ), was isolated from cultures of Nigrospora oryzae CF-298113 and exhibited interesting activity against P. falciparum and Trypanosoma cruzi Tulahuen C4 parasites . The activity against the T. cruzi Tulahuen C4 parasites was the most remarkable, with an IC 50 of 8.46 µM, comparable to that of the standard drug benznidazole, currently used in the treatment of Chagas disease (IC 50 in the same assay of 2.21 µM) ( and ). The activity against P. falciparum was also in the micromolar range, with an IC 50 of 3.21 µM . It was not cytotoxic to the human cancer cell lines A549 (lung carcinoma), A2058 (metastatic melanoma), MCF7 (breast adenocarcinoma), MIA PaCa-2 (pancreatic carcinoma), and HepG2 (hepatocyte carcinoma) at the highest concentration tested of 50 µM . Koshidacins A ( 20 ) and B ( 21 ) demonstrated in vitro antiplasmodial activity and cytotoxicity against human MRC-5 cells . Compounds 20 and 21 demonstrated antiplasmodial activity against a chloroquine-sensitive P. falciparum FCR3 strain with IC 50 values of 17.1 and 0.89 μM, respectively. Similarly, compounds 20 and 21 also exhibited antiplasmodial activity against chloroquine-resistant P. falciparum K1 strain with IC 50 values of 12.5 and 0.83 μM, respectively. Compounds 20 and 21 showed cytotoxicity against human MRC-5 cells, with IC 50 values of 6.8 and 14.7 μM, respectively, suggesting selectivity indices ranging from 0.4 to 18.4 . In addition, when given intraperitoneally at a dose of 30 mg/kg/day for 4 days, compound 21 inhibited 41% of malaria parasites in vivo . Shi et al. (2018) evaluated georatusin ( 22 ) produced by a soil fungus Geomyces auratus . It had no obvious cytotoxicity, but displayed antiparasitic activities against Leishmania donovani (IC 50 = 9.1 μM) and P. falciparum (IC 50 = 1.6 μM) . This discovery offers new insight into the metabolic potential and ecological importance of Geomyces and may encourage further exploration of this genus. New highly N-methylated linear peptides, friomamaride B ( 23 ) and shagamides A-F ( 24 – 29 ), exhibited activity against three strains of blood-stage P. falciparum . All compounds were tested for their effectiveness against three blood-stage of P. falciparum using the blood-stage antiplasmodial and cytotoxicity assay. Friomamaride B ( 23 ), shagamides C ( 26 ) and D ( 27 ) with values less than 10 µg/mL all possess potential activity. The N-terminal of phenylalanine residue is essential for this activity. None of the isolated compounds demonstrated cytotoxic activity. The carboxypeptidase inhibitor NpCI peptide ( 30 ), which is related to the model enzymes of bovine carboxypeptidase A (bCPA) and porcine carboxypeptidase B (pCPB), was discovered by Cabrera-Muñoz et al. (2023) . The kinetic characterization of NpCI revealed that it was a slow inhibitor of bCPA and pCPB. While pCPB inhibition was not significant, the evaluation of NpCI inhibition was also performed by comparing the decrease in bCPA inhibitory activity caused by the substrate with the increase in substrate concentration. The Dd2 strain was used for in vitro antiplasmodial activity against P. falciparum , showing that the cycle was significantly delayed with an IC 50 of 5.5 µmol/L. Parasite growth can be slowed down by increasing inhibitor concentration. The growth inhibition by NpCI occurs during parasite development. The 3D7 strain displayed comparable inhibition with a delayed maturation mechanism. NpCI was not cytotoxic to human cells (IC 50 < 25%). Kiefer et al. (2019) synthesized 17 cyclomarin analogues ( 96 – 100 , 106 – 110 , 119 – 125 ) and evaluated their biological activities against chloroquine-sensitive Pfalcp strain 3D7 and multi-resistant strain Dd2, as well as against Mtb wild-type strain Erdma . The antitubercular activity was also evaluated, and the resazurin reduction microtiter assay (REMA) was used to measure the growth inhibition of Mtb. Three compounds ( 96 , 97 , 99 ) demonstrated excellent parasitic growth suppression . Compound 99 consists of a simplified γ,δ-unsaturated side chain, and the N’-methyl tryptophan unit. Desoxycyclomarin C, which was inspired by a natural product compound, has a similar range of bioactivity, but it is significantly shortened. The anti-mycobacterial activity was demonstrated by the five derivatives ( 96 , 98 , 99 , 108 and 110 ) when applied to Erdman wild-type strain. Due to a remarkable simplification by replacing two of the four non-canonical amino acids by L-valine and L-tryptophan, compound 110 represents a highly appealing natural product-derived lead structure for battling Mtb. Fagundez et al. (2018) revealed that most of the synthesized peptides have the potential to be antimalarials . The corresponding cyclopeptide effectiveness was evaluated against the chloroquine-resistant K1 strain of P. falciparum . Cyclo-Cys(Trt)-Gly-Thr(tBu)-Gly-Cys(Trt)-Gly (compound 140 ) showed potent in vitro and selective activity against this parasite with an EC 50 = 28 µM. The inclusion of a carboxylic group from Glu ( 151 , 152 ) can increase solubilization, and the substitution of hydrophobic amino acids can boost biological activity . Fagundez et al. (2018) also developed cyclopeptides containing N-methyl amino acids that demonstrated promising antiplasmodial activity, compounds 156 – 161 and 163 – 165 . In addition, a new class of antimalarial cyclopeptides that contain N-methyl Gly has been developed that exhibits enhanced antiplasmodial activity. The in vitro evaluation of the compounds against P. falciparum revealed that N-Me-Gly is required to maintain the activity in the presence of a Glu with a free carboxyl group. Moreover, none of the active compounds are toxic against HepG2 cells. Compounds 158 and 160 were assessed for their antimalarial activity in comparison to other antimalarial drugs . These results point to an antimalarial mode of action that does not immediately affect parasite viability. The prophylactic potential of compounds 158 and 160 is demonstrated by their low and submicromolar EC 50 values (0.018 and 0.355 µM) in the liver stage of the parasite. Compounds 157 and 165 are effective against P. berghei parasites, reducing parasitemia by 71 and 66% on day 5. To evaluate the oral bioavailability, the plasma pharmacokinetic of compound 157 in male Swiss Albino mice, following a single oral dosage, was examined, and shows a considerable half-life of 4.93 h. The hirsutellide A analog, compound 177 , exhibited moderate antiplasmodial activity (IC 50 = 2.3 μM) similar to that reported for hirsutellide A (IC 50 = 4.2 μM). Compound 177 is not cytotoxic (IC 50 > 100 μM) to Hep2G cells. ADME profiling for compound 177 displayed moderate stability in humans, but low stability in mouse microsomes . Additionally, the analogs have little to no activity against Mtb H37Rv. Peptide analogs generally have higher antiplasmodial activity (IC 50 = 1.8−7.7 μM) than the depsipeptide analogs (IC 50 = 7.5−20.1 μM), exhibiting a higher aqueous solubility, a high plasma stability and mouse plasma stabilities. Thus, the ester-to-amide substitution and the membrane permeability of hirsutellide A analogs appear to be dependent on the nature of the amino acid substituents. This study provides insight into the structural features relevant to the cyclic peptide-related drugs. Some compounds exhibited promising antimycobacterial activity and low cytotoxicity, making them potential candidates for further research and development as anti-TB agents. The developed macrocyclic peptides as proteasome inhibitors has potential for antimalarial drugs (Zhang et al., 2022) . Cyclic peptide 201 is a noncovalent inhibitor with strong antimalarial activity and high species selectivity, but has poor pharmacokinetics, therefore a docking model was developed. Compound 201 was acetylated using the N-Terminal amino group to boost its antiparasitic and inhibition activities. Five compounds were created, and P1, P3, P5, and P2–P4 linkers were examined for their contribution to the potency and drug-like properties of macrocyclic peptides ( , , and ) . Compound 220 demonstrated good potency against parasite and pharmacokinetics and made optimal use of P1 amide moiety modification . Compound 223 showed moderate inhibition against hu-β2c (IC 50 = 97.4 μM). Compounds 272 and 270 exhibited moderate inhibition of β1c and β1i. All other compounds showed <50% inhibition against all the four subunits, even at 100 μM. Similar to the observation of the simultaneous inhibition of β5 and β2 in tumors, co-inhibition of β5 and β2 is synergistic and the inhibition of Pf20s β5 is sufficient to kill Plasmodium. The compounds were optimized for aqueous solubility, passive membrane permeability, metabolic stability, and a clean off-target profile against CYP450s and the hERG channel. The proteasome function appears to be critical for parasites to survive. Falcitidin and its analogs ( 296 – 303 ) were synthesized by Kotturi et al. (2014) and showed moderate activity against P. falciparum 3D7, but only when N-tritylated on its histidine residue . The IC 50 activity of the new compounds, which ranged from 1 to 5 µM, was typically modest in whole cells. Compound 302 exhibited the highest activity (IC 50 0.14 µM). These new compounds represent an important new peptide chemotype that may be elaborated into improved antimalarial leads . Conroy et al. (2014) developed a new class of antimalarial drugs based on the natural product gallinamide A, which inhibits the falcipain cysteine proteases essential for malaria parasite survival . These gallinamide A analogs were as effective as chloroquine against the P. falciparum parasite. Gallinamide A analog 309 has strong inhibitory activity against FP-2, FP-3, and P. falciparum in vitro . Reducing the enol moiety in the acyl-pyrrolinone unit in compound 310 resulted in a two-fold reduction in antiplasmodial activity (IC 50 = 210 nM), but a slight improvement in activity against FP-2 and FP-3. Analog 311 showed no measurable inhibition of FP-3 and a 3 orders of magnitude drop in inhibitory effectiveness against FP-2 (IC 50 = 3710 nM), as well as significantly decreased activity against P.falciparum. Analog 312 lost the measurable inhibitory activity against the FPS and the parasite after losing both of its olefinic moieties. The activities of analogs 313 , 315 , and 317 were similar: they inhibited FP-2 at low micromolar concentrations (IC 50 = 3.4–11.5 µM), did not significantly inhibit FP-3 at 25 µM, and inhibited P. falciparum at nanomolar concentrations (IC 50 = 320–5400 nM) . Interestingly, peptide 318 demonstrated less potent antiparasitic activity (IC 50 = 6.6 µM) but inhibited FP-2 and FP-3 by adding an N-methylproline functionality at the N-terminus of the peptide while keeping the C terminal benzylamide. The lack of activity was especially noticeable for 315 because it has the same structure as analog 309 . The addition of a more flexible and highly functionalized hydroxyproline methyl ester to the C-terminus in 314 produced inhibitory activity against FP-2 and P. falciparum , which was comparable to the benzylamide-derived compounds. Low micromolar antiparasitic activity (IC 50 = 1.1 µM) and moderate inhibitory activity against FP-2 and FP-3 were achieved by C-terminal functionalization as a thiazole amide in compound 316 . Like analog 309 , the C-terminal N -acyl-pyrrolinone in 346 – 351 restored strong inhibitory activity against the FPs and P. falciparum . Compound 346 displayed similar inhibitory activity to 309 against FP-2, FP-3, and P. falciparum despite having no modification on the pyrrolinone molecule. The pyrrolinone ring of 347 was modified to include a more hydrophobic substituent, which improved its inhibitory efficacy against FP-3 and P. falciparum . As with compound 347 , the addition of aromatic side chains to the pyrrolinone ring did not modify the inhibitory activity against FP-2 but significantly increased activity against FP-3 and P. falciparum . The most effective inhibitor of FP-3 and P. falciparum was compound 349 , which had an indole side chain attached to the pyrrolinone ring . The compounds were screened against the P. falciparum , namely AP M1, AP M17, and AP M18 aminopeptidase (AP) enzymes. The effective N-acyl pyrrolinone analogs 309 and 346 – 349 , as well as the C-terminal amide derivatives 309 , 315 , and 317 , had IC 50 values of less than 600 nM against the 3D7 strain of P. falciparum . The CQ-resistant, Dd2 strain of P. falciparum was powerfully inhibited by every tested compounds (IC 50 = 29.0–421 nM). The compounds were selective inhibitors of P. falciparum over HEK298 cells, with 309 , 315 , and 317 showing no measurable inhibition of this cell line at a concentration of 50 µM, but having strong inhibitory activity against human cathepsin . The gallinamide A analogs, synthesized by Stoye et al. (2019), were also tested for activity against FP2, FP3 also CQ-sensitive 3D7 and CQ-resistant W2 strains of P. falciparum in vitro . Most substitutions in the compound structure were well tolerated, except for analog 387 which had a more potent activity parasite (IC 50 3D7 = 1 nM; W2 = 4 nM) and substituents on R 1 which markedly enhanced the stability in plasma and blood. Five gallinamide A analogs were assessed in vitro for their stability in mouse blood and plasma . The half-life in plasma and blood was significantly prolonged due to the substituents at R 1 and R 3 . Additional cytotoxicity assays in the HEK293 cell line of the compounds displayed no measurable cytotoxicity at 25 µm. The last three analogs 372 , 373 and 387 were assessed in vivo in a mouse model of cerebral malaria (CM), P. berghei ANKA (PbA) infection. Severe signs of disease as well as the rise in parasitemia were both significantly delayed by analog 387 . The peptides derived from PfSERA5 by Kanodia et al. (2014) inhibited the enzymatic activity of PfSERA5P50 protein, which in turn blocked the development of the parasites in an in vitro culture . Evaluation of PfSERA5 derivative against P. falciparum development and growth revealed that SE5 P1 ( 388 ) and SE5 P2 ( 389 ) have the highest parasite growth inhibition and invasion values of 60–70% . Analysis of the synthetic substrate Suc-LLVY-AMC revealed that PfSERA550′s proteolytic activity was greatly reduced by the two C-terminal peptides SE P1 ( 388 ) and SE5 P2 ( 389 ). Kanodia et al. also describe how they used computational modeling to examine molecular docking studies with the known crystal structure of PfSERA5 to explore the effects of SE5 P1 ( 388 ) and SE5 P2 ( 389 ) peptides. SE5 P1 ( 388 ) and SE5 P2 ( 389 ) occupied 50.1 and 57.5% larger than the substrate when the Suc-LLVY-AMC was applied to see the bond , suggesting that the peptides must interact with Glu638 and Ser640/Ser641 to be inhibitory . P. falciparum -infected RBCs uptake of labeled SE5 P1 and SE5 P2 peptides when the localization of biotinylated peptides therein inhibited the protease activity similarly to the non-biotinylated peptides. This indicates that the peptides have access to the intracellular parasites and are co-localized with the PfSERA5 protein. PfSERA5 plays an important role in parasite development and the final proteolytic cleavage, which can be produced as a new drug design and offers information on the potential use of these peptides as antimalarial therapeutics . Angiotensin II analogs synthesized by Silva et al. (2015) and tested in vitro to identify a short bioactive peptide as well as to verify the hydrophobic cluster’s influence on parasite-membrane interaction on both P. gallinaceum and P. falciparum . Fluoresence microscopy was utilized to examine the effects of the peptides on P. gallinaceum sporozoites produced by the salivary glands and the therapeutic index (MHC/MIC ratio). Higher values in the therapeutic index indicate more antimicrobial specificity. It is a parameter that measures the specificity of an antimicrobial agent and is calculated by the ratio MHC and MIC. The MHC/MIC ratio, which was higher in peptide 401 , indicates that the peptides have varied specificities . New peptides related to Ang II were designed, including the most hydrophobic amino acid residues (Val, Ile, Pro, and Phe), aromatic residues (Tyr, His, Pro, and Phe) and residues from the Ang II hydrophobic cluster (Tyr, Ile and His) in an attempt to verify the peptide–parasite interactions. Due to the influence of hydrophobic clusters, side chain aromatic rings, and hydrophobic residues, these peptides showed antiplasmodial activity in P. gallinaceum sporozoite (64–94%) and activity between 89 and 94% . The three peptides with the highest antiplasmodial activity were 1 ( 397 ), 5 ( 401 ), and 6 ( 402 ) with 94, 89, and 94%, respectively . The effect of the peptides in the P. falciparum erythrocytic cycle was assessed in vitro. All peptides reduced new ring formation at 10 −8 mol L −1 , which was studied by Saraiva et al. as the ideal concentration for these inhibition assays. Four analogs reduced the ring formed in the blood stage, but only analogs 5 ( 401 ) and 6 ( 402 ) showed inhibition that was higher by 50% than control . showed that after 24 h of incubation with 2–3% schizont-infected erythrocyte cultures in the absence (control) or presence of 10 − 8 mol L − 1 peptides, the percentage age of rings was evaluated (* p < 0.05 compared to the control, *** p < 0.001). The result is statistically significant compared to the (mean ± standard deviation, n = 2), as indicated by the dark grey shading . Another technique for determining the hemolytic effect of peptides demonstrated that peptides 401 and 402 had an effect of the Ang II to define the hemolysis. These peptides did not exhibit hemolytic effects . The hydrophobic portion and the Arg, Tyr, Pro, and Phe residues increased the antiplasmodial activity when they were present in the primary sequence. Furthermore, these peptides did not display hemolysis or contractile response activities, as shown in (*** p < 0.05 compared to control, n = 2) . The IC 50 values are more promising than the Ang II results, which were demonstrated by evaluating seven concentrations giving 7–65% inhibition . The results imply that intramolecular interactions cause conformational tendencies that are crucial for antiplasmodial activity. When present on the peptide primary sequence, the hydrophobic portion and the residues of Arg, Tyr, Pro, and Phe increased the antiplasmodial activity. Following in vivo model tests of this class of peptides, this type of research aids the development of new chemotherapeutics, which can be explored as antimalarial drugs . The synthesized RAS octapeptides were tested for their effectiveness against P. falciparum and P. gallinaceum , revealing that only [Ala 5 ]-Ang II showed equipotent Ang II activity with 45% of biological activity . At 10 −8 M concentration, P. gallinaceum erythrocytic cycle [Ala 6 ]-Ang II decreased parasite invasion by 49%. The efficiency of anti-plasmodial activity was significantly impacted by the presence or absence of amino acid substitutions. The anti-plasmodial activity of the Ang II molecule depends on specific amino acid side chains. The biological activities and binding affinities are affected when the biologically active peptide is replaced by an alanine residue. From , the percentage of rings was determined after 24 h of incubation of erythrocyte culture infected with 2–3% schizonts in the absence (control) or in the presence of 10 − 8 M analogs. Dark grey shading denotes that the result is statistically significant compared to the control. Regarding the contractile response, [Ala 5 ]-Ang II and [Ala 6 ]-Ang II did not promote contractile activity compared to Ang II, and carbachol [Ala 5 ]-Ang II and [Ala 6 ]-Ang II reduced parasite invasion in red blood cells in the erythrocyte hemolysis assays, and presented no effect on cells integrity at 10 −8 M concentration. This model directs new possibilities for peptide design that can act more effectively in preventing the erythrocytic cycle of the parasite and other phases of the human–malaria cycle. demonstrates that after erythrocyte cultures were infected for 24 h with 3% parasitemia, the percentage of total parasites was determined in the absence (control) or presence of 10 − 8 M analogs. Dark grey shading denotes that the result is not statistically significant compared with control . Other synthesized angiotensin analogs presented high antiplasmodial activity in the P. gallinaceum test . They have similar hydrophobic interactions between the isolated guanidyl (Arg 2 ) and hydroxyl (Tyr 4 ) groups and the side chains of alkyl and aromatic amino acid. The guanidyl (Arg 2 ) and carboxyl (Asp 1 ) groups seem to be less significant in this case. Lactam bridges have an important role in constricting the conformation of bioactive peptides. When evaluated against P. falciparum , all lactam bridge analogs lacked significant activity against P. gallinaceum and perform better against P. falciparum because of the rigidity of the peptide bridge and possibe increased resistance to degradation under this kind of restriction. Analog 416 and 418 presented significant antiplasmodial activity, however analog 425 , a disulfide bridge analog, was the most active peptide against P. gallinaceum . It was also active against P. falciparum because it has the same restriction size and implications when considering hydrophobic and hydrophilic interactions between side chains. All analogs frequently adopt a β-turn conformation, which is consistent with that of the analog that is believed to be most active against P. gallinaceum . Angiotensin II hormones were also synthesized to investigate erythrocytic cycle invasion . Peptides 426 and 427 exhibited >80% activity on P. gallinaceum and >40% activity on P. falciparum . The hemolytic effect, contractile response, and stability in human serum were determined , demonstrating that the peptides did not present hemolytic effects as they were inactive when compared to positive controls and were resistant to degradation in human serum after a prolonged exposure (6 h), especially peptides 426 and 427 , which displayed excellent stability. The study of synthesized Dec-NH 2 and eight analog peptides reported that certain amino acid substitutions significantly improved the peptide’s antiplasmodial activity, providing new sequence principles for creating potent anti-malaria drugs, such as replacing the original sequence by Arg, Phe, and Trp . The most active peptides examined were [Arg] 1 -Dec-NH 2 ( 438 ), [Pro] 4 -Dec-NH 2 ( 437 ), and [Phe] 2 -Dec-NH 2 ( 439 ); all compounds substitutions improved the effectiveness, with the gihest antiplasmodial activities achieved with mutations to the N-terminus of Dec-NH 2 . A higher antiplasmodial activity was achieved by the introduction of a positive charge, increased hydrophobicity, and introduction of a restrictor residue . This review on antimalarial peptides discussed the natural sources, chemical access and antimalarial properties of the peptides, showing that nature (higher plant, microbes, venom) is a source of these antimalarial peptides. Alternatively, they can be chemically synthesized using a variety of techniques, such as solution phase, solid phase and a combination of solid and solution phase synthesis. The selection of coupling reagent, cyclization site, and protection groups are important factors to consider for the successful synthesis. Most isolated and synthesized peptides are active against P. falciparum with some exhibiting inhibition of P. gallinaceum and P. berghei . Moreover, several peptides were also cytotoxic against MRC5-cells, HEK2931, and HepG2. The information presented in this review may be useful in the design and synthesis of new potent antimalarial agents. |
The impact of major congenital anomalies on obstetric outcomes in the United Arab Emirates: the Mutaba’ah Study | bbd4420b-b538-4381-b676-f6f2e8ba608b | 11751165 | Surgical Procedures, Operative[mh] | Major congenital anomalies (MCAs) are severe structural abnormalities identified during pregnancy, at birth, or later in life , . These anomalies account for 2–4% of all live births , and may develop in isolation, or as part of complex patterns or syndromes . MCAs are significant contributors to perinatal mortality and long-term disability. Furthermore, pregnancies complicated by MCAs have an increased risk of severe maternal morbidity (SMM) . However, research on obstetric outcomes and labor patterns in pregnancies complicated by MCAs is sparse. While the diagnosis of MCAs is not typically associated with increased obstetric risk, there remains a significant gap in understanding obstetric outcomes and patterns. A study has shown that CAs significantly increase the risk of malpresentation, potentially prolonging labor and heighten the risk of cesarean delivery . Furthermore, Rossi et al. (2019) concluded that pregnancies affected by congenital heart diseases (CHDs) have a higher risk of cesarean delivery compared to those without . Nevertheless, evidence remains scarce and contradictory. Moreover, while advanced maternal age (AMA) has been linked to adverse obstetric and neonatal outcomes such as malpresentation, cesarean delivery, and congenital anomalies – , studies on the associations between MCAs and obstetric outcomes in the context of maternal age remain insufficient. There is also a considerable research gap regarding the impact of single and multiple MCAs on obstetric outcomes, alongside a dearth of evidence-based guidelines for labor management in affected pregnant women. In the United Arab Emirates (UAE), MCAs pose a significant public health concern , , influenced by the country’s strict law on termination of pregnancy for fetal anomalies (TOPFA) and the high consanguinity rate , . Abdulrahman et al. (2019) reported that cesarean section rate in the UAE (33%) exceeds the global average . Grasping the intricate relationship between MCAs and obstetric outcomes is pivotal for optimizing prenatal care and management strategies. However, no prospective study has comprehensively explored the obstetric outcomes of pregnancies complicated by MCAs in the UAE. This study aims to investigate the effect of MCAs on fetal presentation and delivery mode in a sample of pregnant women from the Emirati population.
Study design, setting and participants This is a prospective analysis of data from the Mutaba’ah Study, an ongoing mother and child cohort study . The Mutaba’ah study complies with the World Medical Association Declaration of Helsinki regarding the ethical conduct of research, and was approved by the Research Ethics Committees of the United Arab Emirates University (ERH-2017-5512) and the Abu Dhabi Health Research and Technology Ethics Committee (DOH/CVDC/2022/72). In this cohort study, pregnant women from the Emirati population, aged 18 years or older and who provided informed consent were consecutively recruited during their antenatal care visits in Al Ain city, regardless of the gestational age. Eligible participants completed self-administered questionnaires during their pregnancies and additional clinical data were extracted from their medical records. A total of 4,648 pregnant women gave birth to singleton live births between November 2017 and May 2023, and were eligible to be included in this analysis. Variables and measurement MCAs Perinatally diagnosed CAs are those detected during pregnancy or within the first week of life . In this study CAs were identified according to the International Classification of Diseases, 10th revision (ICD-10), chapter XVII , and all the ICDs were extracted from the medical records within one week after delivery. The investigated MCAs encompassed the presence of any MCAs, single MCAs, and multiple MCAs. Single system anomalies included: nervous (Q00-Q07) , eye , ear , face and neck (Q10-Q18) , circulatory (Q20-Q28) , respiratory (Q30-Q34) , cleft lip and cleft palate (Q35-Q37) , digestive (Q38-Q45) , genital (Q50-Q56) , urinary (Q60-Q64) , musculoskeletal (Q65-Q79) , other (Q80-Q89) , and chromosomal abnormalities (Q90-Q99) . Minor anomalies were excluded based on the EUROCAT and the CDC guidelines . Outcomes The main outcome variables included fetal presentation (cephalic vs. breech) and delivery mode (vaginal vs. cesarean). Covariables Maternal characteristics included maternal age (years), body mass index (BMI) at delivery (kg/m 2 ), parity (nulliparous “never delivered a baby of more than 20 weeks” vs. multiparous), and previous cesarean delivery (No vs. Yes). Maternal illnesses and pregnancy complications were classified based on the ICD-10 system (No vs. Yes). These encompassed pregestational diabetes mellitus, preeclampsia, chorioamnionitis, polyhydramnios, abruptio placenta and premature rupture of membranes. Neonatal characteristics included newborn sex (female vs. male), birth weight (grams), and gestational age at birth (weeks). Statistical analysis Descriptive statistics were performed to compare the study population characteristics and obstetric outcomes by MCAs status. Continuous variables were summarized by means and standard deviations, while categorical variables were presented as counts and percentages. Mean differences for continuous variables were assessed using the independent t-tests, whereas categorical variables were tested using the Pearson’s Chi-squared or the Fisher’s exact test. Univariable and multivariable regression models were employed to assess the associations between any MCAs, single MCAs or multiple MCAs, and the obstetric outcomes. Significant baseline covariates in the univariable analysis (p < 0.1) were included in the full model and sequentially removed in a backward stepwise fashion until a parsimonious model of statistically influential and biologically plausible covariates remained. The final confounders for breech presentation were maternal age, parity, pregestational diabetes mellitus, previous cesarean delivery, gestational age, and birth weight. The relevant covariates for cesarean delivery comprised maternal age, parity, BMI, previous cesarean delivery, pregestational diabetes mellitus, preeclampsia, chorioamnionitis, polyhydramnios, abruptio placenta, premature rupture of membranes, newborn sex, gestational age, and birth weight. In each model, interaction terms were generated and tested. Due to the significant interaction between maternal age and MCAs, stratification by maternal age (< 35 vs. ≥ 35 years) was employed in both obstetric outcome models. Additionally, the models were stratified by parity (nulliparous vs. multiparous), as maternal age persisted as a significant effect modifier within the parity groups. Furthermore, given the inverse relationship between breech presentation and gestational age , the fetal presentation model was stratified by gestational age (term “ ≥ 37 and < 42” vs. preterm “ < 37” gestational weeks). This model was further stratified by parity but not maternal age, as the significance of the interaction between maternal age and MCAs vanished across the gestational age groups. Models were tested for collinearity, crude and adjusted odds ratios (COR, AOR) with 95% confidence intervals (CI) were reported, and statistical significance was set at p < 0.05. Stata 16.1 (Stata Corp, College Station, TX, USA) was used for data entry and analyses.
This is a prospective analysis of data from the Mutaba’ah Study, an ongoing mother and child cohort study . The Mutaba’ah study complies with the World Medical Association Declaration of Helsinki regarding the ethical conduct of research, and was approved by the Research Ethics Committees of the United Arab Emirates University (ERH-2017-5512) and the Abu Dhabi Health Research and Technology Ethics Committee (DOH/CVDC/2022/72). In this cohort study, pregnant women from the Emirati population, aged 18 years or older and who provided informed consent were consecutively recruited during their antenatal care visits in Al Ain city, regardless of the gestational age. Eligible participants completed self-administered questionnaires during their pregnancies and additional clinical data were extracted from their medical records. A total of 4,648 pregnant women gave birth to singleton live births between November 2017 and May 2023, and were eligible to be included in this analysis.
MCAs Perinatally diagnosed CAs are those detected during pregnancy or within the first week of life . In this study CAs were identified according to the International Classification of Diseases, 10th revision (ICD-10), chapter XVII , and all the ICDs were extracted from the medical records within one week after delivery. The investigated MCAs encompassed the presence of any MCAs, single MCAs, and multiple MCAs. Single system anomalies included: nervous (Q00-Q07) , eye , ear , face and neck (Q10-Q18) , circulatory (Q20-Q28) , respiratory (Q30-Q34) , cleft lip and cleft palate (Q35-Q37) , digestive (Q38-Q45) , genital (Q50-Q56) , urinary (Q60-Q64) , musculoskeletal (Q65-Q79) , other (Q80-Q89) , and chromosomal abnormalities (Q90-Q99) . Minor anomalies were excluded based on the EUROCAT and the CDC guidelines . Outcomes The main outcome variables included fetal presentation (cephalic vs. breech) and delivery mode (vaginal vs. cesarean). Covariables Maternal characteristics included maternal age (years), body mass index (BMI) at delivery (kg/m 2 ), parity (nulliparous “never delivered a baby of more than 20 weeks” vs. multiparous), and previous cesarean delivery (No vs. Yes). Maternal illnesses and pregnancy complications were classified based on the ICD-10 system (No vs. Yes). These encompassed pregestational diabetes mellitus, preeclampsia, chorioamnionitis, polyhydramnios, abruptio placenta and premature rupture of membranes. Neonatal characteristics included newborn sex (female vs. male), birth weight (grams), and gestational age at birth (weeks).
Perinatally diagnosed CAs are those detected during pregnancy or within the first week of life . In this study CAs were identified according to the International Classification of Diseases, 10th revision (ICD-10), chapter XVII , and all the ICDs were extracted from the medical records within one week after delivery. The investigated MCAs encompassed the presence of any MCAs, single MCAs, and multiple MCAs. Single system anomalies included: nervous (Q00-Q07) , eye , ear , face and neck (Q10-Q18) , circulatory (Q20-Q28) , respiratory (Q30-Q34) , cleft lip and cleft palate (Q35-Q37) , digestive (Q38-Q45) , genital (Q50-Q56) , urinary (Q60-Q64) , musculoskeletal (Q65-Q79) , other (Q80-Q89) , and chromosomal abnormalities (Q90-Q99) . Minor anomalies were excluded based on the EUROCAT and the CDC guidelines .
The main outcome variables included fetal presentation (cephalic vs. breech) and delivery mode (vaginal vs. cesarean).
Maternal characteristics included maternal age (years), body mass index (BMI) at delivery (kg/m 2 ), parity (nulliparous “never delivered a baby of more than 20 weeks” vs. multiparous), and previous cesarean delivery (No vs. Yes). Maternal illnesses and pregnancy complications were classified based on the ICD-10 system (No vs. Yes). These encompassed pregestational diabetes mellitus, preeclampsia, chorioamnionitis, polyhydramnios, abruptio placenta and premature rupture of membranes. Neonatal characteristics included newborn sex (female vs. male), birth weight (grams), and gestational age at birth (weeks).
Descriptive statistics were performed to compare the study population characteristics and obstetric outcomes by MCAs status. Continuous variables were summarized by means and standard deviations, while categorical variables were presented as counts and percentages. Mean differences for continuous variables were assessed using the independent t-tests, whereas categorical variables were tested using the Pearson’s Chi-squared or the Fisher’s exact test. Univariable and multivariable regression models were employed to assess the associations between any MCAs, single MCAs or multiple MCAs, and the obstetric outcomes. Significant baseline covariates in the univariable analysis (p < 0.1) were included in the full model and sequentially removed in a backward stepwise fashion until a parsimonious model of statistically influential and biologically plausible covariates remained. The final confounders for breech presentation were maternal age, parity, pregestational diabetes mellitus, previous cesarean delivery, gestational age, and birth weight. The relevant covariates for cesarean delivery comprised maternal age, parity, BMI, previous cesarean delivery, pregestational diabetes mellitus, preeclampsia, chorioamnionitis, polyhydramnios, abruptio placenta, premature rupture of membranes, newborn sex, gestational age, and birth weight. In each model, interaction terms were generated and tested. Due to the significant interaction between maternal age and MCAs, stratification by maternal age (< 35 vs. ≥ 35 years) was employed in both obstetric outcome models. Additionally, the models were stratified by parity (nulliparous vs. multiparous), as maternal age persisted as a significant effect modifier within the parity groups. Furthermore, given the inverse relationship between breech presentation and gestational age , the fetal presentation model was stratified by gestational age (term “ ≥ 37 and < 42” vs. preterm “ < 37” gestational weeks). This model was further stratified by parity but not maternal age, as the significance of the interaction between maternal age and MCAs vanished across the gestational age groups. Models were tested for collinearity, crude and adjusted odds ratios (COR, AOR) with 95% confidence intervals (CI) were reported, and statistical significance was set at p < 0.05. Stata 16.1 (Stata Corp, College Station, TX, USA) was used for data entry and analyses.
Out of 4,648 pregnant women with singleton live births, complete neonatal diagnostic data were available for 4,252 newborns. Among them, 303 neonates were diagnosed with MCAs, of which 255 (84.2%) were single system, while 48 (15.8%) were multiple-system anomalies. Table presents the population characteristics by MCAs status. Maternal age and parity did not vary significantly between mothers of neonates with and without MCAs, however those with MCAs had higher BMI. Of all pregnancies, 3,969 (93.3%) had cephalic and 172 (4.1%) breech presentation, while 2,908 (68.4%) resulted in vaginal deliveries, and 1,332 (31.3%) ended in cesarean deliveries. Among pregnancies with MCAs, 9.3% ended with breech presentation compared to only 3.8% in those without MCAs. Cesarean deliveries were more frequent (43.1%) among those with MCAs compared to those without MCAs (30.5%). The distribution of outcomes by various MCAs is presented in Figs. and . Table presents the association between MCAs and fetal presentation. Overall, neonates of pregnant women under 35 years had higher odds of breech presentation, with those having any MCAs exhibiting 2.7 times higher odds of breech presentation compared to those without MCAs (95% CI: 1.5-5.0). For both any and single MCAs, the significant associations persisted mostly in pregnant women between 18 and 29 years (Supplementary Table ). Further stratification by parity revealed significant associations for both any and single MCAs with breech presentation among multiparous women younger than 35 years (AOR = 3.0, 95% CI: 1.4–6.3) and (AOR = 2.5, 95% CI: 1.1-6.0), as well as nulliparous aged 35 years and older (AOR = 7.4, 95% CI: 1.3–43.0) and (AOR = 8.0, 95% CI: 1.6–46.8), respectively (Fig. and Supplementary Table S2). Table illustrates the association between MCAs and fetal presentation stratified by gestational age. Among term deliveries, the presence of any (AOR = 2.5, 95% CI: 1.4–4.3) and single MCAs (AOR = 2.4, 95% CI: 1.3–4.3) was significantly associated with breech presentation compared to those without MCAs. Following stratification by parity, the associations with breech remained significant for both MCAs among term multiparous and preterm nulliparous women. (Fig. and Supplementary Table S3). For cesarean delivery, the univariate analysis revealed robust associations with all MCAs. Following adjustment and age stratification, this association persisted among pregnant women under 35 years, with those with any and single MCAs showed 80% and 60% higher odds of cesarean delivery, respectively, compared to those without MCAs (Table ). The significant associations persisted mostly in pregnant women between 18 and 29 years for both MCAs (Supplementary Table S4). Additional stratification by parity showed significant associations between all MCAs and cesarean delivery in young (< 35 years) multiparous women (Fig. and Supplementary Table S5). Conducting analysis among young multiparous women, excluding those with previous cesarean delivery, revealed a robust association between any MCAs and cesarean delivery (AOR = 2.1, 95% CI: 1.2–3.6). Further analysis limited to those with cephalic presentation showed increased, though not significant, odds of cesarean delivery associated with MCAs. However, similar analysis limited to those with breech presentation was not applicable due to small sample size. For multiple MCAs, the significant association with breech presentation persisted among pregnant women between 30 and 34 years (AOR = 9.0, 95% CI: 1.6–52.4), whereas for cesarean delivery the association was significant between 18 and 29 years (AOR = 3.4, 95% CI: 1.1–10.9). Further stratification by parity revealed significant associations with both breech (AOR = 5.7, 95% CI: 1.3–24.9) and cesarean delivery (AOR = 4.5, 95% CI: 1.6–12.7) among multiparous women younger than 35 years. Regarding gestational age, multiple MCAs was significantly associated with breech among term multiparous women (AOR = 4.1, 95% CI: 1.1–16.2) (Supplementary Table S6 and Figure ). For single system anomalies, the association with obstetric outcomes varied. For the nervous system anomalies, a significant association with breech presentation was detected among pregnant women aged 35 years and older. In contrast, the respiratory system and other anomalies were associated with breech among younger age group (< 35 years), while for the musculoskeletal system anomalies, the association persisted in both age groups. When stratified by term status, the musculoskeletal and other anomalies were significantly associated with breech presentation in term deliveries; however, the respiratory system anomalies were associated with breech presentation in preterm deliveries. For cesarean delivery, significant associations were identified with the circulatory , the musculoskeletal and chromosomal abnormalities among pregnant women younger than 35 years (Supplementary Tables S7, S9, S11). A sub-analysis by parity revealed additional insights. The circulatory system anomalies were associated with breech presentation among nulliparous aged 35 years or older and preterm deliveries, while the musculoskeletal anomalies were associated with breech among younger women (< 35 years) at term in both parity groups. Both the circulatory and the musculoskeletal system anomalies showed significant associations with cesarean delivery among multiparous women under 35 years (Supplementary Tables S8, S10, S12).
This study highlights the significant impact of MCAs on two obstetric outcomes: breech presentation and cesarean delivery. The association with cesarean delivery remained significant among multiparous younger than 35 years, whereas the association with breech presentation varied by maternal age and parity. Neonates with any or single MCAs were significantly associated with breech presentation among multiparous women below 35 years and nulliparous aged 35 years or older. Breech presentation can signal fundamental issues in fetal morphogenesis or function. Macharey et al. (2021) found that the prevalence of CA in breech was two-fold compared to cephalic presentation (6.5% vs. 3.7%) . Similarly, Mostello et al. (2014) reported that breech infants were more likely to have at least one CA (11.7%) compared to those with cephalic presentation (5.1%) . Potential causal pathways include fetal neuromuscular dysfunction, neurological deficits, uterine abnormalities, or spatial constraints, all of which may hinder cephalic presentation. Our findings showed that the presence of any or single MCAs was significantly associated with breech presentation among term multiparous and preterm nulliparous women. Breech presentation is inversely proportional to gestational age, with a decline from 23.5% in extremely preterm to 2.5% in term pregnancies . Studies have shown that 15–33% of breech presentations spontaneously correct to cephalic position at term . Although CAs present in 17% of preterm and 9% of term breeches , Mostello et al. (2014) identified a robust association between CAs and breech among full-term births (OR = 2.09, 95% CI 1.96–2.23) . According to Toijonen et al. (2020), nulliparity is a risk factor for breech presentation not only at term, but also in moderate to late preterm pregnancies. This is attributed to the abdominal and uterine wall tightness among nulliparous women, potentially hindering fetal rotation to the cephalic presentation . Parity-related relaxation of the uterine wall has also been proposed to modify the odds of breech, potentially leading to a higher rate of external cephalic version among multiparous women . Cammu et al. (2014) reported a decline in breech presentation frequency with increasing parity, with neonates of older mothers showing higher odds of breech presentation . The absence of associations in older and preterm multiparous women in our study could be attributed to more potent confounding variables within these subgroups, which obscure the impact of MCAs. In this study, the musculoskeletal system anomalies were associated with breech presentation among young term nulliparous and multiparous pregnancies. Musculoskeletal anomalies can lead to breech presentation due to restricted movement or abnormal positioning of the fetus in the uterus. For instance, conditions like arthrogryposis, characterized by joint contractures associated with fetal akinesia, can limit the fetus’s ability to kick strongly enough to turn vertex . The present study revealed a robust association between all MCAs and cesarean delivery among multiparous women under 35 years, with neonates exhibiting any MCAs having twice the odds of cesarean delivery compared to those without MCAs. Walsh et al. (2014) identified a higher risk of emergency cesarean delivery in multiparous pregnancies with CHDs due to fetal distress (OR = 2.4, 95% CI: 1.8–4.6) . However, data on non-cardiac anomalies remain scarce. Our findings indicate a significantly increased risk of cesarean delivery among younger (< 35 years) multiparous women without previous cesarean section. This is a particularly interesting finding for this low-risk group. Previous studies indicate that mothers of neonates with MCAs were more susceptible to cesarean delivery due to factors such as slower labor progress, fetal distress, fetal malpresentation, placental abnormalities, soft tissue trauma and cephalopelvic disproportion , . Furthermore cesarean delivery in such cases can avert birth trauma and facilitate surgical interventions such as ex-utero intrapartum treatment (EXIT) . However, the World Health Organization (WHO) warns that cesarean delivery rates exceeding 10% are detrimental as they heighten maternal risk, contribute to long-term childhood complications and overload the healthcare system . Cesarean delivery rates have tripled globally, with one in ten young women opting for cesarean Sect . In the context of MCAs, cesarean sections are highly debated. Many fetuses with MCAs experience poor neonatal adaptation despite the higher cesarean rate. Conversely, some may significantly benefit from cesarean delivery. For instance, in specific cases of complex CHDs, vaginal delivery might be favoured, reserving elective cesarean sections for obstetric indications or fetal arrhythmias. While, in pregnancies of short-statured women with musculoskeletal anomalies, cesarean delivery might be warranted, regardless of the risk of fetal skeletal dysplasia . Tailoring delivery route to the specific type of anomaly is essential to optimize maternal and neonatal outcomes, while carefully weighing associated risks and benefits. Study strengths and limitations . This large population-based study is the first to explore the obstetric outcomes of MCAs in the UAE. To our knowledge, this is the most comprehensive study that investigated the association among different anomaly subtypes, while adjusting for multiple covariates and stratifying by maternal age, gestational age, and parity. However, in this study MCAs were diagnosed perinatally, which potentially could underestimate the actual burden of MCAs diagnosed later. Furthermore, while we studied the overall association between MCAs and delivery mode, data on specific indications for delivery mode and the subtypes of cesarean or vaginal delivery were not assessed. Additionally, statistical power limitations in multiple and single system anomalies have rendered some results inconclusive, emphasising the need for further studies. Given the high cesarean rate in the UAE and the significant impact of MCAs, future research should prioritise MCAs as risk factors for breech presentation and cesarean delivery. Moreover, it’s prudent to investigate the association between MCAs and prolonged or obstructed labor for a more comprehensive understanding of the impact of MCAs. Furthermore, the study findings necessitate the activation of preventive strategies and screening programs to identify at-risk mothers, investigate risk factors for MCAs, and control the modifiable ones. Advances in fetal ultrasound technology and the widespread adoption of first-trimester screening have significantly improved the timely detection of CAs. This information is pivotal for personalized delivery planning and determining the optimum time and modality of treatment. Addressing discrepancies between antenatal and postnatal diagnoses of MCAs is also crucial for assessing the accuracy of antenatal diagnostic services and enhance research validity. The centralization of multidisciplinary perinatology centres has proven effective in this domain. Moreover, mothers should receive education and counselling, with adherence to autonomy and beneficence. According to Stoll et al. (2017), fear of childbirth declined as women’s knowledge of pregnancy and birth grew . Finally, large-scale studies are essential to evaluate maternal and neonatal morbidity and mortality in these pregnancies, elucidate nuanced associations i.e. for multiple and single system anomalies, and incorporate various pregnancy complications and obstetric outcomes.
This study underscores the significant impact of MCAs on fetal presentation and delivery mode. For both breech presentation and cesarean delivery, the associations with MCAs remained significant in multiparous under 35 years. However, a significant association with breech presentation was also detected in nulliparous aged 35 years and older, as well as in term multiparous and preterm nulliparous women. These findings highlight the need for early detection of MCAs, as well as investigating their risk factors, implementing prevention strategies, and adopting a tailored multidisciplinary approach to ensure optimal outcomes for both women and newborns. Furthermore, large-scale studies are imperative to investigate nuanced associations, inform policy and incorporate additional pregnancy complications and obstetric outcomes.
Below is the link to the electronic supplementary material. Supplementary Material 1
|
A novel iridoplasty suture technique to repair iris defects and traumatic mydriasis | 187a2689-f678-4b89-9c17-b6cb9b9be093 | 10391397 | Suturing[mh] | The U-suture iridoplasty technique has been described for the reconstruction of relatively large iris defects and atonic pupils, with fewer suture–knot complexes and minimal iris gap formation between nodes. The procedure was performed under peribulbar anesthesia, with additional anesthesia when necessary. According to the location of the iris defect and the size of the needle used, two self-closing, opposed, 0.9 mm corneal incisions were made in the direction of the planned suture route. An ocular viscoelastic device (Healon®, Pharmacia, Sweden) was applied to the anterior chamber to protect the corneal endothelium. 10–0 looped polypropylene suture (PC-9, Alcon Surgical, Fort Worth, TX, USA) was inserted into the anterior chamber through one of the two corneal incisions. This suture was passed through the margins of both the anterior and posterior iris leaflets and removed from the anterior chamber through the second corneal incision. The needle was reintroduced into the anterior chamber through the second corneal incision. To create a U-suture, the needle was re-entered into the iris 1–2 mm away from the 10–0 polypropylene suture on the posterior leaflet and passed through the opposite leaflet in the same direction. Then, the needle was passed through the first corneal incision and removed from the anterior chamber again [ - and ]. Finally, the suture was securely tied with a modified Siepser sliding knot technique. The viscoelastic material in the anterior chamber was removed using a manual irrigation/aspiration tip and the pressure was controlled by hydrating the incisions. This U-suture pupilloplasty technique was used in 11 consecutive patients. In the slit-lamp examination findings in all cases, the pupil size and shape were relatively normal, maximum pupillary centralization was present, and no gaps between leaflets had been formed in the iris. Thus, an effective improvement in the iris appearance was achieved for all patients. A decrease in visual symptoms, such as light sensitivity and glare, was reported. Intraoperative complications, postoperative complications, or additional surgical need for iris repair were not observed in any case.
This article describes U-suture iridoplasty, which allows gapless iris repair with fewer suture knot complexes for the repair of large iris defects and atonic pupils. In our practice, we employ the U-suture iridoplasty method, which usually involves two close stitches with a knot, in cases where at least two or more sutures will be required during iris repairs, such as relatively large iris defects and traumatic mydriasis. Our clinical experience using this technique has shown that performing the technique in the nasal and lower nasal quadrants, as it requires the use of the left hand, may be relatively difficult and requires surgical experience. However, this technical difficulty can be overcome by over-deviating the sphere downward. Nevertheless, this technique resulted in satisfactory postoperative visual, functional, and cosmetic outcomes. The publication of McCannel’s technique in 1976, in which the iris was directly sutured, revolutionized techniques for successful iris reconstruction. Briefly, McCannel’s technique consisted of inserting a long needle through the proximal and distal ends of the iris defect and creating an extraocular knot by pulling the suture ends from the paracentesis created between the two ends of the defect. Subsequently, many modifications of the McCannel suture have been published.[ - ] In this technique, the risk of iatrogenic iris damage is high due to stretching of the iris tissue while forming the extraocular knot. However, causing iatrogenic iris damage while knotting should be avoided. The targeted iris node should be able to slide up to the pupillary border without stretching the iris tissue in the closed anterior chamber. In 1994, Siepser developed McCannel’s suture method and published an innovative technique that required minimal intracameral manipulation to create an intraocular knot. Siepser described how the knot formed outside the eye can be tied in the closed anterior chamber by sliding it over the iris. Subsequently, modified techniques have been described that simplified the slipknot tying method due to the technical difficulty of the knot in the Siepser slipknot. Recently, Narang et al . described a self-locking knot technique that prevents the ends of the stitches from slipping through the loops. The cerclage iridoplasty technique is used to create an appropriately sized and round pupil opening, especially in the treatment of traumatic mydriasis, and it is a difficult technique to manipulate. Three separate paracenteses are performed on the limbus or transparent cornea. Three to four bites are made from the periphery of the pupillary edge with the needle inserted into the anterior chamber through these paracenteses, and this process is repeated several times. As a result, continuous suturing occurs surrounding the entire pupillary margin. Currently, when repairing large iris defects, more than one node is usually formed and iris leaflets are brought as close to each other as possible. However, this method has some disadvantages, such as prolonging the surgical time and the formation of spaces between the nodes and atrophy in the iris. Although methods that minimize iris damage, such as sliding knot techniques have been developed, it is possible that each knot created will increase the risk of iris atrophy. Thus, we believe that new strategies should be developed by adapting previously described suture techniques for the repair of large iris defects and atonic pupils. The most important advantage of our technique is the formation of two concurrent suture effects with a single suture–knot complex. The U-suture method brings the iris leaflets closer together by slightly constricting them like a pack without creating excessive traction. Fewer knots are likely to result in fewer tractions in the iris during knot formation. In addition, the duration of the surgery is also reduced. As a result, the reduction in surgical time and manipulation reduces the risk of surgical complications.
In conclusion, multiple sutures may be needed to repair iris defects or atonic pupils but in some cases, some gaps still remain, especially in cases with mydriasis. Cerclage iridoplasty is a difficult technique; we believe U-shaped sutures are an easy and useful alternative. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.
Nil.
There are no conflicts of interest.
https://journals.lww.com/ijo
|
CovCopCan: An efficient tool to detect Copy Number Variation from amplicon sequencing data in inherited diseases and cancer | d5b5a2ae-2c15-4a42-b988-a1cc601fd35d | 7041855 | Pathology[mh] | Identifying mutations responsible for inherited or somatic diseases can be essential to define the appropriate therapy for the efficient treatment of patients. For example, this is true for patients presenting an amyloid neuropathy due to Transthyretin ( TTR) point mutations, who can benefit from new treatments, such as Tafamidis . This is also true for cancer, for which it is important to rapidly detect certain Copy Number Variations (CNVs), such as the 17p deletion, a recurrent abnormality in Chronic Lymphocytic Leukemia (CLL), with major therapeutic implications. Because this acquired chromosomal abnormality directly impairs the TP53 gene , it is now recommended to test this CNV before each treatment for CLL . Indeed, TP53 alterations in CLL are responsible for primary resistance to fludarabine and survival of such patients is clearly improved by new-targeted therapies, such as ibrutinib . High-throughput sequencing techniques allow partial or total sequencing of a patient’s genome. Amplicon sequencing is one of the techniques that enables the sequencing of several thousand exons at a very low cost. Although this method is robust for the discovery of small genetic mutations, such as single-nucleotide polymorphisms or short indels, only a few tools are available for the detection of larger variations, such as deletions or duplications in amplicon sequencing data. Some of these tools require control samples to establish a reference set of data (ONCOCNV ). For others (ExomeDepth , IonCopy , DeviCNV , Cov’Cop ), control samples are not necessary. Indeed, if the CNV is rare, the other patient samples tested in the same run can serve as controls. In this strategy, multiple patients are tested at the same time, potentially shortening the time to diagnosis. Most available tools based on the read depth method to detect CNVs include robust statistical methods. ExomeCopy proposes a hidden Markov model to detect CNVs from raw read count data. CONVector was built on a machine-learning algorithm to associate PCR-efficiency correlations for subsets of amplicons. Here, we propose a new tool, CovCopCan, based on the initial read-depth method developed in Cov’Cop, with additional statistical methods and features that allow the rapid and easily detection of CNVs in inherited diseases, as well as somatic data of patients with cancer, even with a low ratio of cancer cells to healthy cells (data sets described in ). CovCopCan includes heuristic methods to compare the value of each amplicon of a patient to those of other patients sequenced in the same run. CovCopCan focuses on data manipulation and results exploration for the interpretation of CNVs. Users have access to an overview of the results for each patient through an interactive visualization, allowing, for example, the exclusion of low-quality amplification from the analysis and quickly restarting CNV detection. In addition, several statistics methods (Loess regression, Cumulative summary) can help in the interpretation of the results.
CNV-detection algorithm Z-score-based CNV detection: “Z-detection” From the raw read count of each amplicon, CovCopCan applies the same corrections and normalization as the Cov’Cop tool , resulting in a normalized read count value (NRC) for each amplicon (see ). Starting from this point, we developed a new CNV-detection algorithm, based on the z-score. The z-score is calculated for each amplicon in each patient, according to the following formula: z − s c o r e p _ i = N R C p _ i − μ p σ p NRC p _ i is the normalized read count of the amplicon i in the patient p , μ p the NRC average of the patient p , and σ corresponds to the standard deviation of the patient p . The z-score follows a standard normal distribution N (0;1). We fixed a threshold corresponding to a significance level of 0.01 for both deletion and duplication events by a one-tailed test. Thus, a negative z-score with a p-value < 0.01 indicates a deleted amplicon, whereas a positive z-score with a p-value < 0.01 indicates a duplicated amplicon. This algorithm automatically determines the best deletion and duplication thresholds based on the variability of a patient's data. The users are free to determine the minimum number of concurrent amplicons required to call a CNV. No minimum distance between amplicons is required, but they have to be on the same chromosome. By default, a minimum of three successive amplicons on the same chromosome was used for all data in this paper. Two-stage ratio to optimize CNV detection The last normalization step of CovCopCan results in a ratio of standardized patient values that gives a theoretical value of 1 for a gene present in two copies, 0.5 for a deletion event, and 1.5 for a duplication. In this last step, each amplicon value is divided by the median of the same amplicon from the other samples. Once this first ratio is calculated and the first round of CNV detection is performed, a second ratio is calculated excluding all amplicons located inside the initially detected CNVs from each sample, and final CNV detection is achieved. This approach is used to improve standardization in regions in which the same CNV event is present in many patients. Merging CNVs We provide a “merge” option to reduce the impact of false-negative amplicons on CNV detection. If two CNV areas located on the same chromosome are disjointed by only one amplicon with a z-score duplicated or deleted at a significance level of 0.05, CovCopCan will then merge the two CNV areas to easily highlight this global CNV. In addition, the user can also define the maximum distance value between two CNVs to be merged. Reference amplicon selection or exclusion For the normalization step, CovCopCan selects a set of amplicons, consisting of those that are the most stable among the patients of a run. These amplicons are then used to normalize the values of the other amplicons. The user can indicate specific amplicons to use for this normalization step (see ). Inversely, our tool also provides the possibility to manually exclude some amplicon data for the last ratio step of normalizations (see ). Control samples Although CovCopCan works without control samples, it is possible to exploit the presence of controls if they are available. In such a case, the median of the last standardization step is no longer calculated using all the samples but only the controls. Then for each patient, the amplicon values are divided by the median calculated for the controls, according to the following formula: R a t i o i _ p a t j = N R C i p a t j M d ( N R C i c o n t r o l s ) N R C i p a t j is the normalized read count of the amplicon i in the patient j . M d ( N R C i c o n t r o l s ) is the median of the normalized read count of the control samples. CovCopCan can be run with only one control sample but more control samples will improve the result. 2D interactive visualization An interactive 2D visualization is available for each patient . The amplicons are represented by dots over their chromosomal positions on the x-axis and their normalized values on the y-axis. Users can interactively zoom in on specific regions and navigate between data in an intuitive and interactive way, allowing simple navigation. Several types of information described below have also been added to the graphical representation. Local regression curve We introduced the possibility to display regression curves on the presented chart to optimize visual CNV detection. We chose to implement the Loess local regression algorithm to easily visualize a sudden change. The Loess regression is calculated for each chromosome. By default, the bandwidth parameter is fixed to 0.25, but it is possible to interactively fine tune it to more or less smoothen the curve. The Loess regression is represented by a green curve on the chart (see ). CUSUM charts For data generated from cancer or mosaic samples, a sample may simultaneously contain “normal” and deleted/duplicated cells. The deletion/duplication detection accuracy depends on the proportion of deleted/duplicated cells relative to that of the normal cells and the normalized values can be close to 1. CNVs will then be very difficult to detect. Consequently, we added a visual method called CUmulative SUMmary control chart (CUSUM; ) to be able to observe a slight increase or decrease in values. For each chromosome, this algorithm calculates the cumulative sum of the positive deviations (values > patient’s average) for deletions and negative deviations (values < patient’s average) for duplications. It can be useful for detecting a slight deviation of the values due to cancer data or mosaicism, as well as small CNVs in inherited diseases. S n + = max ( 0 , S n − 1 + + x n − ( x ¯ + σ ) ) S n − = m i n ( 0 , S n − 1 − + x n − ( x ¯ − σ ) ) Here, x n corresponds to the value of one amplicon, x ¯ is the mean value of all the patient’s amplicons, and σ is the standard deviation. In the visualization of CovCopCan, a blue shape indicates a possible deletion, whereas a pink shape indicates a potential duplication. Although this method makes it possible to highlight potential CNVs, it does not allow precise definition of their breakpoints (see ).
Z-score-based CNV detection: “Z-detection” From the raw read count of each amplicon, CovCopCan applies the same corrections and normalization as the Cov’Cop tool , resulting in a normalized read count value (NRC) for each amplicon (see ). Starting from this point, we developed a new CNV-detection algorithm, based on the z-score. The z-score is calculated for each amplicon in each patient, according to the following formula: z − s c o r e p _ i = N R C p _ i − μ p σ p NRC p _ i is the normalized read count of the amplicon i in the patient p , μ p the NRC average of the patient p , and σ corresponds to the standard deviation of the patient p . The z-score follows a standard normal distribution N (0;1). We fixed a threshold corresponding to a significance level of 0.01 for both deletion and duplication events by a one-tailed test. Thus, a negative z-score with a p-value < 0.01 indicates a deleted amplicon, whereas a positive z-score with a p-value < 0.01 indicates a duplicated amplicon. This algorithm automatically determines the best deletion and duplication thresholds based on the variability of a patient's data. The users are free to determine the minimum number of concurrent amplicons required to call a CNV. No minimum distance between amplicons is required, but they have to be on the same chromosome. By default, a minimum of three successive amplicons on the same chromosome was used for all data in this paper. Two-stage ratio to optimize CNV detection The last normalization step of CovCopCan results in a ratio of standardized patient values that gives a theoretical value of 1 for a gene present in two copies, 0.5 for a deletion event, and 1.5 for a duplication. In this last step, each amplicon value is divided by the median of the same amplicon from the other samples. Once this first ratio is calculated and the first round of CNV detection is performed, a second ratio is calculated excluding all amplicons located inside the initially detected CNVs from each sample, and final CNV detection is achieved. This approach is used to improve standardization in regions in which the same CNV event is present in many patients. Merging CNVs We provide a “merge” option to reduce the impact of false-negative amplicons on CNV detection. If two CNV areas located on the same chromosome are disjointed by only one amplicon with a z-score duplicated or deleted at a significance level of 0.05, CovCopCan will then merge the two CNV areas to easily highlight this global CNV. In addition, the user can also define the maximum distance value between two CNVs to be merged. Reference amplicon selection or exclusion For the normalization step, CovCopCan selects a set of amplicons, consisting of those that are the most stable among the patients of a run. These amplicons are then used to normalize the values of the other amplicons. The user can indicate specific amplicons to use for this normalization step (see ). Inversely, our tool also provides the possibility to manually exclude some amplicon data for the last ratio step of normalizations (see ). Control samples Although CovCopCan works without control samples, it is possible to exploit the presence of controls if they are available. In such a case, the median of the last standardization step is no longer calculated using all the samples but only the controls. Then for each patient, the amplicon values are divided by the median calculated for the controls, according to the following formula: R a t i o i _ p a t j = N R C i p a t j M d ( N R C i c o n t r o l s ) N R C i p a t j is the normalized read count of the amplicon i in the patient j . M d ( N R C i c o n t r o l s ) is the median of the normalized read count of the control samples. CovCopCan can be run with only one control sample but more control samples will improve the result.
From the raw read count of each amplicon, CovCopCan applies the same corrections and normalization as the Cov’Cop tool , resulting in a normalized read count value (NRC) for each amplicon (see ). Starting from this point, we developed a new CNV-detection algorithm, based on the z-score. The z-score is calculated for each amplicon in each patient, according to the following formula: z − s c o r e p _ i = N R C p _ i − μ p σ p NRC p _ i is the normalized read count of the amplicon i in the patient p , μ p the NRC average of the patient p , and σ corresponds to the standard deviation of the patient p . The z-score follows a standard normal distribution N (0;1). We fixed a threshold corresponding to a significance level of 0.01 for both deletion and duplication events by a one-tailed test. Thus, a negative z-score with a p-value < 0.01 indicates a deleted amplicon, whereas a positive z-score with a p-value < 0.01 indicates a duplicated amplicon. This algorithm automatically determines the best deletion and duplication thresholds based on the variability of a patient's data. The users are free to determine the minimum number of concurrent amplicons required to call a CNV. No minimum distance between amplicons is required, but they have to be on the same chromosome. By default, a minimum of three successive amplicons on the same chromosome was used for all data in this paper.
The last normalization step of CovCopCan results in a ratio of standardized patient values that gives a theoretical value of 1 for a gene present in two copies, 0.5 for a deletion event, and 1.5 for a duplication. In this last step, each amplicon value is divided by the median of the same amplicon from the other samples. Once this first ratio is calculated and the first round of CNV detection is performed, a second ratio is calculated excluding all amplicons located inside the initially detected CNVs from each sample, and final CNV detection is achieved. This approach is used to improve standardization in regions in which the same CNV event is present in many patients.
We provide a “merge” option to reduce the impact of false-negative amplicons on CNV detection. If two CNV areas located on the same chromosome are disjointed by only one amplicon with a z-score duplicated or deleted at a significance level of 0.05, CovCopCan will then merge the two CNV areas to easily highlight this global CNV. In addition, the user can also define the maximum distance value between two CNVs to be merged.
For the normalization step, CovCopCan selects a set of amplicons, consisting of those that are the most stable among the patients of a run. These amplicons are then used to normalize the values of the other amplicons. The user can indicate specific amplicons to use for this normalization step (see ). Inversely, our tool also provides the possibility to manually exclude some amplicon data for the last ratio step of normalizations (see ).
Although CovCopCan works without control samples, it is possible to exploit the presence of controls if they are available. In such a case, the median of the last standardization step is no longer calculated using all the samples but only the controls. Then for each patient, the amplicon values are divided by the median calculated for the controls, according to the following formula: R a t i o i _ p a t j = N R C i p a t j M d ( N R C i c o n t r o l s ) N R C i p a t j is the normalized read count of the amplicon i in the patient j . M d ( N R C i c o n t r o l s ) is the median of the normalized read count of the control samples. CovCopCan can be run with only one control sample but more control samples will improve the result.
An interactive 2D visualization is available for each patient . The amplicons are represented by dots over their chromosomal positions on the x-axis and their normalized values on the y-axis. Users can interactively zoom in on specific regions and navigate between data in an intuitive and interactive way, allowing simple navigation. Several types of information described below have also been added to the graphical representation. Local regression curve We introduced the possibility to display regression curves on the presented chart to optimize visual CNV detection. We chose to implement the Loess local regression algorithm to easily visualize a sudden change. The Loess regression is calculated for each chromosome. By default, the bandwidth parameter is fixed to 0.25, but it is possible to interactively fine tune it to more or less smoothen the curve. The Loess regression is represented by a green curve on the chart (see ). CUSUM charts For data generated from cancer or mosaic samples, a sample may simultaneously contain “normal” and deleted/duplicated cells. The deletion/duplication detection accuracy depends on the proportion of deleted/duplicated cells relative to that of the normal cells and the normalized values can be close to 1. CNVs will then be very difficult to detect. Consequently, we added a visual method called CUmulative SUMmary control chart (CUSUM; ) to be able to observe a slight increase or decrease in values. For each chromosome, this algorithm calculates the cumulative sum of the positive deviations (values > patient’s average) for deletions and negative deviations (values < patient’s average) for duplications. It can be useful for detecting a slight deviation of the values due to cancer data or mosaicism, as well as small CNVs in inherited diseases. S n + = max ( 0 , S n − 1 + + x n − ( x ¯ + σ ) ) S n − = m i n ( 0 , S n − 1 − + x n − ( x ¯ − σ ) ) Here, x n corresponds to the value of one amplicon, x ¯ is the mean value of all the patient’s amplicons, and σ is the standard deviation. In the visualization of CovCopCan, a blue shape indicates a possible deletion, whereas a pink shape indicates a potential duplication. Although this method makes it possible to highlight potential CNVs, it does not allow precise definition of their breakpoints (see ).
We introduced the possibility to display regression curves on the presented chart to optimize visual CNV detection. We chose to implement the Loess local regression algorithm to easily visualize a sudden change. The Loess regression is calculated for each chromosome. By default, the bandwidth parameter is fixed to 0.25, but it is possible to interactively fine tune it to more or less smoothen the curve. The Loess regression is represented by a green curve on the chart (see ).
For data generated from cancer or mosaic samples, a sample may simultaneously contain “normal” and deleted/duplicated cells. The deletion/duplication detection accuracy depends on the proportion of deleted/duplicated cells relative to that of the normal cells and the normalized values can be close to 1. CNVs will then be very difficult to detect. Consequently, we added a visual method called CUmulative SUMmary control chart (CUSUM; ) to be able to observe a slight increase or decrease in values. For each chromosome, this algorithm calculates the cumulative sum of the positive deviations (values > patient’s average) for deletions and negative deviations (values < patient’s average) for duplications. It can be useful for detecting a slight deviation of the values due to cancer data or mosaicism, as well as small CNVs in inherited diseases. S n + = max ( 0 , S n − 1 + + x n − ( x ¯ + σ ) ) S n − = m i n ( 0 , S n − 1 − + x n − ( x ¯ − σ ) ) Here, x n corresponds to the value of one amplicon, x ¯ is the mean value of all the patient’s amplicons, and σ is the standard deviation. In the visualization of CovCopCan, a blue shape indicates a possible deletion, whereas a pink shape indicates a potential duplication. Although this method makes it possible to highlight potential CNVs, it does not allow precise definition of their breakpoints (see ).
Two-stage ratio We visualized the result of the two-stage ratio using sequencing data from panel 2 (see for details). This gene panel, designed by Ion AmpliSeq designer software, includes 1,206 amplicons on 70 genes. The run presented here was performed on an Ion Proton device and included seven patients. A deletion on chromosome 13 was shared by three of the seven patients (verified by karyotyping). Examples of the visualization obtained for two of the patients (patient 1 normal and patient 2 “deleted”) are presented in . Without the two-stage ratio, the region in non-deleted patients was disturbed and a false positive duplication event was detected by CovCopCan in both (highlighted by a vertical red rectangle, as for patient 1, ). The two-stage ratio improved the stability of the values so that no false duplication event was detected by CovCopCan, thus increasing the specificity ( , compare A and B). This method also improved the detection of deletions (highlighted by a vertical orange rectangle) in the true deleted patients, decreasing the number of false-negative amplicons . Merging CNVs To reduce the effect of individual false negative amplicons, CovCopCan relaxes the significance threshold when a single non-significant amplicon is flanked on both sides by significant amplicons. For this specific amplicon, the threshold will be automatically switched to 0.05. If this amplicon becomes significantly duplicated, it will be merged with the initial duplicated detected areas. The grey dot in the graph will stay grey, indicating that it is a merged area. Deletions are treated the same way. Here, we show the results of this merging option on a complete chromosome X duplication. A single duplication covering the entire gene is detected by CovCopCan, whereas six successive duplications would have been found without this merging option . Control samples We tested this method with the Panel 2 data . Seven samples were simultaneously sequenced on an Ion Proton sequencer (three controls and four patients). The four patients share the same region q deletion on chromosome 13. Without defining controls, CovCopCan detected a correct deletion (highlighted by the vertical orange rectangle) for one of the four patients and only a partial deletion for another. In addition, two false-positive duplications (highlighted by the vertical red rectangle) were detected in two controls. When the control samples were defined (here three controls without the chromosome 13q deletion), CovCopCan efficiently detected two total q deletions on chromosome 13 and two partial deletions for the two other positive patients. In addition, no false-positive duplications were detected in the three controls. Performance on germline data Amplicon sensitivity and specificity We first tested our algorithm on germline data. We used several coverage files obtained after Proton sequencing of our “CMT-89” Ampliseq library (see , panel 1). We calculated the sensitivity of CovCopCan, by amplicon, using 22 positive controls confirmed by karyotype, real-time PCR, or Multiplex Ligation-dependent Probe Amplification (MLPA). The detected CNVs were present in 22 patients, sequenced in 11 runs . Of the 22 CNVs, 15 are covered by more than 10 amplicons. We used a range of CNV sizes from 4 (TFG) to 98 amplicons (chromosome X duplication). CovCopCan was used with the default settings, with all options active. Raw read values of less than 20 were deleted. The 22 CNVs are covered by a total of 461 amplicons. CovCopCan correctly detected 403 of 461 deleted/duplicated amplicons, giving an amplicon sensitivity of 0.87. If considering CNV detection, CovCopCan was able to detect 22 of the 22 CNVs tested, leading to a sensitivity of 1. In addition, we analyzed the PMP22 gene to calculate the specificity of CovCopCan by amplicon. Indeed, the PMP22 duplication is the most frequent known mutation responsible for CMT disease and all patients were initially screened by MLPA to detect this gene duplication. The PMP22 region was covered by 10 amplicons and the entire design contains 2,394 amplicons. We used 456 patients who had no CNV on PMP22 to estimate the specificity of the CovCopCan algorithm. Of the 4,560 PMP22 amplicons tested, 4,375 were indeed tagged as “normal” and only 185 were false positives, leading to a specificity of 0.96. Comparison with other tools . We compared CovCopCan with three other tools: IonCopy, DeviCNV, and ExomeDepth. IonCopy and DeviCNV are designed to analyze amplicon sequencing data without a control set. ExomeDepth uses a robust model for the read count data and to build an optimized reference set. We used the shiny version of the software IonCopy (v. 2.1.1), with the gene-wise analysis mode and default parameters. DeviCNV (v. 1.5.1) was launched with the recommended parameters, detailed in the manual. ExomeDepth (v.0.1) was also launched with the default parameters. We tested these tools on the same dataset, already described, containing the 22 CNVs. We only considered CNVs supported by at least three amplicons for all the tools. The results are presented in as the number of CNVs detected. CovCopCan, IonCopy, DeviCNV, and ExomeDepth each detected 22, 20, 18, and 19 CNVs, respectively . Only CovCopCan detected all CNVs for a sensitivity of 1. IonCopy, DeviCNV, and ExomeDepth showed sensitivity of 0.91, 0.82, and 0.86, respectively. It was impossible to verify all the other CNVs found by the various tools. Thus, we could not calculate specificity based on these data. However, a small number of CNVs would be expected, since the data correspond to germline samples. Thus, with only seven CNVs detected in addition to the 22 controls, CovCopCan must have had the best specificity for this dataset. Performance on cancer data Low cell fraction CovCopCan can also process cancer data. The main difference between germline and somatic data is that a cancer tissue sample may simultaneously contain both healthy cells and cancer cells. A low proportion of cancer cells may interfere with the detection of CNVs. We estimated the minimum proportion of cancer cells required for CNV detection by simulating the complete deletion of a gene covered by 80 amplicons using panel 1 (2,394 amplicons). We used a coverage matrix containing the data of 16 patients sequenced by an Ion Proton Sequencer. The deletion of the entire gene was simulated following this method: S R C i = R R C i × ( 1 − C a n c e r C e l l P r o p o r t i o n ) + R R C i 2 × C a n c e r C e l l P r o p o r t i o n SRC i is the simulated value of the amplicon i, RRC i the Raw Read Count of the amplicon i, and CancerCellProportion the proportion of cancer cells (0 < values < 1). We simulated a proportion of cancer cells ranging from 0 to 1, in steps of 0.05. The first CNV was detected by the cumulative summary chart for 15% of cancer cells and clearly identifiable for 20%. Using only “Z-detection”, the CNV was detected when 40% of the cells contained the deletion, whereas almost the entire gene (67/80 amplicons) was detected by “Z-detection” as deleted for 60% of cancer cells . We confirmed the results obtained from these simulated data using real data. We sequenced five patient samples harboring various amounts of positive cancer cells carrying the same ATM gene deletion and previously explored with conventional cytogenetics (karyotype and FISH). The data were obtained using panel 2 without control samples. The cumulative algorithm first detected the deletion from 19.5% cancer cells . These results show that CovCopCan can detect CNVs within a heterogeneous sample if the cancer cells make up at least 15 to 20%. Comparison with other tools We compared the performance of CovCopCan against IonCopy, DeviCNV, and ONCOCNV. First, we used these three tools on the deletion of the ATM gene described above. Like CovCopCan, both IonCopy, and ONCOCNV correctly detected the CNV with 19.5% of cancer cells, but not DeviCNV . In addition, we used another dataset obtained using panel 2. We sequenced the DNA of 54 patients in eight runs. Eighteen patients had a partial deletion of a chromosome arm, whereas two had a complete deletion of this same chromosome arm. The partial deletion was covered by 21 amplicons, whereas the entire deletion involved 39. In this study, we did not consider the percentage of cells presenting the CNVs. CovCopCan was used with the default settings, with all options active. Raw read values of less than 20 were deleted. IonCopy was used in the gene-wise mode with the default parameters. DeviCNV was used with the recommended settings. ONCOCNV (v 6.9) was used with the default settings. As with the germline data, we set the minimum number of amplicons to detect CNVs to three for each tool. DeviCNV failed to analyze a run due to a low number of samples (5) and detected four CNVs from the other patients. IonCopy detected nine CNVs. ONCOCNV correctly detected the 20 CNVs but required at least three controls in a run to call them. CovCopCan was able to detect CNVs, with or without controls. Without defining control samples, CovCopCan automatically detected 13 of 20 CNVs. When defining controls, the number of correct CNVs increased to 15 and using the interactive visualization option, such as the CUSUM chart, CovCopCan clearly indicated the presence of a deletion in at least four of the five additional samples .
We visualized the result of the two-stage ratio using sequencing data from panel 2 (see for details). This gene panel, designed by Ion AmpliSeq designer software, includes 1,206 amplicons on 70 genes. The run presented here was performed on an Ion Proton device and included seven patients. A deletion on chromosome 13 was shared by three of the seven patients (verified by karyotyping). Examples of the visualization obtained for two of the patients (patient 1 normal and patient 2 “deleted”) are presented in . Without the two-stage ratio, the region in non-deleted patients was disturbed and a false positive duplication event was detected by CovCopCan in both (highlighted by a vertical red rectangle, as for patient 1, ). The two-stage ratio improved the stability of the values so that no false duplication event was detected by CovCopCan, thus increasing the specificity ( , compare A and B). This method also improved the detection of deletions (highlighted by a vertical orange rectangle) in the true deleted patients, decreasing the number of false-negative amplicons .
To reduce the effect of individual false negative amplicons, CovCopCan relaxes the significance threshold when a single non-significant amplicon is flanked on both sides by significant amplicons. For this specific amplicon, the threshold will be automatically switched to 0.05. If this amplicon becomes significantly duplicated, it will be merged with the initial duplicated detected areas. The grey dot in the graph will stay grey, indicating that it is a merged area. Deletions are treated the same way. Here, we show the results of this merging option on a complete chromosome X duplication. A single duplication covering the entire gene is detected by CovCopCan, whereas six successive duplications would have been found without this merging option .
We tested this method with the Panel 2 data . Seven samples were simultaneously sequenced on an Ion Proton sequencer (three controls and four patients). The four patients share the same region q deletion on chromosome 13. Without defining controls, CovCopCan detected a correct deletion (highlighted by the vertical orange rectangle) for one of the four patients and only a partial deletion for another. In addition, two false-positive duplications (highlighted by the vertical red rectangle) were detected in two controls. When the control samples were defined (here three controls without the chromosome 13q deletion), CovCopCan efficiently detected two total q deletions on chromosome 13 and two partial deletions for the two other positive patients. In addition, no false-positive duplications were detected in the three controls.
Amplicon sensitivity and specificity We first tested our algorithm on germline data. We used several coverage files obtained after Proton sequencing of our “CMT-89” Ampliseq library (see , panel 1). We calculated the sensitivity of CovCopCan, by amplicon, using 22 positive controls confirmed by karyotype, real-time PCR, or Multiplex Ligation-dependent Probe Amplification (MLPA). The detected CNVs were present in 22 patients, sequenced in 11 runs . Of the 22 CNVs, 15 are covered by more than 10 amplicons. We used a range of CNV sizes from 4 (TFG) to 98 amplicons (chromosome X duplication). CovCopCan was used with the default settings, with all options active. Raw read values of less than 20 were deleted. The 22 CNVs are covered by a total of 461 amplicons. CovCopCan correctly detected 403 of 461 deleted/duplicated amplicons, giving an amplicon sensitivity of 0.87. If considering CNV detection, CovCopCan was able to detect 22 of the 22 CNVs tested, leading to a sensitivity of 1. In addition, we analyzed the PMP22 gene to calculate the specificity of CovCopCan by amplicon. Indeed, the PMP22 duplication is the most frequent known mutation responsible for CMT disease and all patients were initially screened by MLPA to detect this gene duplication. The PMP22 region was covered by 10 amplicons and the entire design contains 2,394 amplicons. We used 456 patients who had no CNV on PMP22 to estimate the specificity of the CovCopCan algorithm. Of the 4,560 PMP22 amplicons tested, 4,375 were indeed tagged as “normal” and only 185 were false positives, leading to a specificity of 0.96. Comparison with other tools . We compared CovCopCan with three other tools: IonCopy, DeviCNV, and ExomeDepth. IonCopy and DeviCNV are designed to analyze amplicon sequencing data without a control set. ExomeDepth uses a robust model for the read count data and to build an optimized reference set. We used the shiny version of the software IonCopy (v. 2.1.1), with the gene-wise analysis mode and default parameters. DeviCNV (v. 1.5.1) was launched with the recommended parameters, detailed in the manual. ExomeDepth (v.0.1) was also launched with the default parameters. We tested these tools on the same dataset, already described, containing the 22 CNVs. We only considered CNVs supported by at least three amplicons for all the tools. The results are presented in as the number of CNVs detected. CovCopCan, IonCopy, DeviCNV, and ExomeDepth each detected 22, 20, 18, and 19 CNVs, respectively . Only CovCopCan detected all CNVs for a sensitivity of 1. IonCopy, DeviCNV, and ExomeDepth showed sensitivity of 0.91, 0.82, and 0.86, respectively. It was impossible to verify all the other CNVs found by the various tools. Thus, we could not calculate specificity based on these data. However, a small number of CNVs would be expected, since the data correspond to germline samples. Thus, with only seven CNVs detected in addition to the 22 controls, CovCopCan must have had the best specificity for this dataset.
We first tested our algorithm on germline data. We used several coverage files obtained after Proton sequencing of our “CMT-89” Ampliseq library (see , panel 1). We calculated the sensitivity of CovCopCan, by amplicon, using 22 positive controls confirmed by karyotype, real-time PCR, or Multiplex Ligation-dependent Probe Amplification (MLPA). The detected CNVs were present in 22 patients, sequenced in 11 runs . Of the 22 CNVs, 15 are covered by more than 10 amplicons. We used a range of CNV sizes from 4 (TFG) to 98 amplicons (chromosome X duplication). CovCopCan was used with the default settings, with all options active. Raw read values of less than 20 were deleted. The 22 CNVs are covered by a total of 461 amplicons. CovCopCan correctly detected 403 of 461 deleted/duplicated amplicons, giving an amplicon sensitivity of 0.87. If considering CNV detection, CovCopCan was able to detect 22 of the 22 CNVs tested, leading to a sensitivity of 1. In addition, we analyzed the PMP22 gene to calculate the specificity of CovCopCan by amplicon. Indeed, the PMP22 duplication is the most frequent known mutation responsible for CMT disease and all patients were initially screened by MLPA to detect this gene duplication. The PMP22 region was covered by 10 amplicons and the entire design contains 2,394 amplicons. We used 456 patients who had no CNV on PMP22 to estimate the specificity of the CovCopCan algorithm. Of the 4,560 PMP22 amplicons tested, 4,375 were indeed tagged as “normal” and only 185 were false positives, leading to a specificity of 0.96. Comparison with other tools . We compared CovCopCan with three other tools: IonCopy, DeviCNV, and ExomeDepth. IonCopy and DeviCNV are designed to analyze amplicon sequencing data without a control set. ExomeDepth uses a robust model for the read count data and to build an optimized reference set. We used the shiny version of the software IonCopy (v. 2.1.1), with the gene-wise analysis mode and default parameters. DeviCNV (v. 1.5.1) was launched with the recommended parameters, detailed in the manual. ExomeDepth (v.0.1) was also launched with the default parameters. We tested these tools on the same dataset, already described, containing the 22 CNVs. We only considered CNVs supported by at least three amplicons for all the tools. The results are presented in as the number of CNVs detected. CovCopCan, IonCopy, DeviCNV, and ExomeDepth each detected 22, 20, 18, and 19 CNVs, respectively . Only CovCopCan detected all CNVs for a sensitivity of 1. IonCopy, DeviCNV, and ExomeDepth showed sensitivity of 0.91, 0.82, and 0.86, respectively. It was impossible to verify all the other CNVs found by the various tools. Thus, we could not calculate specificity based on these data. However, a small number of CNVs would be expected, since the data correspond to germline samples. Thus, with only seven CNVs detected in addition to the 22 controls, CovCopCan must have had the best specificity for this dataset.
Low cell fraction CovCopCan can also process cancer data. The main difference between germline and somatic data is that a cancer tissue sample may simultaneously contain both healthy cells and cancer cells. A low proportion of cancer cells may interfere with the detection of CNVs. We estimated the minimum proportion of cancer cells required for CNV detection by simulating the complete deletion of a gene covered by 80 amplicons using panel 1 (2,394 amplicons). We used a coverage matrix containing the data of 16 patients sequenced by an Ion Proton Sequencer. The deletion of the entire gene was simulated following this method: S R C i = R R C i × ( 1 − C a n c e r C e l l P r o p o r t i o n ) + R R C i 2 × C a n c e r C e l l P r o p o r t i o n SRC i is the simulated value of the amplicon i, RRC i the Raw Read Count of the amplicon i, and CancerCellProportion the proportion of cancer cells (0 < values < 1). We simulated a proportion of cancer cells ranging from 0 to 1, in steps of 0.05. The first CNV was detected by the cumulative summary chart for 15% of cancer cells and clearly identifiable for 20%. Using only “Z-detection”, the CNV was detected when 40% of the cells contained the deletion, whereas almost the entire gene (67/80 amplicons) was detected by “Z-detection” as deleted for 60% of cancer cells . We confirmed the results obtained from these simulated data using real data. We sequenced five patient samples harboring various amounts of positive cancer cells carrying the same ATM gene deletion and previously explored with conventional cytogenetics (karyotype and FISH). The data were obtained using panel 2 without control samples. The cumulative algorithm first detected the deletion from 19.5% cancer cells . These results show that CovCopCan can detect CNVs within a heterogeneous sample if the cancer cells make up at least 15 to 20%. Comparison with other tools We compared the performance of CovCopCan against IonCopy, DeviCNV, and ONCOCNV. First, we used these three tools on the deletion of the ATM gene described above. Like CovCopCan, both IonCopy, and ONCOCNV correctly detected the CNV with 19.5% of cancer cells, but not DeviCNV . In addition, we used another dataset obtained using panel 2. We sequenced the DNA of 54 patients in eight runs. Eighteen patients had a partial deletion of a chromosome arm, whereas two had a complete deletion of this same chromosome arm. The partial deletion was covered by 21 amplicons, whereas the entire deletion involved 39. In this study, we did not consider the percentage of cells presenting the CNVs. CovCopCan was used with the default settings, with all options active. Raw read values of less than 20 were deleted. IonCopy was used in the gene-wise mode with the default parameters. DeviCNV was used with the recommended settings. ONCOCNV (v 6.9) was used with the default settings. As with the germline data, we set the minimum number of amplicons to detect CNVs to three for each tool. DeviCNV failed to analyze a run due to a low number of samples (5) and detected four CNVs from the other patients. IonCopy detected nine CNVs. ONCOCNV correctly detected the 20 CNVs but required at least three controls in a run to call them. CovCopCan was able to detect CNVs, with or without controls. Without defining control samples, CovCopCan automatically detected 13 of 20 CNVs. When defining controls, the number of correct CNVs increased to 15 and using the interactive visualization option, such as the CUSUM chart, CovCopCan clearly indicated the presence of a deletion in at least four of the five additional samples .
CovCopCan can also process cancer data. The main difference between germline and somatic data is that a cancer tissue sample may simultaneously contain both healthy cells and cancer cells. A low proportion of cancer cells may interfere with the detection of CNVs. We estimated the minimum proportion of cancer cells required for CNV detection by simulating the complete deletion of a gene covered by 80 amplicons using panel 1 (2,394 amplicons). We used a coverage matrix containing the data of 16 patients sequenced by an Ion Proton Sequencer. The deletion of the entire gene was simulated following this method: S R C i = R R C i × ( 1 − C a n c e r C e l l P r o p o r t i o n ) + R R C i 2 × C a n c e r C e l l P r o p o r t i o n SRC i is the simulated value of the amplicon i, RRC i the Raw Read Count of the amplicon i, and CancerCellProportion the proportion of cancer cells (0 < values < 1). We simulated a proportion of cancer cells ranging from 0 to 1, in steps of 0.05. The first CNV was detected by the cumulative summary chart for 15% of cancer cells and clearly identifiable for 20%. Using only “Z-detection”, the CNV was detected when 40% of the cells contained the deletion, whereas almost the entire gene (67/80 amplicons) was detected by “Z-detection” as deleted for 60% of cancer cells . We confirmed the results obtained from these simulated data using real data. We sequenced five patient samples harboring various amounts of positive cancer cells carrying the same ATM gene deletion and previously explored with conventional cytogenetics (karyotype and FISH). The data were obtained using panel 2 without control samples. The cumulative algorithm first detected the deletion from 19.5% cancer cells . These results show that CovCopCan can detect CNVs within a heterogeneous sample if the cancer cells make up at least 15 to 20%.
We compared the performance of CovCopCan against IonCopy, DeviCNV, and ONCOCNV. First, we used these three tools on the deletion of the ATM gene described above. Like CovCopCan, both IonCopy, and ONCOCNV correctly detected the CNV with 19.5% of cancer cells, but not DeviCNV . In addition, we used another dataset obtained using panel 2. We sequenced the DNA of 54 patients in eight runs. Eighteen patients had a partial deletion of a chromosome arm, whereas two had a complete deletion of this same chromosome arm. The partial deletion was covered by 21 amplicons, whereas the entire deletion involved 39. In this study, we did not consider the percentage of cells presenting the CNVs. CovCopCan was used with the default settings, with all options active. Raw read values of less than 20 were deleted. IonCopy was used in the gene-wise mode with the default parameters. DeviCNV was used with the recommended settings. ONCOCNV (v 6.9) was used with the default settings. As with the germline data, we set the minimum number of amplicons to detect CNVs to three for each tool. DeviCNV failed to analyze a run due to a low number of samples (5) and detected four CNVs from the other patients. IonCopy detected nine CNVs. ONCOCNV correctly detected the 20 CNVs but required at least three controls in a run to call them. CovCopCan was able to detect CNVs, with or without controls. Without defining control samples, CovCopCan automatically detected 13 of 20 CNVs. When defining controls, the number of correct CNVs increased to 15 and using the interactive visualization option, such as the CUSUM chart, CovCopCan clearly indicated the presence of a deletion in at least four of the five additional samples .
CovCopCan sources are available on GitHub: https://git.unilim.fr/merilp02/CovCopCan/tree/master . Pre-complied binaries can be downloaded from this page of the GitHub repository: https://git.unilim.fr/merilp02/CovCopCan/tree/master . CovCopCan offers a wide range of features to interpret data from amplicon sequencing to detect CNVs. This tool works on data generated from Ion Designer (Life Technologies, CA, USA) as well as that from Illumina DesignStudio (Illumina Inc., San Diego, CA, USA). The user-friendly interface associated with our 2D visualization facilitates data exploration and manipulation allowing complex analyses such as those from cancer data. CovCopCan also offers the possibility to export the results in VCF format or graphical output for publications. It can also be used in command-line mode to be integrated into various pipelines (see ). Future development of CovCopCan will involve the possibility to exploit the variant allele fraction (VAF) to improve the statistical detection of CNVs. We will also improve memory consumption and parallelism to ensure that CovCopCan can work on a minimal configuration.
S1 File Supplementary information of this article. The supplementary document provides information on the panels used in this article, a guideline to create an optimized panel to call CNVs, the workflow of CovCopCan algorithm, information on the possibility to define manually reference amplicons, details on graphical visualization elements and command line interface data. (DOCX) Click here for additional data file.
|
Human Untargeted
Metabolomics in High-Throughput Gut
Microbiome Research: Ethanol vs Methanol | 388f020a-8391-4a83-9083-97de88807251 | 11912123 | Biochemistry[mh] | Humans are colonized at birth by microorganisms,
forming complex
communities known as microbiota, which evolve through the course of
the host life, shaping and influencing its physiology and metabolism. Among others, the gut microbiota has been shown
to actively influence neurodevelopment before , and
after partum ( − ) and to induce distal tumors via
carcinogenic metabolism. For these reasons,
elucidating the functionality of these microbial communities has become
paramount, and mass spectrometry-based untargeted metabolomics has
established itself as the to-go tool for these types of investigations. Multiomics large-scale longitudinal or
cross-sectional microbiome
studies represent a challenge both for sequencing and metabolomics
as samples are usually collected in different settings and cannot
always be immediately snap-frozen in liquid nitrogen and stored at
−80 °C, which is considered the gold standard. Additionally,
when designing large-scale studies, the safety of subjects and shipping
costs, which dramatically increase if refrigeration is involved, should
be taken into consideration. The microbiome field tackled these problems
by showcasing that fecal microbiome collection and storage in 95%
ethanol (EtOH) at room temperature stabilizes the microbial communities
up to 8 weeks. Most importantly, EtOH
is also safer to handle when compared to other alcohols, such as methanol
(MeOH), which is extremely toxic and requires special equipment to
be properly handled. Building on this, we recently introduced the
Matrix Method, which employs a high-throughput
pipeline that leverages sample collection in single barcoded Matrix
tubes containing 95% EtOH and automatized robots. This method not
only reduces costs, time, and well-to-well DNA contamination but also
enables the extraction of metabolites from the same biological sample,
offering an all-in-one solution streamlined multiomics analyses. EtOH
extraction from fecal metabolomics studies has been tested before, − but an in-depth downstream data analysis comparison with another
common extraction method (50% MeOH), − used for the discovery of novel
bile acids conjugates and the recently
introduced reverse metabolomics approach, was missing. Here, we showcase that the 95% EtOH metabolomics
extraction part
of the Matrix Method for bulk microbiome analyses yields equivalent
results of a common, but more laborious and time-consuming, 50% MeOH
extraction for human fecal samples. We also show that 95% EtOH can
stabilize the fecal metabolome at room temperature for up to 1 week,
allowing for safe off-site collection and possibly reduced shipping
costs. Finally, we highlight how a robust center log ratio (RCLR)
transformation data analysis workflow for untargeted fecal metabolomics
data overcomes problems of uneven sampling collection and allows for
direct integration with microbiome data.
Sample
Collection and Extraction Human fecal samples
were collected and immediately stored at −80 °C from healthy
volunteers under approved protocols from the University of California
San Diego (IRB#141853) with informed consent. Three fecal samples
from three different adult subjects (one male and two females) were
randomly selected and aliquoted for untargeted metabolomics analysis.
Multiple aliquots of different weights (10, 20, and 30 mg) were generated
in triplicates from the three different fecal samples. Aliquots were
transferred in either 95% (v/v) ethanol (EtOH) or 50% (v/v) methanol
(MeOH). Samples were immediately extracted, except for a subset of
samples that were left in 95% EtOH for either 24 h or 1 week at room
temperature to check the fecal metabolome stability. Samples to which
400 μL of 95% EtOH was added were extracted via the Matrix Method
pipeline, simply consisting in shaking
samples at 1200 rpm for 2 min in a SpexMiniG plate shaker (SPEX SamplePrep
part #1600, NJ, USA), followed by a 5 min centrifugation step at 2700 g . The supernatant (400 μL) was then collected and
stored at −80 °C for downstream analysis. Samples to which
800 μL of MeOH was added underwent a validated extraction protocol, involving homogenization with a 5 mm stainless-steel
bead in a TissueLyser II (QIAGEN) for 5 min at 25 Hz, incubation at
4 °C for 30 min, and centrifugation at 21,130 g for 10 min. The supernatant (400 μL) was then collected and
dried overnight using a CentriVap instrument. Samples were then stored
at −80 °C until resuspension. All supernatants, from EtOH
and MeOH extractions, were dried overnight using a CentriVap instrument
and then resuspended in 200 μL of 50% MeOH containing 1 μM
of sulfamethazine as the internal standard. A pooled sample (QCpool)
was then generated by collecting and mixing 10 μL from each
biological sample and aliquoting 200 μL. Blank samples, consisting
only of extraction solution, were also prepared. Finally, the samples
were incubated for 1 h at −20 °C, centrifuged at 21,130 g for 10 min, and transferred in 2 mL glass vial (Thermo
Scientific) for ultrahigh-performance liquid chromatography tandem
mass spectrometry (UHPLC-MS/MS) analysis. UHPLC-MS/MS Experiment Samples were randomized and
analyzed using an untargeted metabolomics analysis platform comprising
a Vanquish UHPLC system coupled to a Q-Exactive Orbitrap mass spectrometer
(Thermo Fisher Scientific). The chromatography system consisted of
a Phenomenex C18 column (1.7 μm particle size, 2.1 mm ×
50 mm) and a mobile phase of solvent A (water + 0.1% formic acid)
and solvent B (acetonitrile + 0.1% formic acid). Injections of 5 μL
of samples, with a flow rate of 0.5 mL/min, followed this gradient:
0–1 min, 5% B; 1–7 min, 5–99% B; 7–8 min,
99% B; 8–8.5 min, 99–5% B; 8.5–10 min, 5% B.
MS/MS data were acquired in the data-dependent acquisition mode using
positive electrospray ionization (ESI+). Briefly, ESI parameters were
set as follows: 53 L/min sheath gas flow, 14 L/min aux gas flow rate,
3 L/min sweep gas flow, 3.5 kV spray voltage, 269 °C intel capillary,
and aux gas heater set to 438 °C. MS scan range was set to 100–1500 m / z with a resolution at m / z 200 set to 35,000 with 1 microscans. Automatic
gain control (AGC) was set to 5E4 with a maximum injection time of
50 ms. Up to 5 MS/MS (TopN = 5) spectra per MS1 were collected with
a resolution at m / z 200 set to 17,500
with 1 microscans. Injection time was 50 ms with an AGC target of
5E4. The isolation window was set to 2.0 m / z . Normalized collision energy was set to a stepwise increase
of 20, 30, and 40 eV with an apex trigger set to 2–15 s and
a dynamic exclusion of 10 s. UHPLC-MS/MS Data Processing Obtained
raw files were
converted into .mzML open-access format using ProteoWizard MSConvert and deposited on GNPS/MassIVE under the accession
number MSV000095260. Feature detection and extraction was performed
via MZmine 3.9 via batch processing. The
.xml file used for batch processing can be found on the associated
GitHub page. Briefly, data were imported using MS1and MS2 detector
via factor of lowest signal with noise factors set to 3 and 1.1, respectively.
Sequentially, mass detection was performed, and only ions were acquired
between 0.5 and 8 min, with MS1 and MS2 noise levels set to 5E4 and
1E3, respectively. Chromatogram builder parameters were set at five
minimum consecutive scans, 1E5 minimum absolute height, and 10 ppm
for m / z tolerance. Smoothing was
applied before the local minimum resolver, which had the following
parameters: chromatographic threshold of 85%, minimum search range
retention time of 0.2 min, and minimum ratio of peak top/edge of 1.7.
Then, the 13C isotope filter and isotope finder were applied. Features
were aligned using join aligner with weight for m / z set to 3 and retention time tolerance set to
0.2 min. Features not detected in at least three samples were removed
before performing peak finder. Ion identity networking and metaCorrelate
were performed before exporting the final feature table. The GNPS
and SIRIUS export functions were used to generate the feature table
containing the peak areas and the .mgf files necessary for downstream
analyses. Feature-based molecular networking was performed in GNPS2
( https://gnps2.org/status?task=40d2affb3df544d4a2bbf6841a62d45a# ), and it was used to annotate metabolic
features via MS/MS spectral matching to the GNPS library, which represent
a level 2 annotation according the Metabolomics Standard Initiative.
Annotations were obtained via parent ion mass matching with a 0.02
tolerance, a minimum of five matching fragment ions, and a cosine
score similarity > 0.7. A list of all annotated molecular features
can be found in the “Library Results” tab of the FBMN
job page. Molecular classes of ions with m / z < 800 were predicted using CANOPUS Natural Product
Classifier (NPC) via SIRIUS 5.8. Data Analysis Feature Table was imported in R 4.2.2
(R Foundation for Statistical Computing, Vienna, Austria) for downstream
data analysis. Total extracted peak area per sample was calculated
and correlated with the sample run order to identify possible acquisition
problems during the run. The internal standard (IS) peak present in
each sample (Sulfamethazine [M + H], m / z 279.0908 and retention time 2.26 min) was also extracted and correlated
to the sample run order. The coefficients of variance (CVs) of six
different standards (amitriptyline, sulfadimethoxine, sulfamethazine,
sulfamethizole, sulfachloropyridazine, and coumarin 314) present in
the QCmix sample, which was run every 10 biological samples, were
inspected to evaluate the run quality. The CVs of each acquired feature
were also calculated using the QCpool, which was also run for every 10 biological samples. The CV was calculated
by dividing the mean of the extracted peak areas by the standard deviation.
Feature table was cleaned via blank subtraction. Features only detected
in the blank and QCmix samples or those that mean peak areas were
not at least 10 times the ones observed in the QCpool were discarded.
The package ‘homologueDiscoverer v 0.0.0.9’ was used
to remove detected PEGs (polyethylene glycol) contaminants. Features with near zero-variance were removed
using ‘caret v 6.0’. The
package ‘mixOmics v 6.22’ was used for multivariate
analysis. Principal component analysis
(PCA) and partial least-squares discriminant analysis (PLS-DA) were
performed after RCLR transformation via ‘vegan v 2.6’. In the PCA models, PERMANOVA was used to evaluate
group centroid separation, while PERMDIPS2 was used to evaluate homogeneity
of variance between groups. PLS-DA models’ performances were
evaluated using leave-one-out (loo) cross-validation. Variable importance
(VIP) scores were calculated per feature and features with VIPs >
1 were considered significant. The package ‘UpSetR v 1.4’
was used to generate the upset plots. High density region (HDR) plots were generated using ‘ggdensity
v 1.0’. Log2 fold changes (Log2FC) were calculated by taking
the log2 of divided means of the relative abundance of the peak areas
of the group of interest. When the mean was 0, a pseudocount of 1
× 10 –9 was added. Linear mixed effect models
were obtained using ‘lmerTest v 3.1’ using subject id
as the random effect. The packages ‘tidyverse v 2.0’
and ‘ggpubr v 0.6’ were used for data manipulation and
visualization. Code used for the analysis and to generate the figures
of the manuscript is available on GitHub ( https://github.com/simonezuffa/Manuscript_Matrix_Metabolomics ).
Human fecal samples
were collected and immediately stored at −80 °C from healthy
volunteers under approved protocols from the University of California
San Diego (IRB#141853) with informed consent. Three fecal samples
from three different adult subjects (one male and two females) were
randomly selected and aliquoted for untargeted metabolomics analysis.
Multiple aliquots of different weights (10, 20, and 30 mg) were generated
in triplicates from the three different fecal samples. Aliquots were
transferred in either 95% (v/v) ethanol (EtOH) or 50% (v/v) methanol
(MeOH). Samples were immediately extracted, except for a subset of
samples that were left in 95% EtOH for either 24 h or 1 week at room
temperature to check the fecal metabolome stability. Samples to which
400 μL of 95% EtOH was added were extracted via the Matrix Method
pipeline, simply consisting in shaking
samples at 1200 rpm for 2 min in a SpexMiniG plate shaker (SPEX SamplePrep
part #1600, NJ, USA), followed by a 5 min centrifugation step at 2700 g . The supernatant (400 μL) was then collected and
stored at −80 °C for downstream analysis. Samples to which
800 μL of MeOH was added underwent a validated extraction protocol, involving homogenization with a 5 mm stainless-steel
bead in a TissueLyser II (QIAGEN) for 5 min at 25 Hz, incubation at
4 °C for 30 min, and centrifugation at 21,130 g for 10 min. The supernatant (400 μL) was then collected and
dried overnight using a CentriVap instrument. Samples were then stored
at −80 °C until resuspension. All supernatants, from EtOH
and MeOH extractions, were dried overnight using a CentriVap instrument
and then resuspended in 200 μL of 50% MeOH containing 1 μM
of sulfamethazine as the internal standard. A pooled sample (QCpool)
was then generated by collecting and mixing 10 μL from each
biological sample and aliquoting 200 μL. Blank samples, consisting
only of extraction solution, were also prepared. Finally, the samples
were incubated for 1 h at −20 °C, centrifuged at 21,130 g for 10 min, and transferred in 2 mL glass vial (Thermo
Scientific) for ultrahigh-performance liquid chromatography tandem
mass spectrometry (UHPLC-MS/MS) analysis.
Samples were randomized and
analyzed using an untargeted metabolomics analysis platform comprising
a Vanquish UHPLC system coupled to a Q-Exactive Orbitrap mass spectrometer
(Thermo Fisher Scientific). The chromatography system consisted of
a Phenomenex C18 column (1.7 μm particle size, 2.1 mm ×
50 mm) and a mobile phase of solvent A (water + 0.1% formic acid)
and solvent B (acetonitrile + 0.1% formic acid). Injections of 5 μL
of samples, with a flow rate of 0.5 mL/min, followed this gradient:
0–1 min, 5% B; 1–7 min, 5–99% B; 7–8 min,
99% B; 8–8.5 min, 99–5% B; 8.5–10 min, 5% B.
MS/MS data were acquired in the data-dependent acquisition mode using
positive electrospray ionization (ESI+). Briefly, ESI parameters were
set as follows: 53 L/min sheath gas flow, 14 L/min aux gas flow rate,
3 L/min sweep gas flow, 3.5 kV spray voltage, 269 °C intel capillary,
and aux gas heater set to 438 °C. MS scan range was set to 100–1500 m / z with a resolution at m / z 200 set to 35,000 with 1 microscans. Automatic
gain control (AGC) was set to 5E4 with a maximum injection time of
50 ms. Up to 5 MS/MS (TopN = 5) spectra per MS1 were collected with
a resolution at m / z 200 set to 17,500
with 1 microscans. Injection time was 50 ms with an AGC target of
5E4. The isolation window was set to 2.0 m / z . Normalized collision energy was set to a stepwise increase
of 20, 30, and 40 eV with an apex trigger set to 2–15 s and
a dynamic exclusion of 10 s.
Obtained
raw files were
converted into .mzML open-access format using ProteoWizard MSConvert and deposited on GNPS/MassIVE under the accession
number MSV000095260. Feature detection and extraction was performed
via MZmine 3.9 via batch processing. The
.xml file used for batch processing can be found on the associated
GitHub page. Briefly, data were imported using MS1and MS2 detector
via factor of lowest signal with noise factors set to 3 and 1.1, respectively.
Sequentially, mass detection was performed, and only ions were acquired
between 0.5 and 8 min, with MS1 and MS2 noise levels set to 5E4 and
1E3, respectively. Chromatogram builder parameters were set at five
minimum consecutive scans, 1E5 minimum absolute height, and 10 ppm
for m / z tolerance. Smoothing was
applied before the local minimum resolver, which had the following
parameters: chromatographic threshold of 85%, minimum search range
retention time of 0.2 min, and minimum ratio of peak top/edge of 1.7.
Then, the 13C isotope filter and isotope finder were applied. Features
were aligned using join aligner with weight for m / z set to 3 and retention time tolerance set to
0.2 min. Features not detected in at least three samples were removed
before performing peak finder. Ion identity networking and metaCorrelate
were performed before exporting the final feature table. The GNPS
and SIRIUS export functions were used to generate the feature table
containing the peak areas and the .mgf files necessary for downstream
analyses. Feature-based molecular networking was performed in GNPS2
( https://gnps2.org/status?task=40d2affb3df544d4a2bbf6841a62d45a# ), and it was used to annotate metabolic
features via MS/MS spectral matching to the GNPS library, which represent
a level 2 annotation according the Metabolomics Standard Initiative.
Annotations were obtained via parent ion mass matching with a 0.02
tolerance, a minimum of five matching fragment ions, and a cosine
score similarity > 0.7. A list of all annotated molecular features
can be found in the “Library Results” tab of the FBMN
job page. Molecular classes of ions with m / z < 800 were predicted using CANOPUS Natural Product
Classifier (NPC) via SIRIUS 5.8.
Feature Table was imported in R 4.2.2
(R Foundation for Statistical Computing, Vienna, Austria) for downstream
data analysis. Total extracted peak area per sample was calculated
and correlated with the sample run order to identify possible acquisition
problems during the run. The internal standard (IS) peak present in
each sample (Sulfamethazine [M + H], m / z 279.0908 and retention time 2.26 min) was also extracted and correlated
to the sample run order. The coefficients of variance (CVs) of six
different standards (amitriptyline, sulfadimethoxine, sulfamethazine,
sulfamethizole, sulfachloropyridazine, and coumarin 314) present in
the QCmix sample, which was run every 10 biological samples, were
inspected to evaluate the run quality. The CVs of each acquired feature
were also calculated using the QCpool, which was also run for every 10 biological samples. The CV was calculated
by dividing the mean of the extracted peak areas by the standard deviation.
Feature table was cleaned via blank subtraction. Features only detected
in the blank and QCmix samples or those that mean peak areas were
not at least 10 times the ones observed in the QCpool were discarded.
The package ‘homologueDiscoverer v 0.0.0.9’ was used
to remove detected PEGs (polyethylene glycol) contaminants. Features with near zero-variance were removed
using ‘caret v 6.0’. The
package ‘mixOmics v 6.22’ was used for multivariate
analysis. Principal component analysis
(PCA) and partial least-squares discriminant analysis (PLS-DA) were
performed after RCLR transformation via ‘vegan v 2.6’. In the PCA models, PERMANOVA was used to evaluate
group centroid separation, while PERMDIPS2 was used to evaluate homogeneity
of variance between groups. PLS-DA models’ performances were
evaluated using leave-one-out (loo) cross-validation. Variable importance
(VIP) scores were calculated per feature and features with VIPs >
1 were considered significant. The package ‘UpSetR v 1.4’
was used to generate the upset plots. High density region (HDR) plots were generated using ‘ggdensity
v 1.0’. Log2 fold changes (Log2FC) were calculated by taking
the log2 of divided means of the relative abundance of the peak areas
of the group of interest. When the mean was 0, a pseudocount of 1
× 10 –9 was added. Linear mixed effect models
were obtained using ‘lmerTest v 3.1’ using subject id
as the random effect. The packages ‘tidyverse v 2.0’
and ‘ggpubr v 0.6’ were used for data manipulation and
visualization. Code used for the analysis and to generate the figures
of the manuscript is available on GitHub ( https://github.com/simonezuffa/Manuscript_Matrix_Metabolomics ).
Fecal samples from 3 different
human subjects were aliquoted to
generate 72 replicates . Triplicates of different weights, 10, 20, and 30 mg, were
generated and extracted via two different pipelines. The first one
consisted of a 50% MeOH fecal extraction protocol, previously described, while the second one involved a 95% EtOH extraction
part of the recently introduced Matrix Method pipeline. This consists of the collection of fecal samples
in Matrix tubes containing 400 μL of 95% EtOH, followed by 2
min shaking at 1200 rpm and 5 min centrifugation at 2700 g . Importantly, this automatized high-throughput pipeline allows for
the simultaneous extraction of DNA, allowing multiomics analyses in
microbiome research. Additionally, two batches of triplicates collected
in 95% EtOH were left at room temperature for 1 day and 1 week, respectively,
to replicate possible collection scenarios of storing, shipping, and
assessing fecal metabolome stability in 95% EtOH. 95% EtOH Extraction Part of the Matrix Method
Recapitulates
a 50% MeOH Fecal Extraction Protocol Unsupervised dimensionality
reduction via PCA of 20 mg of triplicates showed clear clustering
of subject fecal metabolic profiles using both 95% EtOH and 50% MeOH
extractions ( A). PERMANOVA identified subject id as the highest source of variance
in the data ( R 2 = 0.47, F = 10.98, p < 0.001), followed by extraction
method ( R 2 = 0.14, F =
6.68, p < 0.001). Out of a total of 4616 unique
metabolic features, 75% (3463) were recovered by both extraction methods
( B). These
included 94% (235) of the total annotated features via the GNPS library.
Interestingly, the Matrix Method (95% EtOH) appeared to capture a
higher number of additional features (849) compared to that of the
MeOH extraction (304). These extraction-specific features were classified
as small peptides and fatty acid conjugates for MeOH and oligopeptides,
glycerolipids, and fatty amides for EtOH according to CANOPUS NPC
superclass predictions ( Supplementary Table 1 ). Pairwise PCA and PLS-DA models were generated for each extraction
method to identify the metabolic features responsible for subject
discrimination. These features were then selected to compare both
pipelines. All PCA and PLS-DA score plots displayed clear clustering
by subject ( Supplementary Figure 1 ). All
PLS-DA models obtained a classification error rate of 0, indicating
a perfect discriminating performance. Extracted VIP scores from the
PLS-DA models displayed significant correlation between the two extraction
pipelines ( Supplementary Figure 2 ), with
the highest concordance as observed in the HDR plots ( C). On average, the PLS-DA
models identified 1764 features with VIPs > 1 involved in subject
discrimination. Notably, when considering exclusively the top 100
features obtained in the MeOH models, 86, 85, and 92% were also recovered
and classified as significant by the 95% EtOH models. Features of
interest identified by the multivariate analysis were also investigated
via univariate analysis. A high degree of correlation ( R > 0.9) was observed for the feature Log2FC between subjects obtained
via the two different extraction pipelines ( D). Focusing on annotated molecules of interest
in microbiome research, such as primary and secondary bile acids,
indole amino acids, vitamin B, heme catabolism, microbial N -acyl lipids, and long chain fatty acids, the 95% EtOH
extraction recovers and returns results comparable to those obtained
via 50% MeOH extraction. Recovery examples of tri- and dihydroxylated
bile acids, lysine, and histidine microbially conjugated bile acids,
indoles such as tryptophan and propionic acid, vitamin B5, stercobilin,
arginine-C5:0, and ω-3 arachidonic acid are available in Supplementary Figure 3 . Fecal Metabolome Remains Stable for Up to 1 Week at Room Temperature
in 95% EtOH PCA of triplicates exclusively extracted via
the 95% EtOH pipeline showed clear clustering of samples based on
subject ID (PERMANOVA, R 2 = 0.50, F = 15.24, p < 0.001) and a smaller
effect of sample storage (PERMANOVA, R 2 = 0.08, F = 2.56, p < 0.001),
which included immediate processing or storage at room temperature
for either a day or a week in 95% EtOH ( A). Out of the 4566 features obtained via
EtOH extraction, 88% (3996) were captured at all time points, comprising
94% (231) of all annotated features ( B). The majority of the features distinctively
characterizing immediate (imm), week, immediate and day, and immediate
and week extractions (156 out of 570) were not classified by CANOPUS
( Supplementary Table 2 ). The subsequent
most affected predicted classes were small peptides (65) and oligopeptides
(48). Additionally, to determine if the storage time at room temperature
could affect the metabolic features responsible for subject classification,
27 subject pairwise PLS-DA models were generated comparing samples
of each subject at each different time point. Comparisons between
immediate extractions were considered as “ground truths,”
and the relative recovery of those significant features was investigated
for the different time points ( C). On average, 95.6, 91.6, and 94.7% of the top 100
“ground truth” features were recovered for each pairwise
subject comparison. This suggested that the fecal metabolome could
remain relatively stable in 95% EtOH for up to 1 week at room temperature
without losing predictive power for subject discrimination. Log2FC
in relation to storage time (immediate vs 1 week) of the metabolic
were also explored for each subject ( Supplementary Figure 4A ). Interestingly, the median fold changes of the discriminating
features of interest identified via the pairwise PLS-DA models were
0.705 (Subject A vs Subject B), 0.430 (Subject B vs Subject C), and
0.530 (Subject A vs Subject C), suggesting a relative stability during
storage time of the features of interest ( Supplementary Figure 4B ). RCLR Transformation Can Address Discrepancies in Fecal Sample
Collection A recurrent issue in off-site fecal sample collection
is the uneven sample amount collected by the individuals and the variability
in the fecal water content. These can affect the per sample metabolite
recovery and introduce bias into the downstream data analysis. Although
lyophilization and weighting of the fecal samples can be implemented
before data acquisition, these are time-consuming and unpractical
when analyzing thousands of samples, and several technical errors
can be introduced, such as repeated freeze–thaw cycles, intersample
contamination, and others. To overcome these limitations, post data
acquisition normalization methods can be deployed. Here, we investigate
the use of RCLR transformation in fecal untargeted metabolomics data
as a method to improve accuracy in data analysis. Originally introduced
to tackle the microbiome data compositionality, the RCLR transformation suits the semicompositionality
nature of fecal untargeted metabolomics data. Additionally, RCLR allows
for easy interpretability and seamless multiomics integration with
associated RCLR-transformed microbiome data using tools such as Joint-RPCA and DIABLO. We
investigated the extraction of different sample weights, 10, 20, and
30 mg, in both 95% EtOH and 50% MeOH extraction pipelines. PCA of
RCLR data shows no effect of sample weight in both extraction methods
(PERMANOVAs, R 2 < 0.05, p > 0.08), with most of the variance explained by subject id ( R 2 = ∼60%) and no difference in the dispersion
of the samples (PERMDISP2, p > 0.45) between the
different subjects ( A). On the contrary, PCA of raw data or relative abundance
transformed data displayed less tightened clusters based on subjects,
significant differences between group variances (PERMDISP2, p < 0.001), and small but still significant effect of
the sample weight ( Supplementary Figure 5 ). Upset plot revealed that the number of detected metabolites was
not affected by sample weight, suggesting that just 10 mg is sufficient
to cover the detectable fecal metabolome in reverse-phase LC-MS/MS
data acquired in the positive ionization mode ( B). Finally, we explored correlations between
weight, total obtained gDNA (genomic DNA, ng/μL), and cumulative
extracted peak areas from samples processed via the Matrix Method
(95% EtOH) and stored for 1 week at room temperature. Linear mixed
effect models, accounting for repeated measures on the same subjects,
found weight to be a significant predictor of total extracted gDNA
(β = 0.25, SE = 0.08, t (23) = 2.92, p = 0.00768; Supplementary Figure 6A ) and the total extracted peak areas (β = 0.01, SE = 0.001, t (23) = 9.23, p = 3.36 × 10 –9 ; Supplementary Figure 6B ). Interestingly,
gDNA recovery appeared also to be significantly correlated with the
total extracted peak areas [β = 0.01, SE = 0.004, t (24.98) = 2.441, p = 0.0221] but with variation
in the slopes and intercepts between subjects ( C). Limitations This study focuses exclusively on human
fecal samples, and it has not been validated for other types of biosamples.
Data were acquired via reverse-phase liquid chromatography and positive
ionization mode. For this reason, nondetected features might display
a different behavior. Provided annotations are obtained via parent
ion m / z and MS/MS spectral matching,
resulting in a level 2 annotation according to the metabolomics standards
initiative. CANOPUS was used to generate
molecular class predictions of unknown MS/MS spectra; as such, these
predictions should be considered putative.
Unsupervised dimensionality
reduction via PCA of 20 mg of triplicates showed clear clustering
of subject fecal metabolic profiles using both 95% EtOH and 50% MeOH
extractions ( A). PERMANOVA identified subject id as the highest source of variance
in the data ( R 2 = 0.47, F = 10.98, p < 0.001), followed by extraction
method ( R 2 = 0.14, F =
6.68, p < 0.001). Out of a total of 4616 unique
metabolic features, 75% (3463) were recovered by both extraction methods
( B). These
included 94% (235) of the total annotated features via the GNPS library.
Interestingly, the Matrix Method (95% EtOH) appeared to capture a
higher number of additional features (849) compared to that of the
MeOH extraction (304). These extraction-specific features were classified
as small peptides and fatty acid conjugates for MeOH and oligopeptides,
glycerolipids, and fatty amides for EtOH according to CANOPUS NPC
superclass predictions ( Supplementary Table 1 ). Pairwise PCA and PLS-DA models were generated for each extraction
method to identify the metabolic features responsible for subject
discrimination. These features were then selected to compare both
pipelines. All PCA and PLS-DA score plots displayed clear clustering
by subject ( Supplementary Figure 1 ). All
PLS-DA models obtained a classification error rate of 0, indicating
a perfect discriminating performance. Extracted VIP scores from the
PLS-DA models displayed significant correlation between the two extraction
pipelines ( Supplementary Figure 2 ), with
the highest concordance as observed in the HDR plots ( C). On average, the PLS-DA
models identified 1764 features with VIPs > 1 involved in subject
discrimination. Notably, when considering exclusively the top 100
features obtained in the MeOH models, 86, 85, and 92% were also recovered
and classified as significant by the 95% EtOH models. Features of
interest identified by the multivariate analysis were also investigated
via univariate analysis. A high degree of correlation ( R > 0.9) was observed for the feature Log2FC between subjects obtained
via the two different extraction pipelines ( D). Focusing on annotated molecules of interest
in microbiome research, such as primary and secondary bile acids,
indole amino acids, vitamin B, heme catabolism, microbial N -acyl lipids, and long chain fatty acids, the 95% EtOH
extraction recovers and returns results comparable to those obtained
via 50% MeOH extraction. Recovery examples of tri- and dihydroxylated
bile acids, lysine, and histidine microbially conjugated bile acids,
indoles such as tryptophan and propionic acid, vitamin B5, stercobilin,
arginine-C5:0, and ω-3 arachidonic acid are available in Supplementary Figure 3 .
PCA of triplicates exclusively extracted via
the 95% EtOH pipeline showed clear clustering of samples based on
subject ID (PERMANOVA, R 2 = 0.50, F = 15.24, p < 0.001) and a smaller
effect of sample storage (PERMANOVA, R 2 = 0.08, F = 2.56, p < 0.001),
which included immediate processing or storage at room temperature
for either a day or a week in 95% EtOH ( A). Out of the 4566 features obtained via
EtOH extraction, 88% (3996) were captured at all time points, comprising
94% (231) of all annotated features ( B). The majority of the features distinctively
characterizing immediate (imm), week, immediate and day, and immediate
and week extractions (156 out of 570) were not classified by CANOPUS
( Supplementary Table 2 ). The subsequent
most affected predicted classes were small peptides (65) and oligopeptides
(48). Additionally, to determine if the storage time at room temperature
could affect the metabolic features responsible for subject classification,
27 subject pairwise PLS-DA models were generated comparing samples
of each subject at each different time point. Comparisons between
immediate extractions were considered as “ground truths,”
and the relative recovery of those significant features was investigated
for the different time points ( C). On average, 95.6, 91.6, and 94.7% of the top 100
“ground truth” features were recovered for each pairwise
subject comparison. This suggested that the fecal metabolome could
remain relatively stable in 95% EtOH for up to 1 week at room temperature
without losing predictive power for subject discrimination. Log2FC
in relation to storage time (immediate vs 1 week) of the metabolic
were also explored for each subject ( Supplementary Figure 4A ). Interestingly, the median fold changes of the discriminating
features of interest identified via the pairwise PLS-DA models were
0.705 (Subject A vs Subject B), 0.430 (Subject B vs Subject C), and
0.530 (Subject A vs Subject C), suggesting a relative stability during
storage time of the features of interest ( Supplementary Figure 4B ).
A recurrent issue in off-site fecal sample collection
is the uneven sample amount collected by the individuals and the variability
in the fecal water content. These can affect the per sample metabolite
recovery and introduce bias into the downstream data analysis. Although
lyophilization and weighting of the fecal samples can be implemented
before data acquisition, these are time-consuming and unpractical
when analyzing thousands of samples, and several technical errors
can be introduced, such as repeated freeze–thaw cycles, intersample
contamination, and others. To overcome these limitations, post data
acquisition normalization methods can be deployed. Here, we investigate
the use of RCLR transformation in fecal untargeted metabolomics data
as a method to improve accuracy in data analysis. Originally introduced
to tackle the microbiome data compositionality, the RCLR transformation suits the semicompositionality
nature of fecal untargeted metabolomics data. Additionally, RCLR allows
for easy interpretability and seamless multiomics integration with
associated RCLR-transformed microbiome data using tools such as Joint-RPCA and DIABLO. We
investigated the extraction of different sample weights, 10, 20, and
30 mg, in both 95% EtOH and 50% MeOH extraction pipelines. PCA of
RCLR data shows no effect of sample weight in both extraction methods
(PERMANOVAs, R 2 < 0.05, p > 0.08), with most of the variance explained by subject id ( R 2 = ∼60%) and no difference in the dispersion
of the samples (PERMDISP2, p > 0.45) between the
different subjects ( A). On the contrary, PCA of raw data or relative abundance
transformed data displayed less tightened clusters based on subjects,
significant differences between group variances (PERMDISP2, p < 0.001), and small but still significant effect of
the sample weight ( Supplementary Figure 5 ). Upset plot revealed that the number of detected metabolites was
not affected by sample weight, suggesting that just 10 mg is sufficient
to cover the detectable fecal metabolome in reverse-phase LC-MS/MS
data acquired in the positive ionization mode ( B). Finally, we explored correlations between
weight, total obtained gDNA (genomic DNA, ng/μL), and cumulative
extracted peak areas from samples processed via the Matrix Method
(95% EtOH) and stored for 1 week at room temperature. Linear mixed
effect models, accounting for repeated measures on the same subjects,
found weight to be a significant predictor of total extracted gDNA
(β = 0.25, SE = 0.08, t (23) = 2.92, p = 0.00768; Supplementary Figure 6A ) and the total extracted peak areas (β = 0.01, SE = 0.001, t (23) = 9.23, p = 3.36 × 10 –9 ; Supplementary Figure 6B ). Interestingly,
gDNA recovery appeared also to be significantly correlated with the
total extracted peak areas [β = 0.01, SE = 0.004, t (24.98) = 2.441, p = 0.0221] but with variation
in the slopes and intercepts between subjects ( C).
This study focuses exclusively on human
fecal samples, and it has not been validated for other types of biosamples.
Data were acquired via reverse-phase liquid chromatography and positive
ionization mode. For this reason, nondetected features might display
a different behavior. Provided annotations are obtained via parent
ion m / z and MS/MS spectral matching,
resulting in a level 2 annotation according to the metabolomics standards
initiative. CANOPUS was used to generate
molecular class predictions of unknown MS/MS spectra; as such, these
predictions should be considered putative.
The
presented study highlights that the
recently introduced Matrix
Method, which implements a 95% EtOH extraction, performs as well as
a commonly used 50% MeOH extraction method for metabolomics analysis
of human fecal samples. Moreover, the use of EtOH is safer compared
to MeOH, and it can be handled by nonscientific personnel in the case
of bulk off-site sample collections. Additionally, we showcase how
the fecal metabolome remains relatively stable when stored in 95%
EtOH for up to 1 week at room temperature, maintaining discriminating
power between the investigated samples. This is important as samples
cannot always immediately be stored at −80 °C, the most
ideal condition, and refrigerated shipping can represent a high economic
burden. Finally, we highlight how data analysis via RCLR transformation
helps to remove variance possibly introduced by uneven sampling and
discuss how this transformation suits better integrative microbiome
studies. In conclusion, the 95% EtOH extraction of the Matrix Method
represents a valid and more economically alternative to other widely
used 50% MeOH extractions.
|
The impact of preoperative handgrip strength on postoperative outcomes following transforaminal lumbar interbody fusion | 34fb73b4-cd49-4f2a-b590-44a61837789a | 11951603 | Surgery[mh] | As lifespans lengthen, a rise in age-related spinal disorders, particularly in the lumbar region, presents a growing challenge . These conditions are often attributed to the degeneration and weakening of bones, discs, and surrounding soft tissues . Transforaminal lumbar interbody fusion (TLIF) surgery is a well-established and effective treatment option for various lumbar spinal issues . TLIF is a spinal fusion procedure that can be performed minimally invasively, involving the removal of the intervertebral disc followed by the insertion of an implant to stabilize the spine [ – ]. This process aims to foster bone growth (osteogenesis) and eventually fuse two or more vertebrae, with the goal of reducing pain and enhancing functionality [ , , ]. Handgrip strength (HGS), a convenient measure of voluntary muscle function, has become recognized as a significant biomarker of our health . Its ease of use, speed, low cost, and simplicity make it a valuable tool. Increasingly, research underscores the strong predictive power of HGS for assessing nutritional status and sarcopenia (muscle loss) . While sarcopenia, as defined by the Asian Working Group for Sarcopenia (AWGS) 2019 , is diagnosed based on a combination of low muscle strength, low muscle quantity/quality, and low physical performance, its assessment may pose challenges in patients with lumbar spine degeneration. Specifically, physical performance metrics, such as gait speed or timed-up-and-go tests, could be influenced by preexisting spinal pathology, potentially introducing bias into the evaluation process. In contrast, HGS offers a practical and accessible alternative, as it directly measures muscle function without being significantly affected by lumbar spine degeneration. This study, therefore, focuses on HGS as a straightforward, reliable indicator to investigate its association with postoperative outcomes following TLIF. Although previous studies have demonstrated a correlation between HGS and overall outcomes following spine surgery [ – ], the predictive value of baseline HGS on specific postoperative outcomes in patients undergoing TLIF remains unclear. This study aims to determine whether preoperative HGS is associated with postoperative functional outcomes, quality of life, and independence in activities of daily living at one-year post-TLIF. To account for the longitudinal nature of the data and potential associations among repeated measurements, we will employ a generalized estimating equation (GEE) model . We hypothesize that higher baseline HGS will be associated with more favorable outcomes.
Study design and setting This prospective observational study recruited patients from a single hospital between January 2020 and June 2023. A total of 103 consecutive patients were scheduled for TLIF surgery based on surgical indications and contraindications described in previous studies . Patients aged 18 years or older with TLIF indications such as lumbar spinal stenosis, lumbar disc herniation, or low-grade lumbar spondylolisthesis (Meyerding I or II) were included. Diagnosis was confirmed using standing radiographs and MRI, and all patients must have experienced lower back and radiating pain without improvement after at least three months of conservative treatment. Exclusion criteria comprised patients with previous lumbar spine surgery or revision, high-grade spondylolisthesis (Meyerding > II), cervical stenosis (tandem stenosis), degenerative scoliosis, and those diagnosed or suspected to have underlying or ongoing diseases such as spondylodiscitis, ankylosing spondylitis, spinal neoplasm, spinal metastasis, and traumatic spine injury. Additionally, individuals diagnosed with cognitive or psychological disorders were excluded. The presence of cervical stenosis was specifically assessed through MRI to prevent any potential confounding effect on handgrip strength measurements. The study adhered to the Declaration of Helsinki and followed the STROBE guidelines for observational studies, with approval from our institute’s Ethics Committee. Data on basic demographics and health metrics were collected, including age, sex, body mass index (BMI), Charlson comorbidity index (CCI), HGS, bone mineral density (T-score), American Society of Anesthesiologists (ASA) physical status classification, fusion levels, and total intraoperative blood loss. HGS measurement and group allocation Preoperative maximum handgrip strength was assessed in each patient using a Jamar Hydraulic Dynamometer (Sammons Preston, USA), following standardized testing protocols. Patients were seated comfortably in a chair or bed, ensuring their feet were flat on the ground for stability. The test was conducted with the elbow flexed at 90 degrees, the shoulder adducted, and the forearm in a neutral position (mid-pronation). The wrist was kept in a neutral position without extension or flexion to prevent bias from wrist positioning. Each patient was given clear verbal instructions and a demonstration before performing the test. They were instructed to squeeze the dynamometer as hard as possible for 3–5 seconds, ensuring maximum effort. Three consecutive trials were performed for each hand, with a 30-second rest period between trials to minimize muscle fatigue . The highest recorded value across the three attempts was used for analysis, as recommended by the Asian Working Group for Sarcopenia (AWGS) 2019 guidelines . To ensure consistency and accuracy, all assessments were performed by a trained examiner using the same dynamometer throughout the study. Patients were encouraged with standardized verbal prompts, such as “Squeeze as hard as you can!” to maintain motivation and effort. Any discomfort or pain was noted, but patients experiencing acute hand pain or neuromuscular disorders affecting grip strength were excluded from the study. The final HGS values were categorized based on AWGS criteria, defining low HGS as < 28 kg for males and < 18 kg for females. Operative techniques All surgeries were performed by a single surgeon from the author’s group, using the TLIF techniques. Patients were positioned prone for TLIF procedures. A paramedian incision was utilized for TLIF. Pedicle screws were inserted bilaterally in this procedure. In TLIF, the facet joint was removed to access the disc space, followed by a partial discectomy. All patients received artificial bone grafts using Rafugen™ DBM (Cellumed Co., Ltd., Seoul, Korea) and CeraMatrix bone graft substitute (Xelite Biomed Ltd., Taiwan). The bone graft was packed into the interbody space along with an interbody cage in all patients. Compression of bone grafts, screw head tightening, and placement of a negative pressure drain were performed before wound closure. Postoperatively, all patients received antibiotics, painkillers, and neurotrophic drugs. Drainage tubes were removed at 48 h based on clinical assessment. Patients initiated brace use three days postoperatively for a duration of three months. Strenuous physical activity was restricted throughout this period . Outcome assessment Clinical outcomes, including the Japanese Orthopaedic Association (JOA) score and quality of life, were assessed by an independent, blinded investigator at baseline and 3, 6, and 12 months postoperatively. The JOA score is a validated measure of functional outcome following TLIF . It assesses subjective symptoms (9 points), clinical signs (6 points), and limitations in daily activities (14 points), yielding a total score of 29. A higher JOA score correlates with improved function and reduced pain. Quality of life was assessed using the EuroQol 5-Dimensions 3-Level version (EQ-5D-3L) questionnaire with quality weights estimated for Taiwan . Higher EQ-5D-3L scores represent a better quality of life. The Barthel index is a widely used assessment tool to measure independence in activities of daily living . It quantifies a patient’s ability to perform self-care tasks, such as feeding, bathing, dressing, toileting, transferring, continence, mobility, and stair climbing. A higher Barthel index score indicates greater independence and a lower level of disability. Statistical analysis A sample size of 48 participants was determined to be necessary for this study. This calculation was based on a correlation coefficient formula: [12pt]{minimal}
$$\:N=\:{(_{\:}+\:{Z}_{\:}}{C})}^{2}+3$$ , targeting a statistical power of 0.8, a type I error rate of 0.05, and a two-tailed test. The expected correlation coefficient of 0.395, as reported by Kwon O. et al. , was utilized to estimate the required sample size. Statistical analysis was conducted using SPSS (version 30; IBM, Armonk, NY, USA). Descriptive statistics were used, including sample size or frequency (n) with percentages (%) for categorical variables and means ± standard deviations (SD) for continuous variables. Comparisons between groups classified according to HGS values were conducted using appropriate statistical tests. Categorical variables were analyzed using the chi-square or Fisher’s exact test of independence. Continuous variables were analyzed using the Student’s t-test for normally distributed data or the Mann-Whitney U test for non-normally distributed data. A Wilcoxon signed-rank test was conducted to compare the differences between JOA, EQ-5D-3L, and Barthel index assessment time points compared with the baseline. The GEE model was used to assess the effects of various factors on outcomes, accounting for repeated measures within participants over time and providing a robust approach to handling missing data . Sensitivity analyses were conducted to assess the findings’ robustness by modifying the GEE model’s correlation structures and by sequentially adding or removing covariates (e.g., BMI, gender, and CCI) to evaluate their impact on the associations between HGS and postoperative outcomes. Two-sided p -values of < 0.05 were considered statistically significant for all tests.
This prospective observational study recruited patients from a single hospital between January 2020 and June 2023. A total of 103 consecutive patients were scheduled for TLIF surgery based on surgical indications and contraindications described in previous studies . Patients aged 18 years or older with TLIF indications such as lumbar spinal stenosis, lumbar disc herniation, or low-grade lumbar spondylolisthesis (Meyerding I or II) were included. Diagnosis was confirmed using standing radiographs and MRI, and all patients must have experienced lower back and radiating pain without improvement after at least three months of conservative treatment. Exclusion criteria comprised patients with previous lumbar spine surgery or revision, high-grade spondylolisthesis (Meyerding > II), cervical stenosis (tandem stenosis), degenerative scoliosis, and those diagnosed or suspected to have underlying or ongoing diseases such as spondylodiscitis, ankylosing spondylitis, spinal neoplasm, spinal metastasis, and traumatic spine injury. Additionally, individuals diagnosed with cognitive or psychological disorders were excluded. The presence of cervical stenosis was specifically assessed through MRI to prevent any potential confounding effect on handgrip strength measurements. The study adhered to the Declaration of Helsinki and followed the STROBE guidelines for observational studies, with approval from our institute’s Ethics Committee. Data on basic demographics and health metrics were collected, including age, sex, body mass index (BMI), Charlson comorbidity index (CCI), HGS, bone mineral density (T-score), American Society of Anesthesiologists (ASA) physical status classification, fusion levels, and total intraoperative blood loss.
Preoperative maximum handgrip strength was assessed in each patient using a Jamar Hydraulic Dynamometer (Sammons Preston, USA), following standardized testing protocols. Patients were seated comfortably in a chair or bed, ensuring their feet were flat on the ground for stability. The test was conducted with the elbow flexed at 90 degrees, the shoulder adducted, and the forearm in a neutral position (mid-pronation). The wrist was kept in a neutral position without extension or flexion to prevent bias from wrist positioning. Each patient was given clear verbal instructions and a demonstration before performing the test. They were instructed to squeeze the dynamometer as hard as possible for 3–5 seconds, ensuring maximum effort. Three consecutive trials were performed for each hand, with a 30-second rest period between trials to minimize muscle fatigue . The highest recorded value across the three attempts was used for analysis, as recommended by the Asian Working Group for Sarcopenia (AWGS) 2019 guidelines . To ensure consistency and accuracy, all assessments were performed by a trained examiner using the same dynamometer throughout the study. Patients were encouraged with standardized verbal prompts, such as “Squeeze as hard as you can!” to maintain motivation and effort. Any discomfort or pain was noted, but patients experiencing acute hand pain or neuromuscular disorders affecting grip strength were excluded from the study. The final HGS values were categorized based on AWGS criteria, defining low HGS as < 28 kg for males and < 18 kg for females.
All surgeries were performed by a single surgeon from the author’s group, using the TLIF techniques. Patients were positioned prone for TLIF procedures. A paramedian incision was utilized for TLIF. Pedicle screws were inserted bilaterally in this procedure. In TLIF, the facet joint was removed to access the disc space, followed by a partial discectomy. All patients received artificial bone grafts using Rafugen™ DBM (Cellumed Co., Ltd., Seoul, Korea) and CeraMatrix bone graft substitute (Xelite Biomed Ltd., Taiwan). The bone graft was packed into the interbody space along with an interbody cage in all patients. Compression of bone grafts, screw head tightening, and placement of a negative pressure drain were performed before wound closure. Postoperatively, all patients received antibiotics, painkillers, and neurotrophic drugs. Drainage tubes were removed at 48 h based on clinical assessment. Patients initiated brace use three days postoperatively for a duration of three months. Strenuous physical activity was restricted throughout this period .
Clinical outcomes, including the Japanese Orthopaedic Association (JOA) score and quality of life, were assessed by an independent, blinded investigator at baseline and 3, 6, and 12 months postoperatively. The JOA score is a validated measure of functional outcome following TLIF . It assesses subjective symptoms (9 points), clinical signs (6 points), and limitations in daily activities (14 points), yielding a total score of 29. A higher JOA score correlates with improved function and reduced pain. Quality of life was assessed using the EuroQol 5-Dimensions 3-Level version (EQ-5D-3L) questionnaire with quality weights estimated for Taiwan . Higher EQ-5D-3L scores represent a better quality of life. The Barthel index is a widely used assessment tool to measure independence in activities of daily living . It quantifies a patient’s ability to perform self-care tasks, such as feeding, bathing, dressing, toileting, transferring, continence, mobility, and stair climbing. A higher Barthel index score indicates greater independence and a lower level of disability.
A sample size of 48 participants was determined to be necessary for this study. This calculation was based on a correlation coefficient formula: [12pt]{minimal}
$$\:N=\:{(_{\:}+\:{Z}_{\:}}{C})}^{2}+3$$ , targeting a statistical power of 0.8, a type I error rate of 0.05, and a two-tailed test. The expected correlation coefficient of 0.395, as reported by Kwon O. et al. , was utilized to estimate the required sample size. Statistical analysis was conducted using SPSS (version 30; IBM, Armonk, NY, USA). Descriptive statistics were used, including sample size or frequency (n) with percentages (%) for categorical variables and means ± standard deviations (SD) for continuous variables. Comparisons between groups classified according to HGS values were conducted using appropriate statistical tests. Categorical variables were analyzed using the chi-square or Fisher’s exact test of independence. Continuous variables were analyzed using the Student’s t-test for normally distributed data or the Mann-Whitney U test for non-normally distributed data. A Wilcoxon signed-rank test was conducted to compare the differences between JOA, EQ-5D-3L, and Barthel index assessment time points compared with the baseline. The GEE model was used to assess the effects of various factors on outcomes, accounting for repeated measures within participants over time and providing a robust approach to handling missing data . Sensitivity analyses were conducted to assess the findings’ robustness by modifying the GEE model’s correlation structures and by sequentially adding or removing covariates (e.g., BMI, gender, and CCI) to evaluate their impact on the associations between HGS and postoperative outcomes. Two-sided p -values of < 0.05 were considered statistically significant for all tests.
Study population selection and patient demographics Out of the 103 patients who underwent TLIF, 89 were enrolled for follow-up. By the 3-month mark, 88 patients completed the follow-up, with one lost to follow-up for unknown reasons. Between 3 and 6 months, an additional two patients were lost to follow-up, and another two passed away, reducing the number of evaluable patients to 84 at 6 months. Unfortunately, by the final 1-year follow-up, the number of patients had further declined, with only 74 completing the evaluation due to the loss of nine more patients and two additional deaths (Fig. ). Of the 89 patients, 46 exhibited normal HGS, and 43 presented with low HGS. The average age of the normal HGS group (65.78 ± 9.80) was significantly younger than that of the low HGS group (72.07 ± 8.40 years, t = -3.239, p = 0.002). The sex distribution between the two HGS groups was not statistically significant ( p = 0.135). The normal HGS group had 45.7% males and 54.3% females, while the low HGS group had 30.2% males and 69.8% females. In addition, patients with normal HGS had significantly higher BMI than those with low HGS (Z = -2.681, p = 0.007). Furthermore, a low T-score was associated with low HGS (Z = -2.290, p = 0.022). Other variables, including the CCI, incidence of osteoporosis, spondylolisthesis, number of fusion levels, ASA classification, surgical time, and total intraoperative blood loss, were detailed in Table . However, no significant differences were found between the normal and low HGS groups for these factors. Surgical outcome analysis JOA scores significantly improved in both HGS groups at 3, 6, and 12 months postoperatively compared to baseline ( p < 0.001, Fig. ). Notably, the normal HGS group consistently demonstrated significantly higher JOA scores than the low HGS group at all time points ( p = 0.036, p = 0.006, p < 0.001, p = 0.001 for baseline, 3 months, 6 months, and 12 months postoperation, respectively). Regarding EQ-5D-3L scores, significant improvements were observed only in the normal HGS group over the 12 months postoperatively ( p = 0.015, p = 0.007, p = 0.003 for 3 months, 6 months, and 12 months, respectively). Additionally, at 3 and 6 months, the normal HGS group showed a significant increase compared to the low HGS group, with p -values of 0.024 and 0.003, respectively. The Barthel Index revealed a statistically significant improvement in the normal HGS group after 12 months ( p = 0.035), while the low HGS group experienced a significant decline at all follow-up time points ( p = 0.001, p = 0.004, p = 0.017 for 3 months, 6 months, and 12 months, respectively). Furthermore, statistically significant differences were noted between the normal and low HGS groups at 3 months ( p = 0.001), 6 months ( p = 0.001), and 12 months ( p = 0.006). Associations of baseline parameters with outcome measurements To examine the associations between the parameters listed in Table and the outcome variables over time, we employed a GEE model. The results of this analysis are presented in Table . JOA scores Results indicated that male sex (β = 11.840, p < 0.001), increased BMI (β = 5.044, p < 0.001), and lower CCI (β = -1.104, p = 0.002) were significantly linked with higher JOA scores. Conversely, lower HGS (β = -2.551, p = 0.008) was associated with lower JOA scores. Additionally, JOA scores demonstrated significant improvement over time compared to baseline at 3 months (β = 2.739, p < 0.001), 6 months (β = 3.753, p < 0.001), and 12 months (β = 4.536, p < 0.001). EQ-5D-3L scores Consistent with JOA findings, EQ-5D-3L scores were significantly positively connected with male sex (β = 0.344, p = 0.041), BMI (β = -0.161, p = 0.019), and normal HGS group (β = 0.142, p = 0.007). Notably, underweight individuals exhibited significantly higher EQ-5D-3L scores compared to obese individuals (β = 0.246, p = 0.022). Postoperative EQ-5D-3L scores were significantly elevated at all time points relative to baseline: 3 months (β = 0.060, p = 0.018), 6 months (β = 0.072, p = 0.004), and 12 months (β = 0.086, p = 0.003). Barthel index A lower Barthel Index was exclusively associated with patients in the low HGS group at 12-month follow-up (β = -5.703, p = 0.037). Additionally, male sex (β = 17.026, p = 0.046) and higher BMI (β = 7.240, p = 0.036) were identified as significant predictors of better functional independence outcomes. In contrast to JOA and EQ-5D-3L, Barthel Index experienced a significant decline after 3 months compared to baseline (β = -4.720, p = 0.020).
Out of the 103 patients who underwent TLIF, 89 were enrolled for follow-up. By the 3-month mark, 88 patients completed the follow-up, with one lost to follow-up for unknown reasons. Between 3 and 6 months, an additional two patients were lost to follow-up, and another two passed away, reducing the number of evaluable patients to 84 at 6 months. Unfortunately, by the final 1-year follow-up, the number of patients had further declined, with only 74 completing the evaluation due to the loss of nine more patients and two additional deaths (Fig. ). Of the 89 patients, 46 exhibited normal HGS, and 43 presented with low HGS. The average age of the normal HGS group (65.78 ± 9.80) was significantly younger than that of the low HGS group (72.07 ± 8.40 years, t = -3.239, p = 0.002). The sex distribution between the two HGS groups was not statistically significant ( p = 0.135). The normal HGS group had 45.7% males and 54.3% females, while the low HGS group had 30.2% males and 69.8% females. In addition, patients with normal HGS had significantly higher BMI than those with low HGS (Z = -2.681, p = 0.007). Furthermore, a low T-score was associated with low HGS (Z = -2.290, p = 0.022). Other variables, including the CCI, incidence of osteoporosis, spondylolisthesis, number of fusion levels, ASA classification, surgical time, and total intraoperative blood loss, were detailed in Table . However, no significant differences were found between the normal and low HGS groups for these factors.
JOA scores significantly improved in both HGS groups at 3, 6, and 12 months postoperatively compared to baseline ( p < 0.001, Fig. ). Notably, the normal HGS group consistently demonstrated significantly higher JOA scores than the low HGS group at all time points ( p = 0.036, p = 0.006, p < 0.001, p = 0.001 for baseline, 3 months, 6 months, and 12 months postoperation, respectively). Regarding EQ-5D-3L scores, significant improvements were observed only in the normal HGS group over the 12 months postoperatively ( p = 0.015, p = 0.007, p = 0.003 for 3 months, 6 months, and 12 months, respectively). Additionally, at 3 and 6 months, the normal HGS group showed a significant increase compared to the low HGS group, with p -values of 0.024 and 0.003, respectively. The Barthel Index revealed a statistically significant improvement in the normal HGS group after 12 months ( p = 0.035), while the low HGS group experienced a significant decline at all follow-up time points ( p = 0.001, p = 0.004, p = 0.017 for 3 months, 6 months, and 12 months, respectively). Furthermore, statistically significant differences were noted between the normal and low HGS groups at 3 months ( p = 0.001), 6 months ( p = 0.001), and 12 months ( p = 0.006).
To examine the associations between the parameters listed in Table and the outcome variables over time, we employed a GEE model. The results of this analysis are presented in Table .
Results indicated that male sex (β = 11.840, p < 0.001), increased BMI (β = 5.044, p < 0.001), and lower CCI (β = -1.104, p = 0.002) were significantly linked with higher JOA scores. Conversely, lower HGS (β = -2.551, p = 0.008) was associated with lower JOA scores. Additionally, JOA scores demonstrated significant improvement over time compared to baseline at 3 months (β = 2.739, p < 0.001), 6 months (β = 3.753, p < 0.001), and 12 months (β = 4.536, p < 0.001).
Consistent with JOA findings, EQ-5D-3L scores were significantly positively connected with male sex (β = 0.344, p = 0.041), BMI (β = -0.161, p = 0.019), and normal HGS group (β = 0.142, p = 0.007). Notably, underweight individuals exhibited significantly higher EQ-5D-3L scores compared to obese individuals (β = 0.246, p = 0.022). Postoperative EQ-5D-3L scores were significantly elevated at all time points relative to baseline: 3 months (β = 0.060, p = 0.018), 6 months (β = 0.072, p = 0.004), and 12 months (β = 0.086, p = 0.003).
A lower Barthel Index was exclusively associated with patients in the low HGS group at 12-month follow-up (β = -5.703, p = 0.037). Additionally, male sex (β = 17.026, p = 0.046) and higher BMI (β = 7.240, p = 0.036) were identified as significant predictors of better functional independence outcomes. In contrast to JOA and EQ-5D-3L, Barthel Index experienced a significant decline after 3 months compared to baseline (β = -4.720, p = 0.020).
Effective treatment outcomes require a thorough understanding and management of associated risks. Our study identified sex, BMI, CCI, and particularly HGS as significant predictors of postoperative outcomes, as measured by the JOA score, EQ-5D-3L, and Barthel index. Our study demonstrates that the Transforaminal Lumbar Interbody Fusion (TLIF) procedure consistently yields significant improvements in functional recovery and quality of life. This is evidenced by substantial enhancements in Japanese Orthopaedic Association (JOA), EuroQol five-dimension three-level (EQ-5D-3L), and Barthel Index scores over time, reinforcing the well-established effectiveness of TLIF in managing lumbar spine disorders through pain alleviation and enhanced functional capacity. Furthermore, postoperative complications were minimal, with only a single patient experiencing a superficial surgical site infection that fully resolved within a week. Notably, advancements in TLIF techniques, such as endoscopic approaches and refined safe operating zone identification, have further contributed to positive patient outcomes and accelerated recovery [ , – ]. These findings collectively underscore TLIF as a reliable surgical intervention for optimizing patient outcomes, particularly when coupled with meticulous preoperative assessments and structured postoperative care. HGS is a critical component in sarcopenia assessment . Our analysis demonstrated a significant association of low baseline HGS with older age, lower BMI, and reduced bone mineral density (Table ). Notably, while not statistically significant, the female sex ratio was twice as high in the HGS group compared to males, suggesting a potential sex-related influence on HGS. This aligns with previous research demonstrating that although age universally impacts muscle structure and function, females tend to exhibit a higher sarcopenia prevalence at earlier ages than males [ – ], often accompanied by osteoporosis . By contrast, the link between BMI and HGS in the elderly is debated ; This ambiguity persists in the context of lumbar spine surgery. While most studies in this area suggest a nonsignificant trend towards higher BMI in individuals with low HGS [ , , ], contradictory findings, such as those reported by F. Shen , underscore the complex relationship between these variables. Further investigation is warranted to clarify the interplay between BMI and HGS in this population. HGS is a recognized predictor of outcomes after various types of surgery [ – ], including lumbar spine surgery. Previous research consistently links low HGS to poorer rehabilitation outcomes, often assessed using the Oswestry disability index (ODI) and EQ-5D [ – ]. To comprehensively assess the impact on daily life, we employed the JOA index, a well-established measure highly correlated with the ODI , in conjunction with the EQ-5D and Barthel indices. Our findings reveal significantly greater improvements in all three indices among patients with normal HGS compared to those with low HGS, aligning with previous research. Furthermore, we demonstrated a significant association between low HGS and decreased postoperative outcomes using the GEE model. This finding is consistent with the link between low HGS and various adverse health conditions, such as sarcopenia, poor bone quality, and frailty , which can hinder recovery and treatment efficacy. Our results suggest that HGS is valuable for assessing preoperative functional status and predicting lumbar interbody fusion surgery outcomes. Our analysis revealed that female sex is a significant predictor of poor prognosis following TLIF surgery, a finding consistent with numerous previous studies, including a systematic review [ – ]. In addition to its association with lower JOA and EQ-5D-3L scores, female sex was also significantly correlated with poorer Barthel Index scores at 12 months (β = 17.026, p = 0.046), suggesting that men had a greater likelihood of maintaining postoperative functional independence. Researchers have suggested that estrogen deficiency during menopause, resulting in decreased bone quality, is a primary factor contributing to this condition. Additionally, studies have indicated that women generally have a lower pain tolerance than men , which may affect the pain-related scores which may influence pain-related scores, which are one of the main criteria of the JOA scale. Obesity has been linked to increased postoperative complications in spine surgery [ – ], yet its impact on functional outcomes remains controversial . While previous meta-analyses have not identified a consistent association between obesity and functional scores , our study suggests a potential positive connection between BMI and postoperative JOA, EQ-5D-3L, and Barthel Index scores. However, a closer examination of BMI subgroups, we did not observe statistically significant associations with these outcomes. In addition, although there was an increase in EQ-5D-3L scores compared with the obesity group, the limited number of patients in the underweight group ( n = 2) underpowered this finding. Larger studies are needed to clarify this finding. CCI is a well-established predictor of postoperative JOA improvement rate , complications, reoperations, and mortality [ – ] in spine surgery. Our findings corroborate previous research by demonstrating a significant inverse relationship between CCI and both JOA and EQ-5D scores. These results underscore the critical role of comprehensive comorbidity assessment and management in optimizing patient outcomes following spine surgery. We acknowledge several limitations in this study. The research was conducted at a single center and involved a single surgeon during the COVID-19 pandemic, which may have restricted the sample size and may not represent the diversity of patient populations and surgical practices across different settings. Additionally, approximately 17% of patients were lost to follow-up at 12 months postoperatively for unknown reasons, potentially biasing our final results. Due to significant sex differences in HGS, continuous data analysis was not feasible, leading us to use AWGS criteria for sarcopenia classification– a method that may not be optimal for our population. Moreover, while AWGS 2019 provides widely accepted cut-off values for HGS, these thresholds are fixed and do not account for age-related variations. Future research should consider age-specific thresholds to better assess sarcopenia risk across different age groups. Furthermore, our focus on preoperative HGS alone precludes an assessment of whether postoperative changes in HGS could predict recovery outcomes. Further research involving larger, multicenter studies is necessary to confirm these findings, explore potential interventions to improve outcomes for patients with low HGS, and establish more precise cut-off values for different subgroups, including male and female populations.
This study demonstrated that preoperative HGS is a significant predictor of postoperative functional outcomes following TLIF surgery. Patients with normal HGS exhibited superior improvements in JOA, EQ-5D-3L, and Barthel Index scores compared to those with low HGS. These findings highlight the importance of preoperative HGS assessment in patient selection and management for TLIF surgery.
Below is the link to the electronic supplementary material. Supplementary Material 1
|
Advanced Hysteroscopic Surgery Training | da4a71d1-149b-4536-a52e-6d9a30bc3264 | 4216174 | Gynaecology[mh] | Hysteroscopic surgery is pivotal in the management of many gynecologic pathologies. The skills required to perform advanced hysteroscopic surgery (AHS)—for example, transcervical hysteroscopic endometrial resection (TCRE), hysteroscopic polypectomy, and myomectomy in the management of menorrhagia ; hysteroscopic septolysis in patients with fertility-related gynecologic problems ; hysteroscopic removal of chronically retained products of conception (placenta accreta) ; and excision of intramural ectopic pregnancy —ought to be practiced by contemporary gynecologic surgeons in their day-to-day clinical practice. AHS is a minimally invasive procedure that preserves the uterus in most cases. We suggest a logarithm of training in workshops, including virtual reality (VR), before embarking on operations in the operating room ( ).
Good training is conducive to sound clinical practice; this is particularly true in AHS because the margin for error is rather narrow. There is a learning curve; the operating time decreases as one goes through the learning curve so that, ultimately, the gynecologic surgeon will grasp knowledge, manual dexterity, and training to enable him or her to perform AHS competently in both emergency and elective settings to the benefit of the patient.
The term registrar in Australia and United Kingdom is a synonym of resident in the United States, that is, a trainee in the obstetrics and gynecology integrated training program accredited by the Royal Australian and New Zealand College of Obstetricians and Gynaecologists. Many aspects of the integrated training program in Australia are shared by other countries in the Western world because there is a constant effort to internationalize and update the integrated professional training programs in Australia. Hysteroscopic operations are a primary component of gynecologic surgery in many teaching hospitals around the world. However, opportunities to perform advanced hysteroscopic procedures vary widely among gynecologic training programs, as well as trainees. , The required skills are not more difficult to acquire during registrar (residency) training than conventional surgical procedures. , It is the duty of the consultant (specialist) gynecologist to be in the operating room to personally monitor the progress of the operation and demonstrate leadership skills in the operating room in a teaching capacity. Failure to do so may enhance the risk of litigation if the patient sustains any intraoperative or postoperative complications. , Before contemplating AHS, the registrar has to be familiar with assembling the elements of the operative hysteroresectoscope; with the power sources, as well as their effects and limitations; and with the use of uterine distending fluids and their possible complications, in addition to being on the alert for any intraoperative or postoperative complications. Being aware of the registrar's own limits is important. Technically difficult operations such as division of intractable intrauterine synechiae (adhesions), excision of intramural myomata, and hysteroscopic excision of abnormally invasive placenta residuals should be carried out only by an experienced gynecologic endoscopist.
The main educational components are as follows: Anatomy of the female pelvis. Thorough knowledge, both theoretical and practical, regarding hysteroscopic use of energy sources, as well as their physics and different effects on human cells and tissues. Pathophysiology of disease and diagnoses including differential diagnoses, for example, recognizing myomata and adenomyosis as causes of menorrhagia and uterine enlargement; obtaining magnetic resonance imaging preoperatively to differentiate among the different pathologies ; and understanding that adenomyosis may decrease the amenorrhea rate after TCRE. Operative indications, contraindications, limitations, and possible complications of every AHS, together with full knowledge of the prevention, early recognition, and management of complications. The relative advantages and disadvantages of AHS as opposed to conventional laparotomy and vaginal approaches.
There are skills required to perform both simple hysteroscopic surgery and AHS. These skills are often acquired by attending 2- or 3-day courses that comprise didactic lectures and hands-on components. However, registrar training has at least 2 main problems pertaining to training and its adequacy or otherwise in training programs: First, there is often limited AHS teaching in operating rooms during residency and sometimes at the level of fellowship training. Second, there are no universally accepted standards to accredit registrars. A well-planned structured training program is probably superior to ad hoc opportunity. , Perusal of the aforementioned and other teaching principles in gynecologic endoscopy and related specialties will inevitably lead to the question, Which is the best program? This is a good question to which there is no easy answer. , One way around this problem is for each hospital's accreditation committee to set several realistic criteria of training and accreditation, in accordance with the broad recommendations of professional bodies, such as the Royal Australian and New Zealand College of Obstetricians and Gynaecologists, and recognized conclusions of risk management in national and international settings.
Advanced Level Satisfactory performance of AHS, such as combined hysteroscopic myomectomy and endometrial resection, hysteroscopic septolysis, and excision of pathologically adherent placenta (accrete), requires the expertise and manual dexterity of an experienced endoscopic surgeon. The logical sequence of events is that one has to crawl before he or she can walk; basic skills need to be performed well and practiced before one contemplates performing AHS. However, advanced hysteroscopic skills are required for most advanced procedures, and experience in a specific operation enhances the acquisition of skills necessary to perform others. Hence it is the combined experience in advanced procedures that should be emphasized during training rather than the excellent command of any one individual procedure. Hasson and Getzels suggested a workable and easy-to-implement system of credentialing physicians in AHS. Physicians are certified by the Accreditation Council for Gynecologic Endoscopy on receipt of completed documentation and proctorship. The system has stood the test of time and is practicable, and many units worldwide adopt the same. Where Training in operating theaters is a feasible option because it allows interaction with valuable one-to-one practical tuition in real-life situations. For example, a simple procedure may be chosen (eg, laparoscopic sterilization), and different aspects, alternatives (eg, hysteroscopic sterilization), indications, complications, and technical aspects of occlusive device application may all be tested. Trainees' learning curve for hysteroscopic tubal sterilization showed a shorter procedure time in the operating room. The procedure offered a high efficacy. It has been concluded that knowledge of a specific gynecologic endoscopic procedure can be measured and that a carefully calculated and structured learning package can be effective. Office hysteroscopy includes, but is not limited to, TCRE under local infiltration anesthesia and hysteroscopic sterilization. , Supervision and subjective assessment by the consultant endoscopist in the operating room have their perceived problems: There is a discrepancy between subjective in-training evaluations of surgical performance and an objective assessment using a simulation. In a busy gynecology unit, the registrar may be temporarily sleep deprived and this may be perceived to negatively affect his or her surgical endoscopy performance. However, short-term sleep deficits do not appear to hinder the acquisition of endoscopy skills, even after registrars have been on call the night before. Pregnancy during gynecologic registrar training does not seem to have a negative impact on the surgical experience, especially when the attending endoscopist is supportive of the registrar and her pregnancy status. With the gradual implementation of “safe working hours” of junior doctors in public hospitals in the Western world, our trainees have incurred a decrease in the number of gynecologic endoscopy operations they have performed. Morbidity and mortality rates are lowest when procedures are performed by physicians who perform the procedures frequently and in centers that have large volumes of these procedures. , The clinical and educational implications of these changes in working hours for registrars to gain experience need to be further elucidated. However, there is evidence that teaching registrars in the operating room is expensive in terms of operating time and financial cost. , What then is the alternative?
Satisfactory performance of AHS, such as combined hysteroscopic myomectomy and endometrial resection, hysteroscopic septolysis, and excision of pathologically adherent placenta (accrete), requires the expertise and manual dexterity of an experienced endoscopic surgeon. The logical sequence of events is that one has to crawl before he or she can walk; basic skills need to be performed well and practiced before one contemplates performing AHS. However, advanced hysteroscopic skills are required for most advanced procedures, and experience in a specific operation enhances the acquisition of skills necessary to perform others. Hence it is the combined experience in advanced procedures that should be emphasized during training rather than the excellent command of any one individual procedure. Hasson and Getzels suggested a workable and easy-to-implement system of credentialing physicians in AHS. Physicians are certified by the Accreditation Council for Gynecologic Endoscopy on receipt of completed documentation and proctorship. The system has stood the test of time and is practicable, and many units worldwide adopt the same.
Training in operating theaters is a feasible option because it allows interaction with valuable one-to-one practical tuition in real-life situations. For example, a simple procedure may be chosen (eg, laparoscopic sterilization), and different aspects, alternatives (eg, hysteroscopic sterilization), indications, complications, and technical aspects of occlusive device application may all be tested. Trainees' learning curve for hysteroscopic tubal sterilization showed a shorter procedure time in the operating room. The procedure offered a high efficacy. It has been concluded that knowledge of a specific gynecologic endoscopic procedure can be measured and that a carefully calculated and structured learning package can be effective. Office hysteroscopy includes, but is not limited to, TCRE under local infiltration anesthesia and hysteroscopic sterilization. , Supervision and subjective assessment by the consultant endoscopist in the operating room have their perceived problems: There is a discrepancy between subjective in-training evaluations of surgical performance and an objective assessment using a simulation. In a busy gynecology unit, the registrar may be temporarily sleep deprived and this may be perceived to negatively affect his or her surgical endoscopy performance. However, short-term sleep deficits do not appear to hinder the acquisition of endoscopy skills, even after registrars have been on call the night before. Pregnancy during gynecologic registrar training does not seem to have a negative impact on the surgical experience, especially when the attending endoscopist is supportive of the registrar and her pregnancy status. With the gradual implementation of “safe working hours” of junior doctors in public hospitals in the Western world, our trainees have incurred a decrease in the number of gynecologic endoscopy operations they have performed. Morbidity and mortality rates are lowest when procedures are performed by physicians who perform the procedures frequently and in centers that have large volumes of these procedures. , The clinical and educational implications of these changes in working hours for registrars to gain experience need to be further elucidated. However, there is evidence that teaching registrars in the operating room is expensive in terms of operating time and financial cost. , What then is the alternative?
Workshops Workshops with objective goals of improving theoretical knowledge, enhancing clinical judgment, and initiating and up-scaling manual dexterity must be an integral part of professional development of the trainer and trainee alike in gynecologic endoscopic surgery in any teaching hospital with a respected national and international standing. , The theoretical knowledge of registrars may surpass that of the practicing gynecologic endoscopist. This is a serious drawback that may well negate the value of registrars' learning curve. One possible explanation for this may be that busy practicing gynecologists have very little time available to them to update their professional knowledge because they would rather concentrate on practical aspects of patient care. By so doing, their current knowledge may “decay” at an exponential rate, hence the crucial importance of continuous professional development strenuously advocated by major professional bodies worldwide (eg, Royal Australia and New Zealand College of Obstetricians and Gynaecologists ). Animal Laboratory Animal laboratory teaching is best performed on inanimate tissue and readily available animal tissue (eg, sow uterus or pig bladder ). Eye-hand-foot coordination can be developed and assessed in this setting. However, this approach does not exactly simulate the in vivo human condition in that there is little demonstration of injury, such as that shown by bleeding. Visual Reality Simulators Simulations have been used by the airline industry and military, as well as our colleagues in other medical specialties, to educate, evaluate, and prepare for life-threatening scenarios. The gynecologist goes through specific training to develop a different level of psychomotor skills than that required for conventional (laparotomy and vaginal) surgery. To reduce the need for experimental animals and the more expensive operating-room hands-on learning, bench models were introduced to improve not only endoscopic skills from the technical standpoint but also the operative performance of trainees. Objective structured assessment of technical skills is a multistation performance-based examination of surgical skills. It uses visual reality (VR) simulators and has been successfully used in many institutions worldwide to objectively score candidates. These tools serve to objectively validate teaching methodologies. The findings of these programs could well be used in the selection of appropriate training methodologies. , VR simulations are valuable not only in education but in objectively assessing the trainee's learning curve with good reliability, validity, and cost-effectiveness. VR simulation is a feasible system that the trainee may choose to use regularly. However, many current, qualified advanced hysteroscopic surgeons have not been taught the fundamentals through an organized curriculum that included VR training. Endoscopic surgical competency may be judged by gynecologists experienced in the field of operative endoscopy. Because the definitive criteria for assessing competence in gynecologic endoscopy remain elusive, attempts to streamline the objectivity of assessment were made. Trainees were tested on multiple tasks, involving clinical judgment, dexterity, serial/simultaneous complexity, and spatial orientation. The assessors then assessed overall subject competence for each procedure on simulation. Point-biserial correlational analysis and cluster analysis were performed to ascertain the relationships among the different scales. The cluster analysis showed that the surgeon assessors shared a common perception of competence. , However, VR provides a more objective means for evaluating the psychomotor skills needed to perform endoscopic surgery. In addition, medicolegal and financial constraints of training and evaluation in operating rooms have enhanced the use of VR endoscopic surgery. Hysteroscopic surgery is a relatively new technique used to surgically manage uterine pathology in many leading centers of gynecologic endoscopy worldwide. It requires special surgical skills for handling the operating hysteroscope and remote instrument control. In general, VR simulators seem to enrich education in gynecologic endoscopy, in addition to clinical education and active learning in operating rooms. Nevertheless, these systems have been criticized as not being able to lend themselves to the realistic surgical environment. It has been suggested that this method lacks standards defining performance-based endpoints, with neither a predetermined training duration nor an arbitrary number of repetitions of tasks being adequate to ensure endoscopic proficiency after simulator training. However, the Royal College of Obstetricians and Gynaecologists, London, has recommended that surgical training systems, such as VR hysteroscopy, be evaluated, piloted, and introduced into basic surgical skills courses. In addition, skills acquired during VR operative hysteroscopy sessions can be effectively transferred to the patient's care in the operating room. International multicenter studies objectively comparing VR systems and their potential impact on the practice of gynecologic endoscopic surgery are awaited with interest.
Workshops with objective goals of improving theoretical knowledge, enhancing clinical judgment, and initiating and up-scaling manual dexterity must be an integral part of professional development of the trainer and trainee alike in gynecologic endoscopic surgery in any teaching hospital with a respected national and international standing. , The theoretical knowledge of registrars may surpass that of the practicing gynecologic endoscopist. This is a serious drawback that may well negate the value of registrars' learning curve. One possible explanation for this may be that busy practicing gynecologists have very little time available to them to update their professional knowledge because they would rather concentrate on practical aspects of patient care. By so doing, their current knowledge may “decay” at an exponential rate, hence the crucial importance of continuous professional development strenuously advocated by major professional bodies worldwide (eg, Royal Australia and New Zealand College of Obstetricians and Gynaecologists ).
Animal laboratory teaching is best performed on inanimate tissue and readily available animal tissue (eg, sow uterus or pig bladder ). Eye-hand-foot coordination can be developed and assessed in this setting. However, this approach does not exactly simulate the in vivo human condition in that there is little demonstration of injury, such as that shown by bleeding.
Simulations have been used by the airline industry and military, as well as our colleagues in other medical specialties, to educate, evaluate, and prepare for life-threatening scenarios. The gynecologist goes through specific training to develop a different level of psychomotor skills than that required for conventional (laparotomy and vaginal) surgery. To reduce the need for experimental animals and the more expensive operating-room hands-on learning, bench models were introduced to improve not only endoscopic skills from the technical standpoint but also the operative performance of trainees. Objective structured assessment of technical skills is a multistation performance-based examination of surgical skills. It uses visual reality (VR) simulators and has been successfully used in many institutions worldwide to objectively score candidates. These tools serve to objectively validate teaching methodologies. The findings of these programs could well be used in the selection of appropriate training methodologies. , VR simulations are valuable not only in education but in objectively assessing the trainee's learning curve with good reliability, validity, and cost-effectiveness. VR simulation is a feasible system that the trainee may choose to use regularly. However, many current, qualified advanced hysteroscopic surgeons have not been taught the fundamentals through an organized curriculum that included VR training. Endoscopic surgical competency may be judged by gynecologists experienced in the field of operative endoscopy. Because the definitive criteria for assessing competence in gynecologic endoscopy remain elusive, attempts to streamline the objectivity of assessment were made. Trainees were tested on multiple tasks, involving clinical judgment, dexterity, serial/simultaneous complexity, and spatial orientation. The assessors then assessed overall subject competence for each procedure on simulation. Point-biserial correlational analysis and cluster analysis were performed to ascertain the relationships among the different scales. The cluster analysis showed that the surgeon assessors shared a common perception of competence. , However, VR provides a more objective means for evaluating the psychomotor skills needed to perform endoscopic surgery. In addition, medicolegal and financial constraints of training and evaluation in operating rooms have enhanced the use of VR endoscopic surgery. Hysteroscopic surgery is a relatively new technique used to surgically manage uterine pathology in many leading centers of gynecologic endoscopy worldwide. It requires special surgical skills for handling the operating hysteroscope and remote instrument control. In general, VR simulators seem to enrich education in gynecologic endoscopy, in addition to clinical education and active learning in operating rooms. Nevertheless, these systems have been criticized as not being able to lend themselves to the realistic surgical environment. It has been suggested that this method lacks standards defining performance-based endpoints, with neither a predetermined training duration nor an arbitrary number of repetitions of tasks being adequate to ensure endoscopic proficiency after simulator training. However, the Royal College of Obstetricians and Gynaecologists, London, has recommended that surgical training systems, such as VR hysteroscopy, be evaluated, piloted, and introduced into basic surgical skills courses. In addition, skills acquired during VR operative hysteroscopy sessions can be effectively transferred to the patient's care in the operating room. International multicenter studies objectively comparing VR systems and their potential impact on the practice of gynecologic endoscopic surgery are awaited with interest.
Before operating on patients, trainees should be assessed in workshops, including animal laboratories if available; during VR simulation exercises; and in the operating theater. The results of objective assessments are recorded in trainees' log books.
|
Importance of access to epilepsy monitoring units during the COVID-19 pandemic: Consensus statement of the International League against epilepsy and the International Federation of Clinical Neurophysiology | 665f53c6-c78d-4995-97cc-3ed051b7affe | 8294085 | Physiology[mh] | Background and justification The International League Against Epilepsy (ILAE) and the International Federation of Clinical Neurophysiology (IFCN) appointed an ad-hoc taskforce to provide a rapid response to the challenges concerning video-EEG monitoring, encountered during the current pandemic, caused by the coronavirus SARS-CoV-2 disease (COVID-19). Through consensus discussions, review of the published evidence and experience of the experts representing the two societies, the ad-hoc taskforce elaborated this statement. During the restructuring of healthcare services due to the current pandemic, many hospitals closed video-EEG monitoring facilities, referred to in this document as Epilepsy Monitoring Units (EMU) . A European survey showed that in most centers, inpatient video-EEGs monitoring had been stopped (61.7% for adults, 36.2% for children) or was restricted (38.3% for adults, 53.2% for children) , with detrimental effects on patients with complex and severe epilepsy and other paroxysmal events , such as lack of optimizing the medical treatment and lack of evaluation for epilepsy surgery. There is limited likelihood of triggering seizure emergencies in patients with epilepsy and neurological complications. The measures of closing EMUs were adopted by healthcare providers to focus on re-allocation of resources to services considered more important, more immediately required, and to prevent spreading the disease. Long-term video-EEG monitoring in EMUs was regarded by the healthcare providers as an elective procedure that could be postponed without significant consequences, a categorization that we challenge as incorrect for the following reasons. Long-term video-EEG monitoring is an essential diagnostic tool in patients with complex and severe epilepsy . The main indications are: diagnostic and presurgical evaluation . While video-EEG monitoring is diagnostic, it has direct implications on treatment of epileptic seizures, co-morbidities and important differential diagnoses (arrhythmia and cardiac death, psychogenic non-epileptic seizures and the risk of suicide). Reasoning for continuing the diagnostic monitoring : approximately one third of patients referred to specialized centers on suspicion of drug-resistant epilepsy, do not have epilepsy . Persistent misdiagnosis of paroxysmal events, often cardiac or psychogenic in origin, has severe consequences for them . In patients with drug resistant epilepsy, misclassification of the seizure-types can lead to inadequate choice of antiseizure medication . Video-EEG recording of the patientś habitual clinical episodes is the diagnostic gold standard for patients with unclear paroxysmal events . Reasoning for continuing presurgical evaluation : epilepsy surgery is the evidence-based treatment in patients with drug-resistant focal epilepsy . This requires video-EEG recording of the seizure and in around one third of patients invasive monitoring . Failure to proceed towards surgery, unnecessarily exposes the patients to further seizures, injuries associated with seizures and the risk of Sudden Unexpected Death in Epilepsy (SUDEP) . The appropriate and unrestricted utilization of EMUs in comprehensive epilepsy centers has been shown to reduce mortality of patients with epilepsy . High quality epilepsy care, including video-EEG monitoring has decreased morbidity and mortality . Hence, increasing waiting times can cause considerable problems, with increasing morbidity and mortality. These patients often have worsening epilepsy, co-morbidities, and prioritizing care with restricted resources becomes more and more challenging. Some EMUs managed to continue video-EEG monitoring during the pandemic . Using measures of prevention and protection generally adopted in the hospitals, these EMUs were able to continue this important diagnostic function, without causing local outbreaks . Recommendations for neurophysiology staff with risk factors for COVID-19, and for mental health of the staff have been proposed by the Latin American chapter of the IFCN Task Force – COVID-19 .
Summary statements The ILAE-IFCN ad-hoc taskforce issues the following statement, related to functioning of the EMUs, during the COVID-19 pandemic: 1. Long-term video-EEG monitoring is an essential diagnostic service. 2. Access to video-EEG monitoring of the patients in the EMUs must be given high priority. 3. Patients should be screened for COVID-19, before admission, according to the local regulations. 4. Local policies for COVID-19 infection control should be adhered to during the video-EEG monitoring. 5. In cases of differential diagnosis where reduction of antiseizure medication is not required, consider home video-EEG monitoring as an alternative in selected patients.
Conclusion The ILAE-IFCN ad-hoc taskforce calls for action to ensure that healthcare providers understand the importance of providing diagnostic services for patients with epilepsy and paroxysmal events, and that EMUs continue functioning during emergency situations like the COVID-19 pandemic, while adhering to local healthcare policies.
Disclaimer This report was written by experts selected by the International League Against Epilepsy (ILAE) and the International Federation of Clinical Neurophysiology (IFCN) and was approved for publication by the ILAE and IFCN. Opinions expressed by the authors, however, do not necessarily represent the policy or position of the ILAE and IFCN.
Conflict of interest statement Sándor Beniczky reports speaker honoraria and personal fees from: Natus Neuro, Philips EGI, Epihunter, UCB Pharma, Eisai and Bial-Portela, outside this work. Aatif Husain has received personal compensation for consulting from: UCB Pharma, Jazz Pharma, BlackThorn Therapeutics, Eisai Pharmaceuticals, Marinus Pharmaceuticals, Neurelis Pharmaceuticals, and Merck. He has also received royalties from Wolters Kluwer, Springer and Demos Publishers. Additionally, he has received stipend for editorship from American Clinical Neurophysiology Society. Akio IKEDA belongs to Department of Epilepsy, Movement Disorders and Physiology which is the Industry-Academia Collaboration Courses, supported by Eisai Co., Ltd., Nihon Kohden Corporation, Otsuka Pharmaceutical Co., and UCB Japan Co., Ltd. He also receives honorarium from Eisai Co., Ltd and UCB Japan Co., Lt J. Helen Cross has acted as an investigator for studies with GW Pharma, Zogenix, Vitaflo and Marinius. She has been a speaker and on advisory boards for GW Pharma, Zogenix, and Nutricia; all remuneration has been paid to her department. Her research is supported by the National Institute of Health Research (NIHR) Biomedical Research Centre at Great Ormond Street Hospital. She holds an endowed chair at UCL Great Ormond Street Institute of Child Health; she holds grants from NIHR, EPSRC, GOSH Charity, ERUK, and the Waterloo Foundation Jo Wilmshurst receives an honorarium for her role of associate editor for Epilepsia. M Seeck received speaker’s fees of EGI Philips, holds shares of Epilog and received consulting fees from the Wyss Foundation. N. K. Focke received speaker bureaus and consultancy fees from Bial, Eisai and EGI/Phillips outside the submitted work. Patricia Braga reports no conflicts of interest. Samuel Wiebe, through his institution, has received unrestricted educational grants from UCB Pharma, Sunovion and Eisai. Dr. Schuele has received personal compensation in form of a stipend from Wolters Kluwer as Associate Editor of the Journal of Clinical Neurophysiology. He receives royalties from a book published from Demos Publishers Inc., Springer Publishing on Stereo-encephalography. Honoraria for speaking engagements have been received from the American Academy of Neurology, American Epilepsy Society, American Clinical Neurophysiology Society, Sunovion Inc., SK Life Science, Neurelis, and Greenwich. Eugen Trinka reports personal fees from EVER Pharma, Marinus, Arvelle, Argenix, Medtronic, Bial-Portela & Cª, NewBridge, GL Pharma, GlaxoSmithKline, Boehringer Ingelheim, LivaNova, Eisai, UCB, Biogen, Genzyme Sanofi, and Actavis; his institution received grants from Biogen, UCB Pharma, Eisai, Red Bull, Merck, Bayer, the European Union, FWF Osterreichischer Fond zur Wissenschaftsforderung, Bundesministerium für Wissenschaft und Forschung, and Jubiläumsfond der Österreichischen Nationalbank outside the submitted work.
|
European health regulations reduce registry-based research | 75b8c1f1-3847-49a7-a94a-42af00ea4581 | 11443657 | Psychiatry[mh] | Registry-based medical research forms an integral pillar of modern healthcare. Medical registries are invaluable resources for generating evidence-based medical guidelines and developing innovations such as effective diagnostic and monitoring tools. However, this progress has been subjected to a growing regulatory burden in the European Union (EU), particularly in the realm of data privacy and security, which are also part of the Charter of Fundamental Rights of the EU. The General Data Protection Regulation (GDPR), implemented in 2018, has strengthened personal data protection rights by providing individuals with greater control over their data and imposing stricter obligations on data controllers and processors. The European Health Data Space (EHDS) has been recently approved for establishing a common framework for the governance and sharing of health data across the European Union. While this initiative holds promise for facilitating cross-border access to health data and common policies for data privacy in health research, it also introduces additional layers of bureaucracy and potential delays in research and innovation, raising questions of its effectiveness . The GDPR, as an EU regulation, is directly applicable to all member states, as will be the case with the EHDS. However, prior to the applicability of the EHDS, registry-based medical research was governed by member state law. In Finland, the legislation governing registry-based medical research consists primarily of the following: the Data Protection Act (1050/2018), and more importantly, the Act on the Secondary Use of Health and Social Data (552/2019; Secondary Use Act). The Data Protection Act is a Finnish supplement to the GDPR (Fig. A). It parallels the Federal Data Protection Act (BDSG) in Germany, the Data Protection Act (Loi Informatique et Libertés) in France and many other national legislations implemented by member states. The Data Protection Act in Finland specifies that the public interest under point (e) of Article 6(1) GDPR is a valid legal basis for processing personal data for the purpose of scientific research. The processing of special categories of personal data, such as health data and genetic data, is also permitted but subjected to appropriate safeguards as further specified in the Act and as required under Article 89(1) GDPR. This act is crucial for registry-based medical research, as, together with the GDPR, it defines the core data protection framework under which health data are processed in Finnish scientific research. However, it does not specify how registry data are obtained for such purposes. The Finnish national legislative model in this instance focusses primarily on controlling who may have access to sensitive data and under what circumstances, such as by regulating the relevant permit and consent processes and the relevant documentary requirements. This is largely the result of the relevant lex generalis, the Act on the Openness of Government Activities (21.5.1999/621; Openness Act). This approach is conceptually very different from the model of EU data protection legislation, which is largely agnostic as to how personal data are obtained but instead focusses on regulating the planned or ongoing data processing activities themselves. As a result, Finnish registry-based research is regulated under two different legislative paradigms with different targets. Prior to the Secondary Use Act, registry-based medical research involving medical data of a single public health care entity required a permit granted by that entity as set out in i.a. Section 28 of the Openness Act. As such, studies could be combined with interventions related to the personal integrity of study participants, such as the collection of biological samples, which requires prior ethical approval, and administrative processes could be combined into a single permitting process. This process resulted in a study permit, which was granted by the relevant public healthcare entity. If registry data were required from multiple registries, the Finnish Institute for Health and Welfare acted as the permit authority for the relevant registry data. The Ministry of Social Affairs and Health introduced the Secondary Use Act in Finland, which entered into force in 2019. Its primary purpose was to facilitate the effective and safe processing of personal social and health data for secondary uses, such as research, policymaking and healthcare management, in contrast to primary uses, such as the provision of healthcare services. Individual public health care entities, such as university hospitals, retained the right to grant data permits to their own data in single-registry studies. In contrast, Findata was established as the centralized data permit authority, that is, a one-stop agency granting data permits for secondary use, granting data permits for private healthcare providers, the National Patient Data Repository, or a combination of multiple social and health registries, with multiple other tasks, such as defining policies for safe data processing and assessing the anonymity of data. With the exception of the Health Data Hub in France and Denmark Statistics in Denmark, the scope and extent of the responsibilities are internationally unique. Despite these positive objectives, two independent surveys conducted in 2021 highlighted the discontent of the research community with the Secondary Use Act. The first study focussed on medical doctors and revealed that, out of 430 responders, 79% experienced the data permit process to have become more complex than before, 55% that research projects were delayed and 64% that research costs had increased . Owing to these issues, 42% of the responders reported that they had not initiated a study . Similar results were reported in a second survey commissioned by the Ministry of Social Affairs and Health, where responders ( n = 260) reported that the Secondary Use Act simplified neither the application process nor the combination of multiple registries . Compared with prior practices, the Secondary Use Act contains very specific provisions on documentation requirements; how registry data are collected, combined, transferred and processed prior to their disclosure to the study that has received the data permit; the costs associated with such provisions; and the regulation of secure cloud-based computing instances in which the recipient project may conduct research on such data. In cases where Findata is the authority granting the data permit, researchers of the registry-based study cannot actively participate in the pre-transfer data processing activities. In such cases, the collection and processing work is conducted by the registry authorities and Findata, which may obligate the researchers to both longer processing times and costs related to data collection and processing. While the EHDS regulation is expected to harmonize the principles of health data processing in the EU, there is little retrospective evidence on the impact of increasing privacy regulations on registry-based social and medical research. Given its considerable overlap with the Secondary Use Act in terms of shared objectives, implementation of national data permit authorities and establishing audited data secure environments, we sought to evaluate the impact of the regulation on registry-based research. Here, we examine data permit counts before and after the implementation of the act to reflect the administrative and procedural hurdles faced by researchers, providing a simple measure of how regulatory changes affect access to health data for research purposes.
Research registries Population-based, healthcare and social registries represent retrospective research registries as information has been collected for another primary purpose but may be subject to secondary use in research registries. In contrast, in cohort studies in which subjects are first consented to a medical study, data are recorded and stored longitudinally for their primary use, representing prospective research registries. Retrospective registries have become increasingly popular in Finland due to their extent, availability of high-quality information in digital format and low cost, as these are maintained and managed by organizations or the public sector. Study versus data permits There are five university hospitals in Finland, which are located in Helsinki, Tampere, Kuopio, Turku and Oulu. When we reference these cities, we are specifically alluding to the respective university hospitals. University hospitals serve as regional hubs for specialized care and provide comprehensive medical services to their respective catchment areas. Hospitals are well equipped for registry-based research because of their advanced IT infrastructure and larger patient populations, enabling them to collect, manage and analyse large-scale health data effectively both for primary and secondary purposes. After entry into force of the Secondary Use Act, both a study permit and a data permit are required to conduct registry-based research projects in Finland. Prior to this, only study permits were needed, as explained above. The study permit is issued by the research organization, for example, the university hospital, where the research project is based. This ensures that the research is scientifically justified and ethical and that the rights of the participants are protected, although registry research permits are rarely handled in separate ethical research boards. The data permit is issued by either Findata or GDPR-defined data controllers (e.g. university hospitals or primary healthcare registries) whose registry data are used in the research to evaluate the purpose of data use, the security measures in place and the potential impact on individual privacy. While this dual-permit system ensures that both the data and the study are subject to rigorous scrutiny, data privacy measures have been an integral part of the study permit process before the Secondary Use Act. For simplicity, in this study, we employ the term data permit, as university hospitals can grant both study and data permits, but Findata can grant only data permits. Data collection In January 2024, we solicited counts of new data permits for registry-based research, specifically involving university hospital registries, from the research departments of all five university hospitals (Helsinki, Turku, Tampere, Oulu and Kuopio). To account for variations in research funding and regulatory changes, we also requested data on study permits for medical research involving subjects, human tissue or medical devices. Owing to varying archiving practices, data from 2016 to 2023 were available from Tampere, Oulu, Kuopio and Helsinki. As data were accessible only from 2020 to 2023, Turku was excluded from analyses requiring data prior to 2020. We also solicited counts of new data permits ( n = 375) involving clinical data from university hospitals and grants from Findata from 2020 to 2023. Recognizing that Findata-approved permits cover data from 1–5 hospitals, we integrated details on the annual mean count of university hospitals covered by these permits (mean 2.2–2.6 from 2021 to 2023). Statistical analysis We fitted univariate linear regression analyses using data permit counts as covariate and year as a predictor and examined the slopes of the curves by extracting the coefficient. We used the Mann‒Kendall test to estimate trend over time. On the basis of prior findings, we tested only whether the registry-based study count had decreased since its implementation with the one-sided Mann‒Kendall test . Elsewhere, we applied two-sided tests. We performed statistical analyses and visualizations with R 4.0 via the packages base, sf, mapsFinland and ggplot2.
Population-based, healthcare and social registries represent retrospective research registries as information has been collected for another primary purpose but may be subject to secondary use in research registries. In contrast, in cohort studies in which subjects are first consented to a medical study, data are recorded and stored longitudinally for their primary use, representing prospective research registries. Retrospective registries have become increasingly popular in Finland due to their extent, availability of high-quality information in digital format and low cost, as these are maintained and managed by organizations or the public sector.
There are five university hospitals in Finland, which are located in Helsinki, Tampere, Kuopio, Turku and Oulu. When we reference these cities, we are specifically alluding to the respective university hospitals. University hospitals serve as regional hubs for specialized care and provide comprehensive medical services to their respective catchment areas. Hospitals are well equipped for registry-based research because of their advanced IT infrastructure and larger patient populations, enabling them to collect, manage and analyse large-scale health data effectively both for primary and secondary purposes. After entry into force of the Secondary Use Act, both a study permit and a data permit are required to conduct registry-based research projects in Finland. Prior to this, only study permits were needed, as explained above. The study permit is issued by the research organization, for example, the university hospital, where the research project is based. This ensures that the research is scientifically justified and ethical and that the rights of the participants are protected, although registry research permits are rarely handled in separate ethical research boards. The data permit is issued by either Findata or GDPR-defined data controllers (e.g. university hospitals or primary healthcare registries) whose registry data are used in the research to evaluate the purpose of data use, the security measures in place and the potential impact on individual privacy. While this dual-permit system ensures that both the data and the study are subject to rigorous scrutiny, data privacy measures have been an integral part of the study permit process before the Secondary Use Act. For simplicity, in this study, we employ the term data permit, as university hospitals can grant both study and data permits, but Findata can grant only data permits.
In January 2024, we solicited counts of new data permits for registry-based research, specifically involving university hospital registries, from the research departments of all five university hospitals (Helsinki, Turku, Tampere, Oulu and Kuopio). To account for variations in research funding and regulatory changes, we also requested data on study permits for medical research involving subjects, human tissue or medical devices. Owing to varying archiving practices, data from 2016 to 2023 were available from Tampere, Oulu, Kuopio and Helsinki. As data were accessible only from 2020 to 2023, Turku was excluded from analyses requiring data prior to 2020. We also solicited counts of new data permits ( n = 375) involving clinical data from university hospitals and grants from Findata from 2020 to 2023. Recognizing that Findata-approved permits cover data from 1–5 hospitals, we integrated details on the annual mean count of university hospitals covered by these permits (mean 2.2–2.6 from 2021 to 2023).
We fitted univariate linear regression analyses using data permit counts as covariate and year as a predictor and examined the slopes of the curves by extracting the coefficient. We used the Mann‒Kendall test to estimate trend over time. On the basis of prior findings, we tested only whether the registry-based study count had decreased since its implementation with the one-sided Mann‒Kendall test . Elsewhere, we applied two-sided tests. We performed statistical analyses and visualizations with R 4.0 via the packages base, sf, mapsFinland and ggplot2.
In 2020–2023, 1768 registry-based research data permits were granted by university hospitals (Fig. B). Most of these ( n = 595) were approved by Helsinki, followed by Tampere ( n = 367), Turku ( n = 355), Oulu ( n = 231) and Kuopio ( n = 220). Following a stable period between 2016 and 2019 (median 517, range 497–573; Fig. C), new data permit counts decreased rapidly across hospitals (tau −1, p = 0.042, one-sided Mann‒Kendall test). Compared with 2019, there was a median decrease of 22.4 (range 9.7–31.7) data permits per year, representing an annual median reduction of 5.5% (range 3.4–10.5%). At the same time, 375 registry-based data permits were approved by Findata (Fig. C). However, the percentage of Findata-approved data permits also decreased from 141 to 99 (29.8%) in 2023 (Fig. C). The initial increase could be partly explained by an accumulation of applications due to new systems and policies being established at Findata during 2020–2022. The following decrease in 2023 could hence reflect the pruning of the backlog. Next, we examined data permit counts approved by university hospitals before and after (2016–2017 versus 2018–2019) the implementation of the GDPR. An annual median of 152 new registry research permits were approved pre-GDPR (Fig. C). While the corresponding median in the following 2 years was 131, the count increased from 121 in 2018 to 138 permits in 2019, implying that the GDPR had no conclusive impact on data permit counts (Fig. C). To estimate the decline in data permits, we examined the era before the implementation of the Secondary Use Act. In 2015–2019, a median of 1.9 (range 1.0–7.3) additional data permits were approved, corresponding to a 0.70% (range 0.20–1.3%) yearly accumulation. The finding demonstrates that new permits approved for registry-based research remained stable prior to the implementation of the Secondary Use Act. Assuming that similar progress would have continued until 2023, we fitted a linear regression curve for each hospital and predicted that a total of 586.8 new data permits would have been granted in 2023 (Fig. D). The figure contrasts with the 234 new data permits actually approved. Given that a portion of multi-centre permits have been directed to Findata and that Turku was excluded from the analyses due to missing data, we included data permits accorded by Findata in 2023 ( n = 99) multiplied by 78.3%, reflecting the percentage of data permits approved by university hospitals other than Turku in 2023. The estimated reduction was 275.3 [= 586.8 − 234.0 − (99.0 × 0.78)] data permits, corresponding to a relative reduction of 314.2/586.8 = 46.9%. To exclude other confounding factors, we examined the data permit counts for other types of medical research (Fig. E). While the proportion of approved permits decreased in 2016–2023 (tau −0.54, p = 0.061, two-sided Mann‒Kendall test), the changes did not coincide with either the enactment of the GDPR or the Secondary Use Act.
This retrospective observational cohort study demonstrated the negative impact of the Secondary Use Act on registry-based studies only 3 years after its implementation. The findings sharply contrast with the generally positive attitude towards the secondary use of personal data in medical research . This study has several implications for efforts to understand the potential impact of EHDS regulation on research and innovations. First, while research with retrospective registries could be previously performed with a minimal budget, solid funding has become necessary following the enactment of the Secondary Use Act. Data collection and application fees for submitting a study plan, adding researchers to a valid permit and processing data in secure cloud-based computing instances accumulate significant research costs. Second, stringent data privacy laws may hinder individual countries from participating in multicenter registry studies and potentially impede global health by slowing responses to pandemics, such as coronavirus disease 2019 (COVID-19) . Finland has a long tradition of accessing nationwide healthcare registries, such as the care registry for specialized healthcare, the cancer registry and implant registries, which have all been acknowledged for their coverage and quality [ – ]. Integrating Finnish registry holders in international studies is possible only if data are processed in a computing environment that has been audited to meet the legal requirements. Currently, only nine such environments are eligible, and all are based in Finland, implying that international registry studies would require the transfer of all data to one of these environments . The main source of error in this study stems from the prediction of new data permits in 2023 in a scenario without the Secondary Use Act. However, we anticipate that registry-based research would have gained even more popularity than before, given the substantial investments in healthcare information technology infrastructures facilitating the release of electronic health records to cloud-based computing instances. Thus, many researchers have reported that the regulatory environment hampers not only conventional registry-based studies, but also promising progress towards automatic computer-assisted data curation and disease phenotyping . The enactment of the Secondary Use Act coincided with the COVID-19 pandemic, which could confound the results. Multiple reports have indicated an increase in scientific publishing following the COVID-19 outbreak . The increase in approved non-registry permit counts in 2020–2022 parallels with the increase in approved registry-based research permits in 2020, suggesting that the pandemic likely would not have reduced research activity. However, the exact influence of the pandemic on registry-based research would require examining study permit counts in other countries. In addition, it is plausible that increasing bureaucratic requirements would have motivated researchers to apply for fewer and larger study and data permits to avoid administrative burden related to preparing applications. However, this speculation can be investigated in the future by comparing the number of publications per permit by their year of approval. The enactment of the Secondary Use Act is grounded in the need to balance data accessibility with privacy protection, which is an uncompromised requirement of professional and ethical data use. This balance aims to safeguard personal information while promoting scientific advancements. However, the mechanisms by which this legislation operates can inadvertently stifle research. Even if a registry-based study is fully GDPR compliant, this does not guarantee that it is possible to conduct the study under the Secondary Use Act. As noted, the Secondary Use Act and the proposal for the EHDS exhaustively regulate the processing procedures both prior to and after the transfer of health-registry data for use in the research project. These procedures are largely mandatory and one-size-fits-all in nature, thus not accounting for the specific risk profile of the research project in question or the registry data involved, as would be the approach warranted by the GDPR and as was the case in Finland prior to the entry into force of the Secondary Use Act. This inflexibility and the resulting administrative hurdles delay research timelines and inflate costs. We posit that the specific inflexible processing and procedural requirements and the associated costs resulting from the Secondary Use Act are reasons for the reduction in data permits. These regulatory demands, although designed to enhance data protection and data security, may thus paradoxically hinder progress in medical research and patient care. In conclusion, the results emphasize the need to balance effective and secure data research, that is, protecting the privacy of research subjects and providing research results benefitting the society as a whole. Medical researchers should be involved in planning, interpreting and assessing health data regulations. In addition to their complexity, the cumulative effect of European and national regulations has created a challenging environment for medical researchers. Instead of improving patient privacy rights, increased administrative work and excessive technology requirements to ensure security may delay research projects and increase costs. Ultimately, the regulatory burden may turn against its objectives and impede progress in patient care .
|
Avenues for Strengthening PCORnet’s Capacity to Advance Patient-Centered Economic Outcomes in Patient-Centered Outcomes Research (PCOR) | 0d75c46c-e353-43ce-bf1e-abe71b0fd028 | 10635342 | Patient-Centered Care[mh] | Linking patients represented in a PCORnet site’s curated CDM with other individual-level sources, such as insurance claims, has 2 main advantages: (1) ascertaining health care received and outcomes that occur outside the PCORnet partner health system; and (2) incorporating data not available from the site’s CDM such as insurance payments to the provider. PCORnet CRNs have established the ability to perform such linkage without sharing personally identifiable information outside of network health systems by implementing privacy-preserving record linkage (PPRL) technology network-wide provided by Datavant. Queries of the distributed data network generally do not resolve, or de-duplicate, the identity of a patient who has been treated by 2 or more participating health care systems so that they are counted once instead of 2 or more times in analyses. Instead, the PPRL infrastructure enables matching and linkage of patient records for research purposes, both among multiple health care systems and with commercial health plans to aggregate EHR and claims data. As shown in Figure , PCORnet partners have also linked with data from the Centers for Medicare and Medicaid Services (CMS) by submitting finder files of patient identifiers (upper left) appended with a unique “Site Coded ID,” which are used by Medicare’s honest broker contractor to match patients with beneficiaries; generating a cross-walk file that enables analyses of the linked EHR data and the financial activity reflected in claims records as a limited data set. Further advancements to PCORnet linkage mechanisms may include: (1) improving matching accuracy through referential patient matching which “uses large collections of demographic records, such as information from credit reporting agencies or address change records, providing a multirecord benchmark to match identities” ; and (2) embedding PPRL within cloud-based data ecosystems using external tokenization which enables data linkage when analyses are conducted, thereby reducing the costs and delays associated with generating PPRL before analyses. While some CRNs maintain linked databases, such infrastructure may have limited scalability for organizations using an older, on-site data infrastructure because it can become prohibitively expensive to maintain as the volume of data increases. Other CRNs have standing regulatory agreements with various data partners that enable linkage on a project-specific basis. The data remain at the respective institutions until there is a request for a subset of data to be shared and linked for a given research purpose. Standing agreements may simply be amended for each use case, while the main terms of the agreement are already established. These arrangements are relatively inexpensive to maintain and represent trust among organizations that underscores long-standing, sustainable partnerships. This reusable regulatory framework enables efficient data sharing and linkage without requiring substantial infrastructural investment and maintenance. The regulatory framework, developed in conjunction with patient advisors, addresses patient privacy and security concerns throughout.
CRNs conduct research through data linkage with health plans as exemplified by REACHnet’s (one of the PCORnet CRNs) partnerships with 3 commercial health plans—Humana, CVS Health, and Blue Cross Blue Shield of Louisiana (BCBS-LA). Regulatory agreements and PPRL infrastructure are in place to link claims with EHR data on a study-specific basis to enhance the completeness of outcomes and include economic data exemplified by a study of the impacts of health plan benefits on clinical and cost-related type II diabetes outcomes. REACHnet linked clinical data from 3 PCORnet participating health systems in Louisiana with claims from BCBS-LA. The primary aim is to evaluate BCBS-LA’s zero-dollar copayment benefit for generic prescription drugs on medication adherence, hemoglobin A1c control, out-of-pocket spending, and overall medical costs. While A1c measures are obtained from EHR data, medication dispensing information is incomplete in EHRs so pharmacy claims are used to assess medication adherence. Out-of-pocket and overall medical costs are also evaluated from the health plan data. This study directly addresses patients’ prescription costs as a potential barrier to medication adherence and A1c control in diabetes management. Standing partnerships between PCORnet and health plans enable opportunities to evaluate the impacts of benefit design on patients’ day-to-day clinical outcomes.
CMS makes data available for research supported by the Research Data Assistance Center (ResDAC). Some CRNs also work with state offices to access Medicaid claims more rapidly than through ResDAC. Several PCORnet CRNs have conducted projects that link CMS claims with the PCORnet CDM following ResDAC’s processes as outlined in a recent PCORI white paper. An example of how Medicare data has been used to study patient-centered economic outcomes is the Louisiana Experiment Addressing Diabetes Outcomes (LEAD study) linking clinical data from 3 PCORnet participant health systems in Louisiana with Medicare claims to evaluate reimbursement for non–face-to-face chronic care management (NFFCCM) services, a CMS policy instituted in 2015. The analyses found that use of NFFCCM was not associated with additional copayments for Medicare beneficiaries but was associated with a reduction in total medical expenditures. The Greater Plains Collaborative (GPC) CRN uses current ResDAC processes to populate a cloud-based data-sharing environment for multistate CMS claims, site PCORnet CDMs, tumor registries, and external sources (Fig. ). The GPC study also describes the complementary nature of EHRs and claims in that diagnoses codes in claims underreport obesity in comparison to body mass index consistently collected in EHRs, while common comorbidities such as diabetes are underreported in individual health systems EHRs in comparison with claims. Cloud-based data sharing enables dynamic linkage to CRN data enabling investigators to access study-specific data in a secure environment, reduce the storage and time required to replicate claims at each distributed site thus lowering costs thereby facilitating PCOR to improve care for CMS beneficiaries. To address privacy concerns about linking claims and EHR data, CMS as well as federal funders are increasingly requiring researchers to adopt standardized federal information security practices available consistently from cloud service providers. Additional avenues for CMS claims integration include the CMS Virtual Data Research Center (VDRC) and the Data at the Point of Care pilot Application Programmer Interface (API) for Medicare Claims.
Since the 1990s adult and pediatric health systems have engaged in clinical and financial benchmarking; including consistent length of stay/mortality models, financial comparators, the Joint Commission core measures, and other patient characteristics important to learning health systems. Federally Qualified Health Centers have also studied , Medicaid expansion’s impact upon these practices devoted to underserved communities encompassed by the ADVANCE CRN. Extending benchmarking with other data can further economic analyses. For example, linking the Visient Clinical Database to area-level measures revealed significant disparities in access to chimeric antigen receptor T-cell therapy. The underlying billing systems that support Visient benchmarking also record patient copays and out-of-pocket expenses and could provide consistent measures of financial burden instead of patient survey-based estimates.
Health outcomes are partially associated with the environment in which patients live, and the PCORnet CDM stores patients’ addresses at various levels of resolution (county, 5 and 9 digit postal codes, street address, and detail) and address changes over time to support geographic linkage (ie, geocoding ). Publicly available community-level data provides a way to characterize social and economic factors affecting patients (eg, social deprivation indices incorporating Census tract median incomes) as used in a study of the effect of the Affordable Care Act on diabetes care. PCORnet participating sites linked with the US Census Bureau’s American Community Survey (ACS) data , and may also leverage more recent investments such as the Agency for Healthcare Research and Quality’s (AHRQ) database on Social Determinants of Health (SDOH) and environmental exposure databases on water, soil, air, and climate released by Environmental Protection Agency (EPA). Economic variables at the zip code or census tract level provide contextual information for studying associations between health outcomes and socioeconomic factors of day-to-day life in communities, like computer and internet access/use, labor force participation, and home ownership. Bidirectional relationships may exist between SDOH and health status, whereby social and economic factors impact health outcomes, but health status may also affect economic outcomes like the ability to participate in the workforce.
For decades, database marketing companies have sold individual consumer and household insights for targeted marketing but have recently developed social determinants of health packages. Personal or household attributes include income, home/car ownership, educational attainment, and health/personal care/diabetes purchasing interest and expenditures in these categories. Credit reports are also linked to understand medical debt. , These insights hold promise for more accurate SDOH than area-level measures and also provide insights regarding family and household characteristics. Family linkages are not consistently recorded and extracted from EHRs although a current effort within PCORnet is resolving mother-baby linkage to support PCORI’s maternal morbidity and mortality research priority. While health systems and payors are integrating individual-level SDOH data into practice, their accuracy across diverse patient populations is largely unreported in peer-reviewed literature warranting research into their suitability for use in PCOR. Patient-reported data is arguably the most patient-centered of all data sources, particularly if the information collected is identified by the patient community as especially salient to their health care and economic experiences and needs, such as out-of-pocket costs and financial stresses upon the household. However, the health care industry lacks a large-scale, systematic, and standardized way of collecting these data. Recording individual-level social and financial determinants in the EHR may provide relevant, patient-oriented, standardized measures for use in pragmatic comparative effectiveness research. In conjunction with AHRQ, the Gravity Project is developing standard representations of SDOH data and terminology to support consistency and sharing of measures between EHR systems. This direct integration of standardized and systematically collected patient-focused economic factors into EHR data would be especially advantageous for incorporating into an EHR-based infrastructure like PCORnet as it could support real-time economically informed shared decision-making with patients during their care.
Drug exposure is fundamental to PCOR. Dispensing information via claims (CMS, commercial noted above) can also be obtained from Pharmacy Benefits Managers connected to Surescripts (a company that enables e-prescribing) which in turn can directly update EHRs and support individual-level linkage. However, which health systems receive Surescripts updates is not consistently characterized across PCORnet sites and if addressed would improve our ability to understand drug exposure, adherence, medication safety, and costs. The National Retail Data Mart (NRDM) , is a complementary area-level resource underutilized for PCOR. National retailers provided geocoded data used by health departments and the Center for Disease Control to confirm the level of disease for ongoing and outbreaks of COVID and influenza. Linking to these open-source repositories would enable PCORnet to use data on out-of-pocket costs to assess patient-centered economic outcomes.
Ultimately, comparative effectiveness research using data derived from available EHR data may suffer from uncertainties affecting confidence in results. , The necessary data transformations may censor incomplete records in a systematic way (eg, claims analyses exclude the uninsured/underserved). The data sources may contain gaps in coverage in a systematic way based on funding models and business priorities. Linkages to community-based datasets for social and economic outcomes through the various privacy-preserving methods are often incomplete and may magnify existing biases in the source data. Another area of ongoing concern is the impact of record linkages on the privacy of individuals, groups of individuals, and participating health systems. Ongoing efforts by research teams associated with PCORnet participants are addressing these issues in a systematic way. , It is incumbent on the PCORnet research community to clarify these limitations and invest in engagement efforts to advance trustworthy research that might now holistically evaluate health choices in relation to the financial stability and livelihoods of our patients, their families, and communities.
PCORnet is designed to attract stakeholders and collaborators that can use its infrastructure to conduct research addressing factors that matter most to patients’ health care decisions and experiences. Initially, PCORnet capitalized on the rapid adoption of EHRs incentivized by the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 which had led to nationally available interoperable clinical data. In addition to clinical factors, patient-centered economic outcomes may be relevant to the interests of patients, their families, and health care providers. The enhancements discussed in this commentary highlight ways in which PCORnet can enable research that includes patient-centered economic outcomes. We see this as an opportunity for the health economics research community to inform our infrastructure’s evolution and partner with PCORnet in conducting studies that include salient economic factors in PCOR that improves our national health care system.
|
Added diagnostic value of routinely measured hematology variables in diagnosing immune checkpoint inhibitor mediated toxicity in the emergency department | 99108535-3383-40ae-bb00-1daf96c7f1ed | 10278460 | Internal Medicine[mh] | INTRODUCTION Within the immunotherapeutic field of cancer treatment, multiple new and promising treatment options have emerged over the past years. Among these, immune checkpoint inhibitors (ICI) are increasingly being used as an oncologic treatment strategy for multiple types of cancer and have drastically improved survival of responding patients. For example, patients with advanced melanoma treated with combined nivolumab and ipilimumab therapy have shown to result in a median overall survival of over 60 months, whereas the median survival of patients with metastatic melanoma used to be less than 1 year before the introduction of checkpoint inhibitors. The proportion of cancer patients benefiting from ICI is increasing rapidly, with now over 40% of cancer patients qualifying for ICI treatment. However, their use is associated with a wide variety of immune‐related adverse events (irAE), such as auto‐immune colitis and pneumonitis. Because of overlap in clinical presentation, it can be difficult to differentiate these irAE from progressive disease or other inflammatory conditions, such as infections. Especially in the emergency department (ED) where time and resources are limited, this may lead to diagnostic delay, inappropriate treatment, and a considerable amount of (unnecessary) diagnostic testing. , Accurate and early diagnosis of patients presenting in the ED with irAE is therefore key to start adequate treatment as soon as possible. , Currently, there are only a few biomarkers available that can aid in diagnosing irAE. , A solution to this problem might be found in routinely measured hematological variables. Bacterial infection and viral infections are commonly characterized by high neutrophil and lymphocyte counts respectively, whereas auto‐immune diseases and allergies typically show high eosinophil counts. Previous research has found associations between irAE and increased counts of standard hematology measurements (e.g., absolute lymphocyte count and eosinophil count). In addition, changes in B‐ and T‐cell receptor repertoire show associations with irAE onset and prognosis. However, none of these biomarkers have been extensively validated or are used in clinical practice. Most modern hematology analyzers not only provide blood cell counts, but also measure morphologic characteristics, such as cell size, intrinsic properties and cell viability that carry diagnostic and prognostic value. This raises the question whether they may also be of use in the setting of immunological toxicity. , , To answer these types of questions, scrutinizing complex datasets with conventional statistical methods, such as logistic regression, do not provide stable estimates of the variable's coefficients as models contain too many variables and a low number of samples. New advanced statistical and machine learning (ML) methods are able to remove irrelevant variables thereby reducing the number of variables. In addition, variables of high importance, also known as predictors, can be identified by evaluating the trained coefficient of the trained model. This way, ML allows for the possible identification of new biomarkers and exploration of new horizons in research to aid irAE diagnosis. The aim of this study was therefore to determine the added value of routinely measured hematology characteristics, modeled through ML, as compared to the standard diagnostic practice. This may aid in the diagnosis of irAE in the ED and understanding of the pathophysiology.
METHODS 2.1 Study population This retrospective observational study included all visits to the ED of the University Medical Center Utrecht (UMC Utrecht) between 2013 and 2020 of patients who were being treated with any type of ICI for any type of cancer, until 3 months after cessation of treatment. Because irAE can occur even after cessation of treatment, we chose to include ED visits up to 3 months after treatment with ICI ended. The cutoff of 3 months was chosen after discussion between the authors. If patients had more than one disease episode (defined as a consecutive period with infection‐like symptoms), all patient's ED visits were included separately, whereas for patients with multiple ED visits during one disease episode, only the first visit was included. If patients visited the ED multiple times for the same condition (e.g., due to worsening of symptoms), only the first visit was included. 2.2 Data collection For all ED visits, demographic (age and sex), medication, and hematology data were extracted from the Utrecht Patient Orientated Database (UPOD). In brief, UPOD is a relational database combining clinical characteristics, medication, and laboratory measurements of patients in the UMC Utrecht since 2004. We used hematological variables measured by the CELL‐DYN Sapphire hematology analyzer ( Abbott diagnostics ). The CELL‐DYN Sapphire is a cell counter equipped with a 488‐nm blue diode laser and uses multiple techniques, such as electrical impedance, spectrophotometry, and laser light scattering, to measure morphological characteristics of leukocytes (incl. 5‐part differential), red blood cells (RBCs), and platelets for both classification and enumeration. Each time a component of a complete blood cell count (CBC) is requested, all data generated by the hematology analyzer are automatically stored in UPOD, including a substantial number of raw and research‐only values and background data on cell characteristics which are made available for research purposes. Only visits with available Sapphire data within the first 4 h after ED presentation were included in this study to ensure we only used data from patients with infection‐like symptoms during the ED visit. UPOD data acquisition and management is in accordance with current regulations concerning privacy and ethics. 2.3 irAE label definition A manual chart review was done for all ED visits within our study population by two of the authors (TVtH and BV). Visits for evidently unrelated conditions were excluded. We recorded both the preliminary and definite diagnosis. The preliminary diagnosis was defined as the diagnosis made by the treating physician in the ED and was characterized as either suspected irAE or other . The definitive diagnosis was defined as the diagnosis made by the treating physician at discharge from the hospital or at the end of treatment and was characterized as irAE or other . Ambiguous cases were resolved through consensus. 2.4 Model development Two models were trained to evaluate the added diagnostic performance of the hematology variables for irAE diagnosis. The first model (base) assessed the preliminary diagnosis, sex, and age with logistic regression thereby imitating clinical practice at the ED, whereas the second model (extended) also included the 77 additional hematology variables. A quality control protocol was performed to remove variables with no additional predictive value during model development: hematology variables with a Pearson correlation of >0.80 or with low number of unique ( n = 5) values were removed. The extended model was trained using lasso machine learning that can automatically reduce the number of variables, thereby reducing the risk of overfitting and aiding the interpretability of the model. Means and standard deviations are shown for normally distributed variables whereas medians and inter‐quartile ranges (IQR) are shown for non‐normally distributed variables. Model performance was assessed using cross validation (CV). With CV, the data are split in K number of partitions (folds), of which K‐1 folds are used for training and 1 for testing. This exercise is repeated K times resulting in K models with K performance estimates. Contrary to the conventional train‐and‐test split, multiple models are trained on multiple data splits, thereby using all data to assess the model's performance. The lasso algorithm performs shrinkage of coefficients that can get as small as 0, thereby removing variables. The lambda hyper‐parameter of lasso determines the degree of shrinkage and was optimized in a double loop cross‐validation (DLCV) scheme, also known as nested cross validation (Figure ). A K of 10 was used for both the CV and DLCV schemes. 2.5 Model evaluation The discrimination of models was assessed by plotting receiver operator characteristic (ROC) curves. The area under the ROC (AUROC) is a measure of discrimination, an AUROC of 1 indicates a perfect model, whereas an AUROC 0.5 indicates a random model. The 95% confidence interval (CI) of the AUROC was computed with the R cvAUC package by evaluating the test performances of the two model configurations trained in both CV schemes. Variable coefficients of the ten models trained in the DLCV were evaluated as variable importance (predictors). The clinical application and value of the trained models was evaluated with both calibration plots and net benefit curves. Calibration plots portray the agreement between predicted probabilities and the observed frequency of irAE. A calibration with an intercept 0 and slope of 1 shows perfect calibration, whereas a slope of >1 shows a model that overestimates outcome and a slope of <1 underestimates diagnosis. 80% and 95% CI intervals of the calibration plots were generated with the R givitR package. Net benefit is a measure to evaluate the clinical benefit of a prediction model by comparing the benefit [treating diseased, true positives (TP)] and cost [treating non‐diseased, false‐positive (FP)]. Net benefit is assessed by subtracting the cost from the benefit for the complete range of predictions values ( p t ). Formula 1 shows that the net benefit increases by the number of TP and is penalized by the number of non‐diseased (FP), especially when the prediction threshold value increases p t 1 − p t . Besides the net benefit, the number needed to treat (NNT) is shown as a comparison to how healthcare professionals consider whether the patient has a specific illness or that treatment is required. All analyses were performed in R version 4.1.2. (1) net benefit p t = TP n − FP n p t 1 − p t 2.6 Post hoc subgroup analysis To assess the independence of the identified biomarkers we adjusted for the baseline clinical variables, we performed a multivariate analysis including the identified biomarkers, age, sex, cancer type, and ICI medication. To reduce the number of coefficients and to remove groups with low prevalence, various cancer types, and ICI medications were grouped. A second post hoc analysis was performed to check whether the identified biomarkers were associated with disease severity as measured by CTCAE grade.
Study population This retrospective observational study included all visits to the ED of the University Medical Center Utrecht (UMC Utrecht) between 2013 and 2020 of patients who were being treated with any type of ICI for any type of cancer, until 3 months after cessation of treatment. Because irAE can occur even after cessation of treatment, we chose to include ED visits up to 3 months after treatment with ICI ended. The cutoff of 3 months was chosen after discussion between the authors. If patients had more than one disease episode (defined as a consecutive period with infection‐like symptoms), all patient's ED visits were included separately, whereas for patients with multiple ED visits during one disease episode, only the first visit was included. If patients visited the ED multiple times for the same condition (e.g., due to worsening of symptoms), only the first visit was included.
Data collection For all ED visits, demographic (age and sex), medication, and hematology data were extracted from the Utrecht Patient Orientated Database (UPOD). In brief, UPOD is a relational database combining clinical characteristics, medication, and laboratory measurements of patients in the UMC Utrecht since 2004. We used hematological variables measured by the CELL‐DYN Sapphire hematology analyzer ( Abbott diagnostics ). The CELL‐DYN Sapphire is a cell counter equipped with a 488‐nm blue diode laser and uses multiple techniques, such as electrical impedance, spectrophotometry, and laser light scattering, to measure morphological characteristics of leukocytes (incl. 5‐part differential), red blood cells (RBCs), and platelets for both classification and enumeration. Each time a component of a complete blood cell count (CBC) is requested, all data generated by the hematology analyzer are automatically stored in UPOD, including a substantial number of raw and research‐only values and background data on cell characteristics which are made available for research purposes. Only visits with available Sapphire data within the first 4 h after ED presentation were included in this study to ensure we only used data from patients with infection‐like symptoms during the ED visit. UPOD data acquisition and management is in accordance with current regulations concerning privacy and ethics.
irAE label definition A manual chart review was done for all ED visits within our study population by two of the authors (TVtH and BV). Visits for evidently unrelated conditions were excluded. We recorded both the preliminary and definite diagnosis. The preliminary diagnosis was defined as the diagnosis made by the treating physician in the ED and was characterized as either suspected irAE or other . The definitive diagnosis was defined as the diagnosis made by the treating physician at discharge from the hospital or at the end of treatment and was characterized as irAE or other . Ambiguous cases were resolved through consensus.
Model development Two models were trained to evaluate the added diagnostic performance of the hematology variables for irAE diagnosis. The first model (base) assessed the preliminary diagnosis, sex, and age with logistic regression thereby imitating clinical practice at the ED, whereas the second model (extended) also included the 77 additional hematology variables. A quality control protocol was performed to remove variables with no additional predictive value during model development: hematology variables with a Pearson correlation of >0.80 or with low number of unique ( n = 5) values were removed. The extended model was trained using lasso machine learning that can automatically reduce the number of variables, thereby reducing the risk of overfitting and aiding the interpretability of the model. Means and standard deviations are shown for normally distributed variables whereas medians and inter‐quartile ranges (IQR) are shown for non‐normally distributed variables. Model performance was assessed using cross validation (CV). With CV, the data are split in K number of partitions (folds), of which K‐1 folds are used for training and 1 for testing. This exercise is repeated K times resulting in K models with K performance estimates. Contrary to the conventional train‐and‐test split, multiple models are trained on multiple data splits, thereby using all data to assess the model's performance. The lasso algorithm performs shrinkage of coefficients that can get as small as 0, thereby removing variables. The lambda hyper‐parameter of lasso determines the degree of shrinkage and was optimized in a double loop cross‐validation (DLCV) scheme, also known as nested cross validation (Figure ). A K of 10 was used for both the CV and DLCV schemes.
Model evaluation The discrimination of models was assessed by plotting receiver operator characteristic (ROC) curves. The area under the ROC (AUROC) is a measure of discrimination, an AUROC of 1 indicates a perfect model, whereas an AUROC 0.5 indicates a random model. The 95% confidence interval (CI) of the AUROC was computed with the R cvAUC package by evaluating the test performances of the two model configurations trained in both CV schemes. Variable coefficients of the ten models trained in the DLCV were evaluated as variable importance (predictors). The clinical application and value of the trained models was evaluated with both calibration plots and net benefit curves. Calibration plots portray the agreement between predicted probabilities and the observed frequency of irAE. A calibration with an intercept 0 and slope of 1 shows perfect calibration, whereas a slope of >1 shows a model that overestimates outcome and a slope of <1 underestimates diagnosis. 80% and 95% CI intervals of the calibration plots were generated with the R givitR package. Net benefit is a measure to evaluate the clinical benefit of a prediction model by comparing the benefit [treating diseased, true positives (TP)] and cost [treating non‐diseased, false‐positive (FP)]. Net benefit is assessed by subtracting the cost from the benefit for the complete range of predictions values ( p t ). Formula 1 shows that the net benefit increases by the number of TP and is penalized by the number of non‐diseased (FP), especially when the prediction threshold value increases p t 1 − p t . Besides the net benefit, the number needed to treat (NNT) is shown as a comparison to how healthcare professionals consider whether the patient has a specific illness or that treatment is required. All analyses were performed in R version 4.1.2. (1) net benefit p t = TP n − FP n p t 1 − p t
Post hoc subgroup analysis To assess the independence of the identified biomarkers we adjusted for the baseline clinical variables, we performed a multivariate analysis including the identified biomarkers, age, sex, cancer type, and ICI medication. To reduce the number of coefficients and to remove groups with low prevalence, various cancer types, and ICI medications were grouped. A second post hoc analysis was performed to check whether the identified biomarkers were associated with disease severity as measured by CTCAE grade.
RESULTS 3.1 Patient characteristics Between 2013 and 2020, 409 ED visits of 257 patients who were treated with ICI and had available blood counts were included in this study (mean ED visits per patient 1.6). The irAE diagnosis of 91 visits were inconclusive from the medical records, of which the diagnosis was later adjusted in 24 cases. In both the other ( n = 268) and irAE ( n = 141) sub‐groups there were more males, 63.1% and 64.5%, respectively (Table ). Mean age did not differ between the other (62.2) and irAE group (61.7). The use of both ipilimumab and nivolumab were significantly higher in the irAE group ( p < 0.01), whereas the use of nivolumab and pembrolizumab were significantly lower in the irAE group ( p < 0.01). An overview of the irAE diagnoses is shown in Table . 3.2 Model performance After removing variables that did not meet our quality control criteria, 53 of the 77 Sapphire variables were used in the extended model (Table and Figure ). The base model had an AUROC of 0.67 (0.60–0.79 95% CI) and the extended model had an AUROC of 0.79 (0.75–0.84 95% CI), a difference in 0.13. The training performance was marginally higher for both the base and extended model as compared to the test performance, 0.74 (0.72–0.76 95% CI) and 0.86 (0.84–0.87 95% CI), respectively, providing evidence there was no overfitting. In line with the AUROC metrics, the extended model trained on all data shows the best ROC and PRC curves (Figure ). 3.3 Discriminative metrics To assess the potential value in clinical practice of the extended model, predictions of the base and extended models were evaluated with both calibration and net benefit plots. The extended model showed better calibration than the base model (Figure ). The 95% CI of the base model are very wide compared to the extended model and the predictions of the extended model are more equally distributed. In addition, decision curve analysis showed improved net benefit of the extended model as compared to the base model over the complete threshold probability range (Figure ). 3.4 Variable importance Variables' coefficients, as well as the number of times a variable was selected by the extended model, were documented during training, and are shown in Figure and Table . The preliminary diagnosis was highly predictive for irAE diagnosis in both the base and extended model with a coefficient of 3.53 ± 0.14 and 2.88 ± 0.18, respectively. The extended model also identified the following Sapphire variables as predictors for irAE diagnosis: number of eosinophils (eos), red blood cell count measured with impedance (rbci), coefficient of variance neutrophil depolarization (ndcv), and red blood cell distribution width (rdw), of which the latter was negatively associated with irAE. Eos was highly correlated with percentage of eosinophils (peos) and rbci with other red blood cell measurements variables (rbco, hgb, and hct) (Table ). The sex and age variables were not selected by lasso in any of the ten iterations in the DLCV scheme. 3.5 Post hoc subgroup analysis After adjusting for age, sex, cancer type (grouped as skin, lung, urological or other) and ICI medication (grouped as ipilimumab, nivolumab, pembrolizumab, ipilimumab, and nivolumab, or other) we found that three of the four identified variables were still significantly associated with irAE, namely: eos ( p ‐value 0.0144), rbci ( p ‐value 0.0035), and rdw ( p ‐value 0.0003). In this model we did not find a significant association for ndcv ( p ‐value 0.0781). Furthermore, we did not find an association between the values of the identified variables and the irAE severity as measured by CTCAE grade (Supplementary Figure).
Patient characteristics Between 2013 and 2020, 409 ED visits of 257 patients who were treated with ICI and had available blood counts were included in this study (mean ED visits per patient 1.6). The irAE diagnosis of 91 visits were inconclusive from the medical records, of which the diagnosis was later adjusted in 24 cases. In both the other ( n = 268) and irAE ( n = 141) sub‐groups there were more males, 63.1% and 64.5%, respectively (Table ). Mean age did not differ between the other (62.2) and irAE group (61.7). The use of both ipilimumab and nivolumab were significantly higher in the irAE group ( p < 0.01), whereas the use of nivolumab and pembrolizumab were significantly lower in the irAE group ( p < 0.01). An overview of the irAE diagnoses is shown in Table .
Model performance After removing variables that did not meet our quality control criteria, 53 of the 77 Sapphire variables were used in the extended model (Table and Figure ). The base model had an AUROC of 0.67 (0.60–0.79 95% CI) and the extended model had an AUROC of 0.79 (0.75–0.84 95% CI), a difference in 0.13. The training performance was marginally higher for both the base and extended model as compared to the test performance, 0.74 (0.72–0.76 95% CI) and 0.86 (0.84–0.87 95% CI), respectively, providing evidence there was no overfitting. In line with the AUROC metrics, the extended model trained on all data shows the best ROC and PRC curves (Figure ).
Discriminative metrics To assess the potential value in clinical practice of the extended model, predictions of the base and extended models were evaluated with both calibration and net benefit plots. The extended model showed better calibration than the base model (Figure ). The 95% CI of the base model are very wide compared to the extended model and the predictions of the extended model are more equally distributed. In addition, decision curve analysis showed improved net benefit of the extended model as compared to the base model over the complete threshold probability range (Figure ).
Variable importance Variables' coefficients, as well as the number of times a variable was selected by the extended model, were documented during training, and are shown in Figure and Table . The preliminary diagnosis was highly predictive for irAE diagnosis in both the base and extended model with a coefficient of 3.53 ± 0.14 and 2.88 ± 0.18, respectively. The extended model also identified the following Sapphire variables as predictors for irAE diagnosis: number of eosinophils (eos), red blood cell count measured with impedance (rbci), coefficient of variance neutrophil depolarization (ndcv), and red blood cell distribution width (rdw), of which the latter was negatively associated with irAE. Eos was highly correlated with percentage of eosinophils (peos) and rbci with other red blood cell measurements variables (rbco, hgb, and hct) (Table ). The sex and age variables were not selected by lasso in any of the ten iterations in the DLCV scheme.
Post hoc subgroup analysis After adjusting for age, sex, cancer type (grouped as skin, lung, urological or other) and ICI medication (grouped as ipilimumab, nivolumab, pembrolizumab, ipilimumab, and nivolumab, or other) we found that three of the four identified variables were still significantly associated with irAE, namely: eos ( p ‐value 0.0144), rbci ( p ‐value 0.0035), and rdw ( p ‐value 0.0003). In this model we did not find a significant association for ndcv ( p ‐value 0.0781). Furthermore, we did not find an association between the values of the identified variables and the irAE severity as measured by CTCAE grade (Supplementary Figure).
DISCUSSION Accurate identification of irAE in patients using ICI in the ED is of vital importance to guide treatment decisions. With new statistical methods and ML, we explored the possible added diagnostic value of 77 hematological variables measured by the CELL‐DYN Sapphire in diagnosing irAE in patients using ICI as compared to standard clinical practice. The extended model showed improvement in discrimination, calibration, and net benefit as compared to the base model, indicating that the hematological variables indeed have added value in the diagnostic process of identifying irAE in patients using ICI in the emergency department setting. Our extended model showed better performance as well as calibration over the base model. However, due to the low number of values of the base model and the good predictive performance of the preliminary diagnosis, the predictions of the base model were not equally distributed. The net benefit of the extended model was better than the base model, especially in the therapeutic range around 25%. The exact threshold for the number needed to treat will vary depending on the characteristics of the individual patient and the severity of the symptoms. A false‐positive diagnosis of irAE will lead to cessation of the checkpoint inhibitor, which would possibly withhold a life‐saving therapy from the patient. On the other hand, a false‐negative diagnosis will lead to a delayed treatment for irAE, which is potentially fatal. Of all variables, the preliminary diagnosis was deemed highly important by both the base and extended models indicating that the first diagnosis of the physician is a very good proxy for irAE diagnosis. Both age and sex showed low importance in the base model and were not selected by the lasso algorithm in any of the 10 DLCV iterations, which is in line with existing evidence. Interestingly, only a few of the 77 hematological variables were selected by the lasso algorithm in each iteration. This diagnostic study cannot not determine causality. However, a causal relationship can be postulated based on the literature. Eosinophiles are thought to play a pathogenic role in auto‐immune disorders and are known to be associated with irAE. Neutrophil depolarization is a feature of neutrophil activation, which has also been associated with auto‐immunity, but this has not been studied extensively. We found the red blood cell distribution width (rdw) to be negatively associated with irAE. Increased rdw is known to be associated with infections, which are arguably the most likely alternative diagnosis when considering irAE. Our study has some limitations. The population is highly heterogeneous, with multiple types of tumors and treatments. This may have hampered the identification of a specific predictor for a particular subset of patients. Unfortunately, we did not have enough data to stratify patients based on either cancer type or medication. Even though the post hoc group analysis showed significant results for 3 of the 4 identified variables after adjusting for the baseline characteristics, future research is needed to validate these results. Moreover, the diagnoses were retrospectively defined or changed as our data was collected on routine basis. To our knowledge, this study is one of the first of its kind in exploring the diagnostic potential of these raw and research‐only hematological variables using ML in the emergency department setting. Since the raw data from this type of hematology analyzer are not ubiquitously available, we were not able to externally validate our results. As a result, this study has to be viewed as exploratory and more research is required before these hematological variables, either individually or in a model, can be used in clinical practice. The diagnostic performance of such a model might be improved by combining hematological variables with other new sets of biomarkers, as well as the preliminary diagnosis. This study raises the question if the hematological variables might also have diagnostic value in the setting of other diseases and treatments. , , As they are inexpensive and relatively easily and rapidly obtained in general blood counts, they could be an interesting new tool in future diagnostic research. As shown here, a clinical diagnostic model may aid the clinical decision‐making process of a physician by providing a continuous prediction score that can be combined with the professional interpretation by a clinical chemist to accommodate integral diagnostics of a patient's clinical state. Instead of looking at differences between patients using cross‐sectional data, within‐patient differences may be a better approximation of a patient's health trajectory potentially allowing for predicting the incidence of irAE at the start of ICI treatment. Overall, we show that hematological variables show diagnostic performance in the identification of irAE in patients using ICI at the ED and that they have added value compared to standard diagnostic practice. Our results suggest new directions for further research using (advanced) hematological variables for irAE diagnosis in the emergency setting.
Michael S. A. Niemantsverdriet: Conceptualization (equal); formal analysis (equal); methodology (equal); writing – original draft (equal). Bram E. L. Vrijsen: Conceptualization (equal); data curation (equal); formal analysis (equal); writing – original draft (equal). Thérèse Visser 't Hooft: Data curation (equal). Karijn P. M. Suijkerbuijk: Data curation (equal). Wouter W. van Solinge: Supervision (equal). Maarten J. ten Berg: Conceptualization (equal); formal analysis (equal). Saskia Haitjema: Conceptualization (equal); data curation (equal); formal analysis (equal).
None.
MN is employed by SkylineDx, Rotterdam and receives a PhD fellowship from SkylineDx, Rotterdam. KS: Consulting/advisory relationship: Bristol Myers Squibb, Merck Sharp and Dome, Abbvie, Pierre Fabre, Novartis. Honoraria received: Novartis, Roche, Merck Sharp and Dome. Research funding, TigaTx, Bristol Myers Squibb, Philips, unrelated to this project. All paid to institution and outside the submitted work. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
This study was performed in accordance with the Declaration of Helsinki and the ethical guidelines of our institution. The institutional review board of the UMC Utrecht approved this study (reference number 20–591/C) and waived the need for informed consent as only pseudonymized data were used for this study. Data collection and handling was conducted in accordance with European privacy legislation (GDPR).
Data S1: Click here for additional data file.
|
Polymeric versus lipid nanocapsules for miconazole nitrate enhanced topical delivery: | b76c32f2-034a-488f-bdd2-92d01714171e | 8765242 | Pharmacology[mh] | Introduction The incidence of skin fungal infections is growing nowadays affecting millions of people every year worldwide (Sousa et al., ). Superficial fungal infections, commonly known to target skin, hair, nails, and mucosal tissues, are mostly caused by Cryptococcus, Candida, Aspergillus, and Pneumocystis. They are more frequent and life threatening than systemic fungal infections (Gupta et al., ; Bolla et al., ). They are more observable in patients suffering from immunosuppressing diseases like AIDS, tuberculosis, cancer, and chronic obstructive pulmonary disease (Bolla et al., ). Recently, fungal infections were highly observed in patients suffering from severe COVID-19 symptoms followed by post COVID 19 syndrome. The most common fungal infections in patients with COVID-19 include aspergillosis or invasive candidiasis (Hoenigl, ; El-Kholy et al., ). About half of the patients suffering from immune-compromised issues are susceptible to be infected with Candida albicans , the most abundant fungal pathogen affecting humans (Maródi & Johnston, ). Symptoms of candidiasis are various ranging from severe redness, swelling, itching resulting in skin fissures or sores, white patches to pelvic pain and bloody urine according to site of infection (Mayer et al., ). Topical anti-fungal drug delivery is looked up to as the most preferable pathway in treatment of major superficial skin fungal infections, ensuring its direct access and higher retention rate at the target (Garg et al., ). Topical delivery further subsidizes limitation of pre-systemic metabolism of the drug to enhance bioavailability. Moreover, it increases the efficacy of treatment, allows potential self-medication, and increased patient compliance (Bolla et al., ). However, topical preparations showing poor skin penetration caused adverse skin reactions such as skin irritation, allergic reactions, and itching (Kulawik-Pióro & Miastkowska, ). They also showed variable drug levels at the site of infection with unfortunate dermal and ungual bioavailability. Therefore, novel drug delivery system is envisaged to address the problems associated with topical preparations reducing local side effects and increasing their therapeutic efficacy. Some of these novel carriers are liposomes, niosomes, solid lipid nanoparticles (SLNs), nano-lipid carriers, and polymeric nanoparticles which highly evolve the advent of targeted delivery and drug stability (Verma & Utreja, ; Montoto et al., ). Miconazole nitrate (MN) is a broad-spectrum antifungal drug of the azole derivatives extensively used for the treatment of dermatophytosis, cutaneous mycosis, and fungal infections affecting the vagina, mouth and skin, including candidiasis (Aljaeid & Hosny, ). Miconazole is a weak base (pKa = 6.7) with poor aqueous solubility (1.03 µg/mL) and rapid clearance hindered its systemic efficacy (Jain et al., ). MN acts by dual pathways: impedes the synthesis of ergosterol on fungal cell and causes accretion of reactive oxygen species (ROS) in the fungal cell, triggering oxidative damage, and cell death (Amaral et al., ). Currently, miconazole is available as conventional topical formulations such as cream, lotions, spray liquids, and suppository for vaginal use. Previously published reports stated that miconazole topical applications exhibited poor skin penetration and therefore higher doses are required to recompense low permeability (Qushawy et al., ). Moreover, it suffers from vast side effects on the application site like burning, redness, and swelling (Kenechukwu et al., ). Therefore, various approaches have been attempted to overcome these hitches and consequently improve the therapeutic efficacy of drugs such as SLNs (Aljaeid & Hosny, ), nano-suspensions (Cerdeira et al., ), transdermal films (Ofokansi et al., ), ethosomes, liposomes, and nanostructured lipid carriers (NLCs) (Firooz et al., ). Recently, nanocapsules attracted more interest in drug delivery applications, profiting from their core–shell nanostructure. Nanocapsules are distinctive class of nanoparticles, nanosized carrier composed of two main parts (core) which is oily in nature and a protective thin polymer matrix (shell) constrained therapeutic substance (Nilewar et al., ). They are fruitfully encapsulating both hydrophilic and lipophilic drugs according to their natures with high drug loading (%DL) capacity. Moreover, they enhanced drug solubility, controlled drug release declining burst release induced by pH, temperature, enzymes, and other factors and improve drug stability (Deng et al., ). Additionally, they reported to improve the efficacy of candidiasis pharmacological treatment. Nanocapsules were classified into polymeric nanocapsules (PNCs) and lipid nanocapsules (LNCs), which are different in their compositions. PNCs are composed of a core enclosed by drugs inside and enfolded by polymeric membrane (Nagaich, ), while LNCs are hybrid between both liposomes and nano-emulsions which are composed of a liquid, oily core (medium-chain triglycerides) enclosed by both hydrophilic and lipophilic surfactants (Saliou et al., ). Both are used for the delivery of peptides, proteins, genes, and several compounds. The most commonly used polymers for forming the shell are biocompatible and biodegradable such as poly lactic acid, polylactide-co-glycolide, and poly (Ɛ-caprolactone). Polycaprolactone (PCL) is widely used to prepare different types of nanocapsules comparing to other biodegradable polymers owing to its talented characters; safety, elasticity, cytocompatibility, and very slow biodegradation which were approved by US-FDA (El-Hesaisy & Swidan, ; Rahat et al., ). It is also useful to capture a wide range of drugs such as anti-cancer drugs, anti-inflammatories, and immune modulators presenting diverse flexibility with potential applications in therapy (Mahareek et al., ). PCL nanocapsules are malleable formulations used in the liquid form as well as can be incorporated into semi-solid or solid dosage forms. In LNCs, medium and long chain fatty acids are commonly used such as propylene glycol (PG) dicaprylocaprate (Labrafac ® ) and oleic acids (Pohlmann et al., ). They are synthesized by nanoprecipitation, emulsification, solvent displacement, and solvent evaporation techniques (Govender et al., ). Among these methods, emulsification technique was elected in this study to prepare MN loaded nanocapsules as it is simple, inexpensive, and most widely used method in preparing nanocapsules suspension system (Deng et al., ). Our study aimed to prepare and evaluate both polymeric and LNCs for topical delivery of MN to provide an innovative way to enhance its antifungal activity with minimal side effects, reducing the dose and dosing frequency associated with other conventional topical drug delivery system. Developed nanocapsules were characterized with respect to particle size, drug entrapment, surface morphology, in vitro diffusion and stability studies, and ex vivo retention and permeation. The effect on cell viability and antifungal activity was also investigated.
Materials and methods 2.1. Materials Miconazole nitrate was kindly donated from Sedico Pharmaceutical Company (6th of October City, Egypt). Polycaprolactone was purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). Propylene glycol dicaprylocaprate (Labrafac ® ) was given as a gift by Gattefosse (Saint-Priest, France). Sodium dihydrogen phosphate and disodium hydrogen phosphate were obtained from El-Gomhouria Chemicals Pharmaceutical Company (Cairo, Egypt). Propylene glycol, dichloromethane, and Tween 80 were all pharmaceutical grades and were attained from El-Nasr Chemicals Pharmaceuticals Company (Cairo, Egypt). WISH cell line (normal human epithelial amniotic cells) was obtained at National Cancer Institute (NCI) (Cairo, Egypt). RPMI1640 nutritional media was purchased from Lonza Bioscience (Biological Products Company) (Morristown, NJ). 2.2. Preparation of MN loaded polymeric nanocapsules Polymeric nanocapsules (PNCs) were prepared by emulsification/nanoprecipitation method with limited modifications (Xia et al., ; Jahangir et al., ; Deng et al., ; Oliveira et al., ). It is the most simple and cost effective method. PCL biodegradable polymer is employed in PNCs. The organic and aqueous phases were prepared separately. The organic phase containing MN and poly (ε-caprolactone) (PCL) dissolved in 10 mL dichloromethane at ratios 1:2, 1:3, and 1:4 under vigorous stirring (1200 rpm) at room temperature using magnetic stirrer for 15 min . Meanwhile, the aqueous phase was composed of aqueous solution of Tween 80 (5% w/v) and 10% glycerol (based on preformulation studies to screen the effect of surfactant type) (Jahangir et al., ). The aqueous phase was added dropwise using syringe to the previous organic phase and the mixture was stirred at 2000 rpm for 30 min at room temperature to obtain the required PNCs suspension. The prepared PNCs were centrifuged at 4000 rpm for 15 min and the obtained pellets were re-dispersed again in de-ionized water. Centrifugation cycle was repeated three times to get rid of organic solvents that have to be strictly eliminated from the formulations. The effect different ratios of MN:PCL (1:2, 1:3, and 1:4) on PS, ZP, and %EE were further studied. 2.3. Preparation of MN loaded lipid nanocapsules Based on preformulation studies (data unseen) and previous reports, MN loaded LNCs were prepared using lipid matrix of both oleic acid and Labrafac ® oil with ratio 1:1 (Kamel & Basha, 2013; Eissa et al., ; Kiani et al., ). The previously mentioned oil mixture has shown high solubilization power for MN referred to the long chain of oleic acid and the HLB value. Briefly, 20 mg of MN was mixed with lipid matrix (melted in water bath at 80 °C) at different ratios (1:2, 1:3, and 1:4) where soy phosphatidylcholine was used as surfactant . A hot aqueous solution was prepared by dissolving soy phosphatidylcholine (5 g) in PG (10 g) in a mass ratio of 1:2. The aqueous solution was added using syringe to the previous lipid phase and the mixture was stirred at 1000 rpm for 15 min at room temperature to obtain the required nanocapsules suspension. The obtained LNCs suspensions were ultrasonicated for 5 min using probe sonicator (UP50H, Hielscher, Teltow, Germany). The composition of prepared polymeric and LNCs formulations is shown in . The effect of different ratios of MN:lipid matrix (1:2, 1:3, and 1:4) on PS, ZP, and %EE was further studied. 2.4. Characterization of developed PNCs and LNCs 2.4.1. Particle size, polydispersity index, and zeta potential The average particle size, polydispersity, and zeta potential of the MN loaded nanocapsules were determined using dynamic light scattering integrated in a zeta-sizer Nano-ZS (Malvern Instruments Ltd., Worcestershire, UK). Five milligrams of samples were diluted with a fixed amount of de-ionized water (10 mL) to obtain a suitable scattering intensity, filtered using a 0.22 μm filter (Millipore Co., Billerica, MA) and placed into disposal cuvettes (size). Three measurements were performed for each sample at an angle of 90° at room temperature (25 °C) using two refractive indexes (1.46 and 1.63) for polymer PCL and MN, respectively. Polydispersity index (PDI) was determined for assessing the particle size distribution and the homogeneity of the nanocapsules. Zeta potential was also determined to confirm the stability of the nanocapsules (Danaei et al., ). 2.4.2. Determination of the drug loading and entrapment efficiency (%EE) Drug loading and %EE of MN in the prepared nanocapsules were determined indirectly (El-Leithy & Abdel-Rashid, 2017). The concentration of free MN was measured in aqueous supernatant solution after separation of nanocapsules by centrifugation for 20 min at 10,000 rpm at 4 °C in high-speed cooling centrifuge (XCHR20, Bio Lion, Shanghai, China). The supernatant was filtrated through 0.22 µm membrane filter and the amount of MN entrapped was detected spectrophotometrically (Perkin Elmer UV, Yokohama, Japan) at λ max 280 nm after suitable dilution with methanol. Each experiment was carried out in triplicate and the mean value was deduced. The %DL and %EE were calculated by the following equations: % Drug loading content = weight of the drug in NCs weight of the NCs × 100 % Encapsulation efficiency = initial amount of MN added to NCs − free MN in supernatant initial amount of MN added × 100 2.5. Transmission electron microscopy (TEM) The shape and outlines of prepared NCs were inspected using TEM (Jeol, JEM, Tokyo, Japan). Freshly prepared samples (diluted appropriately with 0.1 M phosphate buffer) were deposited onto the surface of carbon coated copper grids; natively stained with 2% phosphotungstic acid and dried at room temperature. The stained sample was then probed and visualized using TEM (Mora-Huertas et al., ). 2.6. In vitro release study for prepared MN nanocapsules The release of MN from PNCs and LNCs formulations was investigated in phosphate buffer (pH = 7.4) solution using dialysis bag (regular dialysis) method (Govender et al., ; Dar et al., ). The dialysis bag (molecular weight cut off: 12–14 kDa, Livingstone, Sydney, Australia) was first soaked in phosphate-buffered saline (PBS) at pH 7.4 overnight before use. Nanocapsules suspension of selected optimized formulations equivalent to 50 µg of MN was placed inside the dialysis bag, tied at both ends and immersed in 100 mL of PBS (pH 7.4, 37 °C) release medium containing 0.25% sodium lauryl sulfate. The sinking conditions were taken into consideration. The solution was stirred at 100 rpm with the help of the magnetic stirrer at 37 ± 0.5 °C. At scheduled time intervals, 2 mL of the release media was removed, filtered through a 0.22 μm Cameo Acetate membrane filter (Millipore Co., Billerica, MA) and replaced by fresh release medium. The withdrawn samples were analyzed spectrophotometrically (Perkin Elmer UV, Yokohama, Japan) at λ max 280 nm for MN content. For the sake of comparison, release pattern of 5 mg free MN suspension from dialysis bag was also conducted at the same conditions. The release data were subsequently fitted to different release kinetic models representing (zero-order, first-order, Higuchi diffusion, Hixon and Baker equations) to determine the release kinetics. Correlation coefficient ( R 2 ) values were compared for selection of the most appropriate release model that best fits the data (Nasr et al., ). 2.7. Effect of storage on stability of prepared nanocapsules The stability of optimum MN loaded nanocapsules suspensions was studied by storage of three samples in sealed vials at room temperature 25 °C for 3 months. During this period, %EE, PS, PDI, and ZP of the nanocapsules were measured as described previously. Nanocapsules were examined visually for aggregation and change in their appearance. Statistical significance was analyzed by Student’s t -test using SPSS ® software 22.0 (SPSS, Chicago, IL). Difference at p >.05 was considered insignificant. 2.8. Antifungal potential of MN-loaded NCs The antifungal effect of the optimized MN-loaded NCs was studied using diffusion agar method (Dudhipala & Ay, ). One milliliter of standard strain of Candida albicans ATCC 76615 (1 × 10 6 CFU/mL) was cultivated, then inoculated in petri dishes of 150 mm diameter containing 50 mL of the Müller–Hinton agar. Holes of 10 mm diameter were made and filled with 100 µL containing 5 mg of drug suspension or the equivalent amount optimized MN-loaded NCs. The petri dishes were incubated for 4 h at 37 °C. The area where there is disappearance of fungal growth around the holes (inhibition zone) was measured using a caliper. For a full view, the antifungal activity of a positive control (free MN suspension) was implemented (Ahmed & Aljaeid, ). 2.9. Effect of nanocapsules on cell viability (cytotoxicity and genotoxicity) 2.9.1. Cell culture and treatment Remembering that cell culture is a mirror environment for the bio-internal one, the WISH cell line (normal human epithelial amniotic cells) was kept at 37 °C under 5% CO 2 using a water jacketed carbon dioxide incubator. In 96-well microliter plastic plates at concentration of 10 × 10 3 cells/well, the cells were cultivated for five days at sterile area using a laminar flow cabinet biosafety class II level in a specific nutritional medium (RPMI 1640), with 1% antibiotic–antimycotic mixture supplement (10,000 µg/mL potassium penicillin, 10,000 µg/mL streptomycin sulfate, and 25 µg/mL amphotericin B), 10% fetal bovine serum and 1% l -glutamine (El-Leithy et al., ). The media of different plates were aspirated and replaced with fresh medium. The WISH cells incubated in fresh medium were taken as a negative control. However, other cell line plates were treated with free MN, and the equivalent weights of selected nanocapsules formulations. The samples were prepared to reach various concentrations of drugs (0.25–100 µg/mL). 2.9.2. Determination of cell viability (cytotoxicity) The effect of free and encapsulated MN on viability of WISH amniotic cells was studied using MTT assay (Qi et al., ). MTT salt (2.5 μg/mL) was added to each well to be incubated for 4 h at 37 °C under 5% CO 2 . Unbound MTT and dead cells found in each well were removed by suction and subsequently exchanged with 200 μL of 10% sodium dodecyl sulfate. All experimental assays were carried out in triplicate. A cytotoxic natural agent that gives 100% lethality positive control was used as positive control under the same conditions. The plates were then read at λ max 595 nm using a microplate multi-well reader. The percentage of change in viability was calculated according to the formula: Viability % = ( optical density of sample / optical density of control ) × 100 A probit analysis was conducted to determine LC50 using SPSS 11 program (SPSS, Chicago, IL) and the microplates were photographed using inverted microscope. The LC50 is the lethal concentration of the sample which causes the death of 50% of cells in 48 h. 2.9.3. Comet assay (genotoxicity) Comparing to other assays, the comet response in detecting DNA damages was elected as it was the more sensitive, rapid, and reproducible assay (Gunasekarana et al., ). Comet assay analysis was implicated to investigate the effect on DNA of WISH amniotic cells which may give an indication for death pathway. Trevigen's Comet Assay ® kit (Trevigen, Inc., Gaithersburg, MD) and Comet Image Analysis System software (Comet Scores software; TriTek, Sumerduck, VA) were used to analyze; tail length, % of DNA in the tail, and tail moment. 2.10. Ex vivo skin permeation and retentivity This study evaluated the ability of PNCs and LNCs as drug delivery systems to enhance topical permeation/retention of MN (Jain et al., ; Nasr et al., ). The protocol of the study was approved by the Animal Ethics Committee of Faculty of Pharmacy, Helwan University. The experiment was conducted using excised full thickness dehaired abdominal rat skin (male albino rats, Sprague-Dawley; 100 g). The skin sections were installed on modified Franz diffusion cells (Crown Glass Co., Somerville, NJ) with an available permeability surface of 1.76 cm 2 and a receptor volume of 50 mL such that the dermal side of the skin faced to the receptor fluid (PBS; pH 7.4). One milliliter of optimized MN-loaded NCs formulas (equivalent to 5 mg MN) were placed in the donor compartment at 37 °C. The sampling was performed at various intervals up to 24 h, where MN content was estimated spectrophotometrically at 280 nm (Salah et al., ). Subsequently, the skin was removed and washed 10 times with a cotton swab followed by weighing and homogenizing in methanol. The produced solution was centrifuged for 10 min at 5000 rpm and supernatant was filtered then analyzed for drug amount spectrophotometrically to determine percentage drug skin deposition (Nasr et al., ). A similar study was also performed for free MN suspension for sake of comparison. The study was carried out in triplicate for both. 2.11. Statistical analysis Data inspection was accomplished using GraphPad InStat 3 program (GraphPad Software, La Jolla, CA). Results were stated as a mean ± standard deviation. Statistically significant difference was determined using one-way ANOVA test and paired and un-paired Student’s t -test with p <.05 as a minimal level of significance.
Materials Miconazole nitrate was kindly donated from Sedico Pharmaceutical Company (6th of October City, Egypt). Polycaprolactone was purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). Propylene glycol dicaprylocaprate (Labrafac ® ) was given as a gift by Gattefosse (Saint-Priest, France). Sodium dihydrogen phosphate and disodium hydrogen phosphate were obtained from El-Gomhouria Chemicals Pharmaceutical Company (Cairo, Egypt). Propylene glycol, dichloromethane, and Tween 80 were all pharmaceutical grades and were attained from El-Nasr Chemicals Pharmaceuticals Company (Cairo, Egypt). WISH cell line (normal human epithelial amniotic cells) was obtained at National Cancer Institute (NCI) (Cairo, Egypt). RPMI1640 nutritional media was purchased from Lonza Bioscience (Biological Products Company) (Morristown, NJ).
Preparation of MN loaded polymeric nanocapsules Polymeric nanocapsules (PNCs) were prepared by emulsification/nanoprecipitation method with limited modifications (Xia et al., ; Jahangir et al., ; Deng et al., ; Oliveira et al., ). It is the most simple and cost effective method. PCL biodegradable polymer is employed in PNCs. The organic and aqueous phases were prepared separately. The organic phase containing MN and poly (ε-caprolactone) (PCL) dissolved in 10 mL dichloromethane at ratios 1:2, 1:3, and 1:4 under vigorous stirring (1200 rpm) at room temperature using magnetic stirrer for 15 min . Meanwhile, the aqueous phase was composed of aqueous solution of Tween 80 (5% w/v) and 10% glycerol (based on preformulation studies to screen the effect of surfactant type) (Jahangir et al., ). The aqueous phase was added dropwise using syringe to the previous organic phase and the mixture was stirred at 2000 rpm for 30 min at room temperature to obtain the required PNCs suspension. The prepared PNCs were centrifuged at 4000 rpm for 15 min and the obtained pellets were re-dispersed again in de-ionized water. Centrifugation cycle was repeated three times to get rid of organic solvents that have to be strictly eliminated from the formulations. The effect different ratios of MN:PCL (1:2, 1:3, and 1:4) on PS, ZP, and %EE were further studied.
Preparation of MN loaded lipid nanocapsules Based on preformulation studies (data unseen) and previous reports, MN loaded LNCs were prepared using lipid matrix of both oleic acid and Labrafac ® oil with ratio 1:1 (Kamel & Basha, 2013; Eissa et al., ; Kiani et al., ). The previously mentioned oil mixture has shown high solubilization power for MN referred to the long chain of oleic acid and the HLB value. Briefly, 20 mg of MN was mixed with lipid matrix (melted in water bath at 80 °C) at different ratios (1:2, 1:3, and 1:4) where soy phosphatidylcholine was used as surfactant . A hot aqueous solution was prepared by dissolving soy phosphatidylcholine (5 g) in PG (10 g) in a mass ratio of 1:2. The aqueous solution was added using syringe to the previous lipid phase and the mixture was stirred at 1000 rpm for 15 min at room temperature to obtain the required nanocapsules suspension. The obtained LNCs suspensions were ultrasonicated for 5 min using probe sonicator (UP50H, Hielscher, Teltow, Germany). The composition of prepared polymeric and LNCs formulations is shown in . The effect of different ratios of MN:lipid matrix (1:2, 1:3, and 1:4) on PS, ZP, and %EE was further studied.
Characterization of developed PNCs and LNCs 2.4.1. Particle size, polydispersity index, and zeta potential The average particle size, polydispersity, and zeta potential of the MN loaded nanocapsules were determined using dynamic light scattering integrated in a zeta-sizer Nano-ZS (Malvern Instruments Ltd., Worcestershire, UK). Five milligrams of samples were diluted with a fixed amount of de-ionized water (10 mL) to obtain a suitable scattering intensity, filtered using a 0.22 μm filter (Millipore Co., Billerica, MA) and placed into disposal cuvettes (size). Three measurements were performed for each sample at an angle of 90° at room temperature (25 °C) using two refractive indexes (1.46 and 1.63) for polymer PCL and MN, respectively. Polydispersity index (PDI) was determined for assessing the particle size distribution and the homogeneity of the nanocapsules. Zeta potential was also determined to confirm the stability of the nanocapsules (Danaei et al., ). 2.4.2. Determination of the drug loading and entrapment efficiency (%EE) Drug loading and %EE of MN in the prepared nanocapsules were determined indirectly (El-Leithy & Abdel-Rashid, 2017). The concentration of free MN was measured in aqueous supernatant solution after separation of nanocapsules by centrifugation for 20 min at 10,000 rpm at 4 °C in high-speed cooling centrifuge (XCHR20, Bio Lion, Shanghai, China). The supernatant was filtrated through 0.22 µm membrane filter and the amount of MN entrapped was detected spectrophotometrically (Perkin Elmer UV, Yokohama, Japan) at λ max 280 nm after suitable dilution with methanol. Each experiment was carried out in triplicate and the mean value was deduced. The %DL and %EE were calculated by the following equations: % Drug loading content = weight of the drug in NCs weight of the NCs × 100 % Encapsulation efficiency = initial amount of MN added to NCs − free MN in supernatant initial amount of MN added × 100
Particle size, polydispersity index, and zeta potential The average particle size, polydispersity, and zeta potential of the MN loaded nanocapsules were determined using dynamic light scattering integrated in a zeta-sizer Nano-ZS (Malvern Instruments Ltd., Worcestershire, UK). Five milligrams of samples were diluted with a fixed amount of de-ionized water (10 mL) to obtain a suitable scattering intensity, filtered using a 0.22 μm filter (Millipore Co., Billerica, MA) and placed into disposal cuvettes (size). Three measurements were performed for each sample at an angle of 90° at room temperature (25 °C) using two refractive indexes (1.46 and 1.63) for polymer PCL and MN, respectively. Polydispersity index (PDI) was determined for assessing the particle size distribution and the homogeneity of the nanocapsules. Zeta potential was also determined to confirm the stability of the nanocapsules (Danaei et al., ).
Determination of the drug loading and entrapment efficiency (%EE) Drug loading and %EE of MN in the prepared nanocapsules were determined indirectly (El-Leithy & Abdel-Rashid, 2017). The concentration of free MN was measured in aqueous supernatant solution after separation of nanocapsules by centrifugation for 20 min at 10,000 rpm at 4 °C in high-speed cooling centrifuge (XCHR20, Bio Lion, Shanghai, China). The supernatant was filtrated through 0.22 µm membrane filter and the amount of MN entrapped was detected spectrophotometrically (Perkin Elmer UV, Yokohama, Japan) at λ max 280 nm after suitable dilution with methanol. Each experiment was carried out in triplicate and the mean value was deduced. The %DL and %EE were calculated by the following equations: % Drug loading content = weight of the drug in NCs weight of the NCs × 100 % Encapsulation efficiency = initial amount of MN added to NCs − free MN in supernatant initial amount of MN added × 100
Transmission electron microscopy (TEM) The shape and outlines of prepared NCs were inspected using TEM (Jeol, JEM, Tokyo, Japan). Freshly prepared samples (diluted appropriately with 0.1 M phosphate buffer) were deposited onto the surface of carbon coated copper grids; natively stained with 2% phosphotungstic acid and dried at room temperature. The stained sample was then probed and visualized using TEM (Mora-Huertas et al., ).
In vitro release study for prepared MN nanocapsules The release of MN from PNCs and LNCs formulations was investigated in phosphate buffer (pH = 7.4) solution using dialysis bag (regular dialysis) method (Govender et al., ; Dar et al., ). The dialysis bag (molecular weight cut off: 12–14 kDa, Livingstone, Sydney, Australia) was first soaked in phosphate-buffered saline (PBS) at pH 7.4 overnight before use. Nanocapsules suspension of selected optimized formulations equivalent to 50 µg of MN was placed inside the dialysis bag, tied at both ends and immersed in 100 mL of PBS (pH 7.4, 37 °C) release medium containing 0.25% sodium lauryl sulfate. The sinking conditions were taken into consideration. The solution was stirred at 100 rpm with the help of the magnetic stirrer at 37 ± 0.5 °C. At scheduled time intervals, 2 mL of the release media was removed, filtered through a 0.22 μm Cameo Acetate membrane filter (Millipore Co., Billerica, MA) and replaced by fresh release medium. The withdrawn samples were analyzed spectrophotometrically (Perkin Elmer UV, Yokohama, Japan) at λ max 280 nm for MN content. For the sake of comparison, release pattern of 5 mg free MN suspension from dialysis bag was also conducted at the same conditions. The release data were subsequently fitted to different release kinetic models representing (zero-order, first-order, Higuchi diffusion, Hixon and Baker equations) to determine the release kinetics. Correlation coefficient ( R 2 ) values were compared for selection of the most appropriate release model that best fits the data (Nasr et al., ).
Effect of storage on stability of prepared nanocapsules The stability of optimum MN loaded nanocapsules suspensions was studied by storage of three samples in sealed vials at room temperature 25 °C for 3 months. During this period, %EE, PS, PDI, and ZP of the nanocapsules were measured as described previously. Nanocapsules were examined visually for aggregation and change in their appearance. Statistical significance was analyzed by Student’s t -test using SPSS ® software 22.0 (SPSS, Chicago, IL). Difference at p >.05 was considered insignificant.
Antifungal potential of MN-loaded NCs The antifungal effect of the optimized MN-loaded NCs was studied using diffusion agar method (Dudhipala & Ay, ). One milliliter of standard strain of Candida albicans ATCC 76615 (1 × 10 6 CFU/mL) was cultivated, then inoculated in petri dishes of 150 mm diameter containing 50 mL of the Müller–Hinton agar. Holes of 10 mm diameter were made and filled with 100 µL containing 5 mg of drug suspension or the equivalent amount optimized MN-loaded NCs. The petri dishes were incubated for 4 h at 37 °C. The area where there is disappearance of fungal growth around the holes (inhibition zone) was measured using a caliper. For a full view, the antifungal activity of a positive control (free MN suspension) was implemented (Ahmed & Aljaeid, ).
Effect of nanocapsules on cell viability (cytotoxicity and genotoxicity) 2.9.1. Cell culture and treatment Remembering that cell culture is a mirror environment for the bio-internal one, the WISH cell line (normal human epithelial amniotic cells) was kept at 37 °C under 5% CO 2 using a water jacketed carbon dioxide incubator. In 96-well microliter plastic plates at concentration of 10 × 10 3 cells/well, the cells were cultivated for five days at sterile area using a laminar flow cabinet biosafety class II level in a specific nutritional medium (RPMI 1640), with 1% antibiotic–antimycotic mixture supplement (10,000 µg/mL potassium penicillin, 10,000 µg/mL streptomycin sulfate, and 25 µg/mL amphotericin B), 10% fetal bovine serum and 1% l -glutamine (El-Leithy et al., ). The media of different plates were aspirated and replaced with fresh medium. The WISH cells incubated in fresh medium were taken as a negative control. However, other cell line plates were treated with free MN, and the equivalent weights of selected nanocapsules formulations. The samples were prepared to reach various concentrations of drugs (0.25–100 µg/mL). 2.9.2. Determination of cell viability (cytotoxicity) The effect of free and encapsulated MN on viability of WISH amniotic cells was studied using MTT assay (Qi et al., ). MTT salt (2.5 μg/mL) was added to each well to be incubated for 4 h at 37 °C under 5% CO 2 . Unbound MTT and dead cells found in each well were removed by suction and subsequently exchanged with 200 μL of 10% sodium dodecyl sulfate. All experimental assays were carried out in triplicate. A cytotoxic natural agent that gives 100% lethality positive control was used as positive control under the same conditions. The plates were then read at λ max 595 nm using a microplate multi-well reader. The percentage of change in viability was calculated according to the formula: Viability % = ( optical density of sample / optical density of control ) × 100 A probit analysis was conducted to determine LC50 using SPSS 11 program (SPSS, Chicago, IL) and the microplates were photographed using inverted microscope. The LC50 is the lethal concentration of the sample which causes the death of 50% of cells in 48 h. 2.9.3. Comet assay (genotoxicity) Comparing to other assays, the comet response in detecting DNA damages was elected as it was the more sensitive, rapid, and reproducible assay (Gunasekarana et al., ). Comet assay analysis was implicated to investigate the effect on DNA of WISH amniotic cells which may give an indication for death pathway. Trevigen's Comet Assay ® kit (Trevigen, Inc., Gaithersburg, MD) and Comet Image Analysis System software (Comet Scores software; TriTek, Sumerduck, VA) were used to analyze; tail length, % of DNA in the tail, and tail moment.
Cell culture and treatment Remembering that cell culture is a mirror environment for the bio-internal one, the WISH cell line (normal human epithelial amniotic cells) was kept at 37 °C under 5% CO 2 using a water jacketed carbon dioxide incubator. In 96-well microliter plastic plates at concentration of 10 × 10 3 cells/well, the cells were cultivated for five days at sterile area using a laminar flow cabinet biosafety class II level in a specific nutritional medium (RPMI 1640), with 1% antibiotic–antimycotic mixture supplement (10,000 µg/mL potassium penicillin, 10,000 µg/mL streptomycin sulfate, and 25 µg/mL amphotericin B), 10% fetal bovine serum and 1% l -glutamine (El-Leithy et al., ). The media of different plates were aspirated and replaced with fresh medium. The WISH cells incubated in fresh medium were taken as a negative control. However, other cell line plates were treated with free MN, and the equivalent weights of selected nanocapsules formulations. The samples were prepared to reach various concentrations of drugs (0.25–100 µg/mL).
Determination of cell viability (cytotoxicity) The effect of free and encapsulated MN on viability of WISH amniotic cells was studied using MTT assay (Qi et al., ). MTT salt (2.5 μg/mL) was added to each well to be incubated for 4 h at 37 °C under 5% CO 2 . Unbound MTT and dead cells found in each well were removed by suction and subsequently exchanged with 200 μL of 10% sodium dodecyl sulfate. All experimental assays were carried out in triplicate. A cytotoxic natural agent that gives 100% lethality positive control was used as positive control under the same conditions. The plates were then read at λ max 595 nm using a microplate multi-well reader. The percentage of change in viability was calculated according to the formula: Viability % = ( optical density of sample / optical density of control ) × 100 A probit analysis was conducted to determine LC50 using SPSS 11 program (SPSS, Chicago, IL) and the microplates were photographed using inverted microscope. The LC50 is the lethal concentration of the sample which causes the death of 50% of cells in 48 h.
Comet assay (genotoxicity) Comparing to other assays, the comet response in detecting DNA damages was elected as it was the more sensitive, rapid, and reproducible assay (Gunasekarana et al., ). Comet assay analysis was implicated to investigate the effect on DNA of WISH amniotic cells which may give an indication for death pathway. Trevigen's Comet Assay ® kit (Trevigen, Inc., Gaithersburg, MD) and Comet Image Analysis System software (Comet Scores software; TriTek, Sumerduck, VA) were used to analyze; tail length, % of DNA in the tail, and tail moment.
Ex vivo skin permeation and retentivity This study evaluated the ability of PNCs and LNCs as drug delivery systems to enhance topical permeation/retention of MN (Jain et al., ; Nasr et al., ). The protocol of the study was approved by the Animal Ethics Committee of Faculty of Pharmacy, Helwan University. The experiment was conducted using excised full thickness dehaired abdominal rat skin (male albino rats, Sprague-Dawley; 100 g). The skin sections were installed on modified Franz diffusion cells (Crown Glass Co., Somerville, NJ) with an available permeability surface of 1.76 cm 2 and a receptor volume of 50 mL such that the dermal side of the skin faced to the receptor fluid (PBS; pH 7.4). One milliliter of optimized MN-loaded NCs formulas (equivalent to 5 mg MN) were placed in the donor compartment at 37 °C. The sampling was performed at various intervals up to 24 h, where MN content was estimated spectrophotometrically at 280 nm (Salah et al., ). Subsequently, the skin was removed and washed 10 times with a cotton swab followed by weighing and homogenizing in methanol. The produced solution was centrifuged for 10 min at 5000 rpm and supernatant was filtered then analyzed for drug amount spectrophotometrically to determine percentage drug skin deposition (Nasr et al., ). A similar study was also performed for free MN suspension for sake of comparison. The study was carried out in triplicate for both.
Statistical analysis Data inspection was accomplished using GraphPad InStat 3 program (GraphPad Software, La Jolla, CA). Results were stated as a mean ± standard deviation. Statistically significant difference was determined using one-way ANOVA test and paired and un-paired Student’s t -test with p <.05 as a minimal level of significance.
Results and discussion 3.1. Particle size, PDI, and zeta potential The results of particle size, PDI, and ZP of the prepared nanocapsules are displayed in . The measured particle size was in the nanometric size range, which is preferred for drug delivery and penetrability to biological cell (Subramaniam et al., ). The average particle size of PNCs suspensions (F1, F2, and F3) ranged from 108 ± 3.63 nm and 180 ± 2.11 nm. The results obviously showed a statistically significant directly proportional function between drug–polymer ratio (MN:PCL) and average particle size of nanocapsules suspensions at constant surfactant concentration Tween 80 (5% w/v). These results were in line with findings of Sathyamoorthy et al. ; who reported that the particle size increases by increasing PCL concentration which results in increased incidence of collisions between particles during emulsification consequently aggregation of small sized particles yielded large particle size. Moreover, the increase of PCL concentration increases viscosity and resistance of the organic phase to diffuse and distribute into the aqueous phase that prompts the formation of large nanodroplets at the interface (Ajiboye et al., ). Additionally, the high viscosity could decline the drug diffusion from the nanocapsules as another cause for increasing particle size (Tavares et al., ). Tween 80 (HLB 15) was chosen as a surfactant after preliminary studies compared to span 60 (4.7) (data unseen). It was found that hydrophilic surfactants strongly participate in decreasing particle size of nanocapsules by causing high droplets stabilization and more flexibility (Zhu et al., ). Furthermore, the zeta potential of PNCs suspensions (F1, F2, and F3) was −31 ± 3.10, −35.03 ± 4.80, and −40 ± 5.21, respectively. PCL is known for yielding spherical nanocapsules with negative charge on the surface (Rahat et al., ). They displayed negative charges owing to PCL structure and its hydrophobic nature which may cause ionization of the carboxylic groups resulting in a negative potential to the interface (Michels et al., ; Lino et al., ). Therefore, increased concentration of PCL influences negativity of zeta potential of prepared nanocapsules. The high negative zeta potential owing to the charges repulsion toward the natural tendency of aggregation affords stability to nanocapsules from the hostile pH of the biological system (Alves et al., ; Rahat et al., ). On the other hand, the results displayed a negative correlation between MN:lipid ratio and average particle size of LNCs at constant soy phosphatidylcholine concentration (5 g) with values ranging between 89 ± 3.63 and 116 ± 1.63 nm. By increasing lipid content, the average particle size decreased; this could be revealed to the combination of labrafac and oleic acid which attained good solubilization for MN and proper self-emulsification yielded small droplet size (Kamel & Basha, ). Previously, published reports stated that short chain fatty acid ester of labrafac attained small droplets (Atef & Belmonte, ). Moreover, oleic acid can reduce the interfacial tension creating smaller and smoother particles (Sanad et al., ). Furthermore, the high phosphatidylcholine concentration was reported to produce smaller particles by stabilizing the system formation more efficiently (Kassem et al., ). The results also showed that LNCs exhibited relatively low average particle size compared to PNCs referred to low viscosity of the lipid melt unlike PCL dispersion . The negative zeta potential of LNCs (–25.23 ± 3.72 to −33.22 ± 2.1) may be attributed to the amphiphilic surfactant soy phosphatidylcholine as well as presence of oleic acid in the oil core which has strong negative charge (Xia et al., ; El-Hesaisy & Swidan, ). Consequently, by increasing lipid content with respect to MN, zeta potential increased . The fabricated nanocapsules had PDI values extended from 0.2 to 0.35 indicating narrow size distribution, excellent sample size homogeneity and reproducible method of preparation which was suitable for topical delivery applications (Xia et al., ). 3.2. Entrapment efficiency of prepared MN nanocapsules The effect of the MN:PCL ratio and MN:lipid ratio on the %EE of MN in prepared nanocapsules is displayed in . The %EE was in the range between 80 ± 5.21 and 98 ± 5.21. Generally, the high %EE of MN may be attributed to its high lipophilic nature (log P 5.96). Increasing PCL concentration with respect to MN had a considerable effect on %EE, by increasing its concentration the %EE increased (Rahat et al., ). The relatively high EE value was owing to the high chemotactic between MN and PCL as both are highly hydrophobic; hence, both entities have great affinity toward each other (Govender et al., ). Moreover, by increasing PCL concentration, the viscosity of primary emulsion formed increased hindering the distribution of MN into the external phase and consequently increases the entrapment of the drug (Tinca et al., ). The results in also revealed that the %EE values for LNCs (F4, F5, and F6) were higher than PNCs (F1, F2, and F3) and this could be related to the inclusion of lipids in the core of nanocapsules structure where highly lipophilic drugs could be concentrated in (El-Hesaisy & Swidan, ). As well as they donate thick shell for nanocapsules; allow higher amount of drug entrapment as well as slower release (Radwan et al., ). Moreover, oleic acid and Labrafac ® (medium chain fatty acid ester) were reported previously as good solubilizers for lipophilic drugs profiting high %DL (Kamel & Basha, ). Therefore, by increasing lipid contents the %EE increased, F6 > F5 > F4. Based on the previous results, PS, zeta potential, %DL, and %EE , F1 and F6 showed insignificance; therefore, they were subjected for further studies. 3.3. Selection of best formula Based on the previous results, it was concluded that the formula coded F6 could be selected as optimum LNCs formula. It showed small particle size, PDI, and ZP; 89 nm, 0.2 and −31.22, respectively that synchronized with high %EE (98%) . This could be attributed to LNCs structural features which can entrap lipophilic drugs in their oil core that provide high space capacity for drug encapsulation and barrier for drug diffusion. Meanwhile, formula coded F1 was selected as optimum PNCs formula as it showed least particle size and PDI accompanied with high zeta potential and best %EE as shown in . For further comparison, both formulations were subjected to subsequent investigations. 3.4. Transmission electron microscope A representative TEM photomicrograph was conducted for F1 (optimum PNCs) and F6 (optimum LNCs) as illustrated in . For F1, the photos showed small elliptical particles. The PNCs were light in color referred to poor absorption of phosphotungstic acid stain compared to LNCs (Abbas et al., ). The TEM examination showed spherical particles displaying a core–shell structure. It also revealed that the optimized formula (F6) had an almost homogeneous small-sized spherical appearance with a narrow size distribution and not aggregated. The particle size of nanocapsules appears to match the results obtained previously by zeta sizer. 3.5. In vitro release study The in vitro release profile is a mirror image for predicting in vivo drug performance. The release profiles of MN from prepared PNCs (F1) and LNCs (F6) are illustrated in . The pure drug revealed a quicker release within the first two hours (45%), followed by a plateau pattern. However, encapsulated MN showed slow release from PNCs and LNCs formulations over a period up to 48 h. From the results, both PNCs (F1) and LNCs (F6) showed a biphasic drug release profile; displayed an initial rapid % drug release of 30 ± 2.4 and 15 ± 1.9, respectively, in the first 3 h. The statistically significant difference in amount of drug released from PNCs and LNCs in the starting 3 h could be explained by the difference in the structural composition of the two types of nanocapsules. The polymeric type (F1) of nanocapsules showed nearly two folds burst effect compared to the lipid type (F6) owing to deposition of some of the loaded amount of MN on the surface of PNCs (shell). Whereas, the LNCs may have offered more room for %DL in the core part due to the oily nature of LNCs core (Dubey et al., ; Lino et al., ). Another reason, suggested for the previously resulted release pattern of PNCs is the degradation of the thin polymeric outer shell membrane (Cauchetier et al., ). The initial burst effect was followed by more controlled MN release reaching 78 and 90%, after 48 h. The controlled release behavior attained by PNCs (F1) was correlated to the slow diffusion of MN from the viscous PCL matrix as a result of its degradation which is considered as an obstacle for its distribution into the external dissolution medium (Midhun et al., ). However, the low concentration of PCL with respect to MN in F1 makes the release of MN more rapid and completed within 24 h with maximum amount released 78%. On contrast, the release of MN from LNCs (F6) was controlled and extended to 48 h with maximum amount released 90% . This finding was attributed to the high lipophilicity of MN as well as the structural composition of lipid core matrix (high lipid content 1:4) and their high chemotactic which enclosed drug deeply and thus delays its diffusion (Kiani et al., ). Moreover, the rigid external shell of LNCs presents an obstacle for drug diffusion from oil core to the external phase; therefore, it is less probable at the nanocapsules surface. Consequently, LNCs exhibited sustained release functions more than PNCs. Results of cumulative drug diffused were subjected to release kinetics models. The results portrayed that free MN, F1, and F6 release have followed zero order, Higuchi diffusion and Baker and Lonsdale models, respectively, based on regression coefficient values . Pure MN displayed a continuous release of the same amount of drug per unit time (zero order) which correlated to its lipophilicity that is characterized by release kinetics of this nature (Rodrigues et al., ). In turn, PNCs (F1) are best fitted to Higuchi’s diffusion model where the drug released from the nanocapsules after the degradation of thin polymeric outer shell membrane is governed by the slow diffusion of MN from the viscous PCL matrix rather than erosion mechanism (El-Hesaisy & Swidan, ). Moreover, the LNCs (F6) were fixed to the Baker and Lonsdale model which designates controlled drug release manner from spherical-core shell matrices combined with diffusion (Mircioiu et al., ). This result was agreed with rigid external shell of LNCs and MN high lipophilicity which delay drug diffusion from oil core to the external phase and hence achieving controlled drug release mechanism. 3.6. Stability study After 3 months of storage at 25 °C, the selected formulas (F1 and F6) kept its physicochemical properties . The results were found to be statistically insignificant ( p >.05, paired t -test) with those obtained before storage, indicating the stability of MN-loaded nanocapsules. However, the PNCs (F1) seem to be more stable than LNCs (F6) during and after storage owing to the presence of PCL with regard to synthetic and biodegradable polymer keeping stability, while LNCs structural features and their oil core tend to aggregation and deformation (Deng et al., ). 3.7. Antifungal activity of MN-loaded nanocapsules The results of antifungal activity study, presented as the inhibition zone diameter, were in direct correlation with the in vitro release data. The antifungal activity of the optimized PNCs (F1) was high 19.07 mm compared to 11.4 mm for optimum LNCs (F6) and only 5 mm with a drug suspension (positive control). These results highlighted the enhancement of antifungal activity produced by MN after loading into different types of nanocapsules. To study effect on the minimum inhibition concentration (MIC), different dilutions of tested samples were applied. It was found that nanocapsules as drug delivery system can improve the anti-fungal activity of MN by decreasing the needed MIC from 2 µg/mL to nearly 0.75 µg/mL (Ahmed et al., ). 3.8. Effect of nanocapsules on cell viability (cytotoxicity and genotoxicity) The results of MTT assay revealed the low cytotoxicity of MN in either free or encapsulated forms ( and ). The optical images captured by inverted microscope showed samples treated by free MN that completely lost integrity of cells as a sign of cytotoxicity . On the contrary, the optical images showed the ability of nanocapsules to keep cells in a healthy condition . The results also indicated the effect of nanocarrier on cytotoxicity of MN, where LC50 of MN was increased from 45.4 µg/mL for free MN to 83.5 µg/mL and 89.2 for PNCs and LNCs, respectively. Based on a previously published report, iron oxide nanoparticles coated with chitosan nanocarrier did not reduce the cytotoxic potential compared with free MN (Caldeirão et al., ). This finding privileged the effect of using nanocapsules over other nanocarrier systems. On studying effect on DNA damage, it was found out that there was slightly significant difference in tail moment of WISH amniotic cells for MN loaded in nanocapsules compared to free MN (1.2 ± 0.01 and 2.13 ± 0.07, respectively) . The cytotoxic effect of MN may be referred to the production of ROS in human keratinocytes, which may induce oxidative stress and cause cell death (Lam et al., ). Additionally, MN is known of causing inhibition in the growth of cells by activating extrinsic and intrinsic apoptotic pathways (Caldeirão et al., ). These findings, in conjunction with the antifungal activity results, indicate a synchronized advantage of MN loaded nanocapsules; lower cytotoxic potential and better antifungal effect. 3.9. Ex vivo skin permeation and retentivity The results showed a very small amount of free MN could pass through skin samples reaching approximately 20% after 24 h. On the other hand, F1 (optimum PNCs) showed 71% permeability compared to 89% for F6 (LNCs) as observed in . The results agreed with reports confirmed that only a negligible amount of MN could pass through skin (De Brum et al., ). It could be concluded that LNCs reached the dermis, while PNCs were more retained at the outermost layers of the skin. The result was in accordance with the flexibility in a way that higher flexibility gives deeper penetration and high compatibility of phosphatidylcholine in LNCs with skin composition (Coverdale et al., ; El-Leithy & Abdel-Rashid, ). Based on retentivity results as shown in , it was found that F1 (PNCs optimum formula) showed more amount drug deposition compared to F6 (LNCs optimum formula). The results illustrated that free MN suspension; F1 and F6 were 65 ± 2.95 µg/cm 2 and 32.2 ± 3.52 µg/cm 2 , and 12.7 ± 1.52 µg/cm 2 , respectively, after 24 h. Hence, the findings suggested that LNCs can reach the dermis and PNCs can act as reservoir systems at the epidermis (De Brum et al., ). Deposition of drug on the upper skin layers, reducing drug flux and creating a reservoir able to prolong skin residence time could provide better treatment to upper skin fungal infections (Peira et al., ). Moreover, PCL was used widely to control drug release owing to its high permeability in numerous pharmacological compounds as well as its extensive biocompatibility with living tissues and biodegradability through the hydrolytic rupture of ester bonds (Balcucho et al., ). The results confirmed that PNCs may have an advantage over LNCs by offering dual action for both superficial and deep fungal infections synchronized with biphasic release pattern.
Particle size, PDI, and zeta potential The results of particle size, PDI, and ZP of the prepared nanocapsules are displayed in . The measured particle size was in the nanometric size range, which is preferred for drug delivery and penetrability to biological cell (Subramaniam et al., ). The average particle size of PNCs suspensions (F1, F2, and F3) ranged from 108 ± 3.63 nm and 180 ± 2.11 nm. The results obviously showed a statistically significant directly proportional function between drug–polymer ratio (MN:PCL) and average particle size of nanocapsules suspensions at constant surfactant concentration Tween 80 (5% w/v). These results were in line with findings of Sathyamoorthy et al. ; who reported that the particle size increases by increasing PCL concentration which results in increased incidence of collisions between particles during emulsification consequently aggregation of small sized particles yielded large particle size. Moreover, the increase of PCL concentration increases viscosity and resistance of the organic phase to diffuse and distribute into the aqueous phase that prompts the formation of large nanodroplets at the interface (Ajiboye et al., ). Additionally, the high viscosity could decline the drug diffusion from the nanocapsules as another cause for increasing particle size (Tavares et al., ). Tween 80 (HLB 15) was chosen as a surfactant after preliminary studies compared to span 60 (4.7) (data unseen). It was found that hydrophilic surfactants strongly participate in decreasing particle size of nanocapsules by causing high droplets stabilization and more flexibility (Zhu et al., ). Furthermore, the zeta potential of PNCs suspensions (F1, F2, and F3) was −31 ± 3.10, −35.03 ± 4.80, and −40 ± 5.21, respectively. PCL is known for yielding spherical nanocapsules with negative charge on the surface (Rahat et al., ). They displayed negative charges owing to PCL structure and its hydrophobic nature which may cause ionization of the carboxylic groups resulting in a negative potential to the interface (Michels et al., ; Lino et al., ). Therefore, increased concentration of PCL influences negativity of zeta potential of prepared nanocapsules. The high negative zeta potential owing to the charges repulsion toward the natural tendency of aggregation affords stability to nanocapsules from the hostile pH of the biological system (Alves et al., ; Rahat et al., ). On the other hand, the results displayed a negative correlation between MN:lipid ratio and average particle size of LNCs at constant soy phosphatidylcholine concentration (5 g) with values ranging between 89 ± 3.63 and 116 ± 1.63 nm. By increasing lipid content, the average particle size decreased; this could be revealed to the combination of labrafac and oleic acid which attained good solubilization for MN and proper self-emulsification yielded small droplet size (Kamel & Basha, ). Previously, published reports stated that short chain fatty acid ester of labrafac attained small droplets (Atef & Belmonte, ). Moreover, oleic acid can reduce the interfacial tension creating smaller and smoother particles (Sanad et al., ). Furthermore, the high phosphatidylcholine concentration was reported to produce smaller particles by stabilizing the system formation more efficiently (Kassem et al., ). The results also showed that LNCs exhibited relatively low average particle size compared to PNCs referred to low viscosity of the lipid melt unlike PCL dispersion . The negative zeta potential of LNCs (–25.23 ± 3.72 to −33.22 ± 2.1) may be attributed to the amphiphilic surfactant soy phosphatidylcholine as well as presence of oleic acid in the oil core which has strong negative charge (Xia et al., ; El-Hesaisy & Swidan, ). Consequently, by increasing lipid content with respect to MN, zeta potential increased . The fabricated nanocapsules had PDI values extended from 0.2 to 0.35 indicating narrow size distribution, excellent sample size homogeneity and reproducible method of preparation which was suitable for topical delivery applications (Xia et al., ).
Entrapment efficiency of prepared MN nanocapsules The effect of the MN:PCL ratio and MN:lipid ratio on the %EE of MN in prepared nanocapsules is displayed in . The %EE was in the range between 80 ± 5.21 and 98 ± 5.21. Generally, the high %EE of MN may be attributed to its high lipophilic nature (log P 5.96). Increasing PCL concentration with respect to MN had a considerable effect on %EE, by increasing its concentration the %EE increased (Rahat et al., ). The relatively high EE value was owing to the high chemotactic between MN and PCL as both are highly hydrophobic; hence, both entities have great affinity toward each other (Govender et al., ). Moreover, by increasing PCL concentration, the viscosity of primary emulsion formed increased hindering the distribution of MN into the external phase and consequently increases the entrapment of the drug (Tinca et al., ). The results in also revealed that the %EE values for LNCs (F4, F5, and F6) were higher than PNCs (F1, F2, and F3) and this could be related to the inclusion of lipids in the core of nanocapsules structure where highly lipophilic drugs could be concentrated in (El-Hesaisy & Swidan, ). As well as they donate thick shell for nanocapsules; allow higher amount of drug entrapment as well as slower release (Radwan et al., ). Moreover, oleic acid and Labrafac ® (medium chain fatty acid ester) were reported previously as good solubilizers for lipophilic drugs profiting high %DL (Kamel & Basha, ). Therefore, by increasing lipid contents the %EE increased, F6 > F5 > F4. Based on the previous results, PS, zeta potential, %DL, and %EE , F1 and F6 showed insignificance; therefore, they were subjected for further studies.
Selection of best formula Based on the previous results, it was concluded that the formula coded F6 could be selected as optimum LNCs formula. It showed small particle size, PDI, and ZP; 89 nm, 0.2 and −31.22, respectively that synchronized with high %EE (98%) . This could be attributed to LNCs structural features which can entrap lipophilic drugs in their oil core that provide high space capacity for drug encapsulation and barrier for drug diffusion. Meanwhile, formula coded F1 was selected as optimum PNCs formula as it showed least particle size and PDI accompanied with high zeta potential and best %EE as shown in . For further comparison, both formulations were subjected to subsequent investigations.
Transmission electron microscope A representative TEM photomicrograph was conducted for F1 (optimum PNCs) and F6 (optimum LNCs) as illustrated in . For F1, the photos showed small elliptical particles. The PNCs were light in color referred to poor absorption of phosphotungstic acid stain compared to LNCs (Abbas et al., ). The TEM examination showed spherical particles displaying a core–shell structure. It also revealed that the optimized formula (F6) had an almost homogeneous small-sized spherical appearance with a narrow size distribution and not aggregated. The particle size of nanocapsules appears to match the results obtained previously by zeta sizer.
In vitro release study The in vitro release profile is a mirror image for predicting in vivo drug performance. The release profiles of MN from prepared PNCs (F1) and LNCs (F6) are illustrated in . The pure drug revealed a quicker release within the first two hours (45%), followed by a plateau pattern. However, encapsulated MN showed slow release from PNCs and LNCs formulations over a period up to 48 h. From the results, both PNCs (F1) and LNCs (F6) showed a biphasic drug release profile; displayed an initial rapid % drug release of 30 ± 2.4 and 15 ± 1.9, respectively, in the first 3 h. The statistically significant difference in amount of drug released from PNCs and LNCs in the starting 3 h could be explained by the difference in the structural composition of the two types of nanocapsules. The polymeric type (F1) of nanocapsules showed nearly two folds burst effect compared to the lipid type (F6) owing to deposition of some of the loaded amount of MN on the surface of PNCs (shell). Whereas, the LNCs may have offered more room for %DL in the core part due to the oily nature of LNCs core (Dubey et al., ; Lino et al., ). Another reason, suggested for the previously resulted release pattern of PNCs is the degradation of the thin polymeric outer shell membrane (Cauchetier et al., ). The initial burst effect was followed by more controlled MN release reaching 78 and 90%, after 48 h. The controlled release behavior attained by PNCs (F1) was correlated to the slow diffusion of MN from the viscous PCL matrix as a result of its degradation which is considered as an obstacle for its distribution into the external dissolution medium (Midhun et al., ). However, the low concentration of PCL with respect to MN in F1 makes the release of MN more rapid and completed within 24 h with maximum amount released 78%. On contrast, the release of MN from LNCs (F6) was controlled and extended to 48 h with maximum amount released 90% . This finding was attributed to the high lipophilicity of MN as well as the structural composition of lipid core matrix (high lipid content 1:4) and their high chemotactic which enclosed drug deeply and thus delays its diffusion (Kiani et al., ). Moreover, the rigid external shell of LNCs presents an obstacle for drug diffusion from oil core to the external phase; therefore, it is less probable at the nanocapsules surface. Consequently, LNCs exhibited sustained release functions more than PNCs. Results of cumulative drug diffused were subjected to release kinetics models. The results portrayed that free MN, F1, and F6 release have followed zero order, Higuchi diffusion and Baker and Lonsdale models, respectively, based on regression coefficient values . Pure MN displayed a continuous release of the same amount of drug per unit time (zero order) which correlated to its lipophilicity that is characterized by release kinetics of this nature (Rodrigues et al., ). In turn, PNCs (F1) are best fitted to Higuchi’s diffusion model where the drug released from the nanocapsules after the degradation of thin polymeric outer shell membrane is governed by the slow diffusion of MN from the viscous PCL matrix rather than erosion mechanism (El-Hesaisy & Swidan, ). Moreover, the LNCs (F6) were fixed to the Baker and Lonsdale model which designates controlled drug release manner from spherical-core shell matrices combined with diffusion (Mircioiu et al., ). This result was agreed with rigid external shell of LNCs and MN high lipophilicity which delay drug diffusion from oil core to the external phase and hence achieving controlled drug release mechanism.
Stability study After 3 months of storage at 25 °C, the selected formulas (F1 and F6) kept its physicochemical properties . The results were found to be statistically insignificant ( p >.05, paired t -test) with those obtained before storage, indicating the stability of MN-loaded nanocapsules. However, the PNCs (F1) seem to be more stable than LNCs (F6) during and after storage owing to the presence of PCL with regard to synthetic and biodegradable polymer keeping stability, while LNCs structural features and their oil core tend to aggregation and deformation (Deng et al., ).
Antifungal activity of MN-loaded nanocapsules The results of antifungal activity study, presented as the inhibition zone diameter, were in direct correlation with the in vitro release data. The antifungal activity of the optimized PNCs (F1) was high 19.07 mm compared to 11.4 mm for optimum LNCs (F6) and only 5 mm with a drug suspension (positive control). These results highlighted the enhancement of antifungal activity produced by MN after loading into different types of nanocapsules. To study effect on the minimum inhibition concentration (MIC), different dilutions of tested samples were applied. It was found that nanocapsules as drug delivery system can improve the anti-fungal activity of MN by decreasing the needed MIC from 2 µg/mL to nearly 0.75 µg/mL (Ahmed et al., ).
Effect of nanocapsules on cell viability (cytotoxicity and genotoxicity) The results of MTT assay revealed the low cytotoxicity of MN in either free or encapsulated forms ( and ). The optical images captured by inverted microscope showed samples treated by free MN that completely lost integrity of cells as a sign of cytotoxicity . On the contrary, the optical images showed the ability of nanocapsules to keep cells in a healthy condition . The results also indicated the effect of nanocarrier on cytotoxicity of MN, where LC50 of MN was increased from 45.4 µg/mL for free MN to 83.5 µg/mL and 89.2 for PNCs and LNCs, respectively. Based on a previously published report, iron oxide nanoparticles coated with chitosan nanocarrier did not reduce the cytotoxic potential compared with free MN (Caldeirão et al., ). This finding privileged the effect of using nanocapsules over other nanocarrier systems. On studying effect on DNA damage, it was found out that there was slightly significant difference in tail moment of WISH amniotic cells for MN loaded in nanocapsules compared to free MN (1.2 ± 0.01 and 2.13 ± 0.07, respectively) . The cytotoxic effect of MN may be referred to the production of ROS in human keratinocytes, which may induce oxidative stress and cause cell death (Lam et al., ). Additionally, MN is known of causing inhibition in the growth of cells by activating extrinsic and intrinsic apoptotic pathways (Caldeirão et al., ). These findings, in conjunction with the antifungal activity results, indicate a synchronized advantage of MN loaded nanocapsules; lower cytotoxic potential and better antifungal effect.
Ex vivo skin permeation and retentivity The results showed a very small amount of free MN could pass through skin samples reaching approximately 20% after 24 h. On the other hand, F1 (optimum PNCs) showed 71% permeability compared to 89% for F6 (LNCs) as observed in . The results agreed with reports confirmed that only a negligible amount of MN could pass through skin (De Brum et al., ). It could be concluded that LNCs reached the dermis, while PNCs were more retained at the outermost layers of the skin. The result was in accordance with the flexibility in a way that higher flexibility gives deeper penetration and high compatibility of phosphatidylcholine in LNCs with skin composition (Coverdale et al., ; El-Leithy & Abdel-Rashid, ). Based on retentivity results as shown in , it was found that F1 (PNCs optimum formula) showed more amount drug deposition compared to F6 (LNCs optimum formula). The results illustrated that free MN suspension; F1 and F6 were 65 ± 2.95 µg/cm 2 and 32.2 ± 3.52 µg/cm 2 , and 12.7 ± 1.52 µg/cm 2 , respectively, after 24 h. Hence, the findings suggested that LNCs can reach the dermis and PNCs can act as reservoir systems at the epidermis (De Brum et al., ). Deposition of drug on the upper skin layers, reducing drug flux and creating a reservoir able to prolong skin residence time could provide better treatment to upper skin fungal infections (Peira et al., ). Moreover, PCL was used widely to control drug release owing to its high permeability in numerous pharmacological compounds as well as its extensive biocompatibility with living tissues and biodegradability through the hydrolytic rupture of ester bonds (Balcucho et al., ). The results confirmed that PNCs may have an advantage over LNCs by offering dual action for both superficial and deep fungal infections synchronized with biphasic release pattern.
Conclusions MN loaded nanocapsules were successively prepared by simple and cost-effective technique; emulsification/nanoprecipitation using PCL biodegradable polymer in PNCs and lipid matrix (labrafac:oleic acid) at ratio 1:1 in LNCs. Both types of nanocapsules displayed small particle size, slow biphasic release manner, high %EE, and improved stability expressing a good approach for the delivery of MN. These findings, in conjunction with the antifungal activity results, indicate a synchronized advantage of MN loaded nanocapsules; lower cytotoxic potential and better antifungal effect. PNCs were more promising than LNCs offering dual action for both superficial and deep fungal infections synchronized with biphasic release pattern. Thus, PNC is an innovative way with promising results; can enhance antifungal activity with minimal side effects, reducing the dose and dosing frequency.
|
Quantitative Evaluation of Changes in Retinal and Choroidal Blood Flow Following Strabismus Surgery | 967e3a1d-0866-40ce-8692-ee3578b70b15 | 11918063 | Surgical Procedures, Operative[mh] | Strabismus surgery is performed to treat strabismus or anomalous eye positions, and nystagmus. During strabismus surgery, the anterior ciliary artery that is the main blood supply to the anterior segment of the eye is often cut off with the rectus muscle. Hence, although rare, anterior segment ischemia is a potential adverse effect of strabismus surgery. Simultaneous surgery on multiple rectus muscles is associated with an increased risk of anterior segment ischemia. Thus several studies have reported that at least one rectus muscle should be left intact in each eye during strabismus surgery. , Some studies have focused on the pathogenesis of anterior segment ischemia and anterior segment blood flow changes after strabismus surgery. – Nevertheless, only a few studies have examined the effects of strabismus surgery on posterior ocular blood flow. The choroid is commonly nourished by the short posterior ciliary arteries. However, the anterior ciliary arteries also contribute to some blood flow. Therefore, strabismus surgery with extraocular muscle resection may decrease choroidal and anterior ocular blood flow. Some reports have investigated posterior ocular blood flow. – However, a consensus regarding the effect of strabismus surgery on choroidal blood flow has not been reached. The choroid is responsible for the blood supply to the retina and sclera, and it provides nutrition and facilitates thermoregulation, and, particularly, retinal oxygenation. Hence, whether strabismus surgery will affect choroidal blood flow is an important clinical question. The effects of strabismus surgery on posterior ocular blood flow are challenging to evaluate. One of the reasons why is that it is not easy to quantitatively assess posterior ocular blood flow. In recent years, with the development of novel ophthalmic techniques such as optical coherence tomography angiography (OCTA) and laser speckle flowgraphy (LSFG), it has become possible to more easily and quantitatively assess the retinal and choroidal circulation. OCTA is a novel and noninvasive imaging modality that was recently developed based on optical coherence tomography (OCT). Furthermore, it can be used to visualize the retinal vasculature and the choroidal vasculature, to some extent, without the need for dye injection. LSFG is a noninvasive method for measuring relative blood flow in real time. To the best of our knowledge, no studies have used LSFG in strabismus surgery. However, several reports have revealed the use of LSFG for evaluating changes in choroidal blood flow after scleral buckling surgery. – The current study aimed to investigate the effects of extraocular muscle surgery on posterior ocular blood flow, as measured on OCTA and LSFG. We hypothesized that strabismus surgery with extraocular muscle recession can reduce both posterior and anterior ocular blood flow.
Participants This prospective, single-center study included patients who underwent strabismus surgery at Nagoya University Hospital from January 2019 to November 2022. The study protocol was in accordance with the tenets of the Declaration of Helsinki. It was approved by the Ethics Committee of Nagoya University Hospital (2019–0072), Nagoya, Japan, and registered in the University Hospital Medical Information Network Clinical Trials Registry (ID: UMIN000035995). Informed consent or assent was obtained from the patients themselves or the patients’ parents prior to study participation. Patients <3 years of age were excluded from this study. Additionally, patients with other conditions that interfere with accurate examination, such as nystagmus and amblyopia, those with a history of retinal or choroidal disease, those with a history of internal eye surgery and strabismus surgery, and those who underwent transposition surgery for rectus muscle palsy were excluded from the analysis. Study Examination The patients underwent anterior eye and fundus examinations before strabismus surgery. The refractive error of children was measured with cycloplegia, and the patients also underwent comprehensive ophthalmologic examinations, including best-corrected visual acuity and intraocular pressure measurements. The participants were instructed to avoid drinking alcoholic and caffeinated beverages on the morning of the examinations. To decrease the effect of diurnal fluctuations, all assessments were conducted while the participants were in the sitting position between 10:00 AM and 12:00 PM. , Orthoptic evaluation, OCT (Spectralis; Heidelberg Engineering, Heidelberg, Germany), OCTA (AngioPlex, CIRRUS HD-OCT model 5000; Carl Zeiss Meditec AG, Oberkochen, Germany), and LSFG (LSFG-NAVI; Softcare Co., Ltd., Fukutsu, Japan) were performed before strabismus surgery and at one week, one month, and four months after the surgery. In addition, blood pressure and heart rate were also measured before and after the surgery using an automated blood pressure monitor (CH-483C; Citizen, Tokyo, Japan). Subfoveal choroidal thickness (SFCT) was measured from the highly reflective zone of the retinal pigment epithelium to the boundary between the choroid and sclera using enhanced-depth imaging OCT, as reported in previous studies. , A radial line scan through the center of the fovea was obtained at an angle of 30°, and we acquired 50 OCT images, which were averaged to reduce speckle noise. Two researchers measured the SFCT. Each researcher conducted six measurements, and the average value was calculated and used for analysis. Retinal blood flow was measured using OCTA. OCTA uses an algorithm called the OCT microangiography complex to generate images of en face microvascular flow using differences in phase and intensity information from successive B-scans performed at the same location. Using the manufacturer's software, the vessel density (VD) (defined as the area occupied by the blood vessels within a particular area measured in mm 2 /mm 2 ), which is an indicator of retinal blood flow, could be determined. In this study, the VD within a 3 × 3 mm circle in the center of the macula was examined. Choroidal blood flow was measured using LSFG. The principles of LSFG have been described in detail elsewhere. , LSFG can detect the speckle contrast pattern caused by the interference of illuminating laser light scattered by the movement of erythrocytes in a blood vessel, and can measure the relative blood flow in the vessel expressed by the mean blur rate (MBR). The MBR of the macula mainly originates from the choroid. Thus the relative choroidal blood flow can be determined by measuring the MBR of the macula. In this study, we set the center of the square to the macula (250 × 250 pixels, power: 6.31° × 6.31°) and measured the MBR. LSFG was performed three times at each time point in all eyes, and the mean value of the variables was calculated. Surgical Technique Recession or plication of the rectus muscles, recession of the inferior oblique muscle (IO), and tendon lengthening of the superior oblique muscle (SO) were performed. The patients underwent surgeries, which were performed by two surgeons (Y.T. and S.Y.) using the same procedure, using 6-0 Vicryl (Ethicon, Somerville, NJ, USA) for muscle suturing in recession or plication, 5-0 Mersilene (Ethicon) for SO tendon lengthening, and 8-0 Vicryl for conjunctival suturing. The patients received postoperative antimicrobial and steroid eye drops for one month. Statistical Analyses For the summary statistics, categorical variables were expressed as numbers and percentages. Moreover, continuous variables were presented as medians and interquartile ranges as these values had a non-normal distribution. The normality of the continuous variables was tested using the Shapiro-Wilk test or was assessed visually with the Q-Q plot. The preoperative and postoperative (one week, one month, and four months) SFCT, VD, and MBR values were compared using the Wilcoxon signed-rank test (matched-paired comparisons) because these date, which had a non-normal distribution, were longitudinal at the four timepoints in the same personnel. All statistical analyses were performed using JMP 17.2.0 (SAS Institute, Cary, NC, USA) and STATA 18.0 (StataCorp, College Station, TX, USA). For each analysis, the null hypothesis was assessed at a two-tailed significance level of 0.05. To address the issue of type 1 error inflation via multiple pairwise comparisons in the three family-wise groups (SFCT, VD, and MBR), Bonferroni correction was used to adjust the P value thresholds. A two-sided P value = 0.016 was used as the criterion for statistical significance in the multiple pair-wise comparisons (preoperative vs. one week, one month, and four months after surgery) in the three groups (SFCT, VD, and MBR). To compare the preoperative and postoperative state of the choroidal thickness and retinal and choroidal blood flow, the postoperative value–to–preoperative value (post-/pre-) ratio was calculated. The post-/pre-ratio was used to perform intragroup comparisons. In addition, because the changes after strabismus surgery may differ based on the surgical procedure, the preoperative and postoperative changes in the retina and choroid in each procedure were examined.
This prospective, single-center study included patients who underwent strabismus surgery at Nagoya University Hospital from January 2019 to November 2022. The study protocol was in accordance with the tenets of the Declaration of Helsinki. It was approved by the Ethics Committee of Nagoya University Hospital (2019–0072), Nagoya, Japan, and registered in the University Hospital Medical Information Network Clinical Trials Registry (ID: UMIN000035995). Informed consent or assent was obtained from the patients themselves or the patients’ parents prior to study participation. Patients <3 years of age were excluded from this study. Additionally, patients with other conditions that interfere with accurate examination, such as nystagmus and amblyopia, those with a history of retinal or choroidal disease, those with a history of internal eye surgery and strabismus surgery, and those who underwent transposition surgery for rectus muscle palsy were excluded from the analysis.
The patients underwent anterior eye and fundus examinations before strabismus surgery. The refractive error of children was measured with cycloplegia, and the patients also underwent comprehensive ophthalmologic examinations, including best-corrected visual acuity and intraocular pressure measurements. The participants were instructed to avoid drinking alcoholic and caffeinated beverages on the morning of the examinations. To decrease the effect of diurnal fluctuations, all assessments were conducted while the participants were in the sitting position between 10:00 AM and 12:00 PM. , Orthoptic evaluation, OCT (Spectralis; Heidelberg Engineering, Heidelberg, Germany), OCTA (AngioPlex, CIRRUS HD-OCT model 5000; Carl Zeiss Meditec AG, Oberkochen, Germany), and LSFG (LSFG-NAVI; Softcare Co., Ltd., Fukutsu, Japan) were performed before strabismus surgery and at one week, one month, and four months after the surgery. In addition, blood pressure and heart rate were also measured before and after the surgery using an automated blood pressure monitor (CH-483C; Citizen, Tokyo, Japan). Subfoveal choroidal thickness (SFCT) was measured from the highly reflective zone of the retinal pigment epithelium to the boundary between the choroid and sclera using enhanced-depth imaging OCT, as reported in previous studies. , A radial line scan through the center of the fovea was obtained at an angle of 30°, and we acquired 50 OCT images, which were averaged to reduce speckle noise. Two researchers measured the SFCT. Each researcher conducted six measurements, and the average value was calculated and used for analysis. Retinal blood flow was measured using OCTA. OCTA uses an algorithm called the OCT microangiography complex to generate images of en face microvascular flow using differences in phase and intensity information from successive B-scans performed at the same location. Using the manufacturer's software, the vessel density (VD) (defined as the area occupied by the blood vessels within a particular area measured in mm 2 /mm 2 ), which is an indicator of retinal blood flow, could be determined. In this study, the VD within a 3 × 3 mm circle in the center of the macula was examined. Choroidal blood flow was measured using LSFG. The principles of LSFG have been described in detail elsewhere. , LSFG can detect the speckle contrast pattern caused by the interference of illuminating laser light scattered by the movement of erythrocytes in a blood vessel, and can measure the relative blood flow in the vessel expressed by the mean blur rate (MBR). The MBR of the macula mainly originates from the choroid. Thus the relative choroidal blood flow can be determined by measuring the MBR of the macula. In this study, we set the center of the square to the macula (250 × 250 pixels, power: 6.31° × 6.31°) and measured the MBR. LSFG was performed three times at each time point in all eyes, and the mean value of the variables was calculated.
Recession or plication of the rectus muscles, recession of the inferior oblique muscle (IO), and tendon lengthening of the superior oblique muscle (SO) were performed. The patients underwent surgeries, which were performed by two surgeons (Y.T. and S.Y.) using the same procedure, using 6-0 Vicryl (Ethicon, Somerville, NJ, USA) for muscle suturing in recession or plication, 5-0 Mersilene (Ethicon) for SO tendon lengthening, and 8-0 Vicryl for conjunctival suturing. The patients received postoperative antimicrobial and steroid eye drops for one month.
For the summary statistics, categorical variables were expressed as numbers and percentages. Moreover, continuous variables were presented as medians and interquartile ranges as these values had a non-normal distribution. The normality of the continuous variables was tested using the Shapiro-Wilk test or was assessed visually with the Q-Q plot. The preoperative and postoperative (one week, one month, and four months) SFCT, VD, and MBR values were compared using the Wilcoxon signed-rank test (matched-paired comparisons) because these date, which had a non-normal distribution, were longitudinal at the four timepoints in the same personnel. All statistical analyses were performed using JMP 17.2.0 (SAS Institute, Cary, NC, USA) and STATA 18.0 (StataCorp, College Station, TX, USA). For each analysis, the null hypothesis was assessed at a two-tailed significance level of 0.05. To address the issue of type 1 error inflation via multiple pairwise comparisons in the three family-wise groups (SFCT, VD, and MBR), Bonferroni correction was used to adjust the P value thresholds. A two-sided P value = 0.016 was used as the criterion for statistical significance in the multiple pair-wise comparisons (preoperative vs. one week, one month, and four months after surgery) in the three groups (SFCT, VD, and MBR). To compare the preoperative and postoperative state of the choroidal thickness and retinal and choroidal blood flow, the postoperative value–to–preoperative value (post-/pre-) ratio was calculated. The post-/pre-ratio was used to perform intragroup comparisons. In addition, because the changes after strabismus surgery may differ based on the surgical procedure, the preoperative and postoperative changes in the retina and choroid in each procedure were examined.
The study included 254 eyes from 127 patients aged three years or older who underwent strabismus surgery at Nagoya University Hospital from January 2019 to November 2022 and who consented to the study. Consent for the study was obtained from adult participants and from parents of participants <20 years of age. Eyes with a previous history of strabismus surgery (n = 8), epiretinal membrane surgery (n = 1), macular disease (n = 2), optic nerve atrophy (n = 2), nystagmus (n = 2), transposition surgery for rectus muscle palsy (n = 2), and aborted surgery (n = 2) were excluded. Finally, 133eyes from 80 patients (operated eyes = 116, unoperated eyes = 17) for which data could be obtained were included in the analysis. shows the preoperative characteristics, and presents the surgical procedures. depicts the representative multimodal images of a patient who underwent lateral rectus muscle recession and medial rectus muscle plication. shows the changes in the post-/pre-ratio of the SFCT, MBR, and VD of the operated eyes during the observation period, and shows the median post-/pre-ratio of the measurements at one week, one month, and four months after surgery. As mentioned in the Methods section, MBR is an indicator of choroidal blood flow and VD is an indicator of retinal blood flow. The post-/pre-ratio of SFCT at one week and one month after surgery were significantly higher than those before surgery ( P < 0.001, P = 0.010). However, there was no significant change in the post-/pre-ratio of SFCT at four months after surgery. The post-/pre-ratio of MBR at one week was significantly higher than that before surgery ( P < 0.001). Nevertheless, the post-/pre-ratio of MBR at one and four months after surgery did not significantly differ. The post-/pre-ratio of VD was significantly lower at one week after surgery ( P = 0.0016), with no significant change at one and four months after surgery ( ). depicts the correlation between the post-/pre-ratio of SFCT and the post-/pre-ratio of MBR. The choroidal thickness and choroidal blood flow significantly increased at one week. To validate the association between choroidal thickness and blood flow at that time, the correlation between choroidal thickness and blood flow was examined. Results showed a significantly positive correlation between choroidal thickness and choroidal blood flow at one week after surgery ( r = 0.41, P < 0.05). To determine whether the current changes were attributed to strabismus surgery, the changes in preoperative and postoperative measurements in the operated and unoperated eye in 17 patients who had undergone surgery in one eye were further examined ( ). In the operated eye, the post-/pre-ratio of SFCT increased significantly at one week after surgery ( P = 0.0010), with no significant change at one and four months after surgery. The post-/pre-ratio of MBR increased significantly at one week after surgery ( P = 0.0031), with no significant change at one and four months after surgery. The post-/pre-ratio of VD also decreased significantly at one week ( P = 0.0097), with no significant change at one and four months after surgery. The operated eye exhibited similar changes to the overall results. Conversely, in the unoperated eyes, there were no significant changes in SFCT, MBR, or VD at all timepoints. Because this study included several surgical procedures, the effect of postoperative retina and choroid might differ based on the surgical procedure. Therefore the patients were divided into groups according to surgical procedure and the changes in retina and choriod at four months after surgery in each group were examined ( ). However, no significant changes were observed in the analysis groups based on the surgical procedure.
In this study, we examined whether strabismus surgery with extraocular muscle recession can reduce anterior ocular and choroidal blood flow. Few studies have examined the effect of strabismus surgery on posterior ocular blood flow, particularly retinal and choroidal blood flow. However, with the use of newly introduced techniques (such as OCT, OCTA, and LSFG), the retinal and choroidal thickness and blood flow can now be evaluated independently. , To the best of our knowledge, this is the first report on quantitative changes in choroidal blood flow measured by LSFG following strabismus surgery. In the current study, the SFCT and choroidal blood flow increased significantly at one week after surgery. Several studies have investigated the effects of strabismus surgery on the choroid; Inan et al. and Yetkin et al. provided basic insight into postoperative changes by showing a significant decrease in SFCT after strabismus surgery. However, choroidal blood flow was not quantitatively assessed. Xiao et al. used OCTA to quantitatively assess changes in choroidal thickness and choroidal blood flow after horizontal rectus muscle surgery, which contributed significantly to our understanding on blood flow changes. OCTA is suitable for assessing structural changes in choroidal details. However, it has a limitation. Particularly, it cannot assess blood flow in the entire choroid postoperatively. This study used LSFG to quantitatively validate changes in the whole choroidal blood flow postoperatively. The association between choroidal thickness and blood flow remains controversial, and not all studies have always shown a correlation between the two. , Therefore choroidal thickness and blood flow must be evaluated independently, and this is what we focused on in this study. A positive correlation was found between SFCT and choroidal blood flow at one week after surgery. Margraf et al. have revealed how surgical trauma triggers the release of damage-associated molecular patterns, which activate immune responses and promote endothelial dysfunction. This leads to increased vascular permeability, leukocyte recruitment, and local tissue swelling. These processes disrupt normal vascular dynamics, and they may explain the increase in choroidal blood flow and thickness postsurgery. , The differences between the changes in the operated eyes and unoperated eyes of patients who had undergone unilateral strabismus surgery indicated that the abovementioned changes were caused by the strabismus surgery. In the long-term postoperative period, the SFCT and choroidal blood flow decreased at four months after surgery; the difference was not significant. The choroidal blood flow might differ based on the surgical techniques used. Hence, the patients were classified according to surgical techniques, and the choroidal blood flow was examined at four months after surgery. However, the results showed no significant differences in the choroidal blood flow in any of groups. These findings suggest that the impact of surgical combinations on choroidal blood flow may be less evident than initially anticipated. Nevertheless, the short follow-up period (four months) and small sample size might have reduced the statistical power of the analysis. These results emphasize the need for more comprehensive investigations on the effects of surgical combinations on posterior ocular blood flow. Thus future studies should focus on long-term follow-up, consider individual variability, and involve multicenter collaborations to increase sample size and ensure robust conclusions. Previous studies have shown that reduced choroidal blood flow may cause retinal ischemia, age-related macular degeneration, and ocular axis elongation. , Thus reduced choroidal blood flow after strabismus surgery may lead to long-term complications. Strabismus surgery is often performed on children. Thus studies focusing on the long-term prognosis of patients undergoing strabismus surgery should be performed. Contrary to choroidal blood flow, retinal blood flow decreased at one week after surgery. The posterior ciliary artery (PCA) is responsible for blood flow to the choroid. Furthermore, PCA hemodynamics is correlated with choroidal blood flow, and retinal blood flow with central retinal artery blood flow. Because the PCA and central retinal artery are branches of the ocular artery, an increase in PCA blood flow may result in a decrease in central retinal artery blood flow. Therefore we hypothesized that choroidal and PCA blood flow increased at one week after surgery, resulting in a decrease in central retinal artery and retinal blood flow. This study had several limitations. First, it was performed at a single center, and only standard procedures performed at our institution were examined. Therefore the effects of procedures performed at other institutions such as recession of two or more rectus muscles and IO myectomy on the retina and choroid were not comprehensively examined. Second, in this study, two surgeons performed the surgery. However, the influence of surgeon differences was not considered. The scanning range of the OCTA used in this study was 3 × 3 mm, which is smaller than that used in glaucoma research and other studies. This size is used to detect minute changes in the macula and to improve measurement accuracy in studies on children and patients with strabismus who have fixation that was difficult to stabilize. However, scanning a wider area may be more effective for evaluating changes in blood flow. Finally, the number of cases in this study was small. Therefore a long-term follow-up study with a larger number of patients should be performed to ensure the generalizability of the study results. In summary, strabismus surgery decreased retinal blood flow and increased choroidal thickness and blood flow in the early postoperative period. However, no significant changes were observed in the long term compared with the preoperative period.
|
How much can we learn from each other? Polish and Hungarian good practices in financing ophthalmology care as a proposal for implementation in Ukraine | 25522169-5007-4ab0-80a2-ed491167ab28 | 11232999 | Ophthalmology[mh] | Eye conditions include a wide and varied array of disorders that impact various parts of the visual system and visual function. Due to their diversity, categorizing these conditions can be difficult; however, one approach is to differentiate between conditions that usually do not cause vision loss and those that potentially lead to visual impairment. Ageing is the primary risk factor for many eye conditions . According to WHO, cataract (33%) and uncorrected refractive errors (42%) are leading causes of avoidable visual impairment. Unoperated cataract and glaucoma are the leading causes of avoidable blindness . In addition to elevated intraocular pressure, which is the only modifiable risk factor, risk factors for glaucoma include heredity, race, age, vascular factors, and myopia . Age-related macular degeneration (AMD) is now acknowledged as a multifaceted genetic disorder where one or multiple genes play a role in determining an individual’s likelihood of developing the condition . There are approximately twenty-one million with diabetic macular edema (DME) worldwide and its incidence increases with the duration of diabetes . Vitrectomy is a surgical procedure used to treat retinal diseases such as trauma, diabetic retinopathy, retinal detachment, macular hole, epiretinal membranes, and vitreoretinal traction syndromes . In the case of corneal diseases, corneal transplants are used . Health needs in ophthalmology pose several challenges to health systems and therefore require searching for the optimal models for financing and organizing healthcare services. The Visegrad countries and Ukraine currently experience problems in the ophthalmology area such as suboptimal number of services and limited access to services . In 2019, according to the State Statistics Service of Ukraine, 24.4% of households faced challenges in accessing medical care or acquiring medicines and medical supplies . The international team of researchers launched research with the support from the Visegrad Fund (“Access to healthcare services in the context of financing mechanisms. The case of ophthalmology,” grant no. 22120107), aimed to understand financing mechanisms that impact the access to healthcare services in ophthalmology in the Visegrad countries and Ukraine. The descriptions of financing mechanisms and comparisons of available data on ophthalmic care between the Czech Republic, Hungary, Poland, Slovakia, and Ukraine were presented in the summary report from the project . This article covers in detail findings that enable to compare financing cataract, glaucoma, vitrectomy, cornea transplantations, DME, and AMD treatments in Hungary, Poland, and Ukraine from the perspective of different stakeholders; and second, to identify practices from Poland and Hungary that were assessed during the qualitative research as helpful for decision makers in Ukraine in the light of ongoing healthcare reforms. The overarching goal of this research is to learn from each other about regulatory and financing solutions to be able to improve, in general, the ophthalmology care and services of these countries. In this article, the term "good practice" is considered as "a technique or methodology that has proven reliably to lead to a desired result through experience and research" and focuses on the existence of a connection between models of financing ophthalmological services and their impact on the volume, range, availability, and quality of ophthalmological interventions.This term is formulated based on data analysis of scientific publications and studies on the payment system based on diagnostic-related groups (DRG). The data from numerous publications show that starting from the 1990s, the DRG systems were implemented at the international level . This method of payment for inpatient medical services, according to the experience of many countries, made it possible to increase the transparency of the services provided, and payment based on them provided incentives for efficient use of resources . The advantages of the DRG payment system are not only reflected in the increased efficiency and transparency, but also reduced average length of stay . DRG is a relatively scientific and advanced medical payment model recognized globally . In this model, the severity and complexity of disease are compatible, as well as the consumption of medical resources. Disease-based payment is thus achieved. Furthermore, the wide implementation of DRG payment in areas like quality control, cost accounting, and human resource management is realized . The authors considered the term "good practice" in terms of the impact of financing models for ophthalmological care on ensuring the availability and quality of interventions in this area, and not as a "gold standard" or "excellence" of financing models . A definition of "best practice" was explored as "knowledge of what works in specific situations and contexts without using excessive resources to achieve desired outcomes and can be used to develop and implement solutions adapted to similar health problems in other situations and contexts" . The obtained results of the project made it possible to single out separate models of financing ophthalmological services or their components, which can be implemented during reforms of the system of financing ophthalmological care in Ukraine.The article comprises descriptions of methods, results, and arguments to consider while analysing good practices in ophthalmic along with conclusions from our research. The mixed-method research was conducted. Based on legal acts analysis and publicly available data on service utilisation for ophthalmology services we were able to identify thirty-five indicators of the six different ophthalmic diseases listed above for five countries participating in the project. For this article data was used only for Hungary, Poland, and Ukraine. Our analysis covered the period between 2010 and 2020. The focus of the comparison between countries is, on the one hand, the annual performance at the national level: the number of ambulatory cases or cases treated in one-day surgery or acute inpatient care, as well as access to appropriate ophthalmic services. On the other hand, to compare the regulatory background and its changes within the period analysed. It includes the main characteristics, types, and innovations of reimbursement and payment incentive systems, as well as the regulatory elements of quality assurance. We concentrated on country comparison of 2019 data as it was the last year before the COVID-19 pandemic, but we gathered the latest possible data published for the healthcare sector in each country. We included publications on disease burden financing ophthalmic care in the countries of our scope. The second phase of our research was to get qualitative information on financing mechanisms assumed to be a good practice so in-depth interviews were run with key stakeholders–healthcare providers and insurance bodies. This altogether allowed us to prepare the final list of good practices in financing ophthalmology services in Poland and Hungary. The last phase of our research was to verify which of the identified good practices in financing mechanisms of ophthalmic services could be implemented in another country. Based on materials from Poland and Hungary on good practices in ophthalmology, Ukrainian researchers developed a separate questionnaire and conducted in-depth interviews among the main stakeholders who ensure the reform, among providers from the public and private sectors and the national procurement agency for medical services regarding the implementation of the identified good practices in Ukraine. Hence, there were two interview guidebooks prepared ( and Tables). It is essential to understand the rationale for recommendations, which were part of our research and widely covered in the summary report from the project for all Visegrad countries and Ukraine. This article is laser focused on features in financing mechanisms only in Poland and Hungary that could be helpful for decision makers in Ukraine and the phase with qualitative research, with findings that have not been published in the summary report . The qualitative research was conducted following the standards for reporting qualitative research . All interviewers had experience in conducting qualitative research. All participants, representatives from healthcare providers and payers, have at least 10 years’ experience in ophthalmology care and knowledge about financing mechanisms in each country of residence. They all agreed voluntarily (informed verbal consent) to participate in interviews and share their perspectives. All interview data was analysed using thematic analysis. Data analysis started after the first interviews were transcribed and checked for accuracy. This was to ensure that the interview guide was adequately eliciting responses on researchers’ experiences and to monitor data saturation. Qualitative data were generated using semi-structured interviews conducted with the participants from Poland and Hungary between June 1 st and September 30 th , 2022, and with the participants from Ukraine between December 1 st , 2022, and March 30 th , 2023. Interviews were conducted by three teams (from Poland, Hungary, and Ukraine), and each team consisted of two researchers–interviewers (MT and CD for Hungary; BW and KB for Poland; OS and TY for Ukraine). All interviews took place online using Microsoft Teams. Interviews lasted between 35 minutes and 91 minutes (mean = 64 minutes). The semi-structured approach allowed for interviewers to follow natural flows and avenues of conversation, the interview guide helped reduce potential interviewer bias by ensuring that certain questions were asked and in a comparable way. All interviews were anonymized and reviewed by BW. All interviewers had access to information that could identify the interviewee. We obtained the resolution from the Committee on Research Ethics of NaUKMA, Registration number: Federal Wide Assurance №00030125. The Committee concluded that the approval of the materials is not required, therefore, the request for examination of the materials was rejected and closed, as the questionnaire for healthcare experts was designed to gather information on their perceptions regarding financing mechanisms in each country based on publicly available information. The study protocol stipulated that no personal health information would be collected. The consent to participate in the interview was provided verbally and was witnessed by two members of the research team from each country. Details about the process of gathering consents were included in the research protocol, which was reviewed by the Ethics Committee. The interviews were conducted with fifteen participants (53% female; 47% male), five per country, were interviewed to gather healthcare provider (80%) and payer (20%) perspectives from each country. The analysis of regulatory changes and healthcare utilisation along with in the insights shared during the interviews has led us to the following observations and findings. The description of the health care delivery and financing mechanisms In Poland DRG system (called Jednorodne Grupy Pacjentów) was implemented in 2008. At the same time, financial products “ambulatory DRG” (defined by procedures) were implemented in ambulatory settings (the same no matter the type of specialists). For both types of healthcare services financing was limited to the contract level (defined for each clinic/out-patient specialist). In 2015 the reform called “hospital network” was implemented. Hospitals listed in the hospital network and belonging to "hospital" ambulatory settings receive a global budget one for all outpatient and inpatient clinics) based on services performed one year before (DRG plays as points and is used to settle a budget for next year). The "satellite" ambulatory providers (not being part of any hospital listed in the hospital network system) and hospitals not included in the hospital network system are still financed with an ambulatory/hospital DRG up to a limit defined separately for each type of outpatient/inpatient clinic. As a result, the ophthalmological cases are financed with two different regimes (i.e. DRG used as a payment method or as a point system) depending on being or not on “hospital network” lists. Above it, there are some exceptions–services that are financed by DRG on a fee-for-service basis (i.e. without any limits for providers). These are for example child delivery and cataracts. There is no co-payment for healthcare services in the Polish healthcare system. If a patient wants a higher quality treatment or device, he must cover the whole cost of treatment. Since 1993 in Hungary the basis of payment mechanism the NHIF has applied the DRGs-based payment scheme (the Hungarian version is Homogén Betegségcsoport: HBCs) both for in-patient and one-day surgery care. For each DRGs, there is a special weight number that expresses the relative cost level as compared to the national average, and it is multiplied by the national equal base rate. For outpatient care since 1993, it has been used the so-called German score system based on the detailed list of activities calculated in scores (points), each activity code has a fixed number of points. Both the ambulance scores (HUF/point) value and the DRGs base rate are predetermined as a national equal tariff announced by the Ministry that is responsible for the Health Affairs (MoH). The MoH announced different ministerial orders that regulate the detailed list of financed out-patient ophthalmology services, including diagnostic examinations (e.g. OTC) and treatment technologies and in-patient services, short-term care distinguishing twenty-five different DRGs (mainly surgeries). In Hungary, in the compulsory health insurance scheme there is no generic co-payment system, just for the pharmaceutical for ophthalmic treatment, the patients have to contribute 10%, 20%, and 45% of the total price. During the COVID-19 pandemic, the abovementioned performance-based payment system was temporarily suspended, but since the 1 st of February 2023. The former system was applied again with smaller modifications in the codes and tariffs it is true for the formerly successful cataract waiting list reduction programme that was launched again. Four periods of changes in the financing mechanisms of the health care system in Ukraine can be identified. The first period took place until 2018 when the financing of the treatment of all ophthalmological cases in the public sector was conducted according to the fixed budget method. The second period began in 2018 and lasted until the first quarter of 2020—the reform of the health care system in Ukraine began. New payment mechanisms were introduced at the level of primary medical care using the capitation method, and the hospital and outpatient sectors, which provided specialized care to the country’s population, were financed based on a fixed budget. The third period began in 2020 and is characterized by the fact that the financing reform was extended to secondary outpatient and inpatient care. Surgical treatments of patients with ophthalmological cases were included in the package of medical guarantees "Surgery for adults and children in hospital". The fourth period started in 2021 when the principles of financing by diagnostic and related groups (DRGs) began to be applied, and reforming according to this method of payment is still taking place. In 2022, ophthalmological cases will be included in the package of medical guarantees "Surgery for adults and children within one-day hospitalization". It is very difficult to compare the prices of healthcare services between Hungary and Poland, as they differ between countries by ICD-10 (disease types) and/or ICD-9 codes even if they are defined by DRGs. In Ukraine, healthcare services only started to be determined by DRGs in 2021, i.e. specialized medical services were paid according to the global budget, considering adjustment factors. From 2022, payment for healthcare services is conducted as a combination of the global budget and the global rate and payment for a treated case of DRGs . The descriptions of the health care delivery and financing ophthalmology services in Hungary and Poland have led us to the conclusion that there are significant differences across these countries. The differences exist although both countries have implemented DRG-based systems for hospital care for a long time. The comparison of practices in each country allowed us to identify these differences and define good practices in treatments for cataract, glaucoma, AMD, DME, cornea transplantations, and vitrectomy. As was found in the summary report, there is no good practice in financing leading to increase access to all services in ophthalmology and financial incentives differ not only across countries but also across diseases . Identification of practices of health care services and financing of ophthalmology Cataract is and will continue to be, the most frequent treatment provided in ophthalmology care. As our populations are aging, ensuring access to treat the problem of visual impairment when needed will become essential. The major difference identified during our research pertains to the treatment scheme. In Hungary cataract surgeries can be financed within the ambulatory care, while in Poland only in hospital care. This now contributes ’only’ to the higher costs of care, as the mechanism of a financial penalty (a 10% decrease in DRG price) in the case of one-day treatments lower than 80% leads to a sharp increase in short hospitalizations. With 98% of daily cases (Zdrowe dane. Choroby narządu wzroku), Poland does not differ much from the ambulatory treatment scheme provided in Hungary. The main barrier to access treatment is also limited (put by third-party payers on healthcare providers) in financing the number of cataract surgeries. No limits on volumes introduced in Poland in 2018 allowed to treat patients without unnecessary delay. Together with separating the payment for qualification visits (with specific medical criteria for cataract surgery) it increased in number of treatments and reducing significantly the waiting time for treatment (from 489 days in 2016 to month to 127 days in 2020 in Poland; Zdrowe dane. Choroby narządu wzroku). The average waiting time for cataract surgery was 104 days in 2014 and 39 days in 2019 after a few years of a conscious continuous waiting time reduction program organized and financed by the NHIF in Hungary . The second disease that was explored by the project team members in each country was glaucoma. During our team discussions and interviews with healthcare providers, we discovered that glaucoma is mostly treated with the use of pharmacotherapy. Therefore, ensuring access to many medicines that correspond to the wide variety of clinical conditions of patients diagnosed with glaucoma was crucial. The wide list of reimbursed drugs allows matching the treatment with health needs while being affordable for patients. According to a Polish medical doctor who works in an ambulatory care facility with the public-payer contract and private facilities: There is great access to modern drug treatments , with huge public funding , especially in the age group 75+ (P2). For glaucoma and vitrectomy also, financial limits were released in Poland for ambulatory as well as for surgical care. During interviews, they were pointed out as “a wish good practice” ( Distinguishing the consultation- qualification for treatment , cataract surgery , and follow-up consultation was the key to providing access with the right financing , including removing the limits for financing surgeries . There is no data about waiting time for glaucoma treatment . There are waiting lists for vitrectomy . These procedures probably require removing limits too . (…) Removing limitations in financing medical procedures in ophthalmology could help to increase access to services . (P3). The requirements and proper financing would have to be adjusted to move more treatments to 1-day hospitalizations . Cataract , glaucoma , and vitrectomy could be mostly conducted during the 1day stay (P5). AMD was another disease that was analysed. Increasing access to health care can be achieved by adjusting the financial product to the complexity of the procedure. The greatest shift in volumes was observed in Poland, where in 2018 introduction of the so-called “drug programme” allowed the gradual transition from inpatient hospitalizations to outpatient care (from 9,790 in 2015 to 27,787 in 2020, Zdrowe dane. Choroby narządu wzroku). It [the program] is one of the few things that works . It is an example of how the pricing influences the market : the better pricing pushed providers to buy medical equipment and equipment allowed to diagnose the patients and this resulted in higher access to treatment (P6). In Hungary this programme is defined as an ambulatory treatment provided by the hospital so it realised capacity resources in hospitals leading to a sharp increase in the number of people treated (from 4,500 in 2014 to 19,300 in 2019 in Hungary’s NHIF financial database), the payment method was the same than for cataracts a special DRGs code was set up exactly for AMD neovascularisation for relatively low tariffs (RW: 0,15122), appr. 80 EUR/case without volume limit for the hospitals. When analysing the financial mechanism for treatment for DME, wide coverage of diagnostics in Hungary stood up as good practice both increasing access and quality of care respectively ( The diabetic ambulance service (the network of diabetic care centres) itself are quite well organized , who are already diagnosed can receive appropriate care , yearly control of the typical complications (among them the blindness) , however , it should be recommendable to launch further national early detection campaign of DME (P9). Regarding DME following diagnostic services are financed: detailed anamnesis and basic examination: eye movement examination, digital eye pressure estimation, slit lamp examination, a subjective determination of refraction, ophthalmoscopy, blood pressure, blood lipids, gonioscopy, fluorescein angiography, OCT optical coherence tomography. A good practice influences many treatments in financing healthcare services, where it is possible, in outpatient care. In Poland, the introduction of a so-called drug programme for DME as outpatient treatment (realized by high-quality hospitals) lead to allow to help these patients most efficiently the program is developing , and more and more patients are being treated so access to care is improving (P4). Highly complicated and relatively low frequency realized procedures should be subject to centralisation (i.e. performed only in a high-quality health centre). In ophthalmology, such procedure is cornea transplantations. Acquiring organs, tissues, and cells for these procedures are the top barriers to access. Functioning tissue banks organising a sufficient number of transplants is the key to covering the demand for treatment. The other thing is payment mechanisms. If the cost of the cornea is the same for all providers DRG must cover the whole cost of the cornea and the cost of the medical procedure, as in Hungary ( The DRGs on cornea transplant cover the following interventions and activities : costs of extraction of donor cornea , detailed anamnesis and basic examination : eye movement examination , digital eye pressure estimation , slit-lamp examination , subjective determination of refraction , ophthalmoscopy (P10). In Poland, there are tissue banks financed from public sources (i.e. the cornea is given for free to the provider) and tissue banks that do not receive public financing, so the cost of the cornea has to be covered by a healthcare provider. Separating the payments for cornea from the payments for transplantations was an important step towards increasing access to this service by covering costs for those health providers that procure cornea from commercial cell and tissue banks if they do not receive cornea from the public ones. Such financial incentives were implemented in Poland. We should conduct about five thousand surgeries per year , we do about 1200 . The longer patients wait , the more invasive transplantation , hence more complications , and higher costs for everyone . There are mechanisms to gather cornea and incentives for gathering cornea (P4). Financing schemes can also influence the quality of treatment. In this type of action, some incentives were identified the same for several types of ophthalmological diseases. The most common were procedures to centralise care in high-quality healthcare centres. This was a case for special glaucoma surgery and vitrectomy in Hungary (provided by the four Medical Universities, and in some other county hospitals), AMD and DME in Poland (separate contracts for hospitals performing certain conditions), and cornea transplant in Hungary and Poland. ( For some procedures , like for glaucoma , it would be good to have centralization , for those that are rather rare , not in thousands , so that we have highly specialized centres to deal with these rare or more complicated cases . [corneal transplantation] It should not be centralized because we need too many transplantations (P4). Several factors would be highlighted for the concentration of special and very expensive treatments and surgeries : the composition of the team , the improving organizational culture , and team building , the learning curve of physicians , and economies of scale all can contribute to a better cost-effectiveness and higher outcome (P9). For cataract treatment, it was stated that mechanisms ensuring flexibility of payment are good practices for increasing better access to treatment as they are the best way of matching patient needs ( The ideal patient pathway : to detect the problem at a lower level , in any secondary care provider let’s be in a smaller city , county hospital , or University Clinic outpatient department . After receiving the referral to the surgery centre the patient must be diagnosed , and examined to the patient should be thoroughly , to decide about the best available and appropriate lens , soon there be a discussion and agreement with the patient about the best option and its financing if it needs additional private payment (co-payment) (P8). In both analysed countries we found some mechanisms of that kind. In Poland, the price for DRG groups is adjusted by a factor of 1.25 when using toric lenses or iridial lenses (implemented in 2018). This 25% additional payment is aimed at covering the difference in costs between “classic” lenses (covered in DRG price) and toric/iridial ones. In Hungary, there has been introduced a new DRG group (since 2014) dedicated only to the costs of surgery. Hospital material, hotel costs, and human resource costs are reimbursed by the public third-party payer (NHIF), while the cost of a new lens implant is paid directly by the patient. So, the patient can choose between a whole public payment (cost of the lens included in DRG price) or part public payment (cos of the lens is covered by out-of-pocket payment). Quality indicators measured while providing healthcare services and making them publicly available result in better healthcare outcomes. In Poland, such indicators were added to the DRG scheme. When reporting cataract DRG healthcare providers are obligated to report an assessment of surgeries. Three indicators are reported with DRG (to each patient case): posterior capsule rupture, endophthalmitis, and change in visual acuity. Especially vision impairment measurement (before and after) surgery appeared to be a patient-choice sensitive indicator. Also, the definition of financial products can influence the treatment quality. The most important practices identified by our research team were payment for each injection separately and financial products for evaluation of achieved outcomes. Payment for each injection imposes elasticity for treatment conditions for each patient. It is implemented in Poland and Hungary for AMD and in Poland for DME. The change in financing with the fee-for-service payments in 2021 can contribute to greater flexibility in adjusting the treatment scheme to patients’ needs and coverage of providers’ costs linked to the treatment. It also enables greater monitoring and keeping up with the clinical excellence guidelines. In Poland, the quality of care was also assured via new financial products to cover the cost of evaluation of achieved health outcomes during a follow-up consultation after the end of treatment. Good practices in financing ophthalmology care to be proposed for implementation in Ukraine The analysis of good practices of Poland and Hungary and the expert opinion of ophthalmologist-surgeons, representatives of the National Health Service as a purchaser of services, and the heads of health care facilities from the public sector in Ukraine showed that the best practice for Ukraine among those proposed by this study, which can be implemented into routine practice at the moment, is separate payment for lenses for the treatment of cataracts. Experts consider separate payment or even co-payment for standard or more individual (rare) lenses as good practice to adopt. The Programme of Medical Guarantees covers most costs, including diagnostics, consumables, surgery, and basic lenses. One of the experts noted the following: These are very rare lenses , so only co-payment or even full payment for lenses because the state cannot cover all lenses , especially such individual ones (P11). The plans for 2023–2024 include a review of pricing and allocation of individual services or medical products for which the patient will have to pay separately: Right now , we are preparing a more detailed list , which will , as it were , break down the general services to understand that we cannot now fully pay for Programme of the Medical Guarantees . These services will be included in a separate Cabinet of Ministries Decree , which will be supplemented with a list of paid services . It will be clearly understood what the state pays for and what it absolutely does not cover by state budget (P14). The proposed practices for cataract, which relate to financing under the DRG, quality indicators for DRG, and limits, were classified as good practices and partially supported by Ukrainian experts, since the financing under the DRG is only being implemented in Ukraine, it is difficult for experts to predict what changes will happen soon since these processes depend on the adoption of relevant legislation and the introduction of relevant changes in financing system. Thus, it was difficult for the experts to answer about good practices such as a decrease in DRG payment in case of a low share of one-day treatment, because two funding mechanisms are currently being implemented, one-day treatment under the Programme of Medical Guarantees, and the DRG financing system: Currently , all ophthalmology is carried out in outpatient settings , or it is 1-day surgery . In rare cases , the patient may be hospitalized (P12). Experts do not know how the healthcare system will be related to DGR soon. However, different package prices for three lens types as a good practice for cataract was not supported by experts: Due to the high specificity , a very small percentage of toric and aniridia lenses are used , some fractions of a percent , so no , it is better to cover average used lenses . The specificity of lenses for these patients is very high , there are very few of them … These lenses are very specific , sometimes even individual , and their production can be more expensive than the expected social result , so I don’t think that covering from the budget is appropriate to these types of lenses (P13). Unequivocal support from the ophthalmological community was received for the good practice of Poland and Hungary in the maximum expansion of the National list of medications (both original medicines and generics) for the treatment of glaucoma. Experts from Ukraine also consider that it is necessary to expand the "Affordable Medicines" reimbursement program and add absolutely all medicines of various classes used for the treatment of glaucoma (P13). This is justified by the fact that glaucoma as a disease has a significant burden on the state because if it is detected or started treatment too late , it can lead to absolute blindness , and therefore increase the level of disability from glaucoma (P12). The experts supported a good practice as wide coverage of diagnostics for DME and ADM: Screening programs are needed , because these are the diseases that , in suppressed stages , can carry a social burden for the state and increase disability due to blindness (P14). ‘There is no systematic screening of the early stage of AMD . Early diagnosis and early treatment would result in a better outcome (P9). Overall, it can be said that in Hungary there is no national protocol for the organization of early detection and effective treatment of AMD among the rapidly aging population. It was also said that primary care prevention in ophthalmology is usually more expensive than treatment; it is very expensive to cover from the state budget , but screening programs are possible to implement (P10). AMD and DME screening programs are needed because these diseases can carry a social burden for the state and increase disability due to blindness . International standards of treatment are supported in the country , including following the rules of primary diagnosis , but perhaps the government and the Ministry of Health should pay more attention to screenings (P15). Financial products for evaluation of achieved outcomes and payment for each injection, including the biological treatment anti-VEGF, separately were recognized as good AMD and DME practices that should be implemented in Ukraine after the adoption of necessary legislation. Experts noted that medications for the treatment of AMD and DME are expensive , so none of them are sure whether patients will be able to cover all injections out of pocket , and whether the state budget can cover everything . Perhaps it is worth periodically launching local state programs , within which such medicines would be purchased in regions or state clinics with the highest demand (P14). Also, it is necessary “ to expand the state list of medicines due to their high cost to increase access to services (P12). As a result, the implementation of this good practice should be postponed. Cornea transplantation is currently not included in a separate package of the Programme of Medical Guarantees, because such surgery is still very difficult to cover fully (P15). However, it is conducted within the "Surgery for adults and children in hospital" package, as it is a high-tech surgery. This package covers all examinations, surgery, and partial consumables. Transplantation in Ukraine is already performed only in specialized centres, particularly in Dnipro, Odesa, and Kyiv, and it cannot be otherwise , because the patient must be prepared by all standards (P15). Experts found the functioning of tissue banks and separate payments for procedures and cornea as a good practice, but at the legislative level, there is currently no permission to take corneas from commercial cells and tissue banks. There is permission both to use only state banks, but the quality of the cornea is not high (P13), and to conduct transplantation from a posthumous donor. Of course, functioning tissue banks to have sufficient transplants is a good practice , but Ukraine must first develop a legislative framework and all the conditions for conducting high-quality corneal transplantation (P12). Some experts said that there is support for implementing a separate package of medical guarantees for transplants, but the circumstances in Ukraine currently do not provide a clear framework for realization. The system of providing medical services is currently constructed in Ukraine in such a way that surgery for almost all nosologies is performed at the inpatient level or as a one-day hospitalization. Therefore, experts from Ukraine did not support dedicated centres for surgeries of all nosologies as a good practice to be implemented in Ukraine. To implement such a practice, it is necessary to change the approach and system from the legislative level, which is currently difficult and almost impossible in the current conditions. During the interviews with experts, it was revealed that it would be a good practice for Ukraine to perform vitrectomy surgery in specialized institutions to increase quality, which also provides greater opportunities. At the same time, experts doubted that such good practice will improve the quality of vitrectomy services in Ukraine. It is commented by the following: This is a good practice , but in our country , everything is regulated by certain laws and decrees (P12). Implementing these good practices requires significant legal work upfront. In Poland DRG system (called Jednorodne Grupy Pacjentów) was implemented in 2008. At the same time, financial products “ambulatory DRG” (defined by procedures) were implemented in ambulatory settings (the same no matter the type of specialists). For both types of healthcare services financing was limited to the contract level (defined for each clinic/out-patient specialist). In 2015 the reform called “hospital network” was implemented. Hospitals listed in the hospital network and belonging to "hospital" ambulatory settings receive a global budget one for all outpatient and inpatient clinics) based on services performed one year before (DRG plays as points and is used to settle a budget for next year). The "satellite" ambulatory providers (not being part of any hospital listed in the hospital network system) and hospitals not included in the hospital network system are still financed with an ambulatory/hospital DRG up to a limit defined separately for each type of outpatient/inpatient clinic. As a result, the ophthalmological cases are financed with two different regimes (i.e. DRG used as a payment method or as a point system) depending on being or not on “hospital network” lists. Above it, there are some exceptions–services that are financed by DRG on a fee-for-service basis (i.e. without any limits for providers). These are for example child delivery and cataracts. There is no co-payment for healthcare services in the Polish healthcare system. If a patient wants a higher quality treatment or device, he must cover the whole cost of treatment. Since 1993 in Hungary the basis of payment mechanism the NHIF has applied the DRGs-based payment scheme (the Hungarian version is Homogén Betegségcsoport: HBCs) both for in-patient and one-day surgery care. For each DRGs, there is a special weight number that expresses the relative cost level as compared to the national average, and it is multiplied by the national equal base rate. For outpatient care since 1993, it has been used the so-called German score system based on the detailed list of activities calculated in scores (points), each activity code has a fixed number of points. Both the ambulance scores (HUF/point) value and the DRGs base rate are predetermined as a national equal tariff announced by the Ministry that is responsible for the Health Affairs (MoH). The MoH announced different ministerial orders that regulate the detailed list of financed out-patient ophthalmology services, including diagnostic examinations (e.g. OTC) and treatment technologies and in-patient services, short-term care distinguishing twenty-five different DRGs (mainly surgeries). In Hungary, in the compulsory health insurance scheme there is no generic co-payment system, just for the pharmaceutical for ophthalmic treatment, the patients have to contribute 10%, 20%, and 45% of the total price. During the COVID-19 pandemic, the abovementioned performance-based payment system was temporarily suspended, but since the 1 st of February 2023. The former system was applied again with smaller modifications in the codes and tariffs it is true for the formerly successful cataract waiting list reduction programme that was launched again. Four periods of changes in the financing mechanisms of the health care system in Ukraine can be identified. The first period took place until 2018 when the financing of the treatment of all ophthalmological cases in the public sector was conducted according to the fixed budget method. The second period began in 2018 and lasted until the first quarter of 2020—the reform of the health care system in Ukraine began. New payment mechanisms were introduced at the level of primary medical care using the capitation method, and the hospital and outpatient sectors, which provided specialized care to the country’s population, were financed based on a fixed budget. The third period began in 2020 and is characterized by the fact that the financing reform was extended to secondary outpatient and inpatient care. Surgical treatments of patients with ophthalmological cases were included in the package of medical guarantees "Surgery for adults and children in hospital". The fourth period started in 2021 when the principles of financing by diagnostic and related groups (DRGs) began to be applied, and reforming according to this method of payment is still taking place. In 2022, ophthalmological cases will be included in the package of medical guarantees "Surgery for adults and children within one-day hospitalization". It is very difficult to compare the prices of healthcare services between Hungary and Poland, as they differ between countries by ICD-10 (disease types) and/or ICD-9 codes even if they are defined by DRGs. In Ukraine, healthcare services only started to be determined by DRGs in 2021, i.e. specialized medical services were paid according to the global budget, considering adjustment factors. From 2022, payment for healthcare services is conducted as a combination of the global budget and the global rate and payment for a treated case of DRGs . The descriptions of the health care delivery and financing ophthalmology services in Hungary and Poland have led us to the conclusion that there are significant differences across these countries. The differences exist although both countries have implemented DRG-based systems for hospital care for a long time. The comparison of practices in each country allowed us to identify these differences and define good practices in treatments for cataract, glaucoma, AMD, DME, cornea transplantations, and vitrectomy. As was found in the summary report, there is no good practice in financing leading to increase access to all services in ophthalmology and financial incentives differ not only across countries but also across diseases . Cataract is and will continue to be, the most frequent treatment provided in ophthalmology care. As our populations are aging, ensuring access to treat the problem of visual impairment when needed will become essential. The major difference identified during our research pertains to the treatment scheme. In Hungary cataract surgeries can be financed within the ambulatory care, while in Poland only in hospital care. This now contributes ’only’ to the higher costs of care, as the mechanism of a financial penalty (a 10% decrease in DRG price) in the case of one-day treatments lower than 80% leads to a sharp increase in short hospitalizations. With 98% of daily cases (Zdrowe dane. Choroby narządu wzroku), Poland does not differ much from the ambulatory treatment scheme provided in Hungary. The main barrier to access treatment is also limited (put by third-party payers on healthcare providers) in financing the number of cataract surgeries. No limits on volumes introduced in Poland in 2018 allowed to treat patients without unnecessary delay. Together with separating the payment for qualification visits (with specific medical criteria for cataract surgery) it increased in number of treatments and reducing significantly the waiting time for treatment (from 489 days in 2016 to month to 127 days in 2020 in Poland; Zdrowe dane. Choroby narządu wzroku). The average waiting time for cataract surgery was 104 days in 2014 and 39 days in 2019 after a few years of a conscious continuous waiting time reduction program organized and financed by the NHIF in Hungary . The second disease that was explored by the project team members in each country was glaucoma. During our team discussions and interviews with healthcare providers, we discovered that glaucoma is mostly treated with the use of pharmacotherapy. Therefore, ensuring access to many medicines that correspond to the wide variety of clinical conditions of patients diagnosed with glaucoma was crucial. The wide list of reimbursed drugs allows matching the treatment with health needs while being affordable for patients. According to a Polish medical doctor who works in an ambulatory care facility with the public-payer contract and private facilities: There is great access to modern drug treatments , with huge public funding , especially in the age group 75+ (P2). For glaucoma and vitrectomy also, financial limits were released in Poland for ambulatory as well as for surgical care. During interviews, they were pointed out as “a wish good practice” ( Distinguishing the consultation- qualification for treatment , cataract surgery , and follow-up consultation was the key to providing access with the right financing , including removing the limits for financing surgeries . There is no data about waiting time for glaucoma treatment . There are waiting lists for vitrectomy . These procedures probably require removing limits too . (…) Removing limitations in financing medical procedures in ophthalmology could help to increase access to services . (P3). The requirements and proper financing would have to be adjusted to move more treatments to 1-day hospitalizations . Cataract , glaucoma , and vitrectomy could be mostly conducted during the 1day stay (P5). AMD was another disease that was analysed. Increasing access to health care can be achieved by adjusting the financial product to the complexity of the procedure. The greatest shift in volumes was observed in Poland, where in 2018 introduction of the so-called “drug programme” allowed the gradual transition from inpatient hospitalizations to outpatient care (from 9,790 in 2015 to 27,787 in 2020, Zdrowe dane. Choroby narządu wzroku). It [the program] is one of the few things that works . It is an example of how the pricing influences the market : the better pricing pushed providers to buy medical equipment and equipment allowed to diagnose the patients and this resulted in higher access to treatment (P6). In Hungary this programme is defined as an ambulatory treatment provided by the hospital so it realised capacity resources in hospitals leading to a sharp increase in the number of people treated (from 4,500 in 2014 to 19,300 in 2019 in Hungary’s NHIF financial database), the payment method was the same than for cataracts a special DRGs code was set up exactly for AMD neovascularisation for relatively low tariffs (RW: 0,15122), appr. 80 EUR/case without volume limit for the hospitals. When analysing the financial mechanism for treatment for DME, wide coverage of diagnostics in Hungary stood up as good practice both increasing access and quality of care respectively ( The diabetic ambulance service (the network of diabetic care centres) itself are quite well organized , who are already diagnosed can receive appropriate care , yearly control of the typical complications (among them the blindness) , however , it should be recommendable to launch further national early detection campaign of DME (P9). Regarding DME following diagnostic services are financed: detailed anamnesis and basic examination: eye movement examination, digital eye pressure estimation, slit lamp examination, a subjective determination of refraction, ophthalmoscopy, blood pressure, blood lipids, gonioscopy, fluorescein angiography, OCT optical coherence tomography. A good practice influences many treatments in financing healthcare services, where it is possible, in outpatient care. In Poland, the introduction of a so-called drug programme for DME as outpatient treatment (realized by high-quality hospitals) lead to allow to help these patients most efficiently the program is developing , and more and more patients are being treated so access to care is improving (P4). Highly complicated and relatively low frequency realized procedures should be subject to centralisation (i.e. performed only in a high-quality health centre). In ophthalmology, such procedure is cornea transplantations. Acquiring organs, tissues, and cells for these procedures are the top barriers to access. Functioning tissue banks organising a sufficient number of transplants is the key to covering the demand for treatment. The other thing is payment mechanisms. If the cost of the cornea is the same for all providers DRG must cover the whole cost of the cornea and the cost of the medical procedure, as in Hungary ( The DRGs on cornea transplant cover the following interventions and activities : costs of extraction of donor cornea , detailed anamnesis and basic examination : eye movement examination , digital eye pressure estimation , slit-lamp examination , subjective determination of refraction , ophthalmoscopy (P10). In Poland, there are tissue banks financed from public sources (i.e. the cornea is given for free to the provider) and tissue banks that do not receive public financing, so the cost of the cornea has to be covered by a healthcare provider. Separating the payments for cornea from the payments for transplantations was an important step towards increasing access to this service by covering costs for those health providers that procure cornea from commercial cell and tissue banks if they do not receive cornea from the public ones. Such financial incentives were implemented in Poland. We should conduct about five thousand surgeries per year , we do about 1200 . The longer patients wait , the more invasive transplantation , hence more complications , and higher costs for everyone . There are mechanisms to gather cornea and incentives for gathering cornea (P4). Financing schemes can also influence the quality of treatment. In this type of action, some incentives were identified the same for several types of ophthalmological diseases. The most common were procedures to centralise care in high-quality healthcare centres. This was a case for special glaucoma surgery and vitrectomy in Hungary (provided by the four Medical Universities, and in some other county hospitals), AMD and DME in Poland (separate contracts for hospitals performing certain conditions), and cornea transplant in Hungary and Poland. ( For some procedures , like for glaucoma , it would be good to have centralization , for those that are rather rare , not in thousands , so that we have highly specialized centres to deal with these rare or more complicated cases . [corneal transplantation] It should not be centralized because we need too many transplantations (P4). Several factors would be highlighted for the concentration of special and very expensive treatments and surgeries : the composition of the team , the improving organizational culture , and team building , the learning curve of physicians , and economies of scale all can contribute to a better cost-effectiveness and higher outcome (P9). For cataract treatment, it was stated that mechanisms ensuring flexibility of payment are good practices for increasing better access to treatment as they are the best way of matching patient needs ( The ideal patient pathway : to detect the problem at a lower level , in any secondary care provider let’s be in a smaller city , county hospital , or University Clinic outpatient department . After receiving the referral to the surgery centre the patient must be diagnosed , and examined to the patient should be thoroughly , to decide about the best available and appropriate lens , soon there be a discussion and agreement with the patient about the best option and its financing if it needs additional private payment (co-payment) (P8). In both analysed countries we found some mechanisms of that kind. In Poland, the price for DRG groups is adjusted by a factor of 1.25 when using toric lenses or iridial lenses (implemented in 2018). This 25% additional payment is aimed at covering the difference in costs between “classic” lenses (covered in DRG price) and toric/iridial ones. In Hungary, there has been introduced a new DRG group (since 2014) dedicated only to the costs of surgery. Hospital material, hotel costs, and human resource costs are reimbursed by the public third-party payer (NHIF), while the cost of a new lens implant is paid directly by the patient. So, the patient can choose between a whole public payment (cost of the lens included in DRG price) or part public payment (cos of the lens is covered by out-of-pocket payment). Quality indicators measured while providing healthcare services and making them publicly available result in better healthcare outcomes. In Poland, such indicators were added to the DRG scheme. When reporting cataract DRG healthcare providers are obligated to report an assessment of surgeries. Three indicators are reported with DRG (to each patient case): posterior capsule rupture, endophthalmitis, and change in visual acuity. Especially vision impairment measurement (before and after) surgery appeared to be a patient-choice sensitive indicator. Also, the definition of financial products can influence the treatment quality. The most important practices identified by our research team were payment for each injection separately and financial products for evaluation of achieved outcomes. Payment for each injection imposes elasticity for treatment conditions for each patient. It is implemented in Poland and Hungary for AMD and in Poland for DME. The change in financing with the fee-for-service payments in 2021 can contribute to greater flexibility in adjusting the treatment scheme to patients’ needs and coverage of providers’ costs linked to the treatment. It also enables greater monitoring and keeping up with the clinical excellence guidelines. In Poland, the quality of care was also assured via new financial products to cover the cost of evaluation of achieved health outcomes during a follow-up consultation after the end of treatment. The analysis of good practices of Poland and Hungary and the expert opinion of ophthalmologist-surgeons, representatives of the National Health Service as a purchaser of services, and the heads of health care facilities from the public sector in Ukraine showed that the best practice for Ukraine among those proposed by this study, which can be implemented into routine practice at the moment, is separate payment for lenses for the treatment of cataracts. Experts consider separate payment or even co-payment for standard or more individual (rare) lenses as good practice to adopt. The Programme of Medical Guarantees covers most costs, including diagnostics, consumables, surgery, and basic lenses. One of the experts noted the following: These are very rare lenses , so only co-payment or even full payment for lenses because the state cannot cover all lenses , especially such individual ones (P11). The plans for 2023–2024 include a review of pricing and allocation of individual services or medical products for which the patient will have to pay separately: Right now , we are preparing a more detailed list , which will , as it were , break down the general services to understand that we cannot now fully pay for Programme of the Medical Guarantees . These services will be included in a separate Cabinet of Ministries Decree , which will be supplemented with a list of paid services . It will be clearly understood what the state pays for and what it absolutely does not cover by state budget (P14). The proposed practices for cataract, which relate to financing under the DRG, quality indicators for DRG, and limits, were classified as good practices and partially supported by Ukrainian experts, since the financing under the DRG is only being implemented in Ukraine, it is difficult for experts to predict what changes will happen soon since these processes depend on the adoption of relevant legislation and the introduction of relevant changes in financing system. Thus, it was difficult for the experts to answer about good practices such as a decrease in DRG payment in case of a low share of one-day treatment, because two funding mechanisms are currently being implemented, one-day treatment under the Programme of Medical Guarantees, and the DRG financing system: Currently , all ophthalmology is carried out in outpatient settings , or it is 1-day surgery . In rare cases , the patient may be hospitalized (P12). Experts do not know how the healthcare system will be related to DGR soon. However, different package prices for three lens types as a good practice for cataract was not supported by experts: Due to the high specificity , a very small percentage of toric and aniridia lenses are used , some fractions of a percent , so no , it is better to cover average used lenses . The specificity of lenses for these patients is very high , there are very few of them … These lenses are very specific , sometimes even individual , and their production can be more expensive than the expected social result , so I don’t think that covering from the budget is appropriate to these types of lenses (P13). Unequivocal support from the ophthalmological community was received for the good practice of Poland and Hungary in the maximum expansion of the National list of medications (both original medicines and generics) for the treatment of glaucoma. Experts from Ukraine also consider that it is necessary to expand the "Affordable Medicines" reimbursement program and add absolutely all medicines of various classes used for the treatment of glaucoma (P13). This is justified by the fact that glaucoma as a disease has a significant burden on the state because if it is detected or started treatment too late , it can lead to absolute blindness , and therefore increase the level of disability from glaucoma (P12). The experts supported a good practice as wide coverage of diagnostics for DME and ADM: Screening programs are needed , because these are the diseases that , in suppressed stages , can carry a social burden for the state and increase disability due to blindness (P14). ‘There is no systematic screening of the early stage of AMD . Early diagnosis and early treatment would result in a better outcome (P9). Overall, it can be said that in Hungary there is no national protocol for the organization of early detection and effective treatment of AMD among the rapidly aging population. It was also said that primary care prevention in ophthalmology is usually more expensive than treatment; it is very expensive to cover from the state budget , but screening programs are possible to implement (P10). AMD and DME screening programs are needed because these diseases can carry a social burden for the state and increase disability due to blindness . International standards of treatment are supported in the country , including following the rules of primary diagnosis , but perhaps the government and the Ministry of Health should pay more attention to screenings (P15). Financial products for evaluation of achieved outcomes and payment for each injection, including the biological treatment anti-VEGF, separately were recognized as good AMD and DME practices that should be implemented in Ukraine after the adoption of necessary legislation. Experts noted that medications for the treatment of AMD and DME are expensive , so none of them are sure whether patients will be able to cover all injections out of pocket , and whether the state budget can cover everything . Perhaps it is worth periodically launching local state programs , within which such medicines would be purchased in regions or state clinics with the highest demand (P14). Also, it is necessary “ to expand the state list of medicines due to their high cost to increase access to services (P12). As a result, the implementation of this good practice should be postponed. Cornea transplantation is currently not included in a separate package of the Programme of Medical Guarantees, because such surgery is still very difficult to cover fully (P15). However, it is conducted within the "Surgery for adults and children in hospital" package, as it is a high-tech surgery. This package covers all examinations, surgery, and partial consumables. Transplantation in Ukraine is already performed only in specialized centres, particularly in Dnipro, Odesa, and Kyiv, and it cannot be otherwise , because the patient must be prepared by all standards (P15). Experts found the functioning of tissue banks and separate payments for procedures and cornea as a good practice, but at the legislative level, there is currently no permission to take corneas from commercial cells and tissue banks. There is permission both to use only state banks, but the quality of the cornea is not high (P13), and to conduct transplantation from a posthumous donor. Of course, functioning tissue banks to have sufficient transplants is a good practice , but Ukraine must first develop a legislative framework and all the conditions for conducting high-quality corneal transplantation (P12). Some experts said that there is support for implementing a separate package of medical guarantees for transplants, but the circumstances in Ukraine currently do not provide a clear framework for realization. The system of providing medical services is currently constructed in Ukraine in such a way that surgery for almost all nosologies is performed at the inpatient level or as a one-day hospitalization. Therefore, experts from Ukraine did not support dedicated centres for surgeries of all nosologies as a good practice to be implemented in Ukraine. To implement such a practice, it is necessary to change the approach and system from the legislative level, which is currently difficult and almost impossible in the current conditions. During the interviews with experts, it was revealed that it would be a good practice for Ukraine to perform vitrectomy surgery in specialized institutions to increase quality, which also provides greater opportunities. At the same time, experts doubted that such good practice will improve the quality of vitrectomy services in Ukraine. It is commented by the following: This is a good practice , but in our country , everything is regulated by certain laws and decrees (P12). Implementing these good practices requires significant legal work upfront. Several definitions of good practice were found in the literature on health economics and financing healthcare services. In the Polish literature attempts to define good practice in financing healthcare services are brought on when researchers discuss healthcare financing reform, changes in healthcare financing in Poland, and good practices in healthcare financing in Poland and the European Union . They provide insights into the challenges and opportunities in healthcare financing in Poland and highlight the importance of efficient and effective resource allocation, universal coverage, and sustainability in healthcare financing. In Hungary, good practices were consciously defined, collected, and disseminated in the frame of an EU funded project focusing on patient safety and effectiveness . It is also mentioned in the context of the effectiveness of special treatments for select diseases in ophthalmology . In Ukraine, similarly to Poland and Hungary, there is no fixed term or definition of “good practice” in legal acts pertaining to financing health care services. We needed the common denominator for Hungary, Poland, and Ukraine therefore we used a very broad definition that comprises all financial mechanisms that impact the volume or quality of ophthalmology care. There are ample reasons why a financial mechanism may be perceived to be a good practice in one country and cannot be defined as a good practice in another country. These reasons can be grouped into four categories: geographic/regional; cultural/historic/regulatory background; financing/organisational (centralization or decentralization); and infrastructure/workforce-related (availability of facilities/doctors) . Geographic factors can influence the effectiveness of a financial mechanism in different countries. For example, in a country with a large rural population or regions with limited access to healthcare facilities, a decentralized financial mechanism may be more suitable. It allows for greater flexibility and localized decision-making to address regional disparities in healthcare access . On the other hand, in densely populated urban areas, a centralized financial mechanism might be more efficient. It can streamline resource allocation, coordination, and service delivery. Cultural, historic, and regulatory factors can shape the perception of a financial mechanism as a good practice. These factors include: cultural attitudes towards healthcare—different cultures may have varying expectations, preferences, and attitudes towards healthcare financing and delivery . A financial mechanism that aligns with cultural norms and values is more likely to be perceived as effective . Historic context and healthcare system development—countries with different historical trajectories may have unique healthcare systems and financing mechanisms that have evolved . The compatibility of a financial mechanism with the existing system plays a significant role in its perceived effectiveness . Regulatory frameworks—regulatory policies, such as healthcare regulations and insurance laws, vary across countries. A financial mechanism that aligns with existing regulations and legal frameworks is more likely to be considered a good practice . The financing and organizational aspects of a healthcare system can impact the suitability of a financial mechanism . Given the results of our research, the key considerations include: Funding availability and allocation. Countries with various levels of financial resources like Hungary, Poland, may require different financial mechanisms to ensure sustainable and equitable healthcare financing. Centralization or decentralization. The degree of centralization or decentralization in a healthcare system, like in Poland and Hungary, can influence the effectiveness of a financial mechanism. Centralized systems might prioritize efficiency and resource allocation, while decentralized systems, may focus on local autonomy and responsiveness. Governance and administrative structures. The organizational structure of healthcare systems, such as the presence of a single-payer system (Poland), multiple insurance providers, or a mix of public and private entities (like in Hungary), can affect the implementation and success of a financial mechanism. Finally, infrastructure and workforce considerations can impact the viability of a financial mechanism . Factors to consider include availability of healthcare facilities and healthcare professionals. Countries with disparate availability or distribution of healthcare facilities may require different financial mechanisms to ensure equal access to care. Similar observation can be made about healthcare workforce—the availability and distribution of healthcare professionals, such as doctors and specialists, can influence the effectiveness of a financial mechanism. A mechanism that accounts for workforce shortages or maldistribution is more likely to be considered a good practice. It is important to note that these categories are not exhaustive, and multiple factors can interplay within each category. Additionally, each country’s unique circumstances and priorities contribute to the perception of a financial mechanism as a good practice . Therefore, when evaluating the suitability of a financial mechanism, it is crucial to consider the specific context and needs of each country. During the interviews with experts from Hungary, Poland, and Ukraine we identified three main differences in perceptions of good practices and the rationale for implementing them in Ukraine. After conducting the interviews, three positions regarding the application of good practices in Ukraine were revealed. In general, all practices are recognized as good, but there are certain barriers and reasons for the implementation of each of them currently in Ukraine . The first category of perception fully supports good practices due to their importance in Ukraine. This is because an appropriate legal framework has already been developed, financing mechanisms have been introduced, and the implementation of DRGs (Diagnosis-Related Groups) has been expanded, which will be conducted in the near future. The second category of perception supports good practices in general. However, due to uncertain financial possibilities, the implementation of these practices is currently in question. This is because the country is facing a tough economic situation, and it is challenging to predict the availability of funds in state and local budgets. The third category of perception involves good practices that are currently challenging to implement. This is due to an undeveloped regulatory and legislative framework, as well as an incomplete reform process. These practices are not currently supported by experts for implementation. However, it is possible for Ukraine to revisit these practices once the situation stabilizes and the hostilities come to an end. Findings from the interviews in Ukraine prompted us to re-examine current developments in financing ophthalmic care. One of the interviewees brought our attention to a good practice in Poland, which involves removing limitations in financing treatments for glaucoma and vitrectomy. We identified several limitations during our research. The limitation of this study from the side of Ukraine is that there were no changes in healthcare financing mechanisms until 2017. The changes that began to take place after led to the transition of ophthalmology more into the private sector. However, since 2020, significant changes have begun in the financing of secondary specialized care in ophthalmology. That is, since 2020, there is still a transition to a new financing model, which includes financing under the DRGs. In Poland, the most important challenge was missing data for some years in reporting for some procedures and scant information for ambulatory care services in ophthalmology . Also, datasets with aggregated data points about the utilisation of services and changes in publishing practices posed a challenge in data analysis. Finally, we came across a particularly important limitation in descriptions of financial products, including enlisting ICD-9 and ICD-10 medical procedures or lack thereof, hence it was impossible to conduct international comparisons for all diseases included in our research as primarily designed. These limitations were also identified in other studies . Using qualitative research methods aimed to achieve the research objective despite the above limitations. For all these three countries with populations that heavily depend on healthcare services funded from public sources, health equity needs to be embedded in the national health policies. Financing mechanisms in healthcare are subject to constant changes and therefore monitoring and evaluation of interventions is a must-have to assess effectiveness and impact on access to healthcare services in ophthalmology. We have also identified additional research areas such as measuring the quality of care in our countries or creating a flexible financing scheme that can incorporate new treatment methods, devices, and surgical techniques. It is difficult to define one financial mechanism that fits all needs in countries of our interest, we believe that it is possible to name practices that could be used to increase the volume of services as well as to define a mechanism to effect on the quality of care. Our research findings suggest that decision-makers should focus not only on drafting interventions that target to increase service volume reducing waiting time but also those that enable monitoring of achieved health outcomes and incentivise increasing quality of care in ophthalmology. Therefore, all partners are interested in the implementation of best practices and experience, because of the most effective use of public funds and resources; moreover, the healthcare system in Ukraine is currently undergoing reform. Based on the Hungarian experience to keep the flexibility of activity and reimbursement code system in different settings like outpatient care, inpatient care, and one-day surgery as well as incorporating new (innovative) technologies as soon as possible—would be an essential lesson for Ukraine healthcare system. In Poland, it has been essential to remove financial barriers such as limitations for financing ophthalmic care. These barriers were removed for cataract glaucoma and vitrectomy, but they are still existing in other eye diseases. It is also important to monitor the availability of resources, including ensuring the training of a sufficient number of specialists and measuring the quality of care—might offer good regulatory direction to Ukrainian decision-makers. The results of this research are significant for Ukraine, because the health care system is currently at the stage of reform, and Ukrainian decision-makers need to focus on good practices to improve the system. Therefore, Ukraine is currently looking for optimal good practices for the country, and it is important to consider the experience of other countries, which can be considered through the prism of the healthcare financing system in the country. S1 Table Interview guidebook for participants based in Poland and Hungary. (DOCX) S2 Table Interview guidebook for participants in Ukraine. (DOCX) S3 Table Financial mechanisms for ophthalmology treatment (2019). (DOCX) S4 Table Cataract volumes (% change to the previous year) (in thousands). (DOCX) |
Expression of interleukin-8 and integrin β3 predicts prognosis of patients with hepatocellular carcinoma after hepatectomy | c3ad524b-802f-4085-b691-186b32a8275c | 11479500 | Anatomy[mh] | Hepatocellular carcinoma (HCC) ranks the sixth common cancer and the fourth leading cause of cancer-related deaths worldwide. Despite the availability of diverse therapies such as radiofrequency ablation, immunotherapy, and molecular targeted therapy, tumor resection surgery remains the primary therapeutic option for patients with HCC. There is increasing evidence that HCC accumulates mutations that stimulate the progression and exacerbate the outcome of the disease through the tumor microenvironment, which comprises tumor cells, stromal cells, cellular interaction proteins, and cytokines. Therefore, a better understanding of the molecular mechanisms involved in the tumor microenvironment of HCC is essential to accurately predict the prognosis of patients and investigate new therapeutic strategies. The tumor microenvironment is the basis for cancer progression and metastasis. Pro-inflammatory cytokines produced in the tumor microenvironment play a critical role in cancer-related inflammation, invasion, and metastasis. Interleukin-8 (IL-8), secreted by both tumor cells and inflammatory cells, is identified as a key regulator in this process. IL-8 promotes tumor angiogenesis and metastasis through binding to high-affinity cell surface receptors CXC chemokine receptor 1 and CXC chemokine receptor 2. Over-expression of IL-8 has been detected in variety of tumors and reported to play a crucial role in carcinogenesis, angiogenesis, invasion, and metastasis of various cancers, including HCC. Within the context of tumors, IL-8 has been described to have dual pro-tumorigenic roles, including directly stimulating proliferation or transformation of tumor cells and recruiting a large number of immunosuppressive cells to the tumor. Recent clinical studies evaluating IL-8 levels in patients receiving immune checkpoint inhibition agents deduced that myeloid tumor infiltration driven by IL-8 contributes to resistance to immune checkpoint inhibition therapy. Targeting IL-8 or its receptors may provide a novel therapeutic strategy for tumor immune-based therapies. Cytokines play various roles in tumorigenesis, tumor progression, and metastasis through cell-cell or cell-matrix interactions mediated by adhesion molecules. Integrins, a large family of cell-extracellular matrix adhesion molecules composed of α and β subunits, regulate whether cell survive, proliferate, and migrate in response to soluble growth factors and cytokines. Integrin expression varies greatly between normal and tumor tissues. It is becoming increasingly clear that tumor cells enhance the expression of integrins that promote their survival, proliferation, and migration. Among different integrins, integrin β3 serves a vital role in the progression and metastasis of various types of cancer, including HCC. Abnormal expression of integrin β3 is often associated with the development of HCC. Targeted therapies to integrin β3 have been reported in different types of cancer. Increasing evidence suggests that the close relationship between IL-8 and integrin β3 is critical for cancer progression. Our previous study has demonstrated that IL-8 promotes integrin β3 upregulation and cell invasion through PI3K/Akt pathway in HCC. However, the value of IL-8 and integrin β3 in prognosis prediction and targeted therapy of HCC remains unclear. In this study, we evaluated the relationship between IL-8 and integrin β3 expression in HCC. The association of IL-8 and integrin β3 expression levels with clinicopathological variables and patient outcome was investigated to determine whether IL-8 and integrin β3 might provide a theoretical basis for predicting the prognosis of patients with HCC after hepatectomy and for further studies on potential targeted therapeutic strategies. 2.1. Patients and samples This study enrolled 130 HCC patients who underwent hepatectomy in Shandong Provincial Hospital from 2013 to 2017 with a median follow-up of 53 months (range, 7–72 months). Clinical, laboratory, and histopathological data were retrospectively collected from the patients’ records. An invasion or tumor thrombus in the portal trunk or major branches of the portal vein was defined as macrovascular invasion. The inclusion criteria were as follows: definite pathological diagnosis, curative liver resection, and complete clinicopathological data. Patients who received preoperative anticancer treatment or had distant metastases were excluded. The study was conducted following the Declaration of Helsinki and approved by the Medical Ethics Committee of Shandong Provincial Hospital. Written informed consents were obtained from all patients. 2.2. Immunohistochemistry After routine deparaffinization, rehydration, and antigen retrieval, the sections were incubated with anti-IL-8 antibody (1:2000, #ab18672, Abcam), or anti-integrin β3 antibody (1:100, #sc-52589, Santa Cruz) at 4°C overnight, followed by incubation with the corresponding secondary antibodies at 37°C for 30 minutes and visualization using diaminobenzidine. Each section was observed under a light microscope to evaluate the expression of IL-8 and integrin β3. 2.3. Evaluation of IL-8 and integrin β3 immunohistochemistry IL-8 staining was mainly observed in the cytoplasm, while integrin β3 staining was visualized on the plasma membrane of tumor cells. Expression levels were evaluated based on the staining intensity and percentage of positively stained cells. The staining intensity was scored as 0 (no staining), 1 (weak staining), 2 (moderate staining), or 3 (strong staining). The percentage of stained cells was scored as 0 (no positive cells), 1 (≤5% positive cells), 2 (5%–50% positive cells), or 3 (≥50% positive cells). The sum of both parameters gave the final scores of each protein marker in each HCC sample, in which a final score <4 was defined as negative expression, and a score ≥4 was defined as positive expression. 2.4. Statistical analysis Statistical analyses were conducted using SPSS 22.0 software. The categorical variables were analyzed by Chi-square test. Survival analyses were performed using the Kaplan–Meier method and the log-rank test. A Spearman correlation was applied to evaluate the relationship between IL-8 and integrin β3 expression. Univariate and multivariate analyses were conducted with the Cox proportional hazard model to investigate those prognostic factors that predicted overall survival. P value < .05 was considered statistically significant. This study enrolled 130 HCC patients who underwent hepatectomy in Shandong Provincial Hospital from 2013 to 2017 with a median follow-up of 53 months (range, 7–72 months). Clinical, laboratory, and histopathological data were retrospectively collected from the patients’ records. An invasion or tumor thrombus in the portal trunk or major branches of the portal vein was defined as macrovascular invasion. The inclusion criteria were as follows: definite pathological diagnosis, curative liver resection, and complete clinicopathological data. Patients who received preoperative anticancer treatment or had distant metastases were excluded. The study was conducted following the Declaration of Helsinki and approved by the Medical Ethics Committee of Shandong Provincial Hospital. Written informed consents were obtained from all patients. After routine deparaffinization, rehydration, and antigen retrieval, the sections were incubated with anti-IL-8 antibody (1:2000, #ab18672, Abcam), or anti-integrin β3 antibody (1:100, #sc-52589, Santa Cruz) at 4°C overnight, followed by incubation with the corresponding secondary antibodies at 37°C for 30 minutes and visualization using diaminobenzidine. Each section was observed under a light microscope to evaluate the expression of IL-8 and integrin β3. IL-8 staining was mainly observed in the cytoplasm, while integrin β3 staining was visualized on the plasma membrane of tumor cells. Expression levels were evaluated based on the staining intensity and percentage of positively stained cells. The staining intensity was scored as 0 (no staining), 1 (weak staining), 2 (moderate staining), or 3 (strong staining). The percentage of stained cells was scored as 0 (no positive cells), 1 (≤5% positive cells), 2 (5%–50% positive cells), or 3 (≥50% positive cells). The sum of both parameters gave the final scores of each protein marker in each HCC sample, in which a final score <4 was defined as negative expression, and a score ≥4 was defined as positive expression. Statistical analyses were conducted using SPSS 22.0 software. The categorical variables were analyzed by Chi-square test. Survival analyses were performed using the Kaplan–Meier method and the log-rank test. A Spearman correlation was applied to evaluate the relationship between IL-8 and integrin β3 expression. Univariate and multivariate analyses were conducted with the Cox proportional hazard model to investigate those prognostic factors that predicted overall survival. P value < .05 was considered statistically significant. 3.1. Clinicopathological information A total of 130 patients (114 males and 16 females) undergoing surgical resection of hepatocellular carcinoma were enrolled in this study. Their median age was 56 years (range, 35–75 years old). The clinicopathological characteristics of all patients are listed in Table . 3.2. Immunohistochemical analysis of IL-8 and integrin β3 expression in HCC We detected the expression of IL-8 and integrin β3 in a cohort of 130 HCC patients using immunohistochemistry. IL-8 staining was observed mainly in the cytoplasm, while integrin β3 staining was detected on the plasma membrane of tumor cells (Fig. ). Positive IL-8 expression was observed in 66 (50.8%) patients and positive integrin β3 expression was found in 58 (44.6%) patients (Fig. , Table ). 3.3. Relationship between IL-8/integrin β3 expression and clinicopathological variables in HCC The clinicopathological significance of IL-8/integrin β3 expression was evaluated by comparing the IL-8/integrin β3 positive group with the IL-8/integrin β3 negative group of patients with HCC (Table ). The results revealed that IL-8 expression was highly associated with macrovascular invasion, tumor size, differentiation, and tumor-node-metastasis (TNM) stage of the tumors. IL-8 positive specimens exhibited a higher proportion of macrovascular invasion, larger tumor size, poor differentiation, and advanced TNM stage ( P < .05, respectively; Table ). There was a significant association between integrin β3 expression and TNM stage. Integrin β3 positive group exhibited a higher proportion of TNM III-staged tumors ( P < .05, Table ). 3.4. Association of relation between IL-8 and integrin β3 expression in HCC Positive integrin β3 expression was observed in 54.5% of tissues positively expressing IL-8, while the percentage of tissues negatively expressing IL-8 was 34.4%. Integrin β3 expression had a positive association of relation with IL-8 expression based on the Spearman correlation analysis ( r = 0.203, P = .021, Table ). 3.5. Prognostic value of IL-8 and integrin β3 expression in patients with HCC after hepatectomy Survival curves were generated via the Kaplan–Meier survival analysis. Patients with positive IL-8 expression (median survival time [MST] = 48 months) had a significantly poorer overall survival rate than those with negative expression (MST = 58 months; P = .025; log-rank test: χ 2 = 5.027). Kaplan–Meier curves of overall survival based on IL-8 expression are illustrated in Figure A. Patients with positive integrin β3 expression (MST = 48 months) had a significantly poorer overall survival rate than those with negative expression (MST = 57.5 months; P = .003; log-rank test: χ 2 = 8.644). Kaplan–Meier curves of overall survival based on integrin β3 expression are presented in Figure B. The 130 patients enrolled in this study were divided into 4 groups according to IL-8 and integrin β3 expression. The Kaplan–Meier analysis indicated that patients with positive IL-8 and integrin β3 expression (MST = 47 months) had a significantly poorer overall survival rate than other groups ( P = .003; log-rank test: χ 2 = 13.737). Kaplan–Meier curves of overall survival based on IL-8 and integrin β3 expression are shown in Figure C. 3.6. Univariate and multivariate analyses of prognostic factors related to survival of patients with HCC after hepatectomy In order to investigate the prognostic value of IL-8 and integrin β3 expression in patients with hepatocellular carcinoma after hepatectomy, univariate and multivariate analyses were performed using the Cox proportional hazards regression model. Besides clinicopathological variables such as macrovascular invasion, tumor number, tumor size, TNM stage, and alpha-fetoprotein level, IL-8 and integrin β3 expression were proved to be able to predict prognosis based on univariate analysis ( P < .05, respectively, Table ). Variables with P < .05 were then selected as candidates for multivariate analysis. The results indicated that macrovascular invasion, advanced TNM stage, and integrin β3 expression were independent unfavorable prognostic factors (relative risk: 2.902, 3.277, and 2.001; P = .010, .034, and .046, respectively; Table ). A total of 130 patients (114 males and 16 females) undergoing surgical resection of hepatocellular carcinoma were enrolled in this study. Their median age was 56 years (range, 35–75 years old). The clinicopathological characteristics of all patients are listed in Table . We detected the expression of IL-8 and integrin β3 in a cohort of 130 HCC patients using immunohistochemistry. IL-8 staining was observed mainly in the cytoplasm, while integrin β3 staining was detected on the plasma membrane of tumor cells (Fig. ). Positive IL-8 expression was observed in 66 (50.8%) patients and positive integrin β3 expression was found in 58 (44.6%) patients (Fig. , Table ). The clinicopathological significance of IL-8/integrin β3 expression was evaluated by comparing the IL-8/integrin β3 positive group with the IL-8/integrin β3 negative group of patients with HCC (Table ). The results revealed that IL-8 expression was highly associated with macrovascular invasion, tumor size, differentiation, and tumor-node-metastasis (TNM) stage of the tumors. IL-8 positive specimens exhibited a higher proportion of macrovascular invasion, larger tumor size, poor differentiation, and advanced TNM stage ( P < .05, respectively; Table ). There was a significant association between integrin β3 expression and TNM stage. Integrin β3 positive group exhibited a higher proportion of TNM III-staged tumors ( P < .05, Table ). Positive integrin β3 expression was observed in 54.5% of tissues positively expressing IL-8, while the percentage of tissues negatively expressing IL-8 was 34.4%. Integrin β3 expression had a positive association of relation with IL-8 expression based on the Spearman correlation analysis ( r = 0.203, P = .021, Table ). Survival curves were generated via the Kaplan–Meier survival analysis. Patients with positive IL-8 expression (median survival time [MST] = 48 months) had a significantly poorer overall survival rate than those with negative expression (MST = 58 months; P = .025; log-rank test: χ 2 = 5.027). Kaplan–Meier curves of overall survival based on IL-8 expression are illustrated in Figure A. Patients with positive integrin β3 expression (MST = 48 months) had a significantly poorer overall survival rate than those with negative expression (MST = 57.5 months; P = .003; log-rank test: χ 2 = 8.644). Kaplan–Meier curves of overall survival based on integrin β3 expression are presented in Figure B. The 130 patients enrolled in this study were divided into 4 groups according to IL-8 and integrin β3 expression. The Kaplan–Meier analysis indicated that patients with positive IL-8 and integrin β3 expression (MST = 47 months) had a significantly poorer overall survival rate than other groups ( P = .003; log-rank test: χ 2 = 13.737). Kaplan–Meier curves of overall survival based on IL-8 and integrin β3 expression are shown in Figure C. In order to investigate the prognostic value of IL-8 and integrin β3 expression in patients with hepatocellular carcinoma after hepatectomy, univariate and multivariate analyses were performed using the Cox proportional hazards regression model. Besides clinicopathological variables such as macrovascular invasion, tumor number, tumor size, TNM stage, and alpha-fetoprotein level, IL-8 and integrin β3 expression were proved to be able to predict prognosis based on univariate analysis ( P < .05, respectively, Table ). Variables with P < .05 were then selected as candidates for multivariate analysis. The results indicated that macrovascular invasion, advanced TNM stage, and integrin β3 expression were independent unfavorable prognostic factors (relative risk: 2.902, 3.277, and 2.001; P = .010, .034, and .046, respectively; Table ). Tumor progression relies on the ability of cancer cells to effectively invade surrounding tissues. Among the various mechanisms that contribute to tumor progression is the increase of the cancer cells’ motility and invasiveness influenced by a variety of soluble factors secreted from the surrounding microenvironment. Factors such as IL-8, transforming growth factor-β, and others are described as crucial molecules affecting various aspects of tumor progression. IL-8 has been shown to influence the biology, such as proliferation, migration, invasion, and angiogenesis of different types of cancer, including HCC. High circulating IL-8 levels are associated with increased tumor malignancy in HCC. Furthermore, serum IL-8 level is a significant and independent prognostic factor of survival in HCC. In tumor tissues, IL-8 expression is related to vascular invasion, tumor differentiation, and TNM stage. Moreover, high IL-8 expression also predicts poor postoperative prognosis in patients with HCC. In this study, we observed positive IL-8 expression in 50.8% of HCC tumor tissues. IL-8 positive specimens exhibited a higher proportion of macrovascular invasion, larger tumor size, poor differentiation, and advanced TNM stage. Patients with positive IL-8 expression had a significantly poorer overall survival rate than those with negative expression after hepatectomy. As the main cell adhesion receptors for a variety of extracellular matrix components, the integrins, are a family of 24 transmembrane heterodimers comprising 18 α and 8 β integrin subunits that linked noncovalently. Upon binding to the extracellular matrix, the integrins organize the cytoskeleton and activate intracellular signaling, regulating complex cellular behaviors, including survival, proliferation, migration, differentiation, and various cell fate transition during inflammation and cancer. Altered integrin expression patterns have been linked to various types of cancer. Integrin β3 (β3, as part of αvβ3 and αIIbβ3 heterodimers) has been reported to be expressed in metastatic HCC cells and function as a modulator to promote metastatic phenotype of nonmetastatic HCC cells. Our previous study also demonstrated that integrin β3 was overexpressed in highly metastatic HCC cell lines compared with low metastatic cell lines. Previous findings suggest that high integrin β3 expression in anaplastic lymphoma kinase-rearranged non-small cell lung cancer is associated with tumor metastasis, more advanced tumor stages and a worse prognosis. Our results indicated that integrin β3 expression was significantly related to TNM stage in HCC. Patients with positive integrin β3 expression had a significantly poorer overall survival rate than those with negative expression and integrin β3 expression was an independent unfavorable prognostic factor in HCC after hepatectomy. Investigating the association between IL-8 and integrin β3 expression may contribute to our understanding of the potential mechanisms underlying the tumor microenvironment involved in tumor progression. Our previous study has demonstrated that IL-8 promotes integrin β3 upregulation and the invasion of HCC cells through PI3K/Akt pathway. In this study, we found a positive relationship between IL-8 and integrin β3 expression in HCC, and patients with positive IL-8 and integrin β3 expression had a significantly shorter overall survival time. In conclusion, our research indicated that increased expression of IL-8 and integrin β3 in HCC correlated with tumor progression and a poor survival. Integrin β3 expression was found to be an independent unfavorable prognostic factor in HCC after hepatectomy. Targeting integrin β3 might be a potential therapeutic approach in preventing tumor progression in HCC. Funding acquisition: Jiao Zhang, Mingze Ma, Fengkai Sun. Project administration: Jiao Zhang. Writing—review & editing: Jiao Zhang. Data curation: Yi Yin. Resources: Jiliang Tang. Software: Jiliang Tang. Formal analysis: Mingze Ma. Investigation: Huimin Shen, Yingrong Zhang. Conceptualization: Fengkai Sun. Supervision: Fengkai Sun. Writing—original draft: Fengkai Sun. |
Patient characteristics and utilization of an online patient portal in a rural academic general internal medicine practice | 0e7bade9-0509-4ec2-a366-e489b1012c50 | 8851702 | Internal Medicine[mh] | Online patient portals provide the opportunity for patients to better understand and engage with their own health care and connect with their clinic and health care providers [ – ]. Utilizing common portal functionalities such as reviewing progress notes, visit summaries, and communicating with providers outside of traditional office visits may encourage patients to follow recommended care plans in order to better manage chronic health conditions and improve their health outcomes [ , , ]. Despite the potential advantages, there remain many barriers to accessing online patient portals. Poor digital and health literacy, lack of internet access, unawareness of or forgetting about a patient portal’s existence, feeling inadequately trained to use the portal, and privacy concerns are all reported barriers to patient portal access [ , , , ]. Black, Latinx, older patients and patients with chronic health conditions are known to have lower enrollment and use of an online patient portal [ – ]. In Northern New England, many rural patients have issues with access to health care providers and can often live far from their primary care clinic. Patients also may have difficulties with transportation to and from their health care clinic due to the distance and lack of public transportation. With increasing broadband and internet access over the past few years, many have turned to the patient portal to reduce their need for in-person clinic visits. Previously published studies about patient portal use have primarily examined urban (or mixed urban and rural) populations that may differ in terms of physical distance to their health care providers, access to health care providers, availability of internet access or had overall low rates of portal enrollment [ , – ]. The objective of this study was to identify characteristics of patient portal users compared to non-users and assess utilization through number of logons and portal messaging in an academic general internal medicine practice in rural New Hampshire with high rates of patient portal enrollment. Understanding utilization among a defined population of patients is important for planning for staffing needs to respond to and address the increasing volume of portal messages that are sent by patients through the patient portal.
This cross-sectional study was conducted at Dartmouth-Hitchcock Medical Center, a rural academic medical center located in Lebanon, New Hampshire that serves a population of approximately 1.9 million patients from New Hampshire and Vermont . All adult patients who were assigned to a primary care provider in the general internal medicine practice between June 2019 and May 2020 were included in this study. Primary care providers in the general internal medicine clinic included physicians, physician assistants and advanced practice nurse practitioners. Patient data and portal use were obtained through a query from our data warehouse. The hospital system uses the EpicCare™ electronic health record for ambulatory care and the MyChart™ patient portal. Information recorded included demographic information (age, sex), clinical information (body mass index and chronic illnesses), and patient portal use over the study period. Active portal users had an account on the patient portal and logged onto the system during the study period. Message counts represented messages sent between the providers/clinic and patients and included messages sent from the clinic (which could include requests to complete surveys before upcoming appointments and reminders of upcoming appointments) and from patients to their providers during the study period. Defined chronic illnesses were based on internal billing codes using International Classification of Disease, Tenth Edition (ICD-10). Identified chronic illnesses were selected from the Charlson Comorbidity Index and excluded conditions for which only a small number of patients had the condition such as paraplegia, rheumatic disease, HIV, and liver disease . They were dichotomized (present/absent) based on the established groupings of codes by condition. A patient was noted to have a specific chronic illness if they had the specified condition listed in their problem list or a visit diagnosis from one of the disease groupings was included in a clinic visit during the study time period. Defined chronic illnesses included congestive heart failure, chronic obstructive pulmonary disease (COPD), cerebrovascular disease, diabetes mellitus with and without complications, history of a myocardial infarction (MI), peripheral vascular disease, and renal disease. Statistical analysis After data extraction, we conducted internal data validity checks and imported the data for further analyses. Descriptive statistics were evaluated on the entire cohort. We then stratified our cohort by active users of the portal and non-active users to evaluate the differences in baseline characteristics that could potentially influence use. We conducted unpaired t-tests of unequal variance or chi-squares (or their non-parametric equivalences) comparing portal and non-portal users. The null hypothesis was that portal users were younger, more often of white race, more often never smokers, and had lower prevalence rates of chronic medical illness. Further, we hypothesized that healthier weight (e.g., normal BMI) rather than underweight, overweight or obesity was associated with higher portal use. All descriptive statistics were two-sided, and P -values < 0.05 were considered statistically significant. All analyses were performed using STATA v.15 (College Station, TX). The study was reviewed and approved by the Dartmouth-Hitchcock Institutional Review Board (Study #00,006,738) and was considered a minimal risk study and did not require review by the institutional Ethics Committee.
After data extraction, we conducted internal data validity checks and imported the data for further analyses. Descriptive statistics were evaluated on the entire cohort. We then stratified our cohort by active users of the portal and non-active users to evaluate the differences in baseline characteristics that could potentially influence use. We conducted unpaired t-tests of unequal variance or chi-squares (or their non-parametric equivalences) comparing portal and non-portal users. The null hypothesis was that portal users were younger, more often of white race, more often never smokers, and had lower prevalence rates of chronic medical illness. Further, we hypothesized that healthier weight (e.g., normal BMI) rather than underweight, overweight or obesity was associated with higher portal use. All descriptive statistics were two-sided, and P -values < 0.05 were considered statistically significant. All analyses were performed using STATA v.15 (College Station, TX). The study was reviewed and approved by the Dartmouth-Hitchcock Institutional Review Board (Study #00,006,738) and was considered a minimal risk study and did not require review by the institutional Ethics Committee.
General characteristics of portal users A total of 28,028 patients in the general internal medicine clinic were evaluated during the study period. Among these, 22,955 (82%) of patients were active portal users (Table ). Patients who were female, aged 41–65, and non-smokers were more likely to be active users of the portal while current smokers were significantly less likely to use the portal. Of all chronic illnesses evaluated, eight out of nine demonstrated lower portal use compared to the overall cohort: Patients with heart failure, cerebrovascular disease, diabetes mellitus, history of MI, peripheral vascular disease and renal disease were significantly less likely to use the portal. We also examined BMI categories and patients who were classified as being underweight (BMI < 18.5) or having obesity (BMI ≥ 30.0) were the least likely to be active users of the portal compared to normal (BMI 18.5–24.9) or overweight patients (BMI 25.0–29.9). Frequency of portal use For the overall cohort, on average, a patient logged onto the patient portal 25 times and sent and received approximately 6 messages over the one-year study period (Table ). We found that females, patients older than 65, former smokers and obese patients (BMI ≥ 30.0) logged on and sent and received more messages compared to the overall cohort. While the examined minority population was small (220 Black patients and 556 Asian/Pacific Islander patients) in our study, on average Black patients logged on 19 times with 3.6 messages sent and received over the study period and Asian patients logged on 17.9 times with 2.7 messages compared to White patients who logged on 25 times with 5.8 messages per year (Table ).
A total of 28,028 patients in the general internal medicine clinic were evaluated during the study period. Among these, 22,955 (82%) of patients were active portal users (Table ). Patients who were female, aged 41–65, and non-smokers were more likely to be active users of the portal while current smokers were significantly less likely to use the portal. Of all chronic illnesses evaluated, eight out of nine demonstrated lower portal use compared to the overall cohort: Patients with heart failure, cerebrovascular disease, diabetes mellitus, history of MI, peripheral vascular disease and renal disease were significantly less likely to use the portal. We also examined BMI categories and patients who were classified as being underweight (BMI < 18.5) or having obesity (BMI ≥ 30.0) were the least likely to be active users of the portal compared to normal (BMI 18.5–24.9) or overweight patients (BMI 25.0–29.9).
For the overall cohort, on average, a patient logged onto the patient portal 25 times and sent and received approximately 6 messages over the one-year study period (Table ). We found that females, patients older than 65, former smokers and obese patients (BMI ≥ 30.0) logged on and sent and received more messages compared to the overall cohort. While the examined minority population was small (220 Black patients and 556 Asian/Pacific Islander patients) in our study, on average Black patients logged on 19 times with 3.6 messages sent and received over the study period and Asian patients logged on 17.9 times with 2.7 messages compared to White patients who logged on 25 times with 5.8 messages per year (Table ).
In our rural academic general internal medicine clinic, we found an overall high uptake in patient portal use with 82% active users who regularly logged on and interacted with the clinic providers using the messaging function through the portal. The high rates of utilization demonstrated that rural patients are seeking out the patient portal to communicate to their health care providers and are able to access online resources with expansion of smart phone use and cellular and high speed internet availability throughout the region . Rural patients often have to travel significant distances to see their providers and an online platform enabling communication with the clinic and their health care providers can be a valuable tool to improve health outcomes outside of traditional office visits. Importantly, we identified disparities in patient portal use with older patients, active smokers, underweight and obese patients, and the great majority of patients with examined chronic conditions demonstrating lower rates of portal utilization. A systematic review also found that multiple studies have shown lower use by racial and ethnic minorities but unlike the findings of our study, they also identified studies that showed greater use with increased numbers of medical problems . We believe that older patients and those with multiple chronic diseases (including obesity) could be the patients who could benefit the most from use of a patient portal. These patients often have multiple specialists, medications and treatment plans and the ability to access their progress notes, review clinic visit summaries and messaging their providers with questions or health updates has the potential to improve their understanding of their treatment plans, communication with their health care team, patient engagement and their health outcomes. A key element of the Chronic Care Model is self-management support and a well-designed clinical practice could use patient portal messaging to educate patients and identify barriers in managing their chronic illnesses . Unfortunately, we were not able to survey or interview patients with chronic diseases for this research project to identify the underlying reasons for the lower utilization. However, we speculate that the lower utilization rates are likely due to many different and interrelated contributors including education and income levels, access to technology, engagement with and trust in their health care team and knowledge about use of the patient portal. Clarifying these important contributors for lack of use is an important area for future research. This study was also unique in that it examined a population of patients with a significantly higher percentage in patient portal use compared to previously published studies that may reflect increasing availability of high speed internet, cellular service and smart phone use since the publication of the previous studies [ , , ]. The high rates of enrollment and use may also be a result of the clinic and providers’ encouragement and recommendations for portal use. Our institution has had an online patient portal for 20 years which was transitioned to the EpicCare™ electronic health record in 2011. The institution has actively encouraged patients to enroll in the patient portal and keeps them engaged with pre-visit questionnaires, messaging and prescription management. Providers within the clinic also encourage patients to enroll in the portal to review their progress notes and lab and radiology results and instructions on how to activate their account are included in every printed after visit summary for patients who have not activated their account. This study was also able to quantify the volume of messages sent and received by patients which is important to quantify for appropriate clinic staffing. Over the one year study period, there were 137,730 patient portal messages sent and received by the clinic from the 22,955 active portal users. Unfortunately, our data was not able to be broken down into the percentage of messages that were initiated by patients and those that were sent by the clinic but as most of the messages sent by the clinic are in response to patient initiated messaging, both are important to characterize as they both require time and resources to address the patient portal messages. As patient portal use increases in use and begins to replace traditional face-to-face clinic visits and patient telephone calls, primary care clinics need to adapt their clinic staffing models to adequately address the needs of patient portal messaging. Limitations of this study include that this was a single site study and a lack of data on social determinants of health such as patient education, income, digital literacy, or access to the internet or cellular service at their residence. All of these factors may significantly impact online patient portal use but unfortunately these data are not routinely available through data extraction from the electronic health record. While we were able to identify differences in patient portal use, we were not able to interview patients to identify causes or barriers that would help explain the observed differences in this study. Additionally, we were not able to validate data about race and confirm the reliability of documentation of chronic conditions with patients as the data that we obtained for this study was de-identified and the accuracy of problem lists and visit diagnoses could not be verified. Further research with focus groups and individual patient interviews would be very helpful to identify potential social determinants of health that are associated with decreased patient portal usage and to clarify barriers to accessing the patient portal in our rural region. More research should also be performed to evaluate the clinical outcomes of patients who utilize the patient portal and investigate whether use of the patient portal has an impact on rates of hospitalizations, emergency room visits, patient satisfaction, and other disease specific outcomes for diabetes or hypertension management. We also had a very small number of minority patients in our study population and additional studies in rural centers with a more diverse patient population are needed to identify racial and economic factors that prevent utilization of the patient portal.
In a rural academic general internal medicine clinic, we found high patient portal usage overall but lower rates among males, active smokers, older patients and those with examined chronic illnesses. Recognizing and addressing barriers to patient portal use is essential for robust and sustained patient portal uptake and ensuring that the benefits of portal use are equally distributed across the population of all patients.
|
Hydrogen Peroxide in the Pulp Chamber and Color Change in Maxillary Anterior Teeth After In-Office Bleaching | d1847dc8-5d42-40a7-b7db-fdf8952c0b7a | 11520489 | Dentistry[mh] | A smile featuring well-aligned maxillary anterior teeth, with the correct color, position, and shape, is considered important for patients . Professionals should take these characteristics into consideration by professionals, especially given the influence of patient demand influenced by social media . Dental bleaching is highly recommended when color changes are desired . Among several techniques available, in-office dental bleaching is often the most appropriate choice, as it provides faster results . In this technique, the patient undergoes the procedure inside the dental office. After protecting soft and gingival tissues, the professional applies a high concentration (30-40%) of hydrogen peroxide (HP) over the surface of the teeth for 30-50 minutes in each session . Generally, due to the lower molecular weight of HP, it can diffuse easily through the enamel and dentin . This chemical agent acts on organic structures, primarily in dentin, promoting breakdown and, consequently, the whitening effect . However, HP is not confined solely to hard tissues; it diffuses through enamel and dentin until it reaches the pulp chamber in smaller but sufficient amounts to produce tooth sensitivity , . This very common and temporary side effect after bleaching can be explained by inflammatory processes, partial local necrosis of pulp cells, and moderate dentin formation (5-8) due to oxidative stress. Various factors can influence the penetration of HP, including different concentrations, pH, viscosity, and composition of bleaching agents , , . Another factor, less explored in the literature, is variation in tooth size. It is expected that teeth of smaller sizes, indicating lower buccal thickness, will exhibit higher penetration of HP into the pulp chamber compared to larger size teeth with higher buccal thickness. Consequently, this may lead to a greater inflammatory process within the pulp chamber in the former . However, there is no consensus regarding tooth sensitivity when teeth of different sizes are evaluated. While some clinical studies have reported more intense intensity of tooth sensitivity in smaller size teeth , , others have not shown any significant difference, regardless of the size of the teeth . However, some studies evaluating bleaching effects in vitro utilize sections of bovine teeth with simulated pulp chambers in varying thicknesses. While these studies have provided valuable insights, the use of human teeth with their natural pulp chambers can offer a closer approximation to clinical scenarios , , , , , , , , , . On the other hand, premolars are commonly chosen for experiments involving human tooth crowns, primarily due to their extraction during orthodontic procedures , , , , , , . However, a closer examination of the literature reveals that tooth sensitivity is predominantly reported in anterior teeth, rather than premolars , . Therefore, it seems appropriate to investigate whether the amount of HP within the pulp chamber varies when different superior anterior teeth (canines, lateral incisors, and central incisors) undergo in-office bleaching. This study represents the first attempt to correlate tooth thickness with HP permeability into the dental pulp. Therefore, the aim of the present study was to evaluate the amount of HP inside the pulp chamber and color change in groups of extracted human maxillary anterior teeth, specifically canines, lateral incisors, and central incisors, after undergoing in-office bleaching with 35% HP. We tested the following primary research hypothesis: [1] there will be a difference in the amount of HP inside the pulp chamber when different teeth are subjected to in-office bleaching. Additionally, we tested the following secondary hypothesis: [2] there will be a difference in color change after in-office bleaching among different teeth evaluated.
Ethics committee approval and selection of teeth and inclusion and exclusion criteria This study was submitted to the local ethics committee, which approved it under agreement number (5.740.189). This study used thirty human maxillary anterior (10 canines, 10 lateral incisors, and 10 central incisors) obtained from Human Teeth Local Bank. The teeth were observed under a microscope at 10 x magnification (Eclipse E200, Nikon, Tokio, Japan). The selected teeth were required to have a baseline Whiteness Index for Dentistry (WI D ) 25 units or smaller. The WI D was measured by taking the color parameters obtained with the digital spectrophotometer Vita EasyShade (Vita Zahnfabrik, Bad Säckingen, Germany). Exclusion criteria included teeth with endodontic treatment, incomplete root formation, presence of previous restorations, caries, and severe tooth discoloration (such as tetracycline stains or fluorosis). Sample size calculation An earlier study showed that the amount of HP detected in the pulp chamber when premolar specimens were subjected to a one-session of in-office bleaching protocol with 35% HP gel was on average 1.16 ± 0.34 µg/mL. Using a two-tailed test with 5% of alpha and 90% power, to detect a 50% difference between groups, sample sizes of at least eight teeth should be tested in each group. To prevent possible loss of teeth during bleaching procedures, two extra teeth were added per group. Specimen preparation The specimens were prepared according to the previous study in the literature, and it was briefly described in this section , , , , . The roots of the teeth were cut approximately three millimeters from the enamel-cement junction. After removing the pulp tissue, the entrance of the pulp chamber (radicular area) was carefully expanded using a spherical drill, without touching the inner vestibular region of the pulp chamber. Following the procedure, the region was thoroughly rinsed with distilled water. It is important to note that this process does not modify the internal region of the pulp chamber and pulp horns. This was done with the purpose to introduce inside the pulp chamber around 25 μL of solution using a micropipette. Thickness of specimens Transverse images from cone beam computed tomography (CBCT) were acquired perpendicular to the longitudinal axes of the teeth, while sagittal and coronal images were obtained parallel to the longitudinal axes of the teeth after software processing. These images were acquired using a CBCT scanner (Phillips Brilliance 64, Philips Medical Systems, Eindhoven, Nederland), with a voxel size of 125 μm, operated at 120 kV and 350mAs by an experienced radiologist, following the manufacturer's recommended protocol to ensure high-quality images, using an exposure time of 2.5 seconds. The acquired data were then converted to the DICOM (Digital Imaging and Communications in Medicine) format for the subsequent measurement of buccal tooth thickness using the Philips Brilliance™ CT software (Philips Medical Systems, Eindhoven, Nederland), measuring the distance from the pulpal horn to the outermost buccal surface . Since previous histopathological studies have shown the presence of tissue damage in the horn area in human teeth treated with in-office bleaching , , , the mentioned distance was selected to represent the path for penetration of HP into the pulp. Obtaining the analytical curve First, a standard analytical curve was obtained from a 5.000 μg/mL stock solution prepared from a concentrated solution (37% HP, Thermo Fisher Scientific, Madrid, Madrid, Spain). This solution was diluted in an acetate buffer solution (pH = 4) and titrated using traditional methods with a potassium permanganate solution to determine the analytical grade and the actual concentration of the solution . From this concentration, serial volumetric dilutions of 0.000-0.397 μg/mL were performed to draw the analytical curve. A UV-Vis spectrophotometer (UV-1280, Shimadzu, Japan) was used to know the concentrations of HP and finally obtain a standard reference line for the extrapolation of the study samples’ results (R = 0.994; not shown data). Treatment bleaching protocols A single calibrated and experienced operator was responsible for the application of all treatment bleaching protocols. Teeth were fixed vertically to the silicone base (Speedex, Coltène/Whaledent AG, Feldwiesenstrasse, Altstätten, Switzerland). The contour of the buccal surface was delimited with a light-cured resin dam, enclosing an area of 6 mm x 6 mm (Top dam; FGM Dental Products, Joinville, SC, Brazil) ( , , , , . The 35% HP bleaching gel was used as an in-office product (Whiteness HP, FGM Dental Products, Joinville, SC, Brazil) in a single session, and it was applied to the enamel buccal surface three times at 15-minute intervals, according to the manufacturer’s recommendation. The application of the whitening gel was sufficient to cover the area to be performed the bleaching procedure, within the delimitation created with the gingival barrier. Consequently, comparable amounts of gel were used in all specimens, irrespective of tooth type. Hydrogen peroxide permeability in the pulp chamber As in previous studies (9,11,26-28), after the bleaching procedure, the acetate buffer solution in the pulp chamber of each sample was removed and transferred to a glass tube immediately after the session. To ensure the complete removal of HP, the pulp chamber was rinsed four times with 25 μL of acetate buffer. This rinse solution was transferred to the same glass tuve. Following this, 100 μL of 0.5 mg/mL (Leucocrystal Violet, Sigma Chemical Co., St Louis, MO, USA), 50 μL of 1 mg/mL of horseradish peroxidase (Peroxidase Type VI-A, Sigma Chemical Co., St. Louis, MO, USA) and deionized water (2.725 μL) were added to the glass tube. This sequence was repeated separately for each tooth at different times. The resulting solution was measured using a UV-Vis spectrophotometer (UV-1280, Shimadzu). According to the Beer-Lambert Law, absorbance is directly related to the concentration of the solute and the optical path length of the light beam through the solution . Therefore, the concentration of HP (μg/mL) was determined by comparing the absorbance with the previously obtained calibration curve. Color change evaluation The color of all groups was measured before starting any procedure and one week after , using a digital spectrophotometer (VITA Easyshade Advance 5.0, VITA Zahnfabrik, Bad Säckingen, Baden-Württemberg, Germany). As described previously , to standardize the position of the spectrophotometer for the different measurements, guides were made using dense condensation silicone (Speedex, light green color, Coltène/Whaledent AG), with a 6 mm diameter window created in the middle one-third of the buccal surface for each specimen using a metal device. Color measurements were performed in triplicate, and the average of each measurement was used for statistical purposes. The color parameters (L*, a*, and b*) were recorded through the tip of the device inserted into the silicone guide. The color change before (baseline) and after bleaching was determined by the difference between the colors measured with the spectrophotometer, using the WI D , ∆E ab , and ∆E 00 . Perceptual changes were considered significant if the differences between the initial and post-bleaching measurements were WI D > 2.6 , ∆E ab > 2.7 and ∆E 00 > 1.8 . Throughout all the experiment, the specimens were immersed in saliva, as previously described ( , . Statistical analysis Firstly, the data were tested for normality using the Shapiro-Wilk test and for equality of variances using Bartlett’s test (data not shown). Subsequently, the data on buccal thickness (mm), HP concentration detected in the pulp chamber (µg/mL), and color change (baseline WID, WID, ∆Eab, and ∆E00) were statistically analyzed using one-way ANOVA followed by the Tukey test (α = 0.05). Direct comparisons between buccal thickness and HP concentration in the pulp chamber were performed using Pearson’s correlation (α = 0.05).
This study was submitted to the local ethics committee, which approved it under agreement number (5.740.189). This study used thirty human maxillary anterior (10 canines, 10 lateral incisors, and 10 central incisors) obtained from Human Teeth Local Bank. The teeth were observed under a microscope at 10 x magnification (Eclipse E200, Nikon, Tokio, Japan). The selected teeth were required to have a baseline Whiteness Index for Dentistry (WI D ) 25 units or smaller. The WI D was measured by taking the color parameters obtained with the digital spectrophotometer Vita EasyShade (Vita Zahnfabrik, Bad Säckingen, Germany). Exclusion criteria included teeth with endodontic treatment, incomplete root formation, presence of previous restorations, caries, and severe tooth discoloration (such as tetracycline stains or fluorosis).
An earlier study showed that the amount of HP detected in the pulp chamber when premolar specimens were subjected to a one-session of in-office bleaching protocol with 35% HP gel was on average 1.16 ± 0.34 µg/mL. Using a two-tailed test with 5% of alpha and 90% power, to detect a 50% difference between groups, sample sizes of at least eight teeth should be tested in each group. To prevent possible loss of teeth during bleaching procedures, two extra teeth were added per group.
The specimens were prepared according to the previous study in the literature, and it was briefly described in this section , , , , . The roots of the teeth were cut approximately three millimeters from the enamel-cement junction. After removing the pulp tissue, the entrance of the pulp chamber (radicular area) was carefully expanded using a spherical drill, without touching the inner vestibular region of the pulp chamber. Following the procedure, the region was thoroughly rinsed with distilled water. It is important to note that this process does not modify the internal region of the pulp chamber and pulp horns. This was done with the purpose to introduce inside the pulp chamber around 25 μL of solution using a micropipette.
Transverse images from cone beam computed tomography (CBCT) were acquired perpendicular to the longitudinal axes of the teeth, while sagittal and coronal images were obtained parallel to the longitudinal axes of the teeth after software processing. These images were acquired using a CBCT scanner (Phillips Brilliance 64, Philips Medical Systems, Eindhoven, Nederland), with a voxel size of 125 μm, operated at 120 kV and 350mAs by an experienced radiologist, following the manufacturer's recommended protocol to ensure high-quality images, using an exposure time of 2.5 seconds. The acquired data were then converted to the DICOM (Digital Imaging and Communications in Medicine) format for the subsequent measurement of buccal tooth thickness using the Philips Brilliance™ CT software (Philips Medical Systems, Eindhoven, Nederland), measuring the distance from the pulpal horn to the outermost buccal surface . Since previous histopathological studies have shown the presence of tissue damage in the horn area in human teeth treated with in-office bleaching , , , the mentioned distance was selected to represent the path for penetration of HP into the pulp.
First, a standard analytical curve was obtained from a 5.000 μg/mL stock solution prepared from a concentrated solution (37% HP, Thermo Fisher Scientific, Madrid, Madrid, Spain). This solution was diluted in an acetate buffer solution (pH = 4) and titrated using traditional methods with a potassium permanganate solution to determine the analytical grade and the actual concentration of the solution . From this concentration, serial volumetric dilutions of 0.000-0.397 μg/mL were performed to draw the analytical curve. A UV-Vis spectrophotometer (UV-1280, Shimadzu, Japan) was used to know the concentrations of HP and finally obtain a standard reference line for the extrapolation of the study samples’ results (R = 0.994; not shown data).
A single calibrated and experienced operator was responsible for the application of all treatment bleaching protocols. Teeth were fixed vertically to the silicone base (Speedex, Coltène/Whaledent AG, Feldwiesenstrasse, Altstätten, Switzerland). The contour of the buccal surface was delimited with a light-cured resin dam, enclosing an area of 6 mm x 6 mm (Top dam; FGM Dental Products, Joinville, SC, Brazil) ( , , , , . The 35% HP bleaching gel was used as an in-office product (Whiteness HP, FGM Dental Products, Joinville, SC, Brazil) in a single session, and it was applied to the enamel buccal surface three times at 15-minute intervals, according to the manufacturer’s recommendation. The application of the whitening gel was sufficient to cover the area to be performed the bleaching procedure, within the delimitation created with the gingival barrier. Consequently, comparable amounts of gel were used in all specimens, irrespective of tooth type.
As in previous studies (9,11,26-28), after the bleaching procedure, the acetate buffer solution in the pulp chamber of each sample was removed and transferred to a glass tube immediately after the session. To ensure the complete removal of HP, the pulp chamber was rinsed four times with 25 μL of acetate buffer. This rinse solution was transferred to the same glass tuve. Following this, 100 μL of 0.5 mg/mL (Leucocrystal Violet, Sigma Chemical Co., St Louis, MO, USA), 50 μL of 1 mg/mL of horseradish peroxidase (Peroxidase Type VI-A, Sigma Chemical Co., St. Louis, MO, USA) and deionized water (2.725 μL) were added to the glass tube. This sequence was repeated separately for each tooth at different times. The resulting solution was measured using a UV-Vis spectrophotometer (UV-1280, Shimadzu). According to the Beer-Lambert Law, absorbance is directly related to the concentration of the solute and the optical path length of the light beam through the solution . Therefore, the concentration of HP (μg/mL) was determined by comparing the absorbance with the previously obtained calibration curve.
The color of all groups was measured before starting any procedure and one week after , using a digital spectrophotometer (VITA Easyshade Advance 5.0, VITA Zahnfabrik, Bad Säckingen, Baden-Württemberg, Germany). As described previously , to standardize the position of the spectrophotometer for the different measurements, guides were made using dense condensation silicone (Speedex, light green color, Coltène/Whaledent AG), with a 6 mm diameter window created in the middle one-third of the buccal surface for each specimen using a metal device. Color measurements were performed in triplicate, and the average of each measurement was used for statistical purposes. The color parameters (L*, a*, and b*) were recorded through the tip of the device inserted into the silicone guide. The color change before (baseline) and after bleaching was determined by the difference between the colors measured with the spectrophotometer, using the WI D , ∆E ab , and ∆E 00 . Perceptual changes were considered significant if the differences between the initial and post-bleaching measurements were WI D > 2.6 , ∆E ab > 2.7 and ∆E 00 > 1.8 . Throughout all the experiment, the specimens were immersed in saliva, as previously described ( , .
Firstly, the data were tested for normality using the Shapiro-Wilk test and for equality of variances using Bartlett’s test (data not shown). Subsequently, the data on buccal thickness (mm), HP concentration detected in the pulp chamber (µg/mL), and color change (baseline WID, WID, ∆Eab, and ∆E00) were statistically analyzed using one-way ANOVA followed by the Tukey test (α = 0.05). Direct comparisons between buccal thickness and HP concentration in the pulp chamber were performed using Pearson’s correlation (α = 0.05).
The buccal thickness was significantly different among the maxillary anterior teeth evaluated, with significant differences observed only when comparing canines, the thicker teeth, with other teeth (p = 0.04; ). HP was detected in the pulp chamber of all types of teeth. However, the concentration was significantly lower for the canine group (p = 0.02; ). Pearson's correlation indicated a weak negative correlation (r = -0.49), suggesting that as one variable increased, the other decreased by a similar magnitude, though this was not statistically significant (p = 0.99). and show that the baseline WI D tooth color was similar across all groups ( p > 0.05). After bleaching ( and ), all parameters evaluated for color change showed significant bleaching effects ( p < 0.05). However, when comparing all groups, no significant differences were observed in the bleaching outcomes for any of the color parameters evaluated (p > 0.36; and ).
The study results support the primary hypothesis, as there was a difference in the HP levels inside the pulp chambers. However, the secondary hypothesis was rejected, as no significant difference in color was detected among the groups. Upper anterior teeth are often subject to aesthetic demands, and bleaching is essential for improving color and appearance. However, HP diffusion into the pulp chamber is an undesirable but inevitable event that occurs during and after bleaching procedures . Several studies have provided valuable information on how application modes , , , composition , , and the physico-chemical characteristics of the bleaching gel , can influence the amount of HP detected in the pulp chamber. Individual characteristics of the teeth are also important to consider for bleaching procedures. Buccal thickness affects the amount of HP that enters the pulp chamber. Central and lateral incisors have thinner enamel-dentine buccal surfaces than canines, and the former showed more HP inside the pulp chamber than canines after in-office bleaching. It can be argued that in thicker teeth, such as canines, a greater area of organic substrate is available for oxidation. The free radicals react more with the organic structure present in thicker substrates, resulting in a lesser amount of HP inside the pulp chamber. Conversely, teeth with thinner enamel are more likely to experience inflammatory processes in the pulp tissue due to the higher amount of HP that can reach it. This can cause oxidative stress in the tissue, leading to alterations in its morphology and a decrease in cell viability and regeneration. Consequently, this can result in varying degrees of tooth sensitivity, depending on the patient , . The relationship between the buccal thickness of the dental substrate and the amount of HP detected in the pulp chamber after in-office bleaching was also observed in previous studies . The authors simulated various enamel/dentin thicknesses by polishing bovine teeth to create 4-mm and 2.3-mm enamel/dentin discs representing maxillary first premolars and mandibular central incisors, respectively . In contrast, we utilized a more realistic scenario by using anterior human extracted teeth without any preparation. The results demonstrated that even small differences in thickness (0.8 mm and 0.6 mm between canines and lateral incisors and central incisors, respectively) could lead to significant differences in the amount of HP detected in the pulp chamber. Despite similar results being observed compared to the previous study , the amount of HP reaching the pulp chamber is markedly different in the present study. While the aforementioned study reported around 7 µg/mL for thicknesses similar to lower incisors and 5 µg/mL for thicknesses similar to premolars, our study found significantly lower concentrations. Several methodological differences, including the type of teeth used and the format of specimens, are likely responsible for the observed differences. When trimming the teeth to achieve the desired thicknesses, the authors compensated for the smaller thickness of bovine enamel in the premolar group by leaving a thicker dentin substrate compared to human premolars . The thickness of dentin might allow more passage of HP than a more mineralized, less permeable substrate such as enamel. Additionally, the previous study involved the fabrication of an artificial pulp chamber where 1 mm of the medium solution was in contact with the enamel-dentin disk . In our study, a smaller amount of buffer solution could be used due to the natural pulp chamber's constraints, necessitating subsequent rinsing to complete the total solution volume for analysis. This is the first study to attempt to correlate tooth thickness with HP permeability into the dental pulp. Although no correlation was found between the thickness of various teeth and their permeability, it appears that thickness may not be a key factor in explaining the amount of HP reaching the dental pulp. Despite controversial results regarding tooth sensitivity and size , , , only one clinical study has correlated the measured tooth sensitivity of maxillary central incisors with their thickness, finding no significant correlation . This lack of correlation may be attributed to various factors, including specific characteristics such as the patient's age. Previous studies have shown that younger teeth typically exhibit thinner dentin and wider dentinal tubules compared to older teeth. Additionally, younger teeth generally have less secondary dentin than older teeth . One characteristic of the reduction in pulp chamber area is the continuous deposition of a dentin matrix rich in physiological collagen (secondary dentin) by odontoblasts , which occurs with age or as a response to occlusal trauma. In our study, the statistical differences observed between central and lateral incisors compared to canines suggest that the path to the pulp chamber was longer in the latter, likely due to ongoing dentin deposition. Furthermore, apart from anatomical differences, canines undergo a longer maturation process and are more subjected to occlusal trauma, as previously mentioned. Consequently, canines tend to appear darker due to the increased amount of dentin present. This increase in thickness, along with decreased permeability from peritubular dentin deposition in older teeth or due to trauma, is expected to offer additional protection to the pulp . Regarding color change, the present study used the WI D ( Whiteness Index for Dentistry ) to assess baseline tooth color and color change efficacy, following more recent recommendations , . The use of baseline WI D and ΔWI D are newer tools , recommended for measuring tooth bleaching effectiveness based on the CIELab color space. This updated formula reduces the probability of error in assessing whiteness , and enables significant observation of color improvement for all teeth after just one session. Traditionally, previous studies have evaluated color change using only ΔE ab , , . The ΔE ab metric permits comparison with data from earlier studies , , ; however, the ΔE 00 can indicate color differences perceived by the human eye more accurately than ΔE ab . The ΔE 00 considers not only chroma, hue, and lightness, weighting functions; it proposes potential interactions between hue differences and chroma to enhance performance for blue and gray colors . While both ΔE ab and ΔE 00 metrics are used in dentistry, they assume equal influence for all color coordinates. Consequently, these values do not indicate whether the color change of dental structures moved towards lighter or darker shades. This is why the ΔWI D , which measures the level of whiteness, is directly relevant in bleaching studies. Canines, being thicker teeth, are expected to be darker and therefore more challenging to whiten due to their inherent characteristics. However, our findings revealed no statistical difference in tooth color at baseline or after bleaching. A recent study highlighted that the color of the silicone guide itself can influence measurement outcomes. The authors suggested (although not evaluated in their design) using transparent guides for future research in the bleaching field to minimize interference in the results. It is important to note that, due to our unawareness of this potential interference, we used light green silicone. Any potential impact of the colored silicone would have been consistent across both groups, not affecting the study’s internal validity. Several limitations should be considered. While the study focused on relevant teeth for bleaching procedures, not all tooth types were included in the analysis. Additionally, due to the challenge of obtaining anterior sound teeth, primarily teeth from older patients were used, which may limit the generalizability of the findings. The age of the patients whose teeth were obtained can also be considered a limitation, as only older teeth were included in the study. Despite efforts made during the application of the bleaching gel, the exact quantity of gel was not measured beforehand, which could be considered a limitation of the present study. In summary, human teeth with smaller buccal thickness and thinner enamel exhibit greater penetration of hydrogen peroxide into the pulp chamber following in-office bleaching. These findings suggest that, in clinical practice, such teeth may be more prone to developing adverse effects such as tooth sensitivity.
The bleaching pattern observed with in-office bleaching appears to be consistent across all tooth sizes. However, it is noteworthy that only maxillary canines, which are thicker, showed a lower amount of hydrogen peroxide (HP) detected in the pulp chamber compared to maxillary central and lateral incisors.
|
Unlocking Cardiac Insights: Displacement of Aortic Root for Calculation of Ejection Fraction in Emergency Department in India | 4308df5a-2fd1-4733-bafb-2a26fd536d4b | 11931714 | Cardiovascular System[mh] | Background Assessing cardiac function, particularly ejection fraction (EF), is crucial for managing acute dyspnea. – Echocardiography is the current standard for calculating EF, but displacement of the aortic root (DAR) has emerged as a potential tool for EF calculation in patients with undifferentiated dyspnea. , The DAR method quantifies alterations in left ventricular (LV) volume throughout the cardiac cycle, providing a surrogate measure for estimating EF. End-point septal separation (EPSS) measurement is a relatively straightforward skill that an emergency physician can acquire with minimal experience, even when confronted with regional wall motion abnormalities. , However, measurement of LV end-systolic and end-diastolic diameters using 2D or M-mode echocardiography can pose challenges to the emergency physician in clinical practice. Tracing the endocardial border of the heart in an echocardiogram during diastole and systole is often difficult and time-consuming, especially where the wall is poorly defined. – This approach provides clinicians with multiple options for assessing LV systolic function, catering to varying levels of expertise and clinical settings. Mitral annular plane systolic excursion (MAPSE) assesses vertical mitral valve motion using M-mode echocardiography, measuring annular displacement towards the apex. Unlike other methods, MAPSE doesn’t require optimal endocardial definition or clear LV apex visualization, enabling broad applicability. Diminished systolic mitral valve excursion, reflected in MAPSE measurements, reliably indicates LV systolic dysfunction. The MAPSE demonstrates strong correlations, particularly in non-critically ill patients, offering effective LV function assessment even in challenging imaging scenarios. – Emergency physicians are accurate at visual LV EF estimation without quantitative measurements, but objective measures can benefit early learners and facilitate communication. However, EF calculation using the DAR method has not been done in an Indian population in the ED setting. This highlights the need for further studies to determine DAR’s reliability and clinical applicability in the context of an Indian setting. Importance Given the current limited research on the utility of DAR in Indian academic ED settings, with this investigation we aimed to fill the gap by assessing DAR’s reliability and clinical applicability. The study specifically focuses on patients with undifferentiated dyspnea, a population where EF estimation is crucial for appropriate management. Goal of this Investigation Our primary objective was to calculate the EF using DAR and then compare it with EF measurements obtained through the modified Simpson method, defined as the criterion reference by the American Society of Echocardiography (ASE). , The secondary objective was to identify the cut-off for DAR, which could predict LV dysfunction based on EF calculation. Additionally, we sought to compare the EF calculated from DAR with those obtained through EPSS and MAPSE. By evaluating DAR in comparison to the established methods, we aimed to provide insights into its potential as a reliable tool for EF estimation in the Indian setting.
Assessing cardiac function, particularly ejection fraction (EF), is crucial for managing acute dyspnea. – Echocardiography is the current standard for calculating EF, but displacement of the aortic root (DAR) has emerged as a potential tool for EF calculation in patients with undifferentiated dyspnea. , The DAR method quantifies alterations in left ventricular (LV) volume throughout the cardiac cycle, providing a surrogate measure for estimating EF. End-point septal separation (EPSS) measurement is a relatively straightforward skill that an emergency physician can acquire with minimal experience, even when confronted with regional wall motion abnormalities. , However, measurement of LV end-systolic and end-diastolic diameters using 2D or M-mode echocardiography can pose challenges to the emergency physician in clinical practice. Tracing the endocardial border of the heart in an echocardiogram during diastole and systole is often difficult and time-consuming, especially where the wall is poorly defined. – This approach provides clinicians with multiple options for assessing LV systolic function, catering to varying levels of expertise and clinical settings. Mitral annular plane systolic excursion (MAPSE) assesses vertical mitral valve motion using M-mode echocardiography, measuring annular displacement towards the apex. Unlike other methods, MAPSE doesn’t require optimal endocardial definition or clear LV apex visualization, enabling broad applicability. Diminished systolic mitral valve excursion, reflected in MAPSE measurements, reliably indicates LV systolic dysfunction. The MAPSE demonstrates strong correlations, particularly in non-critically ill patients, offering effective LV function assessment even in challenging imaging scenarios. – Emergency physicians are accurate at visual LV EF estimation without quantitative measurements, but objective measures can benefit early learners and facilitate communication. However, EF calculation using the DAR method has not been done in an Indian population in the ED setting. This highlights the need for further studies to determine DAR’s reliability and clinical applicability in the context of an Indian setting.
Given the current limited research on the utility of DAR in Indian academic ED settings, with this investigation we aimed to fill the gap by assessing DAR’s reliability and clinical applicability. The study specifically focuses on patients with undifferentiated dyspnea, a population where EF estimation is crucial for appropriate management.
Our primary objective was to calculate the EF using DAR and then compare it with EF measurements obtained through the modified Simpson method, defined as the criterion reference by the American Society of Echocardiography (ASE). , The secondary objective was to identify the cut-off for DAR, which could predict LV dysfunction based on EF calculation. Additionally, we sought to compare the EF calculated from DAR with those obtained through EPSS and MAPSE. By evaluating DAR in comparison to the established methods, we aimed to provide insights into its potential as a reliable tool for EF estimation in the Indian setting.
Study Design and Setting This prospective, cross-sectional study was conducted across a span of two years, from December 2019–December 2021, within the ED of a teaching hospital in India. The hospital provides a broad spectrum of specialties, and its adult ED has approximately 37,200 visits annually. We obtained initial institutional research board/institutional ethics committee approval, with the registration number ECR/146/Inst/ KA/ 2013/RR-19, IEC: 1057/2019, dated May 8, 2020, and approval for study modifications on September 22, 2021. Additionally, the study is registered with the Clinical Trials Registry–India under the number CTRI/2020/10/028704, dated October 28, 2020. We adhered to ethical standards by obtaining informed consent and ensuring the voluntary participation and compliance of all subjects involved in the study. We assessed the EF of 110 patients with undifferentiated dyspnea using different methods. Selection of Participants and Methods of Measurements We enrolled patients ≥18 years of age, presenting with undifferentiated dyspnea and normal sinus rhythm based on a convenience sampling. The following were excluded: patients intubated outside of a hospital; pregnant women; individuals with elevated cardiac biomarkers at presentation; those with atrial fibrillation, known valvular pathology or surgery, primary or metastatic carcinoma in the thorax; patients for whom the time between echocardiography to obtain EF using DAR and the modified Simpson method was more than 30 minutes; and those who did not provide consent. These factors could have influenced the accuracy and reliability of the EF measurements obtained through different methods. Demographic variables, including age and gender, were considered as potential confounding factors in this study. After obtaining written informed consent, the emergency clinician conducted the bedside ultrasonography proctored by the expert in point-of-care ultrasound (POCUS). Using a 3.6-megahertz micro-convex transducer, the investigator, trained in POCUS during residency training as per the curriculum, employed a Philips CX 50 ultrasound machine (Koninklijke Philips NV, Amsterdam, Netherlands) to compute the EF using DAR. Initially, 2D echocardiograms of the parasternal long-axis view were captured for DAR measurement. This view was achieved by positioning the footprint of the transducer perpendicular to the chest wall at the third or fourth intercostal space, just to the left of the sternum with the pointer towards the right shoulder ( ). Optimum image required clear view of mitral valve leaflets and aortic valves. Subsequently, M-mode was placed just above the level of the aortic valve and DAR recordings were taken. The maximum anterior DAR from the horizontal axis at end-systole was measured using the leading-edge technique and recorded in centimeters (cm) ( ). The computation of EF was then done, using the formula 20 + 44 * DAR (cm). Following the DAR measurement, the investigator calculated the EF using EPSS determined by EF = 75.5 – (2.5 × EPSS), and using MAPSE calculated by 4.8 × MAPSE (millimeters [mm]) + 5.8 for men and 4.2 × MAPSE (mm) + 20 for women. , , – An experienced echocardiographer, blinded to the study procedure, evaluated LV EF using the ASE recommended Modified Simpson’s rule for this measurement ( ). , Outcomes The study systematically categorized outcomes into two groups, delineating ‘normal’ EF as 50% to 70% and ‘low EF’ <50%. The primary outcome measured significant difference in calculated EF between the DAR and modified Simpsons methods. The secondary outcome of the study was to determine cut-off value of DAR with high sensitivity and specificity through receiver operating characteristic (ROC) curve analysis. Secondary outcomes also included comparison of EF calculated from DAR with that calculated from EPSS and MAPSE. Sample Size Calculation With a desired margin of error of 10%, alpha error of 5%, and estimated proportion of 0.5, sample size was calculated to be 96. After considering the dropout rate of 15%, the final sample size was 110. Analysis We used SPSS Statistics, version 26.0 (IBM Corp, Armonk, NY) to analyse the data. The Shapiro-Wilk test assessed normality for continuously distributed data, and we executed group comparisons in the subsequent steps. An exact McNemar test was used to identify the statistically significant changes in EF calculated using the DAR and modified Simpson’s methods. We calculated the Pearson correlation coefficient to measure strength and direction of the linear relationship between two tests. The ROC curve played a pivotal role in determining the optimal cut-off value for the validity measure of DAR.
This prospective, cross-sectional study was conducted across a span of two years, from December 2019–December 2021, within the ED of a teaching hospital in India. The hospital provides a broad spectrum of specialties, and its adult ED has approximately 37,200 visits annually. We obtained initial institutional research board/institutional ethics committee approval, with the registration number ECR/146/Inst/ KA/ 2013/RR-19, IEC: 1057/2019, dated May 8, 2020, and approval for study modifications on September 22, 2021. Additionally, the study is registered with the Clinical Trials Registry–India under the number CTRI/2020/10/028704, dated October 28, 2020. We adhered to ethical standards by obtaining informed consent and ensuring the voluntary participation and compliance of all subjects involved in the study. We assessed the EF of 110 patients with undifferentiated dyspnea using different methods.
We enrolled patients ≥18 years of age, presenting with undifferentiated dyspnea and normal sinus rhythm based on a convenience sampling. The following were excluded: patients intubated outside of a hospital; pregnant women; individuals with elevated cardiac biomarkers at presentation; those with atrial fibrillation, known valvular pathology or surgery, primary or metastatic carcinoma in the thorax; patients for whom the time between echocardiography to obtain EF using DAR and the modified Simpson method was more than 30 minutes; and those who did not provide consent. These factors could have influenced the accuracy and reliability of the EF measurements obtained through different methods. Demographic variables, including age and gender, were considered as potential confounding factors in this study. After obtaining written informed consent, the emergency clinician conducted the bedside ultrasonography proctored by the expert in point-of-care ultrasound (POCUS). Using a 3.6-megahertz micro-convex transducer, the investigator, trained in POCUS during residency training as per the curriculum, employed a Philips CX 50 ultrasound machine (Koninklijke Philips NV, Amsterdam, Netherlands) to compute the EF using DAR. Initially, 2D echocardiograms of the parasternal long-axis view were captured for DAR measurement. This view was achieved by positioning the footprint of the transducer perpendicular to the chest wall at the third or fourth intercostal space, just to the left of the sternum with the pointer towards the right shoulder ( ). Optimum image required clear view of mitral valve leaflets and aortic valves. Subsequently, M-mode was placed just above the level of the aortic valve and DAR recordings were taken. The maximum anterior DAR from the horizontal axis at end-systole was measured using the leading-edge technique and recorded in centimeters (cm) ( ). The computation of EF was then done, using the formula 20 + 44 * DAR (cm). Following the DAR measurement, the investigator calculated the EF using EPSS determined by EF = 75.5 – (2.5 × EPSS), and using MAPSE calculated by 4.8 × MAPSE (millimeters [mm]) + 5.8 for men and 4.2 × MAPSE (mm) + 20 for women. , , – An experienced echocardiographer, blinded to the study procedure, evaluated LV EF using the ASE recommended Modified Simpson’s rule for this measurement ( ). ,
The study systematically categorized outcomes into two groups, delineating ‘normal’ EF as 50% to 70% and ‘low EF’ <50%. The primary outcome measured significant difference in calculated EF between the DAR and modified Simpsons methods. The secondary outcome of the study was to determine cut-off value of DAR with high sensitivity and specificity through receiver operating characteristic (ROC) curve analysis. Secondary outcomes also included comparison of EF calculated from DAR with that calculated from EPSS and MAPSE.
With a desired margin of error of 10%, alpha error of 5%, and estimated proportion of 0.5, sample size was calculated to be 96. After considering the dropout rate of 15%, the final sample size was 110.
We used SPSS Statistics, version 26.0 (IBM Corp, Armonk, NY) to analyse the data. The Shapiro-Wilk test assessed normality for continuously distributed data, and we executed group comparisons in the subsequent steps. An exact McNemar test was used to identify the statistically significant changes in EF calculated using the DAR and modified Simpson’s methods. We calculated the Pearson correlation coefficient to measure strength and direction of the linear relationship between two tests. The ROC curve played a pivotal role in determining the optimal cut-off value for the validity measure of DAR.
A total of 135 patients underwent initial screening for participation in the study. Before the POCUS assessment, we excluded 25 patients based on predefined criteria: five due to external intubation; eight with elevated cardiac biomarkers; three with abnormal rhythm; four with valvular pathology; and five who declined to participate. Following that, a POCUS examination was conducted on 110 patients, with 10 excluded due to poor image quality ( ). The demographic and clinical characteristics of 100 patients who underwent POCUS, including age, heart rate, mean arterial pressure, and the mean DAR values in relation to age, gender, and comorbidities are detailed in . In this study we observed a mean DAR measurement of 0.781 cm (SD 0.277 cm) and an average calculated EF of 54.4% (SD 12.2%). The Pearson correlation coefficient was calculated to measure strength and direction of the linear relationship between two tests and was found to be 0.81, which suggests a strong positive relation between the results. The study conducted an exact McNemar test to identify statistically significant variations in abnormal and normal EF distribution between the EF calculated using DAR/MAPSE/ EPSS and the EF measured by an echocardiographer (criterion reference), as outlined in . The statistical analysis revealed a lack of significant differences ( P = 0.39) between the EF calculated using DAR and the EF measured by echocardiography. We conducted ROC curve analysis, which demonstrated DAR’s validity with a high accuracy reflected in an area under the curve (AUC) of 0.958 (95% confidence interval [CI] 0.914–1.000, P < 0.001) for predicting EF. The optimal cut-off point for DAR was identified as 0.706, providing a sensitivity of 88.7%, specificity of 93.1%, LR+ (likelihood ratio) of 12.86, and LR- of 0.12. ( ). The Pearson correlation coefficient calculated for EF calculated by MAPSE and the modified Simpson method was 0.54 and that of EPSS and the modified Simpson method was 0.76. For calculated EF with MAPSE, 48.3% of patients were categorized as having abnormal EF, exhibiting a statistically significant difference compared to EF calculated by the modified Simpson method ( P = 0.01) ( ). Similarly, calculated EF with EPSS demonstrated a comparable discordance, with 58.6% classified as abnormal, significantly differing from EF calculated by the modified Simpson method ( P = 0.01) ( ). presents a comparative assessment of the efficacy of EF measurements using MAPSE, EPSS, and DAR against the criterion reference. The sensitivity of DAR is notably higher than MAPSE and EPSS, which suggests that it is a better screening tool. Calculated EF from DAR obtained highest negative predictive value (NPV), suggesting a better ability to correctly identify patients with normal EF.
Dyspnea is a common presenting complaint in the ED, accounting for approximately 5% of all ED presentations in the Asia-Pacific region. , Emergency physicians frequently face the challenge of making swift diagnoses and developing treatment plans based on limited clinical information. , Point-of-care ultrasound has become a standard component of routine clinical examinations in the ED, enhancing the management of dyspnea by facilitating the diagnosis of its underlying causes. Similarly, evaluating LVEF through echocardiography plays a crucial role in diagnosing and managing a wide range of patients in the ED, further emphasizing the importance of ultrasound in emergency care. Most research in the ED has emphasized visual assessments of LVEF instead of relying on calculations derived from measuring the dimensions of the LV chamber across the cardiac cycle. – This study addresses a crucial aspect of emergency care by exploring the assessment of LV function in patients with undifferentiated dyspnea. While the modified Simpson method remains the criterion reference, investigating the potential of DAR as an alternative method opens avenues for expedited and more accessible evaluations in time-sensitive environments like the ED. As a non-invasive and easily accessible tool, DAR has shown promise in accurately predicting LVEF, making it valuable for identifying patients at risk of LV dysfunction. , The DAR method showed an accuracy rate of 88% in correctly classifying LV dysfunction, demonstrating its clinical applicability in emergency settings. This rate surpasses the accuracy of MAPSE and EPSS assessments for LV dysfunction, including the 75% accuracy reported in a study by Schick et al. This study’s robust methodology and compelling results substantially contribute to establishing the validity and clinical relevance of DAR. The DAR method exhibits good sensitivity (86.2%) and specificity (88.7%) and has a positive correlation with the values of EF obtained through the modified Simpson method. This sensitivity and specificity are consistent with the findings of Ünlüer et al, who reported 94.4% and 94.1%, respectively. The increased sensitivity of DAR compared to EPSS and MAPSE in our study makes it a valuable tool for the early detection of LV dysfunction in emergency settings.These findings indicate that emergency physicians can use DAR as a valuable alternate tool for assessing the LV function at the bedside. In the ED, where rapid decision-making is crucial, DAR can be incorporated as an initial screening tool to identify patients with compromised LV function, guiding further diagnostic testing, management, interventions or specialist referrals. When comparing DAR with traditional methods, MAPSE showed a sensitivity of 48.3% (95% CI 39.2–57.4) and specificity of 95.8% (95% CI 90.3–98.4), while EPSS exhibited a sensitivity of 45.5% (95% CI 36.2–54.8) and specificity of 97.0% (95% CI 94.4–99.4). These results contrast with prior studies, such as that by McKaigney et al, who observed significantly higher EPSS sensitivity (83.3%) but much lower specificity (50.0%), and Schick et al, who reported MAPSE sensitivity of 42% and specificity of 89%. , The higher sensitivity (83.3%) and lower specificity (50.0%) of EPSS reported by McKaigney et al may stem from their comparison of EPSS with EF calculated using the Teichholz method. Folland et al found that EF calculated through the modified Simpson method demonstrated better correlation with radionuclide ventriculography than the Teichholz method, with correlation coefficients (r values) of 0.75 and 0.46, respectively. Furthermore, the ASE no longer recommends the Teichholz method for calculating LV volumes. , The higher specificity of MAPSE and EPSS in our study suggests that these measurements are more effective in confirming LV dysfunction than in detecting it, underscoring the utility of DAR’s higher sensitivity for early identification. The DAR offers a practical advantage in the ED setting due to the straightforward visualization of the aortic root compared to LV structures, making it easier to measure under challenging conditions. Furthermore, the motion of the aortic root resembles the left atrial volume curve, suggesting that its movement, influenced by its attachment to the cardiac skeleton, may reflect the dynamics of left atrial filling and emptying. – The observed correlation between DAR and stroke volume suggests that DAR measurements may calculate LVEF effectively, providing valuable insights into cardiac performance. Lower DAR values were consistently associated with conditions linked to reduced stroke volume and EF, highlighting DAR’s relevance in assessing patients with undifferentiated dyspnea and potentially compromised cardiac function. The DAR’s high NPV enhances its reliability in excluding patients with normal EF, which is crucial for determining appropriate next steps in ED care. The EPSS exhibited the highest positive predictive value, emphasizing its role in confirming reduced EF. However, DAR’s combined sensitivity and NPV make it a more comprehensive tool for initial screening, ensuring that patients with likely normal cardiac function are appropriately triaged. Despite its advantages, DAR should not be seen as a replacement for all echocardiographic assessments but rather as a complementary tool, especially in time-limited environments. Its heightened sensitivity compared to MAPSE and EPSS, combined with its rapid application, makes it a promising option for emergency physicians. However, further research and validation are required to establish DAR’s broader applicability in diverse patient populations and settings.
While the results are promising, this study has limitations. It was conducted within a single-center environment, potentially limiting the generalizability of the findings. A multicenter study involving diverse patient populations would provide more robust validation. Additionally, the study doesn’t delve into the causes of dyspnea, which can vary widely and might influence the applicability of DAR in different scenarios. We excluded 9% of patients from this study due to a poor POCUS window. Patients enrolled in this study exhibited exclusively regular cardiac rhythms. Although each M-mode recording of the aortic root (AR) had the potential to encompass multiple cardiac cycles for DAR calculation, it is crucial to emphasize that the extent of AR displacement consistently remains notable across all cardiac cycles in individuals with regular heart rhythms. When patients exhibit irregular heart rhythms, a potential adaptation could involve calculating the average DAR measurement over three to five cardiac cycles. This adjustment could enhance the accuracy of measurements in such cohorts. Future research initiatives could delve deeper into investigating and addressing this particular aspect.
DAR emerges as an efficient and reliable method for rapid EF assessment, providing emergency physicians with a valuable tool for bedside evaluation of LV function, especially when time and resources are limited. This paves the way for integrating DAR into emergency protocols and routine emergency clinical practice. While these findings are promising, we acknowledge the need for prospective validation in a diverse patient population.
|
The Healthcare Integrated Research Database ( | 4f3df80c-9db4-4dd1-9a58-3f33e4b36faa | 11798679 | Pharmacology[mh] | Introduction Real‐world data (RWD) are increasingly used to generate real‐world evidence (RWE) to support the development and regulation of healthcare products, perform safety and effectiveness studies, and conduct health economics and outcomes research, among other uses . In response to the expanding utility of RWE, there is increasing availability of RWD sources, including national healthcare registries, disease and exposure registries, electronic health record (EHR) systems, and healthcare claims databases . Individuals in the United States (US) can obtain health insurance from their employer, individual exchanges, or government programs (i.e., Medicare and Medicaid). For employer‐sponsored insurance, there is typically a designated open enrollment period each year, during which employees can sign up for or make changes to their health insurance plans, as well as special enrollment periods for individuals who experience qualifying life events (e.g., marriage, birth/adoption of a child). Employees receive information about the available health plans, including details on premiums, coverage options, provider networks, and out‐of‐pocket costs. Employees evaluate the plans and select one that best meets their needs and budget. Individuals who do not have access to employer‐sponsored coverage can buy insurance through individual exchanges. These exchanges have an annual open enrollment period as well as special enrollment periods for individuals who experience a qualifying life event. The exchanges provide a range of health plans to choose from, allowing individuals to compare details on premiums, coverage options, provider networks, and out‐of‐pocket costs. Medicare is generally available to individuals aged 65 and older or to those under 65 with certain disabilities or conditions. Enrollees choose between traditional Medicare, offered from the federal government (Centers for Medicare & Medicaid Services), or Medicare Advantage plans offered through private insurance companies. Those who purchase traditional Medicare can also purchase supplemental coverage from private insurance companies to help cover out‐of‐pocket costs not covered by traditional Medicare. In addition, individuals can choose prescription drug coverage (Medicare Part D) from private insurance companies approved by Medicare. Eligibility for Medicaid is primarily based on income, household size, and state‐specific criteria. If eligible, individuals may choose from various managed care plans offered by private insurers contracted with the state Medicaid program. When a covered individual seeks medical care, healthcare providers (e.g., doctors, hospitals) deliver services and generate a claim containing service details and costs which is submitted to the insurer using standardized forms. Insurers review claims for accuracy, coverage, and medical necessity and determine the amount payable based on contractual agreements. The insurer then sends payments to providers for approved services. Insurers maintain comprehensive records of all submitted claims, whether paid or denied. Healthcare claims databases are an attractive option to obtain RWD, as they can include the healthcare information needed to support many research activities (such as enrollment dates in the health plan, demographic characteristics, diagnoses, treatments, and costs) in large populations. This information, obtained from all eligible members, providers, and facilities, identifies time periods of health plan membership and provides a nearly comprehensive picture of a patient's interactions with the healthcare system during this time of active enrollment. Though research using claims data alone can provide significant value, it does have limitations. Claims data are not captured for services which are not paid by the health plan (i.e., individual pays cash for services), certain clinical details are not contained in claims data, and the accuracy of diagnosis and procedure coding can be variable. There are a growing number of individual healthcare claims databases and multi‐source claims data aggregator databases available to researchers, and each database possesses attributes that impact its suitability for a specific research purpose. Foremost, a database must include the target population in sufficient numbers and the data elements necessary, with sufficient accuracy and completeness, to conduct valid and reliable research. There may be other requirements for a specific research project, such as the ability to trace the claims data to their origins to ensure satisfactory data quality, the ability to link external sources of information (e.g., medical records, National Death Index (NDI)) to supplement or validate claims data, or the ability to link mothers with their infants to conduct pregnancy‐related research. The Healthcare Integrated Research Database (HIRD, formerly the HealthCore Integrated Research Database) is a data environment curated and maintained by Carelon Research that includes “closed” (reviewed and approved by the payer) healthcare claims of individuals from across the US and has been augmented with additional data to support a variety of health‐related research. The HIRD contains data from 2006 and is updated monthly. This article describes the types, elements, timeliness, and quality of the data in the HIRD for health researchers considering RWD for US populations. Data Types in the HIRD The foundation of the HIRD is the enrollment records and associated healthcare claims for individuals enrolled in commercial, Medicare, or Medicaid health plans offered or managed by Elevance Health (Table ). The HIRD includes data from health plans in 33 states in the US, with individual members located throughout all 50 states. Data from individuals in these plans are generally available for research; however, some individuals or employer groups may choose not to make their data available for research. There are no restrictions on the use of data from Medicare Advantage plans (complete coverage through private health plans), supplemental Medicare coverage through the health plans, and Medicare Part D (pharmacy coverage). The HIRD does not contain traditional Medicare data. Medicaid plans require state‐by‐state approval for use in research. 2.1 Enrollment Records Enrollment records are created for individuals for all time segments during which they are enrolled in the health plan. Knowing when individuals are eligible allows researchers to distinguish the absence of data from the absence of a healthcare encounter . These records include unique identifiers for individuals, type of health plan, period of enrollment in the health plan, and demographic and geographic information (Table ). Individuals typically have multiple enrollment records, or lines of enrollment data called segments, for different periods of time. For administrative reasons, there is at least one segment in a calendar year for each time period enrolled, but a single segment may span multiple years. Members may have multiple segments that can overlap or be adjacent to one another owing to changes in the member's life, including multiple addresses in the same enrollment period, change of name, or change of employer who has coverage with the same health plan. Individuals who change plans or plan types within the same overarching health plan (i.e., have multiple insurance identifiers) are linked with a master identifier so they are recognized as the same individual. This includes individuals who transition to non‐traditional Medicare coverage from individual or employee‐sponsored insurance within the same health plan. Continuous enrollment episodes are created by integrating or “rolling up” records with overlapping dates into continuous enrollment segments. Overlapping and adjacent enrollment segments are joined together with a 1‐day gap allowed. Segments separated by more than 2 days are not rolled up together, resulting in multiple enrollment records with gaps in enrollment for individual patients. An individual can be followed across the healthcare system using unique characteristics that enable deterministic and probabilistic linkage between the individual's claims data and other data sources, including those permanently integrated into the HIRD (described below), as well as additional sources that are typically linked for specific research studies, such as medical records, the NDI, and vaccine/disease registries. Appropriate privacy and security protections are implemented to ensure regulatory compliance. The ability to link mothers and their infants within the HIRD supports pregnancy‐related research. Previous research using internally developed methodology has demonstrated that approximately 75% of infants are added to their mothers' insurance plans, allowing their data to be linked to their mothers' data in the HIRD . 2.2 Medical Claims Medical claims submitted by healthcare professionals and facilities for payment of services rendered include service dates, diagnoses, procedures, provider information, service locations, and service costs (Table ). Diagnoses are recorded using International Classification of Diseases 9th and 10th Clinical Modification codes (ICD‐9‐CM and ICD‐10‐CM), with up to 12 diagnosis codes per claim. If multiple services occur on the same date with the same provider, those services are typically combined into a single claim. Procedures, including treatment administrations such as infusions, are recorded using ICD‐9‐CM procedure codes, ICD‐10 Procedure Coding System (ICD‐10‐PCS) codes, Healthcare Common Procedure Coding System (HCPCS) codes, and Current Procedural Terminology (CPT) codes. Provider information includes National Provider Identifier (NPI) and tax identification numbers, practice name, and practice address. Service locations are classified into one of four settings: inpatient, stand‐alone emergency department, outpatient (including telehealth), and skilled nursing facility. Inpatient episodes (time period between admission and discharge) are constructed from multiple claims using a proprietary algorithm informed by service locations and service dates. Outpatient services are classified using the Restructured Berenson‐Eggers Type of Service Classification System . Service costs in medical claims include the amounts paid by the health plan, the individual, and other health plans. Cost data are available for all medical claims, with a small proportion of the cost data being imputed. 2.3 Pharmacy Claims Pharmacy claims for prescription treatments are submitted by outpatient pharmacies, including mail order and specialty pharmacies. Pharmacy claims include National Drug Codes (NDCs) for the treatment dispensed, quantity dispensed, days' supply, medication costs, dispensing date, prescriber information, and information about the dispensing pharmacy (Table ). NDCs provide information about the manufacturer, medication dispensed, strength, package size, route of administration, and dosage form. NDCs are mapped to Generic Product Identifier (GPI) codes, providing further information on the medication's therapeutic category as well as medication class and sub‐class . Pharmacy cost data include amounts paid by the health plan, the individual, and other health plans. For the subset of individuals who receive pharmacy benefits separately from their medical benefits, costs for pharmacy claims are imputed . 2.4 Provenance of Healthcare Claims Healthcare claims that enter the HIRD are generated when healthcare services are provided to individuals by healthcare providers. Providers submit claims for reimbursement to the health plan, which are either accepted or rejected (and potentially resubmitted). Healthcare claims are linked to enrollment records and then processed before being stored in the data warehouse. Enrollment records and associated healthcare claims are then integrated with external data sources (Figure ). Healthcare claims data in the HIRD are updated monthly. Only paid claims are included, with over 97% of pharmacy claims paid within 30 days, more than 90% of outpatient medical claims within 60 days, and over 90% of inpatient claims within 90 days (Figure ). As a complete healthcare claims history for a defined time period is important for many research studies, a 3‐month lag from the most recent data load is typically imposed on healthcare claims data available for research at the study level. Data in the HIRD have been available since January 2006, unless otherwise specified. The entirety of the HIRD is updated monthly, and quality control metrics are reviewed at each update to assess the data accuracy and completeness of both old data and new incremental data (i.e., data since the previous update). Monthly updating allows the old data to be overwritten with the most current data. By reviewing the entire database each month, trends in data quality can be identified and investigated. For example, data from a large Health Maintenance Organization (HMO) plan were originally deemed incomplete due to capitation arrangements and were excluded from the HIRD due to incomplete capture of healthcare services. Through collaborative projects with the health plan in recent years, the capitated plans began submitting claims with increasing consistency, akin to non‐capitated plans. This led to a reassessment of the previous decision to exclude HMO plans. Upon verifying the enhanced completeness of claims from this large HMO, the addition of these claims to the HIRD increased the researchable population by about 5%. Carelon Research's ability to trace data in the HIRD to their origins enables investigations of key data elements that can improve study quality. For example, in a safety study of a new migraine treatment, it was observed that the first dose recorded in the HIRD claims for many patients did not align with dosing recommendations (i.e., the first dose recorded in the HIRD claims was not the expected loading dose). The study team surveyed providers to inquire about real‐world dosing practices and learned that most providers used free medication samples for loading doses. With this knowledge, the study team was able to conduct more valid analyses and offer more meaningful interpretation of the data in the HIRD . Enrollment Records Enrollment records are created for individuals for all time segments during which they are enrolled in the health plan. Knowing when individuals are eligible allows researchers to distinguish the absence of data from the absence of a healthcare encounter . These records include unique identifiers for individuals, type of health plan, period of enrollment in the health plan, and demographic and geographic information (Table ). Individuals typically have multiple enrollment records, or lines of enrollment data called segments, for different periods of time. For administrative reasons, there is at least one segment in a calendar year for each time period enrolled, but a single segment may span multiple years. Members may have multiple segments that can overlap or be adjacent to one another owing to changes in the member's life, including multiple addresses in the same enrollment period, change of name, or change of employer who has coverage with the same health plan. Individuals who change plans or plan types within the same overarching health plan (i.e., have multiple insurance identifiers) are linked with a master identifier so they are recognized as the same individual. This includes individuals who transition to non‐traditional Medicare coverage from individual or employee‐sponsored insurance within the same health plan. Continuous enrollment episodes are created by integrating or “rolling up” records with overlapping dates into continuous enrollment segments. Overlapping and adjacent enrollment segments are joined together with a 1‐day gap allowed. Segments separated by more than 2 days are not rolled up together, resulting in multiple enrollment records with gaps in enrollment for individual patients. An individual can be followed across the healthcare system using unique characteristics that enable deterministic and probabilistic linkage between the individual's claims data and other data sources, including those permanently integrated into the HIRD (described below), as well as additional sources that are typically linked for specific research studies, such as medical records, the NDI, and vaccine/disease registries. Appropriate privacy and security protections are implemented to ensure regulatory compliance. The ability to link mothers and their infants within the HIRD supports pregnancy‐related research. Previous research using internally developed methodology has demonstrated that approximately 75% of infants are added to their mothers' insurance plans, allowing their data to be linked to their mothers' data in the HIRD . Medical Claims Medical claims submitted by healthcare professionals and facilities for payment of services rendered include service dates, diagnoses, procedures, provider information, service locations, and service costs (Table ). Diagnoses are recorded using International Classification of Diseases 9th and 10th Clinical Modification codes (ICD‐9‐CM and ICD‐10‐CM), with up to 12 diagnosis codes per claim. If multiple services occur on the same date with the same provider, those services are typically combined into a single claim. Procedures, including treatment administrations such as infusions, are recorded using ICD‐9‐CM procedure codes, ICD‐10 Procedure Coding System (ICD‐10‐PCS) codes, Healthcare Common Procedure Coding System (HCPCS) codes, and Current Procedural Terminology (CPT) codes. Provider information includes National Provider Identifier (NPI) and tax identification numbers, practice name, and practice address. Service locations are classified into one of four settings: inpatient, stand‐alone emergency department, outpatient (including telehealth), and skilled nursing facility. Inpatient episodes (time period between admission and discharge) are constructed from multiple claims using a proprietary algorithm informed by service locations and service dates. Outpatient services are classified using the Restructured Berenson‐Eggers Type of Service Classification System . Service costs in medical claims include the amounts paid by the health plan, the individual, and other health plans. Cost data are available for all medical claims, with a small proportion of the cost data being imputed. Pharmacy Claims Pharmacy claims for prescription treatments are submitted by outpatient pharmacies, including mail order and specialty pharmacies. Pharmacy claims include National Drug Codes (NDCs) for the treatment dispensed, quantity dispensed, days' supply, medication costs, dispensing date, prescriber information, and information about the dispensing pharmacy (Table ). NDCs provide information about the manufacturer, medication dispensed, strength, package size, route of administration, and dosage form. NDCs are mapped to Generic Product Identifier (GPI) codes, providing further information on the medication's therapeutic category as well as medication class and sub‐class . Pharmacy cost data include amounts paid by the health plan, the individual, and other health plans. For the subset of individuals who receive pharmacy benefits separately from their medical benefits, costs for pharmacy claims are imputed . Provenance of Healthcare Claims Healthcare claims that enter the HIRD are generated when healthcare services are provided to individuals by healthcare providers. Providers submit claims for reimbursement to the health plan, which are either accepted or rejected (and potentially resubmitted). Healthcare claims are linked to enrollment records and then processed before being stored in the data warehouse. Enrollment records and associated healthcare claims are then integrated with external data sources (Figure ). Healthcare claims data in the HIRD are updated monthly. Only paid claims are included, with over 97% of pharmacy claims paid within 30 days, more than 90% of outpatient medical claims within 60 days, and over 90% of inpatient claims within 90 days (Figure ). As a complete healthcare claims history for a defined time period is important for many research studies, a 3‐month lag from the most recent data load is typically imposed on healthcare claims data available for research at the study level. Data in the HIRD have been available since January 2006, unless otherwise specified. The entirety of the HIRD is updated monthly, and quality control metrics are reviewed at each update to assess the data accuracy and completeness of both old data and new incremental data (i.e., data since the previous update). Monthly updating allows the old data to be overwritten with the most current data. By reviewing the entire database each month, trends in data quality can be identified and investigated. For example, data from a large Health Maintenance Organization (HMO) plan were originally deemed incomplete due to capitation arrangements and were excluded from the HIRD due to incomplete capture of healthcare services. Through collaborative projects with the health plan in recent years, the capitated plans began submitting claims with increasing consistency, akin to non‐capitated plans. This led to a reassessment of the previous decision to exclude HMO plans. Upon verifying the enhanced completeness of claims from this large HMO, the addition of these claims to the HIRD increased the researchable population by about 5%. Carelon Research's ability to trace data in the HIRD to their origins enables investigations of key data elements that can improve study quality. For example, in a safety study of a new migraine treatment, it was observed that the first dose recorded in the HIRD claims for many patients did not align with dosing recommendations (i.e., the first dose recorded in the HIRD claims was not the expected loading dose). The study team surveyed providers to inquire about real‐world dosing practices and learned that most providers used free medication samples for loading doses. With this knowledge, the study team was able to conduct more valid analyses and offer more meaningful interpretation of the data in the HIRD . Other Data Types in the HIRD 3.1 Electronic Health Records The HIRD contains structured and unstructured EHR data for a subset of individuals. EHR data are obtained from provider network systems, large health systems and clinics, and state‐level health information exchanges (Table ). While most integrated EHR data come from outpatient primary care providers, some include records from specialists and inpatient providers. Integrated EHR data are available beginning January 2010. Additionally, the HIRD can be used as a sampling pool to identify individuals with characteristics of interest for which EHR data outside of the HIRD can be requested from providers and inpatient facilities. Structured EHR data include anthropometrics, vital signs, behavioral risk factors, medical history, and medications prescribed. In addition to the coding systems used in medical and pharmacy claims, structured data in the EHR may also contain Systematized Nomenclature of Medicine (SNOMED) codes (diagnoses, procedures), clinical drugs normalized (RxNorm) codes (treatments), and vaccinations (CVX codes). Unstructured EHR data include provider office visit notes. Through natural language processing, unstructured EHR data can be queried to identify clinical information and create structured fields, such as ejection fraction values for individuals with a heart failure diagnosis and heart failure classification . 3.2 Laboratory Results Laboratory test results for outpatient laboratory services are integrated within the HIRD. Laboratory test results generated by nationwide laboratory providers or included in EHRs are defined using Logical Observation Identifiers Names and Codes (LOINC), which provide information regarding the specimen source and methods of measurement (Table ). Laboratory results may be reported in inconsistent formats across labs; therefore, prior to inclusion in the HIRD, they are processed and standardized to ensure logically consistent data. The HIRD includes more than 65 of the most common laboratory tests readily available for research; however, this can be expanded to include other labs of interest. The proportion of individuals with laboratory results varies by test and therapeutic area. 3.3 Vital Status Vital status for individuals in the HIRD is obtained from enrollment records (reason for disenrollment), inpatient claims (discharge status), the Death Master File from the Social Security Administration, utilization management data, Center for Medicare and Medicaid Services records, and online obituary information processed by third‐party vendors. Data from these sources are combined to create a composite mortality variable for research use, indicating day of death (Table ). Cause of death is not directly available in the HIRD but can be approximated algorithmically in many cases . In a validation study among patients with advanced cancer between 2010 and 2018 ( n = 40 679), the composite mortality variable had good agreement with the NDI (sensitivity 89%, specificity 89%, positive predictive value 93%, negative predictive value 92%) . 3.4 Oncology The HIRD includes oncology data from the Carelon Cancer Care Quality Program (CCQP). Clinical oncology data for individuals undergoing cancer treatment in outpatient settings are recorded when a healthcare provider requests preauthorization for cancer treatment. Data entered by providers into the CCQP online portal include cancer type, cancer stage, biomarkers, pathology/histology, line of treatment, planned treatment regimen, height and weight, a metric of functional status (Eastern Cooperative Oncology Group Performance Status Scale), and other clinical details (Table ). In a validation study that compared the contents of the CCQP to medical records for breast, lung, and colorectal cancer patients, good agreement was observed for cancer type, cancer stage, histology (lung cancer only), and select cancer biomarkers . CCQP data are available beginning July 2014. 3.5 Social and Health Equity Data Individual‐level race and ethnicity data are obtained from enrollment files, EHR data, member self‐assessments, and proprietary imputation algorithms . These data are combined to create a composite race and ethnicity variable for research use (Table ). The composite race and ethnicity variable is based on Office of Management and Budget standards (White non‐Hispanic, Native Hawaiian/Other Pacific Islander non‐Hispanic, Black/African American non‐Hispanic, Asian non‐Hispanic, American Indian/Alaska Native non‐Hispanic, Hispanic/Latino, and other race non‐Hispanic) . A validation study found high agreement (Kappa = 0.82) between the composite variable and self‐reported race/ethnicity (among commercially insured Asian, Black/African American, Hispanic/Latino, and White individuals) . A variety of Social Drivers of Health (SDoH) data have been integrated into the HIRD . Area‐level information about urbanicity is derived from the National Center for Health Statistics' urban–rural classification scheme. The HIRD includes area‐level data from the American Community Survey, including over 50 variables at the census block group level associated with healthcare resource utilization, such as educational attainment, income, living conditions, family composition, transportation, and employment (Table ) . The HIRD also includes area‐level data from the Food Access Research Atlas, with over 140 variables at the census tract level related to food access and availability, urbanicity and rurality, and income . Social and health equity data in the HIRD are available for up to 95% of individuals. 3.6 Vaccinations Vaccination data from the Immunization Information System (IIS) are included in the HIRD to supplement vaccination data from healthcare claims and EHR data (Table ). Currently, the IIS data are obtained from 16 jurisdictions in 15 states and represent approximately 60% of members in the HIRD, although the data available may differ by state . Electronic Health Records The HIRD contains structured and unstructured EHR data for a subset of individuals. EHR data are obtained from provider network systems, large health systems and clinics, and state‐level health information exchanges (Table ). While most integrated EHR data come from outpatient primary care providers, some include records from specialists and inpatient providers. Integrated EHR data are available beginning January 2010. Additionally, the HIRD can be used as a sampling pool to identify individuals with characteristics of interest for which EHR data outside of the HIRD can be requested from providers and inpatient facilities. Structured EHR data include anthropometrics, vital signs, behavioral risk factors, medical history, and medications prescribed. In addition to the coding systems used in medical and pharmacy claims, structured data in the EHR may also contain Systematized Nomenclature of Medicine (SNOMED) codes (diagnoses, procedures), clinical drugs normalized (RxNorm) codes (treatments), and vaccinations (CVX codes). Unstructured EHR data include provider office visit notes. Through natural language processing, unstructured EHR data can be queried to identify clinical information and create structured fields, such as ejection fraction values for individuals with a heart failure diagnosis and heart failure classification . Laboratory Results Laboratory test results for outpatient laboratory services are integrated within the HIRD. Laboratory test results generated by nationwide laboratory providers or included in EHRs are defined using Logical Observation Identifiers Names and Codes (LOINC), which provide information regarding the specimen source and methods of measurement (Table ). Laboratory results may be reported in inconsistent formats across labs; therefore, prior to inclusion in the HIRD, they are processed and standardized to ensure logically consistent data. The HIRD includes more than 65 of the most common laboratory tests readily available for research; however, this can be expanded to include other labs of interest. The proportion of individuals with laboratory results varies by test and therapeutic area. Vital Status Vital status for individuals in the HIRD is obtained from enrollment records (reason for disenrollment), inpatient claims (discharge status), the Death Master File from the Social Security Administration, utilization management data, Center for Medicare and Medicaid Services records, and online obituary information processed by third‐party vendors. Data from these sources are combined to create a composite mortality variable for research use, indicating day of death (Table ). Cause of death is not directly available in the HIRD but can be approximated algorithmically in many cases . In a validation study among patients with advanced cancer between 2010 and 2018 ( n = 40 679), the composite mortality variable had good agreement with the NDI (sensitivity 89%, specificity 89%, positive predictive value 93%, negative predictive value 92%) . Oncology The HIRD includes oncology data from the Carelon Cancer Care Quality Program (CCQP). Clinical oncology data for individuals undergoing cancer treatment in outpatient settings are recorded when a healthcare provider requests preauthorization for cancer treatment. Data entered by providers into the CCQP online portal include cancer type, cancer stage, biomarkers, pathology/histology, line of treatment, planned treatment regimen, height and weight, a metric of functional status (Eastern Cooperative Oncology Group Performance Status Scale), and other clinical details (Table ). In a validation study that compared the contents of the CCQP to medical records for breast, lung, and colorectal cancer patients, good agreement was observed for cancer type, cancer stage, histology (lung cancer only), and select cancer biomarkers . CCQP data are available beginning July 2014. Social and Health Equity Data Individual‐level race and ethnicity data are obtained from enrollment files, EHR data, member self‐assessments, and proprietary imputation algorithms . These data are combined to create a composite race and ethnicity variable for research use (Table ). The composite race and ethnicity variable is based on Office of Management and Budget standards (White non‐Hispanic, Native Hawaiian/Other Pacific Islander non‐Hispanic, Black/African American non‐Hispanic, Asian non‐Hispanic, American Indian/Alaska Native non‐Hispanic, Hispanic/Latino, and other race non‐Hispanic) . A validation study found high agreement (Kappa = 0.82) between the composite variable and self‐reported race/ethnicity (among commercially insured Asian, Black/African American, Hispanic/Latino, and White individuals) . A variety of Social Drivers of Health (SDoH) data have been integrated into the HIRD . Area‐level information about urbanicity is derived from the National Center for Health Statistics' urban–rural classification scheme. The HIRD includes area‐level data from the American Community Survey, including over 50 variables at the census block group level associated with healthcare resource utilization, such as educational attainment, income, living conditions, family composition, transportation, and employment (Table ) . The HIRD also includes area‐level data from the Food Access Research Atlas, with over 140 variables at the census tract level related to food access and availability, urbanicity and rurality, and income . Social and health equity data in the HIRD are available for up to 95% of individuals. Vaccinations Vaccination data from the Immunization Information System (IIS) are included in the HIRD to supplement vaccination data from healthcare claims and EHR data (Table ). Currently, the IIS data are obtained from 16 jurisdictions in 15 states and represent approximately 60% of members in the HIRD, although the data available may differ by state . Characteristics of the HIRD Population The HIRD population used for research (“researchable population”) consists of individuals/employer groups with commercial and/or managed Medicare health insurance plans who have agreed to be included for research purposes. The specific population available for use in any given research study varies depending on population requirements for that study. Studies using claims data only can utilize the entire researchable population. As of July 2024, the researchable population included over 91 million individuals with medical benefits: over 72 million of these individuals also have pharmacy benefits with at least 1 day of membership in the HIRD (Table ). Of these, approximately 24 million individuals with medical benefits and approximately 17 million individuals with medical and pharmacy benefits were actively enrolled (i.e., had active health plan membership as of July 2024). The median age for the researchable population is 36 years (interquartile range, IQR: 22, 54), with approximately 10% ≥ 65 years old. The researchable population is approximately half male and half female. Individuals with available race and ethnicity data (46% for the entire HIRD starting 2006; ≥ 80% for the most recent 5 years) are approximately 63% White non‐Hispanic, 15% Hispanic/Latino, 9% Black/African American non‐Hispanic, 7% Asian non‐Hispanic, and 6% other races or ethnicities. Individuals most commonly live in the South census region (32%), followed by the West (27%), Midwest (23%), and Northeast (16%). Approximately 92% of individuals have commercial health insurance, whereas approximately 8% have managed Medicare (including Medicare Advantage, Medicare Supplement, and Medicare Part D) (Table ). Preferred provider organizations are the most common plan type (64%), followed by HMOs (19%) and consumer‐directed health plans (17%). The median duration of continuous enrollment for the entire population is approximately 2.0 years (IQR: 0.8, 4.4). Overall, the population with medical benefits is similar to the population with medical and pharmacy benefits, and the actively enrolled population is similar to the enrolled‐at‐any‐time population (Table ). Notably, actively enrolled patients are more likely to have race and ethnicity data (~80% for actively enrolled with medical benefits), and actively enrolled patients are typically enrolled for longer durations (median (IQR) of 3.8 (1.7, 8.3) years for actively enrolled with medical benefits). Strengths and Limitations The HIRD includes an abundance of RWD for a large population dispersed across the US. These data have supported many RWE studies completed exclusively within the HIRD and have made the HIRD a valuable contributor to multi‐database efforts , including the FDA's Sentinel Initiative , Biologics Effectiveness and Safety (BEST) System , Innovation in Medical Evidence and Development Surveillance (IMEDS) program , Biologics and Biosimilars Collective Intelligence Consortium (BBCIC) , and National Evaluation System for Health Technology (NEST) . Carelon Research's unique relationship with Elevance Health enables auditing and investigation of the source of healthcare claims data in the HIRD, providing assurance of data quality and yielding insights that inform study design and interpretation. The data in the HIRD begin in 2006 and are updated monthly, allowing the HIRD to support both historical and ongoing inquiries. The 2020 HIRD population (commercial and managed Medicare) is similar to the 2020 US Census population in terms of sex (male/female; overlap index = 99.2%), age (5‐year age categories; overlap index = 92.0%), and geographic region of residence (Northeast/Midwest/South/West; overlap index = 94.8%). For race/ethnicity (Hispanic or Latino/non‐Hispanic White/non‐Hispanic Black or African American/non‐Hispanic Asian/Other), the overlap index is 86.8%, with the 2020 HIRD population having 13% more non‐Hispanic White members and 11% fewer Hispanic or Latino and non‐Hispanic Black or African American members than the 2020 US Census population . The size of the database, the available data elements, and longitudinal nature of the data allow for a wide variety of pharmacoepidemiologic and health economic and outcomes research studies. The ability to re‐identify individuals (with appropriate approvals) supports many research initiatives, including validation studies , patient and provider surveys , linkage to medical records , linkage to product and disease registries , linkage to the NDI , as well as recruitment into pragmatic clinical trials . The ability to link family individuals supports pregnancy‐related research . The HIRD also has limitations. Claims data in the HIRD are only captured during periods of enrollment in the health plan, which limits analyses to the time an individual is enrolled in a health plan. Of note, a prior study evaluating enrollment time showed individuals in the HIRD with chronic diseases are enrolled for longer periods of time, allowing for longer follow‐up of this clinically important population . Claims data in the HIRD may incompletely or inaccurately reflect a patient's experience. For example, a pharmacy claim for a prescription medication does not guarantee consumption of the medication, or a medical claim with a diagnosis may be inaccurate due to misdiagnosis or coding errors. Certain data useful for research may not be readily available in the HIRD, such as sample medications from providers, individuals who pay for medications or other services without submitting to the health plan for payment, medications received in inpatient settings, markers of disease severity, or lifestyle risk factors. Limitations of both inaccurate and/or missing data may be addressed through a variety of methods, including the addition of data via linkage to external sources , the collection of data directly from patient or provider surveys, and analytically through quantitative bias analysis and other types of sensitivity analyses. Data Access Data within the HIRD are directly available for research through licensing and collaborations with Carelon Research. Conclusion The HIRD, established in 2006, has progressively expanded by incorporating additional data sources, allowing researchers to address more complex questions. This database allows for direct traceability back to the source data, which enhances research integrity and reliability. The HIRD's utility is evidenced by over 1900 peer‐reviewed publications and presentations. The ability to link the HIRD to other datasets amplifies research capabilities and underscores the value of RWD. The HIRD exemplifies how comprehensive data curation can be leveraged for diverse research applications and significant scientific contributions. 7.1 Plain Language Summary This article describes the HIRD, a RWD source that can be used to support health‐related research. The HIRD includes information on individuals with health insurance provided by Elevance Health. These individuals are located throughout the US. The HIRD is built upon the foundation of health insurance enrollment and claims data, which are linked with EHRs, laboratory test results, mortality data, clinical cancer data, social and health equity data, and vaccination data. The data in the HIRD date back to January 2006, and are updated monthly. As of July 2024, the HIRD included over 91 million individuals with medical insurance, including approximately 24 million individuals who were actively enrolled. A strength of the HIRD is that its data can be traced back to their origins. Also, data in the HIRD can be linked to external data sources, and family members within health plans are linked to each other. Because of these and other attributes, the HIRD has been used to support many health‐related research activities over the past 2 decades. The HIRD is directly available for research through licensing and collaborations with Carelon Research. Plain Language Summary This article describes the HIRD, a RWD source that can be used to support health‐related research. The HIRD includes information on individuals with health insurance provided by Elevance Health. These individuals are located throughout the US. The HIRD is built upon the foundation of health insurance enrollment and claims data, which are linked with EHRs, laboratory test results, mortality data, clinical cancer data, social and health equity data, and vaccination data. The data in the HIRD date back to January 2006, and are updated monthly. As of July 2024, the HIRD included over 91 million individuals with medical insurance, including approximately 24 million individuals who were actively enrolled. A strength of the HIRD is that its data can be traced back to their origins. Also, data in the HIRD can be linked to external data sources, and family members within health plans are linked to each other. Because of these and other attributes, the HIRD has been used to support many health‐related research activities over the past 2 decades. The HIRD is directly available for research through licensing and collaborations with Carelon Research. Carelon Research's access, use, and disclosure of protected health information (PHI) complies with the HIPAA Privacy Rule (45 CFR Part 160 and Subparts A and E of Part 164). Carelon Research does not access, use, or disclose PHI other than as permitted by HIPAA. De‐identified or limited datasets for research are created when feasible; however, when that is not feasible, Carelon Research may seek to obtain a specific waiver of the HIPAA authorization requirements from an Institutional Review Board (IRB). Carelon Research also takes into consideration other federal and state laws and regulations that might limit the use of certain types of data beyond HIPAA limitations, including laws related to substance use disorders and other sensitive medical information. All authors were employees of Carelon Research at the time the manuscript was prepared. All authors are shareholders of Elevance Health. Brett Doherty is currently an employee of Daiichi Sankyo Inc., Basking Ridge, NJ. |
Haze Exposure Changes the Skin Fungal Community and Promotes the Growth of | 2ee0df10-e9da-4e64-952f-16e9d0560376 | 10269824 | Microbiology[mh] | Skin is the largest organ and home to millions of microorganisms in humans. The skin microbiota has important roles in the homeostasis as a component of the host defense ( ). Fungi are less abundant than bacteria on the skin ( ), but they play important roles in skin health ( , ). Fungal communities are ignored by most of the studies of skin microbiota using high-throughput sequencing ( ). Generally, the skin microbiota of healthy people is stable over time ( ). However, it is influenced by a variety of factors, including age ( , ), gender ( ), and environmental variables such as climate ( ), season ( ), geography ( ), hygiene practices ( ), and urbanization ( , ). Among the environmental variables, air pollutants are an important factor that disturbs the skin microbiota ( ). Haze exposure has been a threat to human health in developing industrial countries such as China ( ). According to a report of the World Health Organization, approximately seven million people die each year as a result of diseases caused by air pollution exposure ( ). Particulate matter (PM) is the primary pollutant in the air and is a complex combination of inorganic materials, organic molecules, and biological components ( ). PM increases the risk of diseases and mortality in stroke ( ), heart disease ( , ), lung cancer ( , ), respiratory diseases ( ), and skin damage ( , ). Considering that the skin is most directly in contact with PM pollutants, PM-bound microbes and chemicals can cause even higher degrees of harmful health effects. For example, long-term PM exposure causes skin inflammation, and even accelerates aging and wrinkle formation, affecting skin metabolism and destroying the skin barrier ( ). PM is deposited on the skin and constitutes a part of the skin microbiota. The pollutants attached in PM may disturb the skin microbiota, affecting skin health. Several studies have assessed the effects of pollutants on the skin microbiota ( ). Exposure to O 3 and NO 2 significantly reduced the viability of skin bacteria ( , ). Different polycyclic aromatic hydrocarbons and related xenobiotic chemicals have been shown to be degraded by skin bacteria ( , ). Alterations in the composition and functional characteristics of the skin microbiota have been linked to chronic exposure levels of polycyclic aromatic hydrocarbon pollutants according to Leung et al. ( ). However, to the best of our knowledge, the effect of the haze pollutants on skin fungi has never been evaluated. The goal of this study was to (i) determine whether haze exposure changes the skin fungal community composition and diversity, (ii) identify the genera associated with haze and confirm the interaction between the haze pollutants and the growth of representative strains of the genera in vitro , and (iii) reveal the assembly process of skin fungi during haze and nonhaze days.
The composition of the skin fungal community can be disturbed by haze. A total of 74 skin samples were collected, including 19 samples during haze days (March 14 in spring), 19 samples during nonhaze days (April 27 in spring), 18 samples during haze days (November 29 in winter), and 18 samples during nonhaze days (December 30 in winter) in 2018. The levels of PM2.5 and PM10 and the air quality index (AQI) were recorded from the data of Xinxiang sites of the national program for recording urban air quality in China ( https://air.cnemc.cn:18007 ) for 7 days before sample collection ( and Table S1 in the supplemental material). A total of 3,377,033 reads with 45,636 reads per sample on average were quality filtered, denoised, merged, filtered for chimeras, and clustered into 3,136 unique amplicon sequence variants (ASVs) (Table S2). The results of principal-coordinate analysis (PCoA) and analysis of similarities (ANOSIM) showed that the composition of the fungal community on the skin differed significantly between spring and winter samples ( , Fig. S1A), which suggested that the following analysis should be divided into spring and winter groups. According to the PCoA results, there was no significant difference in the skin fungal community composition between haze and nonhaze days ( , Fig. S1B). To further remove interference factors, we assessed other factors which can disturb the composition of the skin fungal community using canonical correspondence analysis (CCA). The CCA ( ) and permutation tests (Tables S3 and S4) showed that gender had significant correlations with skin fungal community composition in spring ( R 2 = 0.268, P = 0.013) and winter ( R 2 = 0.3623, P = 0.001). Thus, we grouped the samples into female and male groups. Surprisingly, there were significant differences between haze and nonhaze days in both the female and male groups in spring and male samples in winter based on unweighted UniFrac distance analysis, confirming that the composition of the skin fungal community can be disturbed by haze ( and and Fig. S1C and D). Alpha diversity and taxonomic composition of the skin fungal community. The alpha diversity was assessed by the Shannon index, phylogenetic diversity (PD) whole-tree index, and observed features. The high-throughput sequencing captured the dominant phylotypes according to the rarefaction curves (Fig. S2). On average, 43 (from 9 to 92) and 48 (from 12 to 149) ASVs were observed during haze and nonhaze days, respectively, in spring samples. On average, 103 (from 40 to 182) and 80 (from 37 to 130) ASVs were observed during haze and nonhaze days, respectively, in winter samples (Table S5). PD whole-tree ( P = 0.028) and observed features ( P = 0.045) indices were significantly increased in male samples during haze days compared to nonhaze days in winter, and Shannon indices also increased with no significant difference ( P = 0.42). These results indicated that haze increased the fungal diversities in male samples in winter. There was no significant difference in other groups between haze and nonhaze days. In addition, compared to female samples, diversities were increased in male samples in both spring and winter ( ). In skin samples, 8 phyla, 87 orders, and 436 genera were discovered. The skin fungal community was dominated by Basidiomycota, followed by Ascomycota, Mucoromycota, Rozellomycota, Mortierellomycota, Chytridiomycota, Entomophthoromycota, and Glomeromycota (Fig. S3). The predominant order was Malasseziales (77.5%), which belongs to the Basidiomycota. Other dominant orders were Pleosporales (8.4%), Capnodiales (6.8%), Eurotiales (2.8%), and Hypocreales (1.4%), which belongs to the Ascomycota (Table S6). Malassezia (72.9%) was the most abundant genus in fungal community, followed by Alternaria (7.4%), Cladosporium (5.3%), Malasseziales unidentified (1.8%), and Talaromyces (2.3%) ( ). Differences in skin fungal composition between haze and nonhaze days. A heatmap diagram showed the distributions of dominant fungal genera (>0.02%) on the skin ( ). For example, Malassezia was dominant in all samples. Talaromyces was increased during haze days; Cladosporium and Mycosphaerella were increased during nonhaze days in spring. In addition, some genera have different patterns depending on season or gender. For example, Malasseziales unidentified was more abundant in spring than in winter, while Talaromyces was less abundant. Dothideales unidentified, Sordariaceae unidentified, and Zygosaccharomyces were more abundant in female samples than male samples. Linear discriminant analysis effect size (LEfSe) demonstrated that some fungal taxa differed significantly between haze and nonhaze samples ( ). In spring, Dothideomycetes, Capnodiales, Mycosphaerellaceae, Mycosphaerella , Cladosporium , Cladosporiaceae, Schizophyllum , Schizophyllaceae, Agaricales, Ustilaginales, and Ustilaginomycetes were significantly increased during nonhaze days, Trichocomaceae and Talaromyces were significantly increased during haze days. In winter, Pezizaceae was significantly increased during haze days. Furthermore, four genera with significant differences were selected for further analysis using boxplots ( ). Talaromyces was increased during haze days, while Mycosphaerella was increased during nonhaze days in spring. Cladosporium was increased during nonhaze days in spring, while it was decreased in winter. It is hard to indicated the differences, because the relative abundance of Schizophyllum was too low. High concentrations of PM promoted the growth of Talaromyces strains. To further reveal the influence of haze exposure on the skin fungi, five Talaromyces strains were isolated (Table S7). The phylogenetic tree was constructed based on all Talaromyces sequences from high-throughput sequencing, isolation, and a type strain of Talaromyces marneffei , which clustered into four groups ( ). The type strain of Talaromyces marneffei (a common pathogen) and XSF7 were found to be in the same group. One representative strain was selected from each group. XSF1, XSF7, XSF10, and XSF103 were selected to investigate the influence of PM on Talaromyces strains in vitro . The spore suspension of Talaromyces strains was cultured with PM at 0 mg/mL (control), 0.08 mg/mL (low-concentration [LC] group), 0.64 mg/mL (medium-concentration [MC] group), and 5.12 mg/mL (high-concentration [HC] group) for 5 days. The dry weights of Talaromyces strains (particularly XSF7, XSF10, and XSF103) were increased greatly in the HC group but not in the MC or LC groups ( ), indicating that the growth of strains was promoted by high concentrations of PM. The results confirmed that haze exposure influenced the growth of several skin fungi. The skin fungal communities and taxa that increased during haze days deviated from the neutral assembly process. Previous studies suggested that the Sloan neutral model can be used to predict the assembly process of the skin microbial community ( , ). To further clarify the mechanism of changes in the skin fungal community, we assessed whether the skin fungi can be explained by the neutral model during haze and nonhaze days. The neutral model with a lower Akaike information criterion (AIC) score performed better than the binomial and Poisson models in predicting the assembly of the skin fungal community ( ). The skin fungal community during haze days exhibited a larger tendency for niche-based assembly than during nonhaze days in both spring and winter, which was demonstrated by a lower R 2 and migration rate (Nm) ( ). In addition, compared to spring, the skin fungal community with the lower R 2 and Nm in winter indicated a larger tendency for niche-based assembly. Most of the fungal species were in the 95% confidence range of the neutral model ( ). However, the fungal taxa significantly increased during haze days in winter, such as Trichocomaceae, Talaromyces , and Pezizaceae and had a higher percentage of species largely deviating from the neutral assembly process compared to the fungal community ( ). The results predicted by the Sloan neutral model suggested that the fungal community assembly was better fitted to a niche-based assembly model during haze days, and some genera with significant differences, such as Talaromyces , deviated from the neutral assembly process.
A total of 74 skin samples were collected, including 19 samples during haze days (March 14 in spring), 19 samples during nonhaze days (April 27 in spring), 18 samples during haze days (November 29 in winter), and 18 samples during nonhaze days (December 30 in winter) in 2018. The levels of PM2.5 and PM10 and the air quality index (AQI) were recorded from the data of Xinxiang sites of the national program for recording urban air quality in China ( https://air.cnemc.cn:18007 ) for 7 days before sample collection ( and Table S1 in the supplemental material). A total of 3,377,033 reads with 45,636 reads per sample on average were quality filtered, denoised, merged, filtered for chimeras, and clustered into 3,136 unique amplicon sequence variants (ASVs) (Table S2). The results of principal-coordinate analysis (PCoA) and analysis of similarities (ANOSIM) showed that the composition of the fungal community on the skin differed significantly between spring and winter samples ( , Fig. S1A), which suggested that the following analysis should be divided into spring and winter groups. According to the PCoA results, there was no significant difference in the skin fungal community composition between haze and nonhaze days ( , Fig. S1B). To further remove interference factors, we assessed other factors which can disturb the composition of the skin fungal community using canonical correspondence analysis (CCA). The CCA ( ) and permutation tests (Tables S3 and S4) showed that gender had significant correlations with skin fungal community composition in spring ( R 2 = 0.268, P = 0.013) and winter ( R 2 = 0.3623, P = 0.001). Thus, we grouped the samples into female and male groups. Surprisingly, there were significant differences between haze and nonhaze days in both the female and male groups in spring and male samples in winter based on unweighted UniFrac distance analysis, confirming that the composition of the skin fungal community can be disturbed by haze ( and and Fig. S1C and D).
The alpha diversity was assessed by the Shannon index, phylogenetic diversity (PD) whole-tree index, and observed features. The high-throughput sequencing captured the dominant phylotypes according to the rarefaction curves (Fig. S2). On average, 43 (from 9 to 92) and 48 (from 12 to 149) ASVs were observed during haze and nonhaze days, respectively, in spring samples. On average, 103 (from 40 to 182) and 80 (from 37 to 130) ASVs were observed during haze and nonhaze days, respectively, in winter samples (Table S5). PD whole-tree ( P = 0.028) and observed features ( P = 0.045) indices were significantly increased in male samples during haze days compared to nonhaze days in winter, and Shannon indices also increased with no significant difference ( P = 0.42). These results indicated that haze increased the fungal diversities in male samples in winter. There was no significant difference in other groups between haze and nonhaze days. In addition, compared to female samples, diversities were increased in male samples in both spring and winter ( ). In skin samples, 8 phyla, 87 orders, and 436 genera were discovered. The skin fungal community was dominated by Basidiomycota, followed by Ascomycota, Mucoromycota, Rozellomycota, Mortierellomycota, Chytridiomycota, Entomophthoromycota, and Glomeromycota (Fig. S3). The predominant order was Malasseziales (77.5%), which belongs to the Basidiomycota. Other dominant orders were Pleosporales (8.4%), Capnodiales (6.8%), Eurotiales (2.8%), and Hypocreales (1.4%), which belongs to the Ascomycota (Table S6). Malassezia (72.9%) was the most abundant genus in fungal community, followed by Alternaria (7.4%), Cladosporium (5.3%), Malasseziales unidentified (1.8%), and Talaromyces (2.3%) ( ).
A heatmap diagram showed the distributions of dominant fungal genera (>0.02%) on the skin ( ). For example, Malassezia was dominant in all samples. Talaromyces was increased during haze days; Cladosporium and Mycosphaerella were increased during nonhaze days in spring. In addition, some genera have different patterns depending on season or gender. For example, Malasseziales unidentified was more abundant in spring than in winter, while Talaromyces was less abundant. Dothideales unidentified, Sordariaceae unidentified, and Zygosaccharomyces were more abundant in female samples than male samples. Linear discriminant analysis effect size (LEfSe) demonstrated that some fungal taxa differed significantly between haze and nonhaze samples ( ). In spring, Dothideomycetes, Capnodiales, Mycosphaerellaceae, Mycosphaerella , Cladosporium , Cladosporiaceae, Schizophyllum , Schizophyllaceae, Agaricales, Ustilaginales, and Ustilaginomycetes were significantly increased during nonhaze days, Trichocomaceae and Talaromyces were significantly increased during haze days. In winter, Pezizaceae was significantly increased during haze days. Furthermore, four genera with significant differences were selected for further analysis using boxplots ( ). Talaromyces was increased during haze days, while Mycosphaerella was increased during nonhaze days in spring. Cladosporium was increased during nonhaze days in spring, while it was decreased in winter. It is hard to indicated the differences, because the relative abundance of Schizophyllum was too low.
Talaromyces strains. To further reveal the influence of haze exposure on the skin fungi, five Talaromyces strains were isolated (Table S7). The phylogenetic tree was constructed based on all Talaromyces sequences from high-throughput sequencing, isolation, and a type strain of Talaromyces marneffei , which clustered into four groups ( ). The type strain of Talaromyces marneffei (a common pathogen) and XSF7 were found to be in the same group. One representative strain was selected from each group. XSF1, XSF7, XSF10, and XSF103 were selected to investigate the influence of PM on Talaromyces strains in vitro . The spore suspension of Talaromyces strains was cultured with PM at 0 mg/mL (control), 0.08 mg/mL (low-concentration [LC] group), 0.64 mg/mL (medium-concentration [MC] group), and 5.12 mg/mL (high-concentration [HC] group) for 5 days. The dry weights of Talaromyces strains (particularly XSF7, XSF10, and XSF103) were increased greatly in the HC group but not in the MC or LC groups ( ), indicating that the growth of strains was promoted by high concentrations of PM. The results confirmed that haze exposure influenced the growth of several skin fungi.
Previous studies suggested that the Sloan neutral model can be used to predict the assembly process of the skin microbial community ( , ). To further clarify the mechanism of changes in the skin fungal community, we assessed whether the skin fungi can be explained by the neutral model during haze and nonhaze days. The neutral model with a lower Akaike information criterion (AIC) score performed better than the binomial and Poisson models in predicting the assembly of the skin fungal community ( ). The skin fungal community during haze days exhibited a larger tendency for niche-based assembly than during nonhaze days in both spring and winter, which was demonstrated by a lower R 2 and migration rate (Nm) ( ). In addition, compared to spring, the skin fungal community with the lower R 2 and Nm in winter indicated a larger tendency for niche-based assembly. Most of the fungal species were in the 95% confidence range of the neutral model ( ). However, the fungal taxa significantly increased during haze days in winter, such as Trichocomaceae, Talaromyces , and Pezizaceae and had a higher percentage of species largely deviating from the neutral assembly process compared to the fungal community ( ). The results predicted by the Sloan neutral model suggested that the fungal community assembly was better fitted to a niche-based assembly model during haze days, and some genera with significant differences, such as Talaromyces , deviated from the neutral assembly process.
Haze pollutants have been a universal public threat issue to human health, which can cause undesirable effects on the skin ( , ). This study highlighted that haze exposure influenced the diversity and composition of the fungal community. Talaromyces strains were isolated and cultured with PM that we collected during heavy-haze days, confirming the influence of the haze pollutants on the fungi. Furthermore, the assembly process of the fungal community was associated with haze exposure. The exploration of the interplay between haze and skin fungi can elucidate new thoughts about how haze pollutants influence skin health by regulating the skin microbiota. The skin fungal community can be influenced by various factors such as gender, body part, urbanization, season, pollutants, and so on ( , , , ). In this study, significant differences in skin fungal composition between haze and nonhaze samples were observed after we removed interfering factors such as season and gender, indicating that the fungal community might be changed by haze, which is consistent with previous studies ( , ). Polycyclic aromatic hydrocarbons (PAHs), a type of organic pollutants often associated with PM and known to induce skin diseases, were related to changes in the composition and functional capacity of the skin microbiota ( ). In a German study, four bacterial strains from human skin degraded PAHs ( ). He et al. observed that the skin microbiota of the forearm was almost halved after being exposed to O 3 ( ). These studies confirmed the relationships between skin microbiota and air pollution. Diversities in male samples from the winter group were higher during haze days than during nonhaze days in this study. Another study in China found that the increase of diversity was associated with PAH level ( ), which was consistent with our study. Our previous study on airborne fungi associated with PM also indicated that Shannon indices were increased during haze days ( ), implying that the increased diversity of the fungal community in the polluted air might be one of the reasons for the higher diversity of the skin fungal community during haze days. In addition, haze exposure increased the fungal diversity in the male group in winter, while no such pattern was observed in the female group. The variation may be mainly related to the sebum level in the retroauricular crease, as adult males have higher levels of sebum than females ( ). The skin fungal community composition observed in this study is consistent with previous surveys of skin fungi ( , ). For example, Malassezia is the predominant skin fungus in humans and is commonly observed on the skin of the face, scalp, and outer ears, which is rich in sebaceous glands ( ). Despite having a clear correlation to skin diseases such as dandruff ( , ), atopic dermatitis ( ), and pityriasis ( ), Malassezia species can also promote skin health like other skin-resident microbes by competing with the pathogen Staphylococcus aureus , secreting proinflammatory cytokines ( , ), dissociating biofilms ( ), and so on ( ). Alternaria and Cladosporium were the dominant fungal genera in the skin samples. Alternaria and Cladosporium were the most prevalent genera in the air ( , ) and dominated the airborne fungal community in China ( ), which indicated that these fungal genera on the skin might come from the air. Moreover, Alternaria and Cladosporium species are considered risk factors to allergic disease due to their high allergenicity ( , ), highlighting the high risk of allergic skin disease caused by the two genera. With the accumulation of PM on the skin during haze days, the skin fungal composition can be changed. The taxa with significant differences were identified using LEfSe analysis. Mycosphaerella was significantly increased during nonhaze days in the spring group. It is always isolated from plants and contains one of the largest generic complexes of the pathogenic ascomycetes ( , ). Mycosphaerella spp. can live in a range of habitats, such as symbionts, endophytes, hyperparasites, and plant pathogens, because it is saprobic ( , ). Therefore, it is supposed that it lowered the rates of Mycosphaerella spp. released from plants during haze days with still air. Talaromyces was the only genus significantly increased during haze days. To further clarify the relationships between the haze and fungi, Talaromyces strains were isolated and cultured with PM. The results demonstrated that haze can promote the growth of Talaromyces at high concentrations. The genus Talaromyces has a worldwide distribution, and its species are food contaminants, mycotoxin producers, and human pathogens ( ). For example, T. marneffei is a serious pathogen that can cause severe systemic infection by inhalation of conidia from the environment, which is the third-most-prevalent opportunistic infection in HIV patients in Southeast Asia and southern China ( ). It can proliferate in macrophages, causing local infection in the skin and lungs as well as severe systemic disease ( ). The sequence of the type strain T. marneffei CBS 388.87 from GenBank was clustered into one group with XSF7, indicating that PM may promote the growth of T. marneffei and increase the infection rate during haze exposure. In addition, Talaromyces species can produce abundant secondary metabolites with diverse bioactivities ( ). For example, Talaromyces pinophilus has been reported to be able to degrade agricultural waste ( ). Therefore, whether the significant changes of Talaromyces observed during haze exposure are associated with (or even cause) adverse health outcomes should be further studied. The fungal community assembly was better suited to a niche-based assembly model during haze days. As demonstrated by Leung et al. ( ), the scalp microbial community in the more heavily polluted city (Baoding, China) was better suited to the niche-based model, which is consistent with our observations. The retroauricular crease and the scalp are both enriched with sebaceous glands and have similar physiological characteristics, which explains the similar assembly process between our results and Leung et al.’s studies ( ), while the cheek microbiota of individuals from the more lightly polluted city (Dalian, China) was better suited to a niche-based model ( ), which suggested that the skin microbiota was influenced by the skin site and environment. According to the findings, microbial assembly is a dynamic process influenced by factors such as host physiology, environment, and season ( , ). In addition, taxa with significant differences between haze and nonhaze days, such as Talaromyces and Pezizaceae, deviated from the neutral model, which indicated that they had greater colonization potential ( , , ). In particular, Talaromyces spp. enriched during haze days largely deviated from the neutral model prediction, suggesting that Talaromyces may be selected by haze pollutants and increased among the skin fungi community. Conclusion. Our study first reported the changes of the skin fungal community during haze and nonhaze days. Haze exposure influenced the composition and diversity of the skin fungal community. Thus, the taxa with significant differences were observed, such as Talaromyces . Furthermore, the in vitro culture experiment revealed that the growth of representative Talaromyces strains was promoted at high concentrations of PM, confirming the high-throughput sequencing results. Finally, compared to nonhaze days, the fungal community assembly was better fitted to a niche-based assembly model during haze days. Several genera with significant differences, such as Talaromyces , deviated from the neutral assembly process. Our work provided a comprehensive characterization of the skin fungal community during haze and nonhaze days. It elucidates new insights on how haze exposure influences the skin fungal community, which provides the basis for further clarifying how the changes are associated with adverse health outcomes. We tried to clarify the interaction between haze and fungi at the strain level by the in vitro culture experiment, yet it is not enough. In the future, a combination of multiomics and the culture of haze pollutants and various fungal taxa will be required to confirm the relationships between haze, fungi, and skin health.
Skin sample collection. This research was approved by the ethics committee of Xinxiang Medical University. Healthy students aged 18 to 28 who resided on the campus of Xinxiang Medical University (35°16′53.18″N, 113°55′37.92″E) for at least 2 years were included in this study ( ). The map was constructed using Pixel Map Generator ( https://pixelmap.amcharts.com/ ). Exclusion criteria included a history of chronic disease, use of antimicrobials or topical steroids within 6 months prior to sampling, and bathing, shampooing, or moisturizing within 24 h before sampling. The skin swab samples were collected from the skin of the retroauricular crease (behind the left and right ears). Sterile FLOQSwabs (Copan Flock Technologies, Italy) were dipped in a buffer with 0.15 M NaCl and 0.1% Tween 20. Regions of 5 cm by 5 cm were sampled. To access the site, the ear was folded forward with one hand to expose the crease. With the other hand, the shaft of the swab was held parallel to the surface of the skin and rubbed back and forth along the retroauricular crease approximately 50 times with firm pressure for 30 s. The head of the swab was inserted into the buffer-containing tube and cut from the handle aseptically. The unsampled swabs that had not been in contact with the skin were included as negative swab samples. A total of 74 skin samples were collected on March 14 (haze days in the spring, 19 samples), April 27 (nonhaze days in the spring, 19 samples), November 29 (haze days in the winter, 18 samples), and December 30 (nonhaze days in the winter, 18 samples) in 2018. The pH, water content, and oil content of the skin were recorded on each sample day and are listed in Table S2. The AQI ( ) was used to determine the haze levels. A day with an AQI of >100 was defined as a haze day, and one with an AQI of <100 was defined as a nonhaze day. The levels of PM2.5, PM10, and AQI were recorded from the data of Xinxiang sites of the national the national program for recording urban air quality in China ( https://air.cnemc.cn:18007 ) for 7 days before sample collection ( and Table S1). DNA extraction and PCR. Three blank swabs were used to generate genomic DNA as a negative control. The swab tubes were vortexed for 10 min before DNA extraction. The DNeasy PowerSoil kit (Qiagen, Germany) was used to extract genomic DNA from all samples, including negative swab samples. The fungal internal transcribed spacer 1 (ITS1) region of the rRNA gene was amplified by PCR (95°C for 3 min, followed by 35 cycles at 95°C for 30 s, 55°C for 40 s, and 72°C for 45 s, and a final extension at 72°C for 10 min) using primers ITS5 and ITS2 ( ). Ex Taq DNA polymerase (TaKaRa, Japan) was used to conduct the PCR amplification. PCRs were performed in triplicate in 20-μL mixtures containing 4 μL of 5× Ex Taq buffer, 2 μL of 2.5 mM deoxynucleoside triphosphate (dNTP), 0.8 μL of each primer (5 μM), 0.4 μL of Ex Taq DNA polymerase, and 10 ng of template DNA. Illumina MiSeq sequencing. Amplicons were purified using the AxyPrep DNA gel extraction kit (Axygen, USA) according to the manufacturer’s instructions and quantified using QuantiFluor-ST (Promega, USA). The purified amplicons were sequenced (2 × 300 bp) on an Illumina MiSeq platform. Sequencing data processing and analyses. Quality-filtering and denoising of raw reads and the taxonomy of ITS1 sequences against the UNITE database (version 04.02.2020, http://unite.ut.ee ) ( ) were conducted using QIIME 2 ( ). PCoA and the analyses of diversities with the Shannon index, the PD whole-tree index, and observed features were performed using QIIME 2 ( ). The PD whole-tree index represents the diversity based on the phylogenetic tree ( ), and observed feature represent the richness of the fungal community. R software (version 4.0.2) was used to create the scatter, box, line, violin, bar, heatmap, and graphs, and CCA ( ). LEfSe was used to identify the taxa with significant differences between haze and nonhaze days ( ). Taxa with a log linear discriminant analysis (LDA) score of >2.5 were considered and plotted. Cultivation, isolation, and identification of Talaromyces strains. The buffer-containing swab sample was diluted six times and vortexed for 10 min before cultivation. Then, 100 μL of solution was spread-plated onto Sabouraud dextrose agar (SDA) medium (Oxoid, UK) and potato dextrose agar medium (Oxoid). To inhibit bacterial growth, the media were both supplemented with streptomycin sulfate and tetracycline (0.05 mg/mL). After incubation at 28°C for 5 days, the phenotypically distinct colonies were transferred onto corresponding agar slant. Chelex-100 was used to extract the genomic DNA of the strains. The fungal ITS regions were amplified by PCR (95°C for 5 min, followed by 35 cycles at 95°C for 30 s, 55°C for 45 s, and 72°C for 45 s, with a final extension at 72°C for 10 min) using primers ITS1 and ITS4 ( ). PCR products were purified and Sanger sequenced at Sangon Biotech (Shanghai, China). The 60 bp of the beginning ITS reads were cut off and cut into 410 bp with BioEdit software to ensure the accuracy of the sequences. To identify fungal species, the top hit from BLAST analysis and GenBank was employed. Five Talaromyces strains were obtained and frozen at −80°C in 20% glycerol. The detailed information of the strains is listed in Table S7. The sequences of the type strain Talaromyces marneffei CBS 388.87 was downloaded from NCBI GenBank (accession number NR_103671 ) and aligned using MAFFT online ( ). The phylogenetic trees were developed using MEGA software (version 7) based on the neighbor-joining method ( ). Four representative strains were selected for further study. Collection and processing of PM. Sampling was conducted using the LB-120F PM sampler (Lubo, China). PM was drawn at 100 L/min and collected on 80-mm glass fiber aerosol collection filters during haze days. The sampling filter was replaced with another one after 2 h, and 20 filters were collected. The filters were cut in half, immersed in 60 mL of deionized water, and sonicated for 30 min after sampling. The filters were immersed in another 60 mL of deionized water for another 30 min. The resulting mixed solution was centrifuged at 1,300 × g for 10 min. The supernatants were sterilized through a 0.22-μm filter, while the precipitations of PM were autoclaved. The two parts were mixed and vacuum freeze-dried overnight before being stored at −80°C. Culture of the representative Talaromyces strains and PM. The representative Talaromyces strains were inoculated onto the SDA slant and cultured for 7 days at 28°C. To prepare the spore suspension, the spores were scraped using a sterile pipette and put into 1 mL of Sabouraud dextrose broth (SDB) in the tube. The suspension was vortexed for 2 min and allowed to stand for 3 min before being transferred to another tube without big particles floating or sinking. The optical density at 570 nm (OD 570 ) of the suspension was adjusted to 0.2. Then, 10 μL of spore suspension was inoculated in 2 mL of SDB and cultured with PM at 0 mg/mL (control), 0.08 mg/mL (LC group), 0.64 mg/mL (MC group), or 5.12 mg/mL (HC group) for 5 days. Three replicates of each strain were conducted in each group. The dry weights were measured, and the ratios of HC, MC, or LC to control were used to estimate the influence of PM pollutants on the Talaromyces representative strains. Prediction by the Sloan neutral model. To explore the mechanism of changes in the skin fungal community, Sloan neutral model prediction ( ) was conducted as described previously ( , ). The migration rate (Nm) is an indication of dispersal limitation, and a higher R 2 value is an indication of fitting to the neutral model. Based on AIC scores, the three common models (neutral, binomial, and Poisson model) were compared with each other. Ethics approval and consent to participate. The research protocol for taking samples from volunteers was approved by the ethics committee of Xinxiang Medical University. Data and materials availability. The raw sequencing data have been deposited at the NCBI Sequence Read Archive (SRA) database under accession number PRJNA796697 .
This research was approved by the ethics committee of Xinxiang Medical University. Healthy students aged 18 to 28 who resided on the campus of Xinxiang Medical University (35°16′53.18″N, 113°55′37.92″E) for at least 2 years were included in this study ( ). The map was constructed using Pixel Map Generator ( https://pixelmap.amcharts.com/ ). Exclusion criteria included a history of chronic disease, use of antimicrobials or topical steroids within 6 months prior to sampling, and bathing, shampooing, or moisturizing within 24 h before sampling. The skin swab samples were collected from the skin of the retroauricular crease (behind the left and right ears). Sterile FLOQSwabs (Copan Flock Technologies, Italy) were dipped in a buffer with 0.15 M NaCl and 0.1% Tween 20. Regions of 5 cm by 5 cm were sampled. To access the site, the ear was folded forward with one hand to expose the crease. With the other hand, the shaft of the swab was held parallel to the surface of the skin and rubbed back and forth along the retroauricular crease approximately 50 times with firm pressure for 30 s. The head of the swab was inserted into the buffer-containing tube and cut from the handle aseptically. The unsampled swabs that had not been in contact with the skin were included as negative swab samples. A total of 74 skin samples were collected on March 14 (haze days in the spring, 19 samples), April 27 (nonhaze days in the spring, 19 samples), November 29 (haze days in the winter, 18 samples), and December 30 (nonhaze days in the winter, 18 samples) in 2018. The pH, water content, and oil content of the skin were recorded on each sample day and are listed in Table S2. The AQI ( ) was used to determine the haze levels. A day with an AQI of >100 was defined as a haze day, and one with an AQI of <100 was defined as a nonhaze day. The levels of PM2.5, PM10, and AQI were recorded from the data of Xinxiang sites of the national the national program for recording urban air quality in China ( https://air.cnemc.cn:18007 ) for 7 days before sample collection ( and Table S1).
Three blank swabs were used to generate genomic DNA as a negative control. The swab tubes were vortexed for 10 min before DNA extraction. The DNeasy PowerSoil kit (Qiagen, Germany) was used to extract genomic DNA from all samples, including negative swab samples. The fungal internal transcribed spacer 1 (ITS1) region of the rRNA gene was amplified by PCR (95°C for 3 min, followed by 35 cycles at 95°C for 30 s, 55°C for 40 s, and 72°C for 45 s, and a final extension at 72°C for 10 min) using primers ITS5 and ITS2 ( ). Ex Taq DNA polymerase (TaKaRa, Japan) was used to conduct the PCR amplification. PCRs were performed in triplicate in 20-μL mixtures containing 4 μL of 5× Ex Taq buffer, 2 μL of 2.5 mM deoxynucleoside triphosphate (dNTP), 0.8 μL of each primer (5 μM), 0.4 μL of Ex Taq DNA polymerase, and 10 ng of template DNA.
Amplicons were purified using the AxyPrep DNA gel extraction kit (Axygen, USA) according to the manufacturer’s instructions and quantified using QuantiFluor-ST (Promega, USA). The purified amplicons were sequenced (2 × 300 bp) on an Illumina MiSeq platform.
Quality-filtering and denoising of raw reads and the taxonomy of ITS1 sequences against the UNITE database (version 04.02.2020, http://unite.ut.ee ) ( ) were conducted using QIIME 2 ( ). PCoA and the analyses of diversities with the Shannon index, the PD whole-tree index, and observed features were performed using QIIME 2 ( ). The PD whole-tree index represents the diversity based on the phylogenetic tree ( ), and observed feature represent the richness of the fungal community. R software (version 4.0.2) was used to create the scatter, box, line, violin, bar, heatmap, and graphs, and CCA ( ). LEfSe was used to identify the taxa with significant differences between haze and nonhaze days ( ). Taxa with a log linear discriminant analysis (LDA) score of >2.5 were considered and plotted.
Talaromyces strains. The buffer-containing swab sample was diluted six times and vortexed for 10 min before cultivation. Then, 100 μL of solution was spread-plated onto Sabouraud dextrose agar (SDA) medium (Oxoid, UK) and potato dextrose agar medium (Oxoid). To inhibit bacterial growth, the media were both supplemented with streptomycin sulfate and tetracycline (0.05 mg/mL). After incubation at 28°C for 5 days, the phenotypically distinct colonies were transferred onto corresponding agar slant. Chelex-100 was used to extract the genomic DNA of the strains. The fungal ITS regions were amplified by PCR (95°C for 5 min, followed by 35 cycles at 95°C for 30 s, 55°C for 45 s, and 72°C for 45 s, with a final extension at 72°C for 10 min) using primers ITS1 and ITS4 ( ). PCR products were purified and Sanger sequenced at Sangon Biotech (Shanghai, China). The 60 bp of the beginning ITS reads were cut off and cut into 410 bp with BioEdit software to ensure the accuracy of the sequences. To identify fungal species, the top hit from BLAST analysis and GenBank was employed. Five Talaromyces strains were obtained and frozen at −80°C in 20% glycerol. The detailed information of the strains is listed in Table S7. The sequences of the type strain Talaromyces marneffei CBS 388.87 was downloaded from NCBI GenBank (accession number NR_103671 ) and aligned using MAFFT online ( ). The phylogenetic trees were developed using MEGA software (version 7) based on the neighbor-joining method ( ). Four representative strains were selected for further study.
Sampling was conducted using the LB-120F PM sampler (Lubo, China). PM was drawn at 100 L/min and collected on 80-mm glass fiber aerosol collection filters during haze days. The sampling filter was replaced with another one after 2 h, and 20 filters were collected. The filters were cut in half, immersed in 60 mL of deionized water, and sonicated for 30 min after sampling. The filters were immersed in another 60 mL of deionized water for another 30 min. The resulting mixed solution was centrifuged at 1,300 × g for 10 min. The supernatants were sterilized through a 0.22-μm filter, while the precipitations of PM were autoclaved. The two parts were mixed and vacuum freeze-dried overnight before being stored at −80°C.
Talaromyces strains and PM. The representative Talaromyces strains were inoculated onto the SDA slant and cultured for 7 days at 28°C. To prepare the spore suspension, the spores were scraped using a sterile pipette and put into 1 mL of Sabouraud dextrose broth (SDB) in the tube. The suspension was vortexed for 2 min and allowed to stand for 3 min before being transferred to another tube without big particles floating or sinking. The optical density at 570 nm (OD 570 ) of the suspension was adjusted to 0.2. Then, 10 μL of spore suspension was inoculated in 2 mL of SDB and cultured with PM at 0 mg/mL (control), 0.08 mg/mL (LC group), 0.64 mg/mL (MC group), or 5.12 mg/mL (HC group) for 5 days. Three replicates of each strain were conducted in each group. The dry weights were measured, and the ratios of HC, MC, or LC to control were used to estimate the influence of PM pollutants on the Talaromyces representative strains.
To explore the mechanism of changes in the skin fungal community, Sloan neutral model prediction ( ) was conducted as described previously ( , ). The migration rate (Nm) is an indication of dispersal limitation, and a higher R 2 value is an indication of fitting to the neutral model. Based on AIC scores, the three common models (neutral, binomial, and Poisson model) were compared with each other.
The research protocol for taking samples from volunteers was approved by the ethics committee of Xinxiang Medical University.
The raw sequencing data have been deposited at the NCBI Sequence Read Archive (SRA) database under accession number PRJNA796697 .
Reviewer comments
|
Effects of COVID-19 pandemic on stress level of residents and fellows during ophthalmology training | 3be7d110-f287-49b7-8fb6-ab240e0eb7ae | 9359222 | Ophthalmology[mh] | The present study was approved by Ethics Committee for Human Research and comprised a 42-questions survey elaborated by the authors and addressed the impact of COVID-19 on the mental health of residents and fellows in ophthalmology. The questionnaire was divided into two sections: The first section contained demographic characteristics of the participants (8 questions) and their working conditions before and during the COVID-19 pandemic (24 questions), and the second part of the questionnaire consisted of Perceived Stress Scale (PSS-10), a 10-question and validated survey. The PSS-10 was developed with the objective of estimate the stress level of each participant with questions about negatives and positives feelings and ability or inability of dealing with stressful situations. Half of the items reflect positively and the other half negatively and five options of answer, each of them with a punctuation that varies from 0 to 4, being: 0 = never, 1 = almost never, 2 = sometimes, 3 = almost always, and 4 = always. Positive questions are calculated inverted (0 = 4, 1 = 3, 2 = 2, 3 = 1, 4 = 0) and higher the sum, greater the stress. The score value obtained from the PSS-10 can range from 0 to 40 points. The higher the level of stress, the higher the score. The participants of the survey were Brazilian residents and fellows in ophthalmology identified by the professional network of the authors including members of their national ophthalmological societies. The online survey was sent by email and WhatsApp (WhatsApp LLC, Menlo Park, California, USA) with a link to access the survey on Google Forms (Google LLC, Mountain View, California, USA). Complete questionnaires were received between June 22 and July 9 of 2020. Responders could answer only once to the survey and the participation was voluntary, anonymous, and offered no financial support. Categorical variables were recorded as absolute and relative frequencies. Comparisons between “before” and “during” the pandemic periods were performed, using the McNemar test for categorical data. Qui-square tests (mainly the Pearson Qui-square test) were applied to categorical data to assess how likely it is that any observed differences could happen at random. Comparison between sex was performed using the Mann-Whitney Test. The Spearman Correlation Test was used to evaluate the association between stress and workload in the emergency room working. The comparation between level of professional training and stress was analyzed using the Kruskal Wallis Test. A P value <0.05 was considered statistically significant. Statistical analysis was performed using the Statistical Package for Social Sciences software version 25.0 (IBM, Armonk, NY).
A total of 271 physicians responded to the survey: 100 (36.9%) fellows in ophthalmology and 171 (63.1%) residents in Ophthalmology. The study sample had representation of all years of residency and fellowship training. The mean age of the participants was 28.9 ± 2.7 (23.0–40.0) years and the study included a higher number of women (73.1%). Most participants were single (67.9%) or married (21.0%). Of the participants, 243 (89.7%) did not have children. According to the geographical distribution of participants, 33 (12.2%) professionals were doing their training in the Midwest, 119 (43.9%) in the north/northeast, and 119 (43.9%) in the south/southeast of Brazil . There was no significant difference in the level of stress detected in our work using the PSS-10 between the groups of fellows (average score 22.5) and residents (average score 22.1). Residents had a higher workload then fellows before the pandemic ( P > 0.001) . During the pandemic, a reduction and equalization of the workload of residents and fellows was noted ( P = 0.195) . No correlation was identified between the workload and stress level of professionals during the pandemic ( P = 0.760). The year of professional training did not influence the level of stress ( P = 0.376) , nor the age of fellows and residents ( P = 0.857 and P = 0.643, respectively) . Higher levels of stress were identified in female participants ( P = 0.001) . Before the pandemic, 131 residents (76.6%) and 91 fellows (91.0%) were performing the surgical training. During the pandemic, 54 residents (31.6%) and 44 fellows (44.0%) were still on surgical training ( P < 0.001). The stress level of residents and fellows during the COVID-19 pandemic that had their surgical training interrupted was significantly higher ( P < 0.001) .
The pandemic has had a negative impact on the mental health of individuals from the most diverse sectors of society, but health workers have the aggravating factor of dealing directly with the infected and are at a greater risk of contracting COVID-19. Data from the World Health Organization (WHO) and the Ministry of Health indicate that 570,000 health workers have been infected with COVID-19 in the Americas, and as of September 2020, the WHO has reported 2,500 health worker deaths on the continent. As of December 26, 2020, 435,872 cases of COVID-19 were confirmed in healthcare professionals in Brazil and 2,736 of these cases were reported as Severe Acute Respiratory Syndrome (SARS), of which they were hospitalized and 441 died. COVID-19 was the cause in 86.6% of these deaths. Although conjunctivitis is an uncommon presentation of COVID-19, infected or suspected patients may be admitted or referred to ophthalmology clinics, which increases the possibility of exposure of ophthalmologists to COVID-19. Routine eye exams, such as slit lamp examination, and treatment procedures that are within the range of aerosol transmission, are performed by close contact. In addition, the short distance between ophthalmologist and patient during direct ophthalmoscopy increases their risk of contracting COVID-19 given their exposure to contaminated ocular secretions. Thus, when trying to cope with this stressful global event, ophthalmologists may also develop psychiatric disorders. Katsurayama et al . showed that medical residency training alone can promote psychological suffering to residents due to an interaction between sleep deprivation, social deprivation, and individual vulnerability. It was also observed that residents of university hospitals were more stressed and depressed than residents of non-university hospitals, which suggests that services with a predominantly academic profile require more and, consequently, doctors have a higher level of stress. Firth-Cozens reported in their work that first-year residents were considered more vulnerable to stress due to less ability in the adaptive process, thus composing a risk group. In addition, Khanna et al . assessed depression in ophthalmologists from India during confinement by COVID-19 using the Patient Health Questionnaire-9 (PHQ-9), a self-report measure used to assess the severity of depression over the previous 2 weeks and found that depression was significantly greater in younger ophthalmologists. However, our results show no significant differences between residents and fellows with respect to year of study or age. The high level of stress observed in female residents and fellows in the present study corroborates with Frank et al . that reported sex-related issues such as stress related to multiple functions, including managing domestic activities and professional life. Additionally, Grover et al . showed that female ophthalmologists present higher levels of stress, depression, and anxiety compared to male ophthalmologists. As expected, the present study showed that residents had an increased workload before the pandemic compared to fellows but during the pandemic the working hours for both residents and fellows dropped, and the workload became equivalent. However, it is noteworthy that the interruption of surgical training has impacted the stress level of residents and fellows. Our finding corroborates with Gondim et al . that showed a workload reduction of 45.6% in Brazilian ophthalmology residency programs and concluded that all medical residency services were affected by the pandemic. Moreover, our findings draw attention to the impact that COVID-19 may have on the quality of training of future ophthalmologists given the possible gaps in knowledge and training that this pandemic may lead to and that are challenging to be remedied. It is undeniable that the COVID-19 pandemic had a negative effect on ophthalmology training. The real challenge is to quantify this impact considering the complexity of training and the multiple factors involved. Thus, our study aimed to evaluate the impact of the pandemic based on the physiological status of residents and fellows. Our study revealed that the pandemic resulted in an increased level of stress among female ophthalmology residents and fellows, as well as among those physicians that had their surgical training suspended during the pandemic. The present study highlights the stress level of these professionals that may be caused by the insecurity and uncertainties raised by the interruption of proper practical activities and surgical training. The results also alert the medical community to the need for specific psychological support for these professionals facing a life-threatening scenario and highlights the need for measures to mitigate the loss of in-person clinical and surgical training.
As a consequence of the COVID-19 pandemic, the workload of residents and fellows decreased. Residents and fellows experienced similar levels of stress during this period. However, higher levels of stress were observed among female trainees and those that had their surgical training interrupted during the pandemic. Ethics committee approval Altino Ventura Foundation (Protocol number: 3.784.484). Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.
Altino Ventura Foundation (Protocol number: 3.784.484).
Nil.
There are no conflicts of interest.
|
Evaluating the stability of nursery-established arbuscular mycorrhizal fungal associations in apple rootstocks | 90d1a3a3-ce76-45c1-bf3d-223d18d75f23 | 11784189 | Microbiology[mh] | Endophytic microorganisms living inside plant tissues often play very specific and crucial roles in promoting the health and growth of their host plant. For example, most plants in the legume family (i.e., Fabaceae) recruit nitrogen-fixing bacteria (rhizobia) to live inside nodules on their roots . Like rhizobia, arbuscular mycorrhizal fungi (AMF) belonging to the phylum Glomeromycota are another group of soilborne endosymbionts generally considered to be beneficial . AMF form diverse relationships with a host of land plants (roughly 80% of all species) . This speaks to the importance of these fungi in the health and functioning of entire terrestrial ecosystems, including agroecosystems. Mycorrhizal hyphal strands (3–7 μm) are considerably smaller in diameter than fine roots (<1–2 mm) or even root hairs (5–20 μm) , enabling them to extend into otherwise inaccessible soil patches. This allows AMF to scavenge resources necessary for plant growth including: phosphorous , nitrogen , and water . In addition to improving plant resource availability, AMF improve soil structure , carbon content , and hydraulic properties while occupying a niche in the root that might otherwise be filled by disease-causing soilborne organisms . Colonization by mycorrhizal fungi has also been shown to reduce pest and pathogen pressure above ground in a number of crops (including apple) . Although mycorrhizal associations have been shown to improve productivity/yield in a diversity of cropping systems , including fruit tree orchards , structuring AMF–plant relationships to improve productivity and sustainability in agroecosystems is inherently difficult. This is because soil tillage , soil compaction , high soil phosphorous content, and the use of systemic fungicides can all negatively impact AMF colonization rates . As a result, little information exists on the survival/effectiveness of AMF inoculants in the soil and their impacts on native or pre-established AMF communities in root tissue. Within the limited body of research that has been done, results are often conflicting, and inoculated isolates are not always distinguished from native AMF. A review of several studies using the inoculant Rhizophagus irregularis illustrates the problems. For example, in a study by Renaut et al. inoculation of field plots after sowing corn, soybean, or wheat seeds did not alter the composition of the indigenous AMF community in any of the crops . Native and inoculated strains of R. irregularis could not be distinguished . In a different study by Buysens et al. , R. irregularis inoculant (applied at planting) was successfully traced in the field . Compared with native R. irregularis strains, the inoculant was detected at low levels in very few plants. In another study, however, inoculation at sowing with a commercially available strain of R. irregularis reduced the diversity of indigenous AMF detected in pea roots relative to non-inoculated controls. Here, the commercial inoculant appeared to be most effective at displacing closely related taxa (e.g., indigenous Rhizophagus spp.) . Similarly, in yet another study, pre-inoculation of plants with R. irregularis was found to suppress colonization by native AMF, especially those of the same species . Finally, in the study by Islam et al. , differences in AMF community structure between inoculated vs control treatments lessened over time (i.e., three growing seasons), suggesting that introduced AMF taxa may not persist long term among locally adapted AMF communities . In addition to environmental compatibility (i.e., the ability of a given taxon to thrive in the habitat in question), AMF community assembly is shaped by a number of other factors. In apple, there is evidence suggesting a selective capacity of rootstock genotype on the fungal endophytic microbiome . Certain rootstock genotypes may be more effectively colonized by or compatible with particular AMF species than with others . The current study was specifically designed to assess the ability of a known community of AMF with a limited number of species (i.e., the AMF inoculant) to compete with pre-existing/nursery-derived AMF contained in apple roots (without the confounding factor of site). To further investigate the selective capacity of rootstock genotype on the AMF community, a variety of commercially available apple rootstock genotypes (G.890, G.935, M.7, and M.26) were used. It was hypothesized that the commercial AMF inoculant would survive and differentially colonize new root tissue depending on the apple rootstock genotype utilized, altering pre-established/nursery-derived AMF communities. In our investigation, rootstocks were cultivated in pasteurized potting soil with or without the commercial AMF consortium. After 1 month, plants were harvested, and new root tissue was collected for DNA extraction. The subsequent changes to the nursery-derived AMF community (established prior to planting) were then characterized using the Glomeromycota-specific primer set AML1/2, which targets the V3–V4–V5 regions of the nuclear 18S rRNA gene . It should be noted that the nuclear ribosomal internal transcribed spacer (ITS) region is the molecular marker used by the scientific community to identify fungi. Although this region can accurately detect AMF, it is not suitable for lower-level taxonomic discrimination . Instead, the nuclear rDNA region (in particular the 18S rRNA gene) is widely used for characterizing AMF diversity. Over the last few decades, the available nuclear rDNA sequence data have grown considerably, enabling molecular studies of AMF communities. At present, however, there are no 18S rRNA databases for molecular identification of AMF that are actively curated. Therefore, as part of this study, a phylogenetic tree for Phylum Glomeromycota was constructed using over 140 high-quality (i.e., from well-identified AMF cultures) 18S rRNA gene sequences . This phylogenetic tree (which included 91 different AMF species from 24 different genera) was used to accurately assign high throughput sequencing data to the species level. Selection and preparation of inoculum The multi-species AMF formulation used in this study (Mycorrhizal Applications; Grants Pass, OR) was desirable because, compared with other commercially available mixtures, it was purported to contain a diverse AMF consortium, including a number of ecologically relevant colonizers previously associated with apple. According to the manufacturer, the product contained the following five mycorrhizal fungal species: Glomus clarum (Order Glomerales); Glomus monosporum (Order Glomerales); Septoglomus deserticola , formerly Glomus deserticola (Order Glomerales); Paraglomus brasilianum (Order Archaeosporales); and Gigaspora margarita (Order Gigasporaceae), 132 ppg/g. The experiment was conducted using pasteurized potting soil (Sunshine Professional Growing Mix #1; Sun Gro Horticulture, Abbotsford, Canada). Soil was pasteurized at 80°C for 2 × 8-h cycles (with a cool-down/aseptic mixing step between cycles). The five-species consortia were hand-mixed into pasteurized potting soil prior to planting at the recommended field rate per the manufacturer’s instructions (500–600 propagules/tree; 4.5 g/tree). This application rate is referred to as “1×” in the figures and text below. Experimental design and planting Rootstocks used in this study included G935, G.890, M.7, and M.26; 0.635 cm trunk diameter. All rootstocks were obtained in a single shipment from TRECO (Woodburn, OR, USA). For each rootstock × treatment combination, five replicates were planted into 2.7-L pots containing pasteurized potting mix with or without the AMF inoculum. Pots were arranged in a completely randomized design on the greenhouse bench. Immediately prior to planting, a small amount of fine root tissue (~2 g) was collected from various locations on the root system in order to obtain a representative sample of the “nursery-derived” AMF community. Root tissue was stored at −80°C until processing. Initial trunk diameter (cm) was measured at planting using a caliper. There were 40 experimental plants in total + additional “sentinel” plants for the different rootstock genotypes. Sentinel plants were used to verify that a sufficient amount of new root tissue had been produced before harvesting the experiment. Rootstocks were grown for a period of 4 weeks at 22°C–28°C under supplemental lighting to maintain a 16-h photoperiod and watered as needed; plants did not receive supplemental nutrients. At 4 weeks post-planting, the rootstocks were removed from pots, and new (white) root tissue was collected from each plant for DNA isolation. To avoid sampling root tissue that may have been present at the nursery, only white roots were collected. Plant growth characteristics were also measured at this time and included: trunk diameter (cm), total plant biomass (g), root volume (mL), and leader shoot length (cm). DNA extraction from root tissue Prior to DNA extraction, all root tissues were thoroughly rinsed in tap water to remove any residual soil particles/debris and blotted dry. Cleaned roots were then ground to a fine powder in liquid nitrogen by using a sterile mortar/pestle. Because the focus of this study is on the root-associated AMF community, roots were not pre-sterilized (as is done for endophytes). In other words, DNA was extracted from root surface + endosphere. Ground root tissue was preserved at −80°C until molecular analysis. The DNeasy Plant Pro Kit (Qiagen, Valencia, CA, USA) was used according to the manufacturer’s instructions to isolate DNA from 50 mg of ground root tissue per plant. DNA extraction from soil Because pasteurized potting mix represents an ablated soil microbiome (with no AMF species present), it was assumed that the commercial inoculum would be the main source of “new” AMF. In order to confirm the absence of AMF in pasteurized potting soil (prior to adding inoculum/planting), DNA was extracted from 3 × 10 g of pasteurized potting soil using the DNeasy PowerMax Soil Kit (Qiagen, Valencia, CA, USA). These samples were sequenced as described below. DNA extraction from commercial AMF inoculum DNA was also directly extracted and sequenced from the commercial AMF mixture in order to confirm the species composition. AMF spores and hyphae were extracted from 125 mL of the granular formulation via wet sieving (500, 250, and 45 μm metal sieves) followed by sucrose-density gradient centrifugation (using 20% and 60% sucrose) as described by INVAM ( https://invam.ku.edu/spore-extraction ). DNA was extracted from the concentrated spore sample using the DNeasy Plant Pro Kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. Two samples from separate DNA extractions were sequenced. 18S amplification and DNA sequencing Prior to sequencing, all DNA samples were quantified using a Qubit Fluorometer (Thermo Scientific, Waltham, MA, USA), and sample integrity was verified as described by Van Horn et al. . Briefly, samples were subjected to PCR amplification of the fungal internal transcribed spacer (ITS) rRNA gene, followed by visualization on a 2% agarose gel. Amplicons were generated using the Glomeromycotan-specific primers AML1 (5′-ATC AAC TTT CGA TGG TAG GAT AGA-3′) and AML2 (5′-GAA CCC AAA CAC TTT GGT TTC C-3′) and sequenced on a PacBio Sequel instrument (average reads per sample = 10,000) using the circular consensus sequencing (CCS) mode. PacBio “long-read” sequencing technology was used because the AML1/2 primer set amplifies a relatively long section of the 18S (SSU) rRNA (amplicon size = ~800 bp). The CCS was used to generate high fidelity consensus sequences by correcting the stochastic errors generated in each round of sequencing. Sequencing results were obtained from the sequencing facility (Molecular Research, Shallowater, TX, USA), demultiplexed in house and then pre-processed using the DADA2 (v1.26.0) pipeline, which included quality filtering, trimming, and dereplicating . Dereplicated sequences were then used for sequencing error estimation, and the error model was used in denoising and chimera removal . The amount and percentage of reads that passed each step are listed in Table S1. The DADA2 algorithm was then used to assign amplicon sequence variants (ASVs) to the processed reads and to summarize the results into an ASV table in which each row represents an ASV and the number of ASVs observed in each sample (column) are listed. It is worth mentioning that many AMF species have a high-level of intra-genomic heterogeneity of ribosomal sequences (i.e., each spore can have dozens of sequence variants). This complicates the ability to accurately determine the number of species based on clustering sequences . In this study, CCS was used to generate ASVs differing by as little as one nucleotide, an approach that is appropriate for not lumping closely related species together. Raw sequencing data are located in the NCBI Sequence Read Archive (SRA) under NCBI BioProject ID PRJNA1124126 . Relative abundance calculation Prior to analysis of ASV read counts, data were edited to remove singletons and doubletons as well as non-Glomeromycotan ASVs. The relative abundance of each ASV was calculated per the total reads in each sample (Table S2). Heatmaps were generated using the python Seaborn heatmap package with average abundance of the biological replicates as input. Phylogenetic analysis and taxonomy assignment of AMF ASVs Before performing phylogenetic reconstruction analyses, a total of 174 well-curated Glomeromycota 18S (SSU) rRNA reference sequences (which included 91 different species from 24 different genera) were gleaned from two different studies (summarized in Supplementary Files S1 and S2, respectively). AMF ASVs from this study were integrated with 1) all 174 sequences from the two reference sets or 2) with only the reference sequences from Krüger et al. , creating two sets of sequences for downstream analyses. Analysis of our AMF ASV with only the Stefani et al. data set was not performed due to the small number of reference sequences (number of sequences = 28) provided by Stefani et al. . Next, nucleotide sequence alignments of the two sets of sequences were executed using the GeneFamilyAligner tool from PlantTribes2 with the MAFFT algorithm. Sequences beyond the AML1/2 primer binding region were trimmed from the alignment using Geneious (v. 9.0.5) . Maximum likelihood phylogenetic trees were computed using the command line version of IQtree2 with automated substitution model selection enabled, 2,000 ultra-fast bootstrapping and bnni refinement selected. Paraglomus was used as an outgroup as it represents the most basal glomeromycotan branch. Phylogenetic trees inferred from the two sequences sets have similar topology; however, the tree with sequences from both reference papers has an overall lower bootstrap support, likely due to a few sequences with divergent nucleotides in regions highly conserved in other sequences. As a result, the tree inferred with our AMF ASV plus reference sequences from Krüger et al., was selected for downstream analysis (Fig. S1: full tree file). Based on information from Krüger et al. , this tree included 21 sequences from type/ex-type cultures. The tree supported five major clades (some are at order level, some family level). The simplified version of the tree showing these five clades was created using FigTree v.1.4.4 . Interestingly, an ASV clade with long branch length was observed in the phylogenetic tree. Further investigation into the sequence alignment showed that those ASVs contain extra nucleotides not shared with sequences outside this clade (Fig. S2A and B). In order to test whether long branches were caused by regions that did not align with other sequences, different alignment trimming stringencies were tested: removal of sites with 90% + gaps, 50% + gaps, 10% + gaps, or 0% + gap (in which any sites with gaps are removed). Additional phylogenetic trees were inferred using the same method described above with the trimmed alignments. The trimming did not affect the tree topology or branch length (except the most rigorously trimmed tree; 0% + gap). The bootstraps were also stable. However, the tree with the highest overall bootstrap value did not receive any post-alignment trimming and was selected for taxonomy assignment (Fig. S1). Taxonomy was assigned to ASVs by identifying monophyletic groups housing ASV and curated reference sequences. For instance, ASVs 177, 178, 179, and 166 were annotated as Funneliformis coronatus because they were found in the same monophyletic group with F. coronatus reference sequences; ASV 174 was annotated only to the genus level, Rhizophagus , as it fell between different Rhizophagus species from the reference. Statistical analyses Significant differences between nursery-derived AMF communities and those existing after planting were assessed using relative abundance data via One-way analysis of similarities (ANOSIM) with the Bray–Curtis dissimilarity coefficient. For each rootstock genotype, significant differences in growth characteristics between AMF-inoculated and non-inoculated control treatments were assessed using Two-way analysis of variance (ANOVA) (with rootstock genotype and treatment as factors) and means were compared by Tukey’s multiple comparisons test ( P ≥ 0.05). Growth data (increase in trunk diameter, shoot biomass, and root volume) were transformed prior to analysis (y = log(y)) and tested for normality. All transformed data sets passed Anderson–Darling, D’Agostino and Pearson, Shapiro–Wilk, and Kolmogorov–Smirnov normality tests. For each rootstock genotype, Mann–Whitney tests (Holm–Sidal method) were used to check for significant differences in root:shoot biomass ratios at harvest. Significant differences in the amount of AMF DNA detected in root tissue (as estimated from qPCR of total fungal DNA) were assessed using Two-way ANOVA followed by Tukey’s multiple comparisons test. Quantitative estimation of AMF in root tissue In this study, because only frozen root tissue was available, the percentage of AMF colonization could not be directly evaluated via microscopy. Instead, AMF abundance was assessed in relation to the total amount of fungi present via a combination of methods. First, the fungal community was sequenced using the universal ITS1f (5′ CTTGGTCATTTAGAGGAAGTAA )/ITS2r (5′- GCTGCGTTCTTCATCGATGC ) primer pair. Briefly, DNA extracted from root surface + endosphere (as described in Section 2.3) was sent to the sequencing facility (Molecular Research, Shallowater TX, USA) where it was PCR-amplified (prior to library preparation) and sequenced using an Illumina NovaSeq platform (20,000 reads per sample). Paired-end sequences were joined, and those <150 bp or with ambiguous base calls were removed. Sequences were quality filtered using a maximum expected error threshold of 1.0, dereplicated, and denoised. Final ASVs were taxonomically classified using BLASTn against a curated database derived from NCBI ( www.ncbi.nlm.nih.gov ). Prior to ASV read count analysis, data were edited to remove singletons and doubletons as well as any non-fungal reads. The abundance of Glomeromycota reads relative to total fungal reads in each sample was then calculated. The absolute amount of fungal DNA present in these samples was also measured using a QuantStudio3 Real-Time PCR System with the NSI1 (5′-GAT TGA ATG GCT TAG TGA GG) and 5.8S (5′-CGC TGC GTT CTT CAT CG) primer pair. Run conditions were performed as described in Somera et al. . Purified genomic DNA from Illonectria robusta (isolate # 14–264) was used to generate the standard curve with a dilution range from 0.01 to 100 pg µL−1. All reactions were performed in triplicate, and each 96-well plate included a no-template control. Relative percentages of AMF were then transformed into absolute values using fungal DNA quantities. The multi-species AMF formulation used in this study (Mycorrhizal Applications; Grants Pass, OR) was desirable because, compared with other commercially available mixtures, it was purported to contain a diverse AMF consortium, including a number of ecologically relevant colonizers previously associated with apple. According to the manufacturer, the product contained the following five mycorrhizal fungal species: Glomus clarum (Order Glomerales); Glomus monosporum (Order Glomerales); Septoglomus deserticola , formerly Glomus deserticola (Order Glomerales); Paraglomus brasilianum (Order Archaeosporales); and Gigaspora margarita (Order Gigasporaceae), 132 ppg/g. The experiment was conducted using pasteurized potting soil (Sunshine Professional Growing Mix #1; Sun Gro Horticulture, Abbotsford, Canada). Soil was pasteurized at 80°C for 2 × 8-h cycles (with a cool-down/aseptic mixing step between cycles). The five-species consortia were hand-mixed into pasteurized potting soil prior to planting at the recommended field rate per the manufacturer’s instructions (500–600 propagules/tree; 4.5 g/tree). This application rate is referred to as “1×” in the figures and text below. Rootstocks used in this study included G935, G.890, M.7, and M.26; 0.635 cm trunk diameter. All rootstocks were obtained in a single shipment from TRECO (Woodburn, OR, USA). For each rootstock × treatment combination, five replicates were planted into 2.7-L pots containing pasteurized potting mix with or without the AMF inoculum. Pots were arranged in a completely randomized design on the greenhouse bench. Immediately prior to planting, a small amount of fine root tissue (~2 g) was collected from various locations on the root system in order to obtain a representative sample of the “nursery-derived” AMF community. Root tissue was stored at −80°C until processing. Initial trunk diameter (cm) was measured at planting using a caliper. There were 40 experimental plants in total + additional “sentinel” plants for the different rootstock genotypes. Sentinel plants were used to verify that a sufficient amount of new root tissue had been produced before harvesting the experiment. Rootstocks were grown for a period of 4 weeks at 22°C–28°C under supplemental lighting to maintain a 16-h photoperiod and watered as needed; plants did not receive supplemental nutrients. At 4 weeks post-planting, the rootstocks were removed from pots, and new (white) root tissue was collected from each plant for DNA isolation. To avoid sampling root tissue that may have been present at the nursery, only white roots were collected. Plant growth characteristics were also measured at this time and included: trunk diameter (cm), total plant biomass (g), root volume (mL), and leader shoot length (cm). Prior to DNA extraction, all root tissues were thoroughly rinsed in tap water to remove any residual soil particles/debris and blotted dry. Cleaned roots were then ground to a fine powder in liquid nitrogen by using a sterile mortar/pestle. Because the focus of this study is on the root-associated AMF community, roots were not pre-sterilized (as is done for endophytes). In other words, DNA was extracted from root surface + endosphere. Ground root tissue was preserved at −80°C until molecular analysis. The DNeasy Plant Pro Kit (Qiagen, Valencia, CA, USA) was used according to the manufacturer’s instructions to isolate DNA from 50 mg of ground root tissue per plant. Because pasteurized potting mix represents an ablated soil microbiome (with no AMF species present), it was assumed that the commercial inoculum would be the main source of “new” AMF. In order to confirm the absence of AMF in pasteurized potting soil (prior to adding inoculum/planting), DNA was extracted from 3 × 10 g of pasteurized potting soil using the DNeasy PowerMax Soil Kit (Qiagen, Valencia, CA, USA). These samples were sequenced as described below. DNA was also directly extracted and sequenced from the commercial AMF mixture in order to confirm the species composition. AMF spores and hyphae were extracted from 125 mL of the granular formulation via wet sieving (500, 250, and 45 μm metal sieves) followed by sucrose-density gradient centrifugation (using 20% and 60% sucrose) as described by INVAM ( https://invam.ku.edu/spore-extraction ). DNA was extracted from the concentrated spore sample using the DNeasy Plant Pro Kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. Two samples from separate DNA extractions were sequenced. Prior to sequencing, all DNA samples were quantified using a Qubit Fluorometer (Thermo Scientific, Waltham, MA, USA), and sample integrity was verified as described by Van Horn et al. . Briefly, samples were subjected to PCR amplification of the fungal internal transcribed spacer (ITS) rRNA gene, followed by visualization on a 2% agarose gel. Amplicons were generated using the Glomeromycotan-specific primers AML1 (5′-ATC AAC TTT CGA TGG TAG GAT AGA-3′) and AML2 (5′-GAA CCC AAA CAC TTT GGT TTC C-3′) and sequenced on a PacBio Sequel instrument (average reads per sample = 10,000) using the circular consensus sequencing (CCS) mode. PacBio “long-read” sequencing technology was used because the AML1/2 primer set amplifies a relatively long section of the 18S (SSU) rRNA (amplicon size = ~800 bp). The CCS was used to generate high fidelity consensus sequences by correcting the stochastic errors generated in each round of sequencing. Sequencing results were obtained from the sequencing facility (Molecular Research, Shallowater, TX, USA), demultiplexed in house and then pre-processed using the DADA2 (v1.26.0) pipeline, which included quality filtering, trimming, and dereplicating . Dereplicated sequences were then used for sequencing error estimation, and the error model was used in denoising and chimera removal . The amount and percentage of reads that passed each step are listed in Table S1. The DADA2 algorithm was then used to assign amplicon sequence variants (ASVs) to the processed reads and to summarize the results into an ASV table in which each row represents an ASV and the number of ASVs observed in each sample (column) are listed. It is worth mentioning that many AMF species have a high-level of intra-genomic heterogeneity of ribosomal sequences (i.e., each spore can have dozens of sequence variants). This complicates the ability to accurately determine the number of species based on clustering sequences . In this study, CCS was used to generate ASVs differing by as little as one nucleotide, an approach that is appropriate for not lumping closely related species together. Raw sequencing data are located in the NCBI Sequence Read Archive (SRA) under NCBI BioProject ID PRJNA1124126 . Prior to analysis of ASV read counts, data were edited to remove singletons and doubletons as well as non-Glomeromycotan ASVs. The relative abundance of each ASV was calculated per the total reads in each sample (Table S2). Heatmaps were generated using the python Seaborn heatmap package with average abundance of the biological replicates as input. Before performing phylogenetic reconstruction analyses, a total of 174 well-curated Glomeromycota 18S (SSU) rRNA reference sequences (which included 91 different species from 24 different genera) were gleaned from two different studies (summarized in Supplementary Files S1 and S2, respectively). AMF ASVs from this study were integrated with 1) all 174 sequences from the two reference sets or 2) with only the reference sequences from Krüger et al. , creating two sets of sequences for downstream analyses. Analysis of our AMF ASV with only the Stefani et al. data set was not performed due to the small number of reference sequences (number of sequences = 28) provided by Stefani et al. . Next, nucleotide sequence alignments of the two sets of sequences were executed using the GeneFamilyAligner tool from PlantTribes2 with the MAFFT algorithm. Sequences beyond the AML1/2 primer binding region were trimmed from the alignment using Geneious (v. 9.0.5) . Maximum likelihood phylogenetic trees were computed using the command line version of IQtree2 with automated substitution model selection enabled, 2,000 ultra-fast bootstrapping and bnni refinement selected. Paraglomus was used as an outgroup as it represents the most basal glomeromycotan branch. Phylogenetic trees inferred from the two sequences sets have similar topology; however, the tree with sequences from both reference papers has an overall lower bootstrap support, likely due to a few sequences with divergent nucleotides in regions highly conserved in other sequences. As a result, the tree inferred with our AMF ASV plus reference sequences from Krüger et al., was selected for downstream analysis (Fig. S1: full tree file). Based on information from Krüger et al. , this tree included 21 sequences from type/ex-type cultures. The tree supported five major clades (some are at order level, some family level). The simplified version of the tree showing these five clades was created using FigTree v.1.4.4 . Interestingly, an ASV clade with long branch length was observed in the phylogenetic tree. Further investigation into the sequence alignment showed that those ASVs contain extra nucleotides not shared with sequences outside this clade (Fig. S2A and B). In order to test whether long branches were caused by regions that did not align with other sequences, different alignment trimming stringencies were tested: removal of sites with 90% + gaps, 50% + gaps, 10% + gaps, or 0% + gap (in which any sites with gaps are removed). Additional phylogenetic trees were inferred using the same method described above with the trimmed alignments. The trimming did not affect the tree topology or branch length (except the most rigorously trimmed tree; 0% + gap). The bootstraps were also stable. However, the tree with the highest overall bootstrap value did not receive any post-alignment trimming and was selected for taxonomy assignment (Fig. S1). Taxonomy was assigned to ASVs by identifying monophyletic groups housing ASV and curated reference sequences. For instance, ASVs 177, 178, 179, and 166 were annotated as Funneliformis coronatus because they were found in the same monophyletic group with F. coronatus reference sequences; ASV 174 was annotated only to the genus level, Rhizophagus , as it fell between different Rhizophagus species from the reference. Significant differences between nursery-derived AMF communities and those existing after planting were assessed using relative abundance data via One-way analysis of similarities (ANOSIM) with the Bray–Curtis dissimilarity coefficient. For each rootstock genotype, significant differences in growth characteristics between AMF-inoculated and non-inoculated control treatments were assessed using Two-way analysis of variance (ANOVA) (with rootstock genotype and treatment as factors) and means were compared by Tukey’s multiple comparisons test ( P ≥ 0.05). Growth data (increase in trunk diameter, shoot biomass, and root volume) were transformed prior to analysis (y = log(y)) and tested for normality. All transformed data sets passed Anderson–Darling, D’Agostino and Pearson, Shapiro–Wilk, and Kolmogorov–Smirnov normality tests. For each rootstock genotype, Mann–Whitney tests (Holm–Sidal method) were used to check for significant differences in root:shoot biomass ratios at harvest. Significant differences in the amount of AMF DNA detected in root tissue (as estimated from qPCR of total fungal DNA) were assessed using Two-way ANOVA followed by Tukey’s multiple comparisons test. In this study, because only frozen root tissue was available, the percentage of AMF colonization could not be directly evaluated via microscopy. Instead, AMF abundance was assessed in relation to the total amount of fungi present via a combination of methods. First, the fungal community was sequenced using the universal ITS1f (5′ CTTGGTCATTTAGAGGAAGTAA )/ITS2r (5′- GCTGCGTTCTTCATCGATGC ) primer pair. Briefly, DNA extracted from root surface + endosphere (as described in Section 2.3) was sent to the sequencing facility (Molecular Research, Shallowater TX, USA) where it was PCR-amplified (prior to library preparation) and sequenced using an Illumina NovaSeq platform (20,000 reads per sample). Paired-end sequences were joined, and those <150 bp or with ambiguous base calls were removed. Sequences were quality filtered using a maximum expected error threshold of 1.0, dereplicated, and denoised. Final ASVs were taxonomically classified using BLASTn against a curated database derived from NCBI ( www.ncbi.nlm.nih.gov ). Prior to ASV read count analysis, data were edited to remove singletons and doubletons as well as any non-fungal reads. The abundance of Glomeromycota reads relative to total fungal reads in each sample was then calculated. The absolute amount of fungal DNA present in these samples was also measured using a QuantStudio3 Real-Time PCR System with the NSI1 (5′-GAT TGA ATG GCT TAG TGA GG) and 5.8S (5′-CGC TGC GTT CTT CAT CG) primer pair. Run conditions were performed as described in Somera et al. . Purified genomic DNA from Illonectria robusta (isolate # 14–264) was used to generate the standard curve with a dilution range from 0.01 to 100 pg µL−1. All reactions were performed in triplicate, and each 96-well plate included a no-template control. Relative percentages of AMF were then transformed into absolute values using fungal DNA quantities. Overview of PacBio sequencing statistics A total of 640,087 AMF 18S rRNA sequence reads were obtained after cleaning/de-noising. The use of PacBio sequencing technology allowed for the generation of relatively long sequence reads of high quality, with sufficient information content for phylogenetic analysis (which is not the case with shorter amplicons; <300 bp). In this study, median read length ranged from 794 to 1,053 bp (Fig. S3). Rarefaction analyses indicated that the sequencing depth (10,000 reads per sample) sufficiently captured AMF diversity in the roots; the curves reached the asymptote around a depth of 7,000 (Fig. S4). A total of 180 AMF amplicon sequence variants (ASVs) was detected across all experimental samples. Phylogeny-based assessment of AMF community composition in apple rootstocks In many studies, AMF taxonomy is determined using the AMF-specific reference sequence database MaarjAM ( https://maarjam.ut.ee/ ) . However, as noted above, this database is no longer actively curated. In this study, tree-based (phylogenetic) taxonomic assignments of sequence data were used to assess AMF community structure in apple root tissue pre- and post-planting. The phylogenetic tree in this study (Fig. S1) shares the same major clades with trees presented in both references (Fig. S5), namely studies by Krüger et al. and Stefani et al. . Our tree shared a more similar clade-to-clade relationship to the Krüger tree and resolved with high-confidence some unresolved relationships in the Kruger tree. Tree-based taxonomy assignment showed that experimental ASVs represented taxa from a handful of genera, including Rhizophagus , Funneliformis , Claroideoglomus, and Ambispora/Archaeospora . Phylogenetic analysis was also consistent with previous studies in which R. irregularis and R. intraradices are well separated . Pre-planting (nursery-derived) AMF No Glomermycota spp. were identified in pasteurized soil, indicating that AMF were not present in the growth media prior to planting. All rootstocks were obtained from the same nursery location; however, prior to planting, significant differences in AMF community composition were identified between M.26 and M.7 ( P = 0.004), M.26 and G.935 ( P = 0.02), and M.7 and G.890 ( P = 0.04) . The AMF taxa detected in apple roots at this time likely represented those naturally occurring in nursery orchard soil and included those in clade 1 ( Rhizophagus spp.), clade 3 (unknown), and clade 5 (Archaeosporales), but not clade 2 ( Funneliformis spp.) or 4 ( Claroideoglomus spp.) . Of these taxa, Rhizophagus spp. (including R. irregularis, R. fasciculatus , and R. vesiculiferus ) represented 91%, 89%, and 86% of the reads in G.935, M.7, and G.890, respectively, but only 8% of the reads in M.26 . In M.26, the nursery-derived AMF community was largely dominated by ASVs in clades 3 (40%) and 5 (50%) . Phylogenetic analysis showed that clade 3 represented a unique, well-supported, monophyletic clade of unknown taxonomy ( ; Fig. S1). ASVs belonging to clade 3 were also detected in G.935 pre-planting, but at a much lower relative abundance (2%) than in M.26 . Comparison of ASV sequences with those deposited in the NCBI 18S rRNA sequence (SSU) database (blastn) identified F. mosseae as the closest match (≤92% identity for ASVs 23, 85, 19, 49, 41, and 139); however, it is clear that clade 3 represents a distinct phylogenetic group with unique sequences that are not shared with the sequences from the literature in the alignment . It should also be noted that, upon closer inspection, the high quality of the sequence data strengthens our confidence that these are unique sequences rather than artifacts. ASVs belonging to clade 5 shared the highest similarity with Ambispora and Archaeospora (which represent sister taxa in our tree; Fig. S1) but contained extra, unique nucleotides in the sequence alignment (Fig. S2B). Upon close inspection, it was determined that these were not artifacts, so we did not trim these regions. Therefore, clade 5 ASVs formed a divergent group of sequences residing on long branches . The large number of ASVs in clade 5 (Order Archaeosporales) likely represents a limited (albeit undetermined) number of AMF species. ASVs from clade 5 were present in all rootstock genotypes prior to planting but represented a much smaller percentage in G890 (14%), M7 (11%), and G.935 (7%) than in M.26 (50%). AMF detected in the commercial inoculum differed greatly from the product description A total of 25 AMF amplicon sequence variants (ASVs) representing three different genera were detected in the DNA directly extracted from the commercial mixture. Phylogeny-based classification revealed that all AMF sequences detected in the CM samples either branched from clade 2 ( Funneliformis/Glomus; 93%), clade 4 ( Claroideoglomus ; 6%), or clade 1 ( Rhizophagus; 1%) (Fig. S1). In the CM samples, clade 1 was represented by R. irregularis (ASV7) and Rhizophagus sp. (ASV 174). ASVs belonging to clade 2 were most closely related to either Funneliformis mosseae or Funneliformis coronatus . One of the AMF species purported to be contained in the commercial mixture was F. monosporum (= Glomus monosporum , F. mosseae; https://invam.ku.edu/mosseae ). Therefore, tree-based taxonomic assignments provided evidence that F. mosseae was in fact present in the mixture. As noted above, this AMF species was only detected in the commercial inoculum and did not appear to successfully colonize apple roots. This result was unexpected considering that Funneliformis species, including F. mosseae , have been identified in the apple orchards of Washington State and Italy . No other AMF species, which were purported to be in the commercial product, were identified . It should be noted that Rhizophagus clarum was purported to be in the commercial mixture, but this species was not detected in the CM samples. Overall, the commercial mixture contained a very different consortium of AMF than expected. Apple rootstocks established relationships with introduced AMF in a genotype-specific manner ASVs in clade 2 or 4 were not detected pre-planting or in plants cultivated in the no-AMF control soils and were considered to represent AMF introduced by the commercial mixture. Claroideoglomus species (clade 4) identified in the commercial inoculum successfully colonized both Malling rootstocks. For example, ASV 32, most closely related to C. luteum / C. claroideum, was detected in the CM and also in M.26 and M.7 (1× CM) samples . In addition, ASVs 79 and 96 (most closely related to C. claroideum ) were present in the commercial mixture and in M.26 (1× CM) samples . Together, these three ASVs represented 5% and 3.5% of the AMF community in M.26 and M.7 (1× CM) treatments, respectively. By comparison, no Claroideoglomus ASVs were detected in the Geneva rootstocks used in this study (G.890 and G.935). It should, however, be noted that rootstock/AMF association patterns do not necessarily differentiate according to rootstock type (e.g., Geneva vs Malling). Introduced AMF may disproportionately affect low-abundance AMF taxa Apple rootstock genotype appears to be an important factor influencing the assembly of AM fungal communities. In this study, the commercial mixture appeared to have a particularly large effect on M.26, especially its ability to maintain pre-established associations with Rhizophagus spp. Overall, Rhizophagus spp. only represented 10% of the initial (pre-established) AMF community in M.26. Rhizophagus spp. became more prominent when M.26 was cultivated in non-inoculated soil (50% relative abundance), but disappeared completely in the presence of the commercial inoculum . One of the most dominant taxa, ASV 1 ( R. irregularis or R. vesciculiferous ) , was present in the nursery-derived root tissue of G.935 (48%), M.7 (45%), and G.890 (14%), but was undetectable in M.26 prior to planting. The relative abundance of ASV1 increased to 11.5% when M.26 was cultivated in no-AMF control soil, but was not detected in 1× CM treatments. This result suggests that ASV1 was present in M.26 root tissue prior to planting but only increased to detectable levels after planting in the absence of the commercial inoculum. In G.935 and M.7 rootstocks, however, the relative abundance of ASV1 remained relatively high (≥30%), regardless of treatment (Control or 1× CM), illustrating the stability of these particular AMF–rootstock associations. It is also interesting to note that, according to one-way ANOSIM analysis, G.935 and M.7 did not differ significantly in AMF community composition in any treatment , a result which, in part, was due to their continued association with ASV 1. Taken together, these results suggest that M.26 may have been more susceptible to the loss of clade one taxa than other rootstock genotypes due to low initial (pre-plant) abundance. Along these same lines, ASV2 ( R. fasciculatus ) was the most abundant AMF taxa (34%) identified in G.890 pre-planting, and this association remained relatively stable after planting (23%–26%), regardless of treatment. ASV2 was also detected in M.26 (7%) and M.7 pre-planting (3%), but these associations were not maintained in the presence of the commercial inoculum. By comparison, the relative abundance of ASV two in M.26 increased to 25% in non-inoculated control soil. One-way ANOSIM analysis also reflected the differential responses by M.26 and G.890 to the 1× CM treatment (but not to the control) . The colonization of M.26 by introduced Claroideoglomus spp. may have contributed to the suppression/exclusion of resident Rhizophagus taxa, even though the abundance of Claroideoglomus spp. in M.26 was relatively low (5%) at the time of sampling. In M.7, colonization by Claroideoglomus sp. (ASV 32; 3.5%) may have also contributed to a slight reduction in the relative abundance of Rhizophagus spp. (no-AMF control = 62%; 1× CM = 53%). It is also worth noting that antagonistic interactions between R. irregularis and C. claroideum have been previously documented . Introduced AMF can influence resident AMF community structure regardless of colonization success This finding is best is illustrated using clade 5 (Archaeosporales). Within this clade, ASVs 3, 4, and 11 (all terminal branches connected to a single node) most likely represent sequence variation within a single species (Fig. S1). These ASVs were detected in root tissue prior to planting in all rootstock genotypes (~1%–9%) and became enriched in all rootstock genotypes following cultivation in the no-AMF control soil (11%–35% relative abundance) . In comparison, this same group of ASVs (3, 4, and 11) became suppressed in both Geneva rootstocks following cultivation in inoculated soil. This was surprising considering that no ASVs associated with the CM were detected in the Geneva rootstocks used in this study. In G.935, ASVs 3, 4, and 11 persisted in the presence of the commercial inoculum at lower relative abundance (5%) than in the control soil (11%). In G.890, however, not a single clade 5 ASV was detected in the 1× CM treatment, even though this group represented 21% of the AMF community when cultivated in the no-AMF control soil. The complete loss of clade 5 taxa from G.890 in 1× CM samples resulted in an increased dominance of clade 1 ( Rhizophagus spp.), which comprised 100% of the AMF community . This result suggests that less efficient AMF can alter or even displace ecologically relevant (indigenous) AMF communities living in symbiosis with the host plant, even if colonization by the introduced inoculant is not effective. Although the commercial mixture appeared to negatively affect the ability of Geneva rootstocks to maintain relationships with indigenous ASVs in clade 5, it had a positive effect on their relative abundance in both Malling rootstocks. In M.26 and M.7, ASVs 3, 4, and 11 greatly increased in the presence of the commercial inoculum, representing 63% and 41% of the total reads . This may be one reason why ordination analysis of relative abundance data indicated that although AMF communities in M.26 and M.7 were significantly different pre-planting ( P = 0.004), they did not differ in the 1× CM treatment . This result suggests that the introduced inoculant stimulated distantly related resident AMF species, an effect that is occasionally observed in studies with commercial inoculants . That said, the effects of the AMF inoculant on resident AMF in clade 5 were highly varied in M.26. Although ASVs 3, 4, and 11 increased in dominance in the 1× CM treatment (as described above), other clade 5 taxa disappeared. For example, a group of closely related sequences likely to represent a single species (ASVs 14, 28, and 16) accounted for 34% of the total AMF community when M.26 was cultivated in no-AMF control soil, but was not detected in the presence of the commercial inoculum (1× M.26). This result highlights the differential effects of the commercial inoculum on closely related AMF species within a particular rootstock. It is also worth noting that treatment was a significant source of variation affecting the absolute abundance of AMF, (Two-way ANOVA; P = 0.0001), which was generally higher prior to planting than after planting (Fig. S6). This reduction was most likely due to high levels of soil phosphorus (186 mg/kg) in potting soil. However, in all rootstock genotypes (except G.890), AMF abundance was further reduced in the presence of the commercial inoculant. Challenges associated with fine-scale differentiation between AMF species It is important to note that genetically different nuclei can coexist within individual AMF spores (i.e., there is high within-spore variation in rRNA gene sequences) . In general, intraspecific sequence variability for Rhizophagus (and R. irregularis in particular) is known to be relatively high . In this study, 27 different ASVs were associated with only three species of Rhizophagus ( ; clade 1). In Glomeromycotan taxa with high intrasporal rRNA sequence variation, sequences from different spores (or even different soil samples) can sometimes be more similar than sequences of the same spore or soil sample . In this study, ASV 7 ( R. irregularis; clade 1) was not detected in Malling or Geneva rootstocks prior to planting but became enriched in Geneva rootstocks in both 1× and no-AMF control treatments; 1× G.890 (13%), 1× G.935 (12%), C.G890 (15%), and C.G935 (5%). ASV 7 was also detected in the commercial mixture. However, even though ASV or OTU-delimiting approaches yield the most biologically relevant taxonomic units for understanding AMF community composition, fine-scale, intrasporal genetic variation makes it difficult to tell whether ASV 7 originated solely from the commercial mixture. Plant growth responses The effect of the commercial AMF inoculant on resident AMF communities was also reflected in plant growth data. No significant differences in trunk diameter or shoot biomass were identified between inoculated and no-AMF control treatments. However, the root volume of M.7 plants cultivated in the inoculated (1× CM) soil was significantly reduced relative to those cultivated in the control treatment ( P = 0.0008). The cause of the reduced root volume in M.7 plants cultivated in the 1× CM treatment is not clear as colonization of M.7 plants by the commercial inoculant Claroideoglomus sp. (ASV 32; 3.5%) did not appear to displace resident taxa or largely impact AMF community structure relative to no-AMF controls. ITS-based sequencing indicated that the fungal replant pathogens Ilyonectria robusta and Rhizoctonia solani (anastomosis groups unknown) were present in the root tissue of all rootstock genotypes, regardless of treatment. However, no significant differences in the relative abundance of either organism were identified between any of the rootstock genotypes in any treatment (one-way ANOVA; P < 0.05). It is possible that the commercial AMF consortia differentially impacted resource allocation relative to no-AMF controls in M.7; shoot:root ratios were close to being significantly different ( ; p adj = 0.07; P = 0.01). Interactions between AMF and host plants have been shown to cause shifts in partitioning of biomass between shoots and roots in other studies . A total of 640,087 AMF 18S rRNA sequence reads were obtained after cleaning/de-noising. The use of PacBio sequencing technology allowed for the generation of relatively long sequence reads of high quality, with sufficient information content for phylogenetic analysis (which is not the case with shorter amplicons; <300 bp). In this study, median read length ranged from 794 to 1,053 bp (Fig. S3). Rarefaction analyses indicated that the sequencing depth (10,000 reads per sample) sufficiently captured AMF diversity in the roots; the curves reached the asymptote around a depth of 7,000 (Fig. S4). A total of 180 AMF amplicon sequence variants (ASVs) was detected across all experimental samples. In many studies, AMF taxonomy is determined using the AMF-specific reference sequence database MaarjAM ( https://maarjam.ut.ee/ ) . However, as noted above, this database is no longer actively curated. In this study, tree-based (phylogenetic) taxonomic assignments of sequence data were used to assess AMF community structure in apple root tissue pre- and post-planting. The phylogenetic tree in this study (Fig. S1) shares the same major clades with trees presented in both references (Fig. S5), namely studies by Krüger et al. and Stefani et al. . Our tree shared a more similar clade-to-clade relationship to the Krüger tree and resolved with high-confidence some unresolved relationships in the Kruger tree. Tree-based taxonomy assignment showed that experimental ASVs represented taxa from a handful of genera, including Rhizophagus , Funneliformis , Claroideoglomus, and Ambispora/Archaeospora . Phylogenetic analysis was also consistent with previous studies in which R. irregularis and R. intraradices are well separated . No Glomermycota spp. were identified in pasteurized soil, indicating that AMF were not present in the growth media prior to planting. All rootstocks were obtained from the same nursery location; however, prior to planting, significant differences in AMF community composition were identified between M.26 and M.7 ( P = 0.004), M.26 and G.935 ( P = 0.02), and M.7 and G.890 ( P = 0.04) . The AMF taxa detected in apple roots at this time likely represented those naturally occurring in nursery orchard soil and included those in clade 1 ( Rhizophagus spp.), clade 3 (unknown), and clade 5 (Archaeosporales), but not clade 2 ( Funneliformis spp.) or 4 ( Claroideoglomus spp.) . Of these taxa, Rhizophagus spp. (including R. irregularis, R. fasciculatus , and R. vesiculiferus ) represented 91%, 89%, and 86% of the reads in G.935, M.7, and G.890, respectively, but only 8% of the reads in M.26 . In M.26, the nursery-derived AMF community was largely dominated by ASVs in clades 3 (40%) and 5 (50%) . Phylogenetic analysis showed that clade 3 represented a unique, well-supported, monophyletic clade of unknown taxonomy ( ; Fig. S1). ASVs belonging to clade 3 were also detected in G.935 pre-planting, but at a much lower relative abundance (2%) than in M.26 . Comparison of ASV sequences with those deposited in the NCBI 18S rRNA sequence (SSU) database (blastn) identified F. mosseae as the closest match (≤92% identity for ASVs 23, 85, 19, 49, 41, and 139); however, it is clear that clade 3 represents a distinct phylogenetic group with unique sequences that are not shared with the sequences from the literature in the alignment . It should also be noted that, upon closer inspection, the high quality of the sequence data strengthens our confidence that these are unique sequences rather than artifacts. ASVs belonging to clade 5 shared the highest similarity with Ambispora and Archaeospora (which represent sister taxa in our tree; Fig. S1) but contained extra, unique nucleotides in the sequence alignment (Fig. S2B). Upon close inspection, it was determined that these were not artifacts, so we did not trim these regions. Therefore, clade 5 ASVs formed a divergent group of sequences residing on long branches . The large number of ASVs in clade 5 (Order Archaeosporales) likely represents a limited (albeit undetermined) number of AMF species. ASVs from clade 5 were present in all rootstock genotypes prior to planting but represented a much smaller percentage in G890 (14%), M7 (11%), and G.935 (7%) than in M.26 (50%). A total of 25 AMF amplicon sequence variants (ASVs) representing three different genera were detected in the DNA directly extracted from the commercial mixture. Phylogeny-based classification revealed that all AMF sequences detected in the CM samples either branched from clade 2 ( Funneliformis/Glomus; 93%), clade 4 ( Claroideoglomus ; 6%), or clade 1 ( Rhizophagus; 1%) (Fig. S1). In the CM samples, clade 1 was represented by R. irregularis (ASV7) and Rhizophagus sp. (ASV 174). ASVs belonging to clade 2 were most closely related to either Funneliformis mosseae or Funneliformis coronatus . One of the AMF species purported to be contained in the commercial mixture was F. monosporum (= Glomus monosporum , F. mosseae; https://invam.ku.edu/mosseae ). Therefore, tree-based taxonomic assignments provided evidence that F. mosseae was in fact present in the mixture. As noted above, this AMF species was only detected in the commercial inoculum and did not appear to successfully colonize apple roots. This result was unexpected considering that Funneliformis species, including F. mosseae , have been identified in the apple orchards of Washington State and Italy . No other AMF species, which were purported to be in the commercial product, were identified . It should be noted that Rhizophagus clarum was purported to be in the commercial mixture, but this species was not detected in the CM samples. Overall, the commercial mixture contained a very different consortium of AMF than expected. ASVs in clade 2 or 4 were not detected pre-planting or in plants cultivated in the no-AMF control soils and were considered to represent AMF introduced by the commercial mixture. Claroideoglomus species (clade 4) identified in the commercial inoculum successfully colonized both Malling rootstocks. For example, ASV 32, most closely related to C. luteum / C. claroideum, was detected in the CM and also in M.26 and M.7 (1× CM) samples . In addition, ASVs 79 and 96 (most closely related to C. claroideum ) were present in the commercial mixture and in M.26 (1× CM) samples . Together, these three ASVs represented 5% and 3.5% of the AMF community in M.26 and M.7 (1× CM) treatments, respectively. By comparison, no Claroideoglomus ASVs were detected in the Geneva rootstocks used in this study (G.890 and G.935). It should, however, be noted that rootstock/AMF association patterns do not necessarily differentiate according to rootstock type (e.g., Geneva vs Malling). Apple rootstock genotype appears to be an important factor influencing the assembly of AM fungal communities. In this study, the commercial mixture appeared to have a particularly large effect on M.26, especially its ability to maintain pre-established associations with Rhizophagus spp. Overall, Rhizophagus spp. only represented 10% of the initial (pre-established) AMF community in M.26. Rhizophagus spp. became more prominent when M.26 was cultivated in non-inoculated soil (50% relative abundance), but disappeared completely in the presence of the commercial inoculum . One of the most dominant taxa, ASV 1 ( R. irregularis or R. vesciculiferous ) , was present in the nursery-derived root tissue of G.935 (48%), M.7 (45%), and G.890 (14%), but was undetectable in M.26 prior to planting. The relative abundance of ASV1 increased to 11.5% when M.26 was cultivated in no-AMF control soil, but was not detected in 1× CM treatments. This result suggests that ASV1 was present in M.26 root tissue prior to planting but only increased to detectable levels after planting in the absence of the commercial inoculum. In G.935 and M.7 rootstocks, however, the relative abundance of ASV1 remained relatively high (≥30%), regardless of treatment (Control or 1× CM), illustrating the stability of these particular AMF–rootstock associations. It is also interesting to note that, according to one-way ANOSIM analysis, G.935 and M.7 did not differ significantly in AMF community composition in any treatment , a result which, in part, was due to their continued association with ASV 1. Taken together, these results suggest that M.26 may have been more susceptible to the loss of clade one taxa than other rootstock genotypes due to low initial (pre-plant) abundance. Along these same lines, ASV2 ( R. fasciculatus ) was the most abundant AMF taxa (34%) identified in G.890 pre-planting, and this association remained relatively stable after planting (23%–26%), regardless of treatment. ASV2 was also detected in M.26 (7%) and M.7 pre-planting (3%), but these associations were not maintained in the presence of the commercial inoculum. By comparison, the relative abundance of ASV two in M.26 increased to 25% in non-inoculated control soil. One-way ANOSIM analysis also reflected the differential responses by M.26 and G.890 to the 1× CM treatment (but not to the control) . The colonization of M.26 by introduced Claroideoglomus spp. may have contributed to the suppression/exclusion of resident Rhizophagus taxa, even though the abundance of Claroideoglomus spp. in M.26 was relatively low (5%) at the time of sampling. In M.7, colonization by Claroideoglomus sp. (ASV 32; 3.5%) may have also contributed to a slight reduction in the relative abundance of Rhizophagus spp. (no-AMF control = 62%; 1× CM = 53%). It is also worth noting that antagonistic interactions between R. irregularis and C. claroideum have been previously documented . This finding is best is illustrated using clade 5 (Archaeosporales). Within this clade, ASVs 3, 4, and 11 (all terminal branches connected to a single node) most likely represent sequence variation within a single species (Fig. S1). These ASVs were detected in root tissue prior to planting in all rootstock genotypes (~1%–9%) and became enriched in all rootstock genotypes following cultivation in the no-AMF control soil (11%–35% relative abundance) . In comparison, this same group of ASVs (3, 4, and 11) became suppressed in both Geneva rootstocks following cultivation in inoculated soil. This was surprising considering that no ASVs associated with the CM were detected in the Geneva rootstocks used in this study. In G.935, ASVs 3, 4, and 11 persisted in the presence of the commercial inoculum at lower relative abundance (5%) than in the control soil (11%). In G.890, however, not a single clade 5 ASV was detected in the 1× CM treatment, even though this group represented 21% of the AMF community when cultivated in the no-AMF control soil. The complete loss of clade 5 taxa from G.890 in 1× CM samples resulted in an increased dominance of clade 1 ( Rhizophagus spp.), which comprised 100% of the AMF community . This result suggests that less efficient AMF can alter or even displace ecologically relevant (indigenous) AMF communities living in symbiosis with the host plant, even if colonization by the introduced inoculant is not effective. Although the commercial mixture appeared to negatively affect the ability of Geneva rootstocks to maintain relationships with indigenous ASVs in clade 5, it had a positive effect on their relative abundance in both Malling rootstocks. In M.26 and M.7, ASVs 3, 4, and 11 greatly increased in the presence of the commercial inoculum, representing 63% and 41% of the total reads . This may be one reason why ordination analysis of relative abundance data indicated that although AMF communities in M.26 and M.7 were significantly different pre-planting ( P = 0.004), they did not differ in the 1× CM treatment . This result suggests that the introduced inoculant stimulated distantly related resident AMF species, an effect that is occasionally observed in studies with commercial inoculants . That said, the effects of the AMF inoculant on resident AMF in clade 5 were highly varied in M.26. Although ASVs 3, 4, and 11 increased in dominance in the 1× CM treatment (as described above), other clade 5 taxa disappeared. For example, a group of closely related sequences likely to represent a single species (ASVs 14, 28, and 16) accounted for 34% of the total AMF community when M.26 was cultivated in no-AMF control soil, but was not detected in the presence of the commercial inoculum (1× M.26). This result highlights the differential effects of the commercial inoculum on closely related AMF species within a particular rootstock. It is also worth noting that treatment was a significant source of variation affecting the absolute abundance of AMF, (Two-way ANOVA; P = 0.0001), which was generally higher prior to planting than after planting (Fig. S6). This reduction was most likely due to high levels of soil phosphorus (186 mg/kg) in potting soil. However, in all rootstock genotypes (except G.890), AMF abundance was further reduced in the presence of the commercial inoculant. It is important to note that genetically different nuclei can coexist within individual AMF spores (i.e., there is high within-spore variation in rRNA gene sequences) . In general, intraspecific sequence variability for Rhizophagus (and R. irregularis in particular) is known to be relatively high . In this study, 27 different ASVs were associated with only three species of Rhizophagus ( ; clade 1). In Glomeromycotan taxa with high intrasporal rRNA sequence variation, sequences from different spores (or even different soil samples) can sometimes be more similar than sequences of the same spore or soil sample . In this study, ASV 7 ( R. irregularis; clade 1) was not detected in Malling or Geneva rootstocks prior to planting but became enriched in Geneva rootstocks in both 1× and no-AMF control treatments; 1× G.890 (13%), 1× G.935 (12%), C.G890 (15%), and C.G935 (5%). ASV 7 was also detected in the commercial mixture. However, even though ASV or OTU-delimiting approaches yield the most biologically relevant taxonomic units for understanding AMF community composition, fine-scale, intrasporal genetic variation makes it difficult to tell whether ASV 7 originated solely from the commercial mixture. The effect of the commercial AMF inoculant on resident AMF communities was also reflected in plant growth data. No significant differences in trunk diameter or shoot biomass were identified between inoculated and no-AMF control treatments. However, the root volume of M.7 plants cultivated in the inoculated (1× CM) soil was significantly reduced relative to those cultivated in the control treatment ( P = 0.0008). The cause of the reduced root volume in M.7 plants cultivated in the 1× CM treatment is not clear as colonization of M.7 plants by the commercial inoculant Claroideoglomus sp. (ASV 32; 3.5%) did not appear to displace resident taxa or largely impact AMF community structure relative to no-AMF controls. ITS-based sequencing indicated that the fungal replant pathogens Ilyonectria robusta and Rhizoctonia solani (anastomosis groups unknown) were present in the root tissue of all rootstock genotypes, regardless of treatment. However, no significant differences in the relative abundance of either organism were identified between any of the rootstock genotypes in any treatment (one-way ANOVA; P < 0.05). It is possible that the commercial AMF consortia differentially impacted resource allocation relative to no-AMF controls in M.7; shoot:root ratios were close to being significantly different ( ; p adj = 0.07; P = 0.01). Interactions between AMF and host plants have been shown to cause shifts in partitioning of biomass between shoots and roots in other studies . Over the past few years, interest in the use of products of microbial origin to promote sustainability in agricultural ecosystems has been growing. This includes interest in AMF inoculants, particularly in light of the increasing emphasis on soil health. The development of systematic and sustainable approaches for engineering the phytobiome as a means to improve crop production has even been referred to as “the next green revolution.” As part of this movement, the global agricultural biologicals market has seen rapid growth . Despite this enthusiasm, however, a lack of understanding surrounding the modes of action of these products and the associated increases in plant yield or performance persists . As referenced previously, the product used in this experiment was specifically selected because the manufacturer indicated it contained a diverse mix of AMF species. However, the product did not contain what it was purported to contain. This highlights the problem regarding the identification of AMF species identities contained in commercially available products. While this is concerning, it is not surprising, given the issues associated with the use of web-based databases for determining AMF identity (in our study, it was not essential to the experiment that the inoculum consist of the particular species purported to be contained in the mixture). Obtaining accurate taxonomic assignments of DNA sequences from Phylum Glomeromycota is challenging . The standard “barcode” for molecular identification of fungi (the nuclear ribosomal internal transcribed spacer) does not provide adequate resolution of AMF taxa. Instead, the scientific community generally sequences the V3–V4 regions of the 18S rRNA gene . At present, there are no actively curated (i.e., up to date) web-based 18S rRNA databases. Therefore, in order to describe arbuscular mycorrhizal fungal communities at the genus and species-level, sequence data must be placed onto an AMF-specific phylogenetic tree. This is an area of research that needs attention because it has important ramifications for the successful application of AMF inoculants. The use of phylogenetic-based tools by product developers and industry practitioners would ensure quality-control in terms of AMF product composition/consistency. Although taxonomic assignment can be greatly improved using phylogenetics, it should be noted that accurate identification of AMF based on nuclear ribosomal DNA remains a challenge. The available molecular data (i.e., rDNA reference sequences with sufficient length) is not yet complete in terms of the diversity of taxon coverage. Furthermore, less than half of the 237 currently described AMF species are propagated by INVAM ( https://invam.ku.edu/species-diversity ). This is partly due to the obligate biotrophic nature of AMF as well as their morphological plasticity, traits which make characterizing biological material from described species difficult. Due to the paucity of data on the interactions of commercial AMF inoculants with established indigenous AMF communities, one of the primary goals of this study was to establish whether or not apple rootstocks serve as a significant source of AMF inoculum from the nursery where they are produced. In general, pre-established taxa would be expected to be favored over introduced species due to priority effects . Thus, we hypothesized that nursery-derived AMF would strongly influence AMF community structure after planting and limit effective colonization by mycorrhizal inoculants. In order to further explore the selective capacity of host genotype on the root-associated AMF community, a diverse group of apple rootstock genotypes was included. The resulting data showed that the ability of a commercial inoculant to effectively alter and/or compete with pre-existing AMF largely depends on the AMF species/rootstock present. In this study, the Claroideoglomus species ( C. claroideum and C. luteum ) contained in the commercial mixture were able to successfully compete with the resident AMF populations contained in Malling but not Geneva rootstocks. Therefore, the ability of an apple plant to maintain its resident AMF community likely depends on rootstock-specific filters, which influence the composition of the species pool. The findings from this study also suggest that strength of the priority effect may depend on the relative abundance of the resident species. For example, although Rhizophagus disappeared completely when M.26 was cultivated in inoculated soil, the group only represented only 10% of the community to begin with (compared to >86% in all other rootstocks). Low-abundant AMF taxa were also found to be prone to disappearance in a study that assessed the effects of agricultural management intensity levels on AMF community assembly . Rare taxa play pivotal roles in maintaining the stability of plant-associated microbiomes. In fact, Xiong et al. showed that “hub” species of fungal co-occurrence networks in rhizosphere soil and root tissue were primarily comprised of rare taxa, including AMF . Rare taxa also contribute to community stability by acting as a reservoir that can rapidly respond to environmental change . Hence, commercially available AMF products have the potential to alter the resident AMF community in ways that negatively impact the entire plant-associated microbiome. Regardless of rootstock genotype, nursery-established AMF communities did appear to strongly limit the plant’s ability to interact with introduced AMF propagules, suggesting that rootstocks serve as a significant source of AMF inoculum from nurseries where they are produced. For example, as previously mentioned, F. mosseae is considered to have a high aptitude for colonization with apple. Therefore, it was surprising that the introduced Funneliformis spp. were unable to infect any of the plants within the experimental timeframe (4 weeks) , especially considering that Funneliformis spp. (clade 2) represented 93% of the AMF contained in the inoculum, while clade 4 ( Claroideoglomus ) represented a much lower percentage (6%). In addition, a number of different rootstock/AMF interactions that were established pre-planting persisted, regardless of treatment. These included G.935/M.7 × R. irregularis/R. vesciculiferous, G.890 × Rhizophagus fasciculatus as well as ASVs 3, 4, and 11 in M.26 and M.7. These results highlight the potential for nursery-established AMF associations to be maintained when transplanted into the field. On the other hand, the data showed that AMF inoculants do not necessarily need to outcompete the existing resident AM fungal community to have an effect on resident AMF communities living in symbiosis with the host plant and/or plant growth. For example, although AMF present in the CM did not effectively colonize the Geneva rootstocks used in this study, the effect of inoculation on AMF community dynamics was evidenced in other ways. The most dramatic example was the complete loss of clade 5 from G.890, even though this group increased from 14% to 21% of the AMF community when cultivated in the no-AMF control soil. These results are mirrored by another study in which inoculated AMF taxa failed to colonize maize plant roots but altered species dominance and reduced diversity of the pre-existing AMF community . Predicting AMF community dynamics in apple rootstocks is complex (even in an AMF-free growth medium). In conventional nursery systems, rootstocks are typically propagated in stoolbeds (a process by which new shoots from a “mother tree” develop roots). In this way, relationships with AMF are naturally established and nursery-derived material is never “clean” to begin with. More recently, plant tissue culture has become a major propagation tool for fruit tree rootstocks and is expected to become the standard going forward as techniques continue to improve. This approach to rootstock propagation may present new opportunities to inoculate “AMF-free” plant material with ecologically relevant AMF species. In a study by Calvet et al., micropropagated “AMF-free” peach rootstocks inoculated with a mixed community of Glomus spp. prior to transplantation into orchard soil significantly reduced nematode populations in roots relative to non-inoculated plants . Manipulating AMF community re-assembly post-fumigation is another potential path forward. Broad spectrum soil fumigation significantly reduces pathogen activity (and improves tree growth) but also greatly depresses the entire soil microbiome (including native AMF communities). The loss of the resident microbial population impairs the ability of the soil to defend against re-infestation. However, if applied to recently fumigated soil, AMF taxa (e.g., Claroideoglomus spp.) could potentially prevent the re-infestation of later-arriving pathogenic organisms via priority effects. In a study by Resendes et al., once an apple root was colonized by AMF, it was no longer accessible to a potential pathogen (and vice versa ) . Therefore, in the context of orchard/nursery management, introduced AMF must be capable of establishing in new root tissue in a short period of time in order to compete with soilborne root pathogens (and other indigenous microbes already present). Moreover, the fastest AMF colonizers are also often the most extensive . Since pasteurized soil is a proxy for fumigation, the results of the current study suggest that commercial AMF inoculants applied to recently fumigated soil can become established in new root tissue alongside AMF, which are pre-established at the nursery. That said, our study also highlights the importance of considering the dependency of AMF community structure in regard to rootstock genotype. It is not clear why the Claroideoglomus spp. contained in the commercial inoculum were unable to successfully colonize the Geneva rootstocks evaluated in this study. The differential colonization of rootstock genotypes by AMF contained in the commercial inoculum provides evidence for the selective capacity of host genotype on the plant-associated AMF microbiome. Few studies have documented the selective capacity of apple rootstock genotype on the endophytic microbiome, especially AMF assemblages . In the study by Cook et al., both apple rootstock genotype and AMF species were found to be significant sources of variation affecting the percentage of colonization . However, significant interaction effects between these two factors were not identified. Among the AMF tested, C. etunicatum and R. irregularis represented the most compatible fungal partners in pasteurized orchard soil, regardless of apple rootstock genotype . In our study, nursery-derived communities were dominated by Rhizophagus spp. in three out of four of the rootstock genotypes tested. It is also interesting to note that in the study by Van Horn et al., G.890 rootstocks consistently harbored the highest percentage of arbuscular mycorrhizal fungal species (>5% of the total endophytic fungal community) . In the current study, although genotype was not a significant source of variation affecting absolute abundance in any of the treatments (two-way ANOVA; P = 0.09), AMF DNA reached the highest levels in G.890 no-AMF control and 1× CM treatments (Fig. S6). These two (G.890) treatments also contained the most AMF in terms of relative abundance (3.3 and 5.7%, respectively). That said, because the initial nursery-derived AMF communities varied significantly between rootstock genotypes, our ability to explore the selective capacity of host genotype on the plant-associated AMF microbiome was limited. Finally, the use of a Glomeromycota-specific phylogenetic tree was critical for our ability to accurately assign taxonomy to AMF sequences and supported the detection of a divergent group of sequences within the Order Archaeosporales, which were present in all rootstocks prior to planting. It is likely that a relatively small proportion of all existing Archaeosporales have been characterized . In addition, our phylogenetic tree supported the detection of a new, undescribed lineage of AMF (clade 3), possibly within the Order Glomerales. These results highlight the need for studies going forward to continue to utilize and broaden the molecular data available for phylogenetic-based classification of AMF communities. A better understanding of what “real” AMF communities look like is an essential component to the successful transfer of compatible rootstock/AMF combinations from the laboratory to field. |
Associations Between Agency and Sexual and Reproductive Health Communication in Early Adolescence: A Cross-cultural, Cross-sectional Study | ea0b3834-8a79-4987-be3c-7d41c708d1b1 | 7456790 | Health Communication[mh] | Study context This comparative study employs data from three urban low-resource settings in Kinshasa, Democratic Republic of the Congo (DRC); Cuenca, Ecuador; and Shanghai, China. The contexts of these study sites vary considerably. In Kinshasa, a city of over 11 million where adolescents comprise a quarter of the population, young people often grow up in challenging environments . Nearly half of Kinshasa's population subsists on less than a dollar a day , and 7% of adolescents have dropped out of school at lower secondary school . The city of Cuenca has approximately 600,000 inhabitants; about 2% of the population meets the criteria for income poverty , and 40% of its population are under age 20 . Five percent of adolescents in the southern region of Ecuador, where Cuenca lies, are estimated to be out of school at lower secondary school . In contrast, Shanghai is a megacity of over 24 million, with a small share of its population receiving governmental financial support . Twelve percent of Shanghai's population is under the age of 17, and rates of school dropout are under 1% . Sampling Adolescents aged 10–14 were surveyed in disadvantaged urban areas of three cities (Kinshasa, DRC; Cuenca, Ecuador; and Shanghai, China), as part of the GEAS between June 2017 and March 2018. In Kinshasa, both in-school and out-of-school adolescents were included in the original study due to interest among stakeholders in studying these issues among out-of-school adolescents. Probability and multistaged sampling used to select in-school and out-of-school participants, respectively. Participants in Cuenca were recruited using probability sampling from schools, stratified by age and sex. In Shanghai, all eligible students in grades 6–8 were recruited from three purposively selected public schools. After obtaining parental consent and adolescent assent, adolescents completed the ninety-minute survey using tablets, by face-to-face interview (Kinshasa, due to low literacy) or computer-assisted self-interview (Cuenca and Shanghai). Research protocols were approved by each site's institutional ethical review committee and approved or deemed exempt for secondary data analysis by the Johns Hopkins School of Public Health institutional review board. The initial samples included 2,842 adolescents in Kinshasa, 704 in Cuenca, and 1,760 in Shanghai. In Kinshasa, we considered only adolescents in the control arm of a broader quasi-experimental study for inclusion (n = 1,381 cases), to keep consistency with the other sites and to avoid any introduction of bias due to self-selection into the intervention. We then excluded individuals missing all three SRH communication outcomes (n = 6 in Kinshasa, n = 2 in Cuenca, n = 152 in Shanghai) or missing more than 30% of items used to construct the agency subscales (n = 8 in Kinshasa, n = 5 in Cuenca, n = 184 in Shanghai). Applying these exclusion criteria, 1.0% of cases were dropped in Kinshasa, 1.0% in Cuenca, and 19.1% in Shanghai. Given the large share of cases excluded in Shanghai, we assessed and noted differences between the included and excluded subsamples by age, sex distribution, educational attainment, caregiver monitoring, and perceived neighborhood safety . After exclusion, our analytical samples were comprised of 1,367 adolescents in Kinshasa, 697 in Cuenca, and 1,424 in Shanghai. Measures Two of the three GEAS cross-cultural domains of agency were considered for this analysis: voice (seven items measuring the extent to which young people can express their opinions and be heard) and decision-making (four items measuring adolescents' ability to make choices autonomously in their daily lives). The items, response options, and internal reliability for each scale are presented in . Each of the scales ranged from 1 to 4, with a higher score indicating greater agency. Due to skewed distributions of agency mean scores and an effort to keep consistency in analytical strategies across sites, we dichotomized the continuous mean scores at their medians within each site to identify adolescents with “high” or “low” voice and decision-making power in logistic regressions. Covariates were self-reported at various levels of the ecological environment. Individual sociodemographic factors included adolescents' age, binary indicators of sex, pubertal onset (prepubertal vs. pubertal), and educational attainment (behind in school or out of school vs. at or above expected grade level for age). Family characteristics included binary variables for parental structure (living with both parents vs. with one parent or other relatives) as well as caregiver closeness, monitoring, and migration. Peer and neighborhood-level covariates included a binary variable indicating time typically spent with close friends (no close friends or no time spent with friends weekly vs. saw friends once a week or more), as well as binary indicators reflecting social cohesion (based on an aggregate measure of four items measuring trust and solidarity among neighbors), and whether or not participants feel safe in their neighborhoods. Three individual items were used to assess whether adolescents had ever discussed three sexual and reproductive health topics with anyone: sexual relationships, pregnancy and how it occurs, or contraception. Data analysis We first conducted exploratory analysis to evaluate patterns of missingness across all items comprising the two agency subscales and excluded observations that met the exclusion criteria (outlined in the section). For the remaining samples, we used k-nearest neighbor (kNN) imputation to impute missing agency responses (with k-values of 31, 25, and 37 in Kinshasa, Cuenca, and Shanghai, respectively) followed by imputation to account for missing data on covariates (kNN with k-values of 36, 22, and 31 in Kinshasa, Cuenca, and Shanghai, respectively) . SRH communication patterns were examined overall by site and by sex, while agency and ecological factors were described by site, sex, and SRH communication outcomes using chi-squared, Fisher exact, and Student t -tests. We examined bivariate associations between ecological factors (individual, family, peer, and neighborhood levels) and the three SRH communication outcomes as well as those between agency (voice and decision-making) and SRH communication. Multivariable logistic regressions assessed the independent effect of each ecological factor on SRH communication and subsequently evaluated the effect of agency levels on SRH communication, adjusting for all ecological covariates. Collinearity among covariates was assessed by the variance inflation factor value and no multicollinearity was noted. All analyses were stratified by site to assess similarities and differences in these associations by context. Interactions between agency and sex were also tested in the latter multivariable models in order to assess sex differences in the relationships between agency and SRH communication. kNN imputation was conducted using RStudio (RStudio, Inc., Boston, MA); all other analyses were conducted using Stata/SE 15.1 (StataCorp LLC, College Station, TX).
This comparative study employs data from three urban low-resource settings in Kinshasa, Democratic Republic of the Congo (DRC); Cuenca, Ecuador; and Shanghai, China. The contexts of these study sites vary considerably. In Kinshasa, a city of over 11 million where adolescents comprise a quarter of the population, young people often grow up in challenging environments . Nearly half of Kinshasa's population subsists on less than a dollar a day , and 7% of adolescents have dropped out of school at lower secondary school . The city of Cuenca has approximately 600,000 inhabitants; about 2% of the population meets the criteria for income poverty , and 40% of its population are under age 20 . Five percent of adolescents in the southern region of Ecuador, where Cuenca lies, are estimated to be out of school at lower secondary school . In contrast, Shanghai is a megacity of over 24 million, with a small share of its population receiving governmental financial support . Twelve percent of Shanghai's population is under the age of 17, and rates of school dropout are under 1% .
Adolescents aged 10–14 were surveyed in disadvantaged urban areas of three cities (Kinshasa, DRC; Cuenca, Ecuador; and Shanghai, China), as part of the GEAS between June 2017 and March 2018. In Kinshasa, both in-school and out-of-school adolescents were included in the original study due to interest among stakeholders in studying these issues among out-of-school adolescents. Probability and multistaged sampling used to select in-school and out-of-school participants, respectively. Participants in Cuenca were recruited using probability sampling from schools, stratified by age and sex. In Shanghai, all eligible students in grades 6–8 were recruited from three purposively selected public schools. After obtaining parental consent and adolescent assent, adolescents completed the ninety-minute survey using tablets, by face-to-face interview (Kinshasa, due to low literacy) or computer-assisted self-interview (Cuenca and Shanghai). Research protocols were approved by each site's institutional ethical review committee and approved or deemed exempt for secondary data analysis by the Johns Hopkins School of Public Health institutional review board. The initial samples included 2,842 adolescents in Kinshasa, 704 in Cuenca, and 1,760 in Shanghai. In Kinshasa, we considered only adolescents in the control arm of a broader quasi-experimental study for inclusion (n = 1,381 cases), to keep consistency with the other sites and to avoid any introduction of bias due to self-selection into the intervention. We then excluded individuals missing all three SRH communication outcomes (n = 6 in Kinshasa, n = 2 in Cuenca, n = 152 in Shanghai) or missing more than 30% of items used to construct the agency subscales (n = 8 in Kinshasa, n = 5 in Cuenca, n = 184 in Shanghai). Applying these exclusion criteria, 1.0% of cases were dropped in Kinshasa, 1.0% in Cuenca, and 19.1% in Shanghai. Given the large share of cases excluded in Shanghai, we assessed and noted differences between the included and excluded subsamples by age, sex distribution, educational attainment, caregiver monitoring, and perceived neighborhood safety . After exclusion, our analytical samples were comprised of 1,367 adolescents in Kinshasa, 697 in Cuenca, and 1,424 in Shanghai.
Two of the three GEAS cross-cultural domains of agency were considered for this analysis: voice (seven items measuring the extent to which young people can express their opinions and be heard) and decision-making (four items measuring adolescents' ability to make choices autonomously in their daily lives). The items, response options, and internal reliability for each scale are presented in . Each of the scales ranged from 1 to 4, with a higher score indicating greater agency. Due to skewed distributions of agency mean scores and an effort to keep consistency in analytical strategies across sites, we dichotomized the continuous mean scores at their medians within each site to identify adolescents with “high” or “low” voice and decision-making power in logistic regressions. Covariates were self-reported at various levels of the ecological environment. Individual sociodemographic factors included adolescents' age, binary indicators of sex, pubertal onset (prepubertal vs. pubertal), and educational attainment (behind in school or out of school vs. at or above expected grade level for age). Family characteristics included binary variables for parental structure (living with both parents vs. with one parent or other relatives) as well as caregiver closeness, monitoring, and migration. Peer and neighborhood-level covariates included a binary variable indicating time typically spent with close friends (no close friends or no time spent with friends weekly vs. saw friends once a week or more), as well as binary indicators reflecting social cohesion (based on an aggregate measure of four items measuring trust and solidarity among neighbors), and whether or not participants feel safe in their neighborhoods. Three individual items were used to assess whether adolescents had ever discussed three sexual and reproductive health topics with anyone: sexual relationships, pregnancy and how it occurs, or contraception.
We first conducted exploratory analysis to evaluate patterns of missingness across all items comprising the two agency subscales and excluded observations that met the exclusion criteria (outlined in the section). For the remaining samples, we used k-nearest neighbor (kNN) imputation to impute missing agency responses (with k-values of 31, 25, and 37 in Kinshasa, Cuenca, and Shanghai, respectively) followed by imputation to account for missing data on covariates (kNN with k-values of 36, 22, and 31 in Kinshasa, Cuenca, and Shanghai, respectively) . SRH communication patterns were examined overall by site and by sex, while agency and ecological factors were described by site, sex, and SRH communication outcomes using chi-squared, Fisher exact, and Student t -tests. We examined bivariate associations between ecological factors (individual, family, peer, and neighborhood levels) and the three SRH communication outcomes as well as those between agency (voice and decision-making) and SRH communication. Multivariable logistic regressions assessed the independent effect of each ecological factor on SRH communication and subsequently evaluated the effect of agency levels on SRH communication, adjusting for all ecological covariates. Collinearity among covariates was assessed by the variance inflation factor value and no multicollinearity was noted. All analyses were stratified by site to assess similarities and differences in these associations by context. Interactions between agency and sex were also tested in the latter multivariable models in order to assess sex differences in the relationships between agency and SRH communication. kNN imputation was conducted using RStudio (RStudio, Inc., Boston, MA); all other analyses were conducted using Stata/SE 15.1 (StataCorp LLC, College Station, TX).
summarizes sample characteristics. On average, adolescents were older in Shanghai than in the other two sites. Parental structure differed between contexts; just over half of adolescents in Kinshasa and over three quarters of those in Shanghai were living with both parents at the time of the survey. Caregiver migration was less common in Cuenca (22.1% compared to about half in the other two sites). Neighborhood perceptions were most positive in Shanghai (56.9% reported high social cohesion and 96.8% felt safe in their neighborhood). SRH communication patterns varied by site and sex . Adolescents in Cuenca were the most likely to have talked to anyone about sexual relationships (43.8%), pregnancy (58.3%), or contraception (39.1%), while only 1 in 10 had ever discussed each topic in Kinshasa. In Kinshasa, girls were more likely than boys to have discussed both pregnancy (12.1% vs. 7.9%, p = .009) and contraception (10.4% vs. 7.2%, p = .042). In Cuenca, more girls than boys had discussed sexual relationships (47.8% vs. 39.9%, p = .036) and pregnancy (63.5% vs. 53.3%, p = .006). No differences by sex were detected in Shanghai. Levels of agency also differed by sex and site . Mean scores for voice and decision-making were lowest in Kinshasa (2.4 for voice and 2.7 for decision-making). Scores for voice were higher for boys than girls in Kinshasa (2.5 vs. 2.4, p < .001) and comparable by sex in the other two sites. Decision-making mean scores were higher for girls than boys in Shanghai (3.5 vs. 3.4, p = .003), while no sex differences were observed in Kinshasa and Cuenca. Distributions of ecological factors by the three SRH communication outcomes are outlined in . We next examined the results of bivariate logistic regression analysis for the individual, family, peer, and community factors predictive of SRH communication. In each of the three sites, older age and pubertal onset were associated with increased odds of communication about all three SRH topics. These results were confirmed in multivariable analysis, with the odds of SRH communication increasing between 17% and 74% across topics and sites with older age and ranging between 2.2- and 3.7-fold across topics in Cuenca and Kinshasa among adolescents with pubertal onset . Additional factors related to SRH communication were heterogeneous across sites in bivariate and multivariable analyses . In Kinshasa, adolescents with higher education performance were more likely to have communicated about pregnancy and about contraception before and after adjustment. Adolescents living with both parents and those who reported feeling close to their parents were less likely to report communication about pregnancy. Perceived neighborhood insecurity was linked to lower likelihood of having discussed sexual relationships. Meanwhile, adolescents whose parents were born in Cuenca were more likely to discuss all SRH topics. In Shanghai, low parental closeness was related to communication about the three SRH topics, while time spent with friends was linked to higher odds of having discussed all three topics. Those in Shanghai who reported low cohesion among their community were more likely to have discussed sexual relationships, and participants who felt threatened in their neighborhood were also more likely to report they had discussed SRH. Bivariate logistic regressions assessing odds of SRH communication by levels of voice and decision-making revealed links between both voice and decision-making and SRH communication in Kinshasa and Cuenca ( and ). After adjustment for socioecological factors at the individual, family, peer, and neighborhood levels, more voice was related to higher odds of communication about all three SRH topics in both Kinshasa and Cuenca with adjusted odds ratios ranging from 1.6 to 2.2. No significant relationship between voice and SRH communication was observed in Shanghai after adjustment. Significance tests of interaction terms revealed possible differential effects of voice on communication about pregnancy by sex in Kinshasa ( p = .017), with a significantly stronger association among girls than boys (adjusted relative odds ratio [aOR] for girls vs. boys, 2.1; 95% confidence interval [CI], 1.2–3.7). Fewer associations were observed between decision-making capacity and SRH communication. Adolescents with high decision-making scores were more likely to have talked about pregnancy (aOR, 1.5; 95% CI, 1.0–2.1) and contraception (aOR, 1.6; 95% CI, 1.1–2.3) in Cuenca. No significant differences in the associations of SRH communication and decision-making were detected by sex.
This analysis found SRH communication among young adolescents was relatively uncommon, associated with developmental factors across sites, and linked in two contexts to adolescents' ability to voice their needs and opinions. Decision-making was related to history of pregnancy and contraception discussions among adolescents in Cuenca. Particularly low levels of SRH communication were observed in Kinshasa, within the context of high fertility and a low modern contraceptive prevalence rate among adolescents in the DRC. Prior research has demonstrated substantial social barriers to these conversations in household and school settings in the DRC . Across sites, pregnancy was the most commonly discussed SRH topic, followed by sexual relationships and contraception. These patterns echo evidence from prior investigations into parent-child SRH communication, which found that these conversations tend to focus on pregnancy risk and abstinence, and less frequently address contraception . In Kinshasa and Cuenca, more girls than boys had discussed most SRH topics, a finding that reflects the social and biological consequences of sexual activity that disproportionately impact girls , which may prompt more SRH conversations with girls. Other studies have highlighted parents' prioritization of SRH discussions with their daughters as a result of a sexual double standard, views that boys' sexual activity is inevitable, and greater fear for their daughters' safety in sexual encounters . Such sex differences in these discussions may be particularly pronounced in highly patriarchal societies. Our analysis also examined the role of the socioecological environment in shaping SRH communication, in recognition its actualization can be promoted or constrained by one's context. Congruent with prior research, increased age and puberty were strongly associated with SRH communication. Older age increased participants' likelihood of having discussed each of the SRH topics across contexts, as these topics grow more relevant to adolescents' emerging sexual lives. Independently, our study found that puberty was the characteristic most strongly linked to SRH communication in Cuenca and Kinshasa. At the same time that young adolescents may be more curious about sex at the onset of puberty than they were in the prior period , they may also be engaged in discussions about sexual and reproductive health topics by their parents, who worry about the health and social consequences of sexual activity as they mature . This study found additional patterns of individual, family, peer, and neighborhood factors linked with SRH communication that were heterogeneous across sites. Factors predominantly on the individual level in Kinshasa, family level in Cuenca, and family, peer, and neighborhood levels in Shanghai were influential upon SRH communication. These findings indicate the varied cultural, social, and structural factors that impact SRH communication in each setting and indicate differential implications for interventions to create an enabling environment. These varying results across sites also underscore the varied contexts in which such discussions can take place, within family or school settings, or in the neighborhood with peers. Our findings additionally call attention to the agents who may transmit SRH information of varied quality to adolescents, information that ranges in accuracy and in its framing of sexuality and reproductive health. Additional research is needed to more extensively explore the nature of SRH communication interactions in early adolescence and how this communication contributes to SRH knowledge. Beyond the role of ecological factors in shaping SRH communication, a key finding to emerge from our study is the relationship between SRH communication and both increased decision-making power in Cuenca and greater voice in Kinshasa and Cuenca, after adjustment for the socioecological environment. Our findings, if corroborated by additional research, suggest a role for agency-promoting interventions in setting positive SRH trajectories for both boys and girls in the adolescent period. While promoting empowerment is a long-term and complex process, in the short-term, interventions such as comprehensive sexuality education should pay special attention to adolescents who are less able to voice their opinions or influence decision-making to ensure they receive important SRH information even if they are less likely than others to initiate such a conversation. On the other hand, the associations found here may indicate that discussion of SRH may play a role in promoting adolescents' perception of their own power by building foundational negotiation skills and knowledge. Regardless of the direction of this relationship, our study builds upon a small body of prior findings that link SRH-specific agency indicators to SRH outcomes during the adolescent period . This study found considerable variation in agency, experiences of SRH communication, and the relationships between the two across three very distinctive urban contexts. While we found links between all SRH communication and voice in Kinshasa and Cuenca, decision-making was only associated with communication about pregnancy and contraception in Cuenca. In addition, no relationships between these factors were found in Shanghai. Differences in these findings across the three sites could be attributed to context-specific factors such as the influence of religiosity in Cuenca and Kinshasa, the existence of Machismo culture in Latin America, and the impact of the One-Child Policy on girls' empowerment in Shanghai . These elements shape gender roles, expressions of empowerment among adolescent boys and girls, norms about SRH communication, and ultimately these findings. Within more conservative social contexts, adolescents may need greater agency to overcome social barriers to the conversations about SRH that young people in other settings can more easily navigate. The present analysis is also novel in its inclusion of boys and comparison of the examined relationships by sex. While the field has focused on girls' preparedness for sexual activity, the present results challenge the assumption that boys arrive at sexual activity both more empowered and more prepared than girls, as levels of SRH communication were largely comparable by sex in these three settings. While we found that voice was more strongly associated with communication about pregnancy for girls than boys in Kinshasa, notably no other associations between SRH communication and agency differed significantly by sex. Our results also contribute novel evidence by employing a validated measure of overall agency among young adolescents to evaluate such associations, suggesting that a broader concept of agency may relate to SRH skills. Several limitations of this study should be noted. First, findings are not generalizable beyond our study areas. Second, our measure of SRH communication only described whether or not adolescents reported having discussed each SRH topic. These measures lacked specificity about the people adolescents talked to, contexts in which they took place, content, temporality, frequency or quality of these discussions, which likely impact their relationship to agency, and to downstream SRH behaviors and outcomes . Indicators were self-reported and therefore subject to social desirability bias. Third, the large share (19%) of the Shanghai sample excluded due to missing data and observed differences between the included and excluded cases may have introduced unaccounted bias into our analyses. Fourth, the cross-sectional associations observed cannot be interpreted as causational. Therefore, a longitudinal assessment of these relationships, which may be carried out using subsequent waves of the GEAS, is necessary to better understand the directionality of observed associations in their contribution to healthy behaviors as adolescents become sexually active. Taken together, our findings suggest that empowerment factors, including the enabling environment and dimensions of agency, are linked to communication about SRH in the early adolescent period in certain contexts. We conclude that further research, with representative samples, longitudinal data and more specific SRH communication items that allow for disaggregation between types of SRH communication, is needed to fully understand the role of empowerment in shaping SRH trajectories during early adolescence.
|
Current status of the analytical validation of next generation sequencing applications for pharmacogenetic profiling | aa2158f4-0ca1-4f10-9576-67b6a190670d | 10635985 | Pharmacology[mh] | In recent years, regulatory adaptations in the United States of America (USA) and the European Union (EU) with regard to the use of in vitro diagnostic applications in the management of drug therapy have been introduced . This also affected the performance requirements of next generation sequencing (NGS). Analytical validity is considered a critical step in test assessment. According to the US in vitro diagnostic (IVD) regulation and the new in vitro diagnostic medical devices regulation (IVDR) in the EU, the analytical and clinical performance of an in vitro diagnostic test to be legally marketed should first be evaluated . Therefore, the suitability of an in vitro diagnostic NGS test needs to be demonstrated for the intended use . Laboratories offering targeted NGS , whole exome sequencing (WES) and whole genome sequencing (WGS) as diagnostic tests are also obliged to evaluate performance characteristics and carry out a full validation prior to clinical application as laboratory developed test (LDT). In the EU, the evaluation of analytical performance of an IVD or LDT should encompass a variety of general criteria also including the collection and handling of according specimen. However, explicit criteria that meet the needs of the assessment of analytical NGS performance are not specified . Furthermore, standardization for performance evaluation is currently controversial . For the analytical validation of targeted NGS based on NGS panels, in a recent publication on the consensus recommendation of the Association for Molecular Pathology and College of American Pathologists, an assessment of accuracy in terms of the positive percent agreement (PPA) and positive predictive value (PPV) was recommended. Furthermore, the precision of variant detection in terms of reproducibility and repeatability, reportable range and reference range, limits of detection, analytical specificity (interfering substances) and carryover was included in the recommendation . In the USA, further guidance for analytical validation of in vitro diagnostics on the basis of NGS is provided by non-binding recommendations of the Food and Drug Administration (FDA). Here, test performance characteristics such as accuracy also in terms of PPA, technical PPV and additionally negative percent agreement (NPA), precision, limit of detection, and analytical specificity in terms of interference, cross-reactivity and cross-contamination are suggested for evaluation . FDA documents such as the “Summary of Safety and Effectiveness” provide publicly available analytical performance evaluations of NGS-based diagnostic tests as companion diagnostics. However, predominantly for personalized therapeutic management for drugs applied in oncology . In the USA, such approved NGS-based tests can therefore be applied in clinical practice for the intended use. However, no NGS-based companion diagnostic is approved to guide therapy with regard to the pharmacogenetic profile of drug metabolizing enzymes. Large-scale analyses such as whole exome and whole genome sequencing are increasingly used as laboratory developed tests . Such data is considered a valuable resource for pharmacogenomic profiling . However, the added value of WES and WGS applications compared to other large scale molecular genetic tests in pharmacogenetics is controversial and limitations were reported with regard to coverage and short read-based assessment . The status of the validation of NGS-based applications for clinical pharmacogenomic (PGx) profiling is unclear, although value for therapeutic management has been reported , and clinical implementation studies have been performed in several countries . Also in the EU, for a few drug prescriptions, e.g. testing for cytochrome P450 2C1 9 ( CYP2C19 ) variants prior to therapy with atazanavir is recommended and testing for cytochrome P450 2D6 ( CYP2D6 ) prior to therapy with eliglustat is even required . However, PGx testing is not applied as companion diagnostics yet . Approaches to repurpose e.g. whole exome sequencing data obtained in clinical settings also for pharmacogenetic profiling suggest that such secondary findings may provide an extraordinary opportunity to integrate valuable information for personalized treatment. However, the feasibility evaluations reported limitations with regard to several pharmacogenes . Furthermore validity of these approaches needs to be established . Here, on the basis of current literature identified at the platforms Pubmed and Pubmed Central, we provide an overview of the available information on the analytical validation status of applied targeted NGS including second and third generation sequencing and furthermore whole exome and whole genome sequencing for targeting relevant pharmacogenetic biomarkers. Additionally, we discuss the potential of third generation sequencing techniques for whole genome sequencing and the assessment of pharmacogenetic information provided by such techniques. The FDA list “Nucleic acid-based tests” was applied in order to analyze the validation status of NGS applications for genetic testing of pharmacogenes coding drug metabolizing enzymes. Here, listed genes of the category “drug metabolizing enzymes” were used for further investigations . The list provides genetic tests that have either been cleared or approved. Such tests are appropriate for comparison in the evaluation of the analytical validation of other testing techniques for pharmacogenetic profiling. Second generation and third generation sequencing techniques were included in the analysis. Analytical validation on the basis of cleared or approved genetic tests Screening for publications providing a performance comparison of FDA approved nucleic acid-based diagnostic tests concerning drug metabolizing enzymes with second and third generation sequencing applications was performed in August 2022. Many of these tests were also CE certified in the European Union. The aim was to identify whether such approved or cleared tests have been used to validate NGS based applications with a focus on relevant pharmacogenes. According to the FDA recommendations, such tests identified as appropriate by the FDA, should be applied as reference tests of choice in an analytical validation . Here, the publication platforms Pubmed and Pubmed Central were screened for according publications using the keywords “test name” (Table , see Trade Name), associated gene(s) (Table , see “Gene”), next generation sequencing or whole exome sequencing or whole genome sequencing or long read sequencing or Nanopore or Pacific Biosciences and validity or validation. Publications identified for second and third generation sequencing applications were evaluated separately. Congress abstracts were not included and duplications were excluded. Publications not providing information on validation results with the according FDA cleared or approved nucleic acid-based diagnostic tests involving the according gene in the method or result section of the publication were excluded. The few publications included were analyzed for the evaluation of NGS performance on the basis of relevant analytical validation criteria recommended for NGS in current literature such as accuracy, precision, analytical specificity (endogenous and exogenous interference, cross-reactions, cross-contamination) and limit of detection . Validation status of NGS based applications To further evaluate the validation status of targeted NGS, WES and WGS in research and the clinic with regard to the previously analyzed pharmacogenes (Table ), a second keyword search was performed. Cytochrome P450 2C9, 2C19 and 2D6, Vitamin K epOxide Reductase Complex Subunit 1 (VKORC1) and UDP-Glucuronosyltransferase 1 Polypeptid A1 (UGT1A1) were also the focus of this evaluation. This search was extended to any test or orthogonal method applied for performance comparison with an NGS based application to increase the yield of findings. Thereby, the keywords next generation sequencing and VKORC1 or CYP2C9 or CYP2C19 or CYP2D6 or UGT1A1 and validity or validation were used to screen the publication platforms Pubmed and Pubmed Central. The same search as for NGS was performed for WES and WGS. Title and abstract were screened for suitability of the publication in terms of information on a performance evaluation of the according method involving the according gene of interest. The included publications were analyzed for the evaluation of analytical validation criteria recommended for NGS in current literature such as limit of detection, accuracy (positive percent agreement, negative percent agreement and positive predictive value), precision (reproducibility and repeatability), and analytical specificity (endogenous and exogenous interference, cross-reactions, cross-contamination). Screening for publications providing a performance comparison of FDA approved nucleic acid-based diagnostic tests concerning drug metabolizing enzymes with second and third generation sequencing applications was performed in August 2022. Many of these tests were also CE certified in the European Union. The aim was to identify whether such approved or cleared tests have been used to validate NGS based applications with a focus on relevant pharmacogenes. According to the FDA recommendations, such tests identified as appropriate by the FDA, should be applied as reference tests of choice in an analytical validation . Here, the publication platforms Pubmed and Pubmed Central were screened for according publications using the keywords “test name” (Table , see Trade Name), associated gene(s) (Table , see “Gene”), next generation sequencing or whole exome sequencing or whole genome sequencing or long read sequencing or Nanopore or Pacific Biosciences and validity or validation. Publications identified for second and third generation sequencing applications were evaluated separately. Congress abstracts were not included and duplications were excluded. Publications not providing information on validation results with the according FDA cleared or approved nucleic acid-based diagnostic tests involving the according gene in the method or result section of the publication were excluded. The few publications included were analyzed for the evaluation of NGS performance on the basis of relevant analytical validation criteria recommended for NGS in current literature such as accuracy, precision, analytical specificity (endogenous and exogenous interference, cross-reactions, cross-contamination) and limit of detection . To further evaluate the validation status of targeted NGS, WES and WGS in research and the clinic with regard to the previously analyzed pharmacogenes (Table ), a second keyword search was performed. Cytochrome P450 2C9, 2C19 and 2D6, Vitamin K epOxide Reductase Complex Subunit 1 (VKORC1) and UDP-Glucuronosyltransferase 1 Polypeptid A1 (UGT1A1) were also the focus of this evaluation. This search was extended to any test or orthogonal method applied for performance comparison with an NGS based application to increase the yield of findings. Thereby, the keywords next generation sequencing and VKORC1 or CYP2C9 or CYP2C19 or CYP2D6 or UGT1A1 and validity or validation were used to screen the publication platforms Pubmed and Pubmed Central. The same search as for NGS was performed for WES and WGS. Title and abstract were screened for suitability of the publication in terms of information on a performance evaluation of the according method involving the according gene of interest. The included publications were analyzed for the evaluation of analytical validation criteria recommended for NGS in current literature such as limit of detection, accuracy (positive percent agreement, negative percent agreement and positive predictive value), precision (reproducibility and repeatability), and analytical specificity (endogenous and exogenous interference, cross-reactions, cross-contamination). Analytical validation of NGS- based tests applying orthogonal FDA cleared or approved genetic tests The present keyword search (Fig. ) aimed to identify published analytical validation studies of NGS-applications with FDA cleared or approved orthogonal tests. FDA cleared or approved genetic tests used for keyword search were identified as PCR- or predominantly microarray-based methods for genotyping of drug metabolizing enzymes (Table ). A keyword search applying the platform Pubmed did not result in any findings. An additional keyword search on the platform Pubmed Central resulted in a few publications addressing some of the cleared or approved nucleic acid-based tests. The keywords validation and validity resulted in overlapping publication lists. However, few suitable publications with regard to test comparison or validation were identified (Table ) . None of the publications in this literature search provided a comparison of WES or WGS with the listed nucleic acid-based tests. Publications of comparison experiments with next generation sequencing were identified in the keyword search for the Invader UGT1A1 Molecular Assay (Third Wave Technologies Inc., Wisconsin, USA) and the xTAG CYP2D6 Kit v3 (Luminex Molecular Diagnostics, Inc., Toronto, ON, Canada) . The published performance comparison with the Invader UGT1A1 Molecular Assay however, was carried out with the OpenArray pharmacogenomics panel, which comprised of 4 customized TaqMan® OpenArray Genotyping Plates. NGS was included in the validation of the OpenArray panel (Thermo Fisher Scientific, Waltham, MA, USA) but was not validated with the Invader UGT1A1 Molecular Assay . Therefore, this publication also did not result in an appropriate finding according to the inclusion criteria for this review. Performance comparison of the Luminex xTAG CYP2D6, with second generation sequencing was detected in only one of the identified publications . Here, Carvalho Henriques et al. provided only an evaluation of concordance for CYP2D6 and CYP2C19 variant detection in their cross-validation of a large variety of different techniques. These also included NGS represented by Ion Torrent™semiconductor sequencing on the basis of the Ion AmpliSeq Pharmacogenomics Panel (Thermo Fisher Scientific, Waltham, MA, USA). Therefore, the accuracy via positive percent agreement, negative percent agreement and positive predictive value and also evaluations in terms of further important validation criteria such as precision, limit of detection, and analytical specificity were not addressed. Performance comparison of the Luminex xTAG CYP2D6 with a third-generation sequencing application was only identified in one publication . Qiao et al. provide a concordance evaluation of the long-read SMRT sequencing of CYP2D6 on the Pacific Biosciences platform with the xTAG CYP2D6 Kit v3 (Luminex Corporation, TX, USA) and in terms of CNVs, the TaqMan® real-time qPCR Copy Number Assays (Applied Biosystems, Carlsbad, CA, USA). Furthermore, quality metrics and precision of the SMRT sequencing was evaluated in terms of intra- and inter-run reproducibility on the basis of triplicates. However, sample size was low and other aspects of accuracy such as positive percent agreement, negative percent agreement and positive predictive value and other criteria of analytical validation were not addressed . Validation status of NGS based applications In the second literature search described previously, about 99% of the publications screened by title and abstract were excluded, as they did not include information on sequencing of the pharmacogenes of interest or the use of NGS applications or any aspects of performance evaluation. Duplications due to overlapping results by applying the keywords validity or validation and due to published overlapping analyses including several of the pharmacogenes of interest were excluded in the screening phase. Further publications were excluded in the assessment for eligibility due to lacking focus on the NGS applications or the pharmacogenes of interest in the performance evaluations (Fig. ). In summary, of the 15 publications included (1 duplicate assessed for WGS and WES), 13 publications on second generation sequencing including also WES and WGS and two publications with a focus on third generation sequencing of CYP2D6 were identified with performance evaluations. They provided information to target at least one of the pharmacogenes of interest (Table ). All of these publications provide concordance evaluations of the next generation sequencing application with other genotyping methods such as TaqMan-based genotyping assays, Sanger sequencing, Agena Bioscience ADME genotyping panels, array-based applications and custom PCR-based assays (Supplementary material ). The third generation sequencing articles identified used the Pacific Biosciences platform . Evaluations focusing on second generation sequencing were mainly based on the Illumina platforms, for WES and WGS the HiSeq 2000 and/or HiSeq 2500 platforms were used . For targeted second generation sequencing, the HiSeq platforms (HiSeq 1500, HiSeq 2000, HiSeq 2500, HiSeq 4000) the MiSeq, NextSeq 500 and MiniSeq system was reported . Library preparation systems used for test evaluations on the Illumina platforms also differed. The Agilent SureSelect kits were used more often than other enrichment kits (Supplementary material ). Other platforms were used when only a few variants or genes were analyzed. The PyroMark Q24 (QIAGEN GmbH, Venlo, Netherlands)), a pyrosequencing platform, was applied in a comparison of CYP2D6*10 identification and the Ion Chef instrument along with the Ion AmpliSeq Pharmacogenomics Panel was used in a cross-validation of several molecular genetic techniques for genotyping of CYP2D6 and CYP2C19 (Supplementary material ) . NGS applications assessed in a majority of the analysed publications (93%), considered the pharmacogene CYP2D6 while CYP2C9 , CYP2C19 , VKORC1 and UGT1A1 were less often covered (Table ). However, in one of the analyzed publications, the NGS application that covered CYP2D6 and UGT1A1 was an oncology sequencing assay and therefore not explicitly intended to be used for pharmacogenetic profiling (Supplementary material ) . In addition to concordance evaluations, about 46.67% (7 of 15) of the analyzed publications provide further data on analytical performance criteria (Table ). Accuracy assessments (33.3%, 5 of 15) were performed, these included overall accuracy or in a majority of cases, as is recommended, accuracy in terms of the PPV along with positive percent agreement (also referred to as sensitivity) and negative percent agreement (referred to as specificity) (26.67%, 4 of 15) (Table ). The limit of detection for variant types, such as SNVs and indels, is reported in one publication and was determined on the basis of detected variant allele fractions . In 3 further publications only assessments in terms of minor allele frequency or fraction or detection rate were performed without determining a limit of detection for the analyzed variant types (Supplementary material ). Only one publication providing an evaluation of a targeted second generation sequencing panel appropriately addressed an evaluation of precision, in terms of intra-run and inter-run precision, that also provided the most comprehensive assessment in terms of FDA validation criteria such as also accuracy and further criteria such as reportable regions along with quality of the NGS metrics. However, analytical validation was based on only three reference samples . Furthermore, one publication evaluating CYP2D6 SMRT third generation sequencing reported an assessment of precision in terms of triplicate intra-run sequencing of a validation sample and inter-run sequencing . In summary, few studies evaluate analytical performance criteria such as limit of detection, accuracy and precision of second generation sequencing panels and fewer still focus on third generation sequencing. Analytical specificity in terms of interference, cross-reactions and cross-contamination has not been assessed or provided by any of these published evaluations. Therefore, according to the information provided on the platforms Pubmed and Pubmed Central, none of the NGS- based tests has an analytical performance evaluation that has been executed according to the current recommendations in literature and of the FDA. Although, several studies compare WES and WGS with other molecular genetic techniques in terms of variant detection in pharmacogenes coding drug metabolizing enzymes, here also no articles were identified that carried out an analytical validation in terms of evaluating performance criteria as recommended for NGS-based IVDs or LDTs . The present keyword search (Fig. ) aimed to identify published analytical validation studies of NGS-applications with FDA cleared or approved orthogonal tests. FDA cleared or approved genetic tests used for keyword search were identified as PCR- or predominantly microarray-based methods for genotyping of drug metabolizing enzymes (Table ). A keyword search applying the platform Pubmed did not result in any findings. An additional keyword search on the platform Pubmed Central resulted in a few publications addressing some of the cleared or approved nucleic acid-based tests. The keywords validation and validity resulted in overlapping publication lists. However, few suitable publications with regard to test comparison or validation were identified (Table ) . None of the publications in this literature search provided a comparison of WES or WGS with the listed nucleic acid-based tests. Publications of comparison experiments with next generation sequencing were identified in the keyword search for the Invader UGT1A1 Molecular Assay (Third Wave Technologies Inc., Wisconsin, USA) and the xTAG CYP2D6 Kit v3 (Luminex Molecular Diagnostics, Inc., Toronto, ON, Canada) . The published performance comparison with the Invader UGT1A1 Molecular Assay however, was carried out with the OpenArray pharmacogenomics panel, which comprised of 4 customized TaqMan® OpenArray Genotyping Plates. NGS was included in the validation of the OpenArray panel (Thermo Fisher Scientific, Waltham, MA, USA) but was not validated with the Invader UGT1A1 Molecular Assay . Therefore, this publication also did not result in an appropriate finding according to the inclusion criteria for this review. Performance comparison of the Luminex xTAG CYP2D6, with second generation sequencing was detected in only one of the identified publications . Here, Carvalho Henriques et al. provided only an evaluation of concordance for CYP2D6 and CYP2C19 variant detection in their cross-validation of a large variety of different techniques. These also included NGS represented by Ion Torrent™semiconductor sequencing on the basis of the Ion AmpliSeq Pharmacogenomics Panel (Thermo Fisher Scientific, Waltham, MA, USA). Therefore, the accuracy via positive percent agreement, negative percent agreement and positive predictive value and also evaluations in terms of further important validation criteria such as precision, limit of detection, and analytical specificity were not addressed. Performance comparison of the Luminex xTAG CYP2D6 with a third-generation sequencing application was only identified in one publication . Qiao et al. provide a concordance evaluation of the long-read SMRT sequencing of CYP2D6 on the Pacific Biosciences platform with the xTAG CYP2D6 Kit v3 (Luminex Corporation, TX, USA) and in terms of CNVs, the TaqMan® real-time qPCR Copy Number Assays (Applied Biosystems, Carlsbad, CA, USA). Furthermore, quality metrics and precision of the SMRT sequencing was evaluated in terms of intra- and inter-run reproducibility on the basis of triplicates. However, sample size was low and other aspects of accuracy such as positive percent agreement, negative percent agreement and positive predictive value and other criteria of analytical validation were not addressed . In the second literature search described previously, about 99% of the publications screened by title and abstract were excluded, as they did not include information on sequencing of the pharmacogenes of interest or the use of NGS applications or any aspects of performance evaluation. Duplications due to overlapping results by applying the keywords validity or validation and due to published overlapping analyses including several of the pharmacogenes of interest were excluded in the screening phase. Further publications were excluded in the assessment for eligibility due to lacking focus on the NGS applications or the pharmacogenes of interest in the performance evaluations (Fig. ). In summary, of the 15 publications included (1 duplicate assessed for WGS and WES), 13 publications on second generation sequencing including also WES and WGS and two publications with a focus on third generation sequencing of CYP2D6 were identified with performance evaluations. They provided information to target at least one of the pharmacogenes of interest (Table ). All of these publications provide concordance evaluations of the next generation sequencing application with other genotyping methods such as TaqMan-based genotyping assays, Sanger sequencing, Agena Bioscience ADME genotyping panels, array-based applications and custom PCR-based assays (Supplementary material ). The third generation sequencing articles identified used the Pacific Biosciences platform . Evaluations focusing on second generation sequencing were mainly based on the Illumina platforms, for WES and WGS the HiSeq 2000 and/or HiSeq 2500 platforms were used . For targeted second generation sequencing, the HiSeq platforms (HiSeq 1500, HiSeq 2000, HiSeq 2500, HiSeq 4000) the MiSeq, NextSeq 500 and MiniSeq system was reported . Library preparation systems used for test evaluations on the Illumina platforms also differed. The Agilent SureSelect kits were used more often than other enrichment kits (Supplementary material ). Other platforms were used when only a few variants or genes were analyzed. The PyroMark Q24 (QIAGEN GmbH, Venlo, Netherlands)), a pyrosequencing platform, was applied in a comparison of CYP2D6*10 identification and the Ion Chef instrument along with the Ion AmpliSeq Pharmacogenomics Panel was used in a cross-validation of several molecular genetic techniques for genotyping of CYP2D6 and CYP2C19 (Supplementary material ) . NGS applications assessed in a majority of the analysed publications (93%), considered the pharmacogene CYP2D6 while CYP2C9 , CYP2C19 , VKORC1 and UGT1A1 were less often covered (Table ). However, in one of the analyzed publications, the NGS application that covered CYP2D6 and UGT1A1 was an oncology sequencing assay and therefore not explicitly intended to be used for pharmacogenetic profiling (Supplementary material ) . In addition to concordance evaluations, about 46.67% (7 of 15) of the analyzed publications provide further data on analytical performance criteria (Table ). Accuracy assessments (33.3%, 5 of 15) were performed, these included overall accuracy or in a majority of cases, as is recommended, accuracy in terms of the PPV along with positive percent agreement (also referred to as sensitivity) and negative percent agreement (referred to as specificity) (26.67%, 4 of 15) (Table ). The limit of detection for variant types, such as SNVs and indels, is reported in one publication and was determined on the basis of detected variant allele fractions . In 3 further publications only assessments in terms of minor allele frequency or fraction or detection rate were performed without determining a limit of detection for the analyzed variant types (Supplementary material ). Only one publication providing an evaluation of a targeted second generation sequencing panel appropriately addressed an evaluation of precision, in terms of intra-run and inter-run precision, that also provided the most comprehensive assessment in terms of FDA validation criteria such as also accuracy and further criteria such as reportable regions along with quality of the NGS metrics. However, analytical validation was based on only three reference samples . Furthermore, one publication evaluating CYP2D6 SMRT third generation sequencing reported an assessment of precision in terms of triplicate intra-run sequencing of a validation sample and inter-run sequencing . In summary, few studies evaluate analytical performance criteria such as limit of detection, accuracy and precision of second generation sequencing panels and fewer still focus on third generation sequencing. Analytical specificity in terms of interference, cross-reactions and cross-contamination has not been assessed or provided by any of these published evaluations. Therefore, according to the information provided on the platforms Pubmed and Pubmed Central, none of the NGS- based tests has an analytical performance evaluation that has been executed according to the current recommendations in literature and of the FDA. Although, several studies compare WES and WGS with other molecular genetic techniques in terms of variant detection in pharmacogenes coding drug metabolizing enzymes, here also no articles were identified that carried out an analytical validation in terms of evaluating performance criteria as recommended for NGS-based IVDs or LDTs . The present literature assessment shows that published studies on the analytical validation of NGS based pharmacogenetic tests concerning drug metabolizing enzymes are scarce. Specifically, this applies for WES, WGS and also third generation sequencing, as currently represented by Single Molecule Real Time sequencing using the PacBio (Pacific Biosciences of California, Inc, California, USA) and increasingly also Oxford Nanopore (Oxford Nanopore Technologies (ONT) Inc, Oxford, UK) platforms. According to currently available, but non-binding, FDA recommendations, the evaluations identified are not sufficiently comprehensive. Furthermore, also, relevant analytical performance metrics according to the College of American Pathologists (CAP) guidance documents and the MM09 guideline of the Clinical and Laboratory Standards Institute (CLSI) were not addressed by the evaluated publications. In addition to the aforementioned analytical performance criteria recommended by the FDA, the CAP/ CLSI analytical performance metrics also include robustness, reportable range and reference interval. Metrics such as interference and cross-reactivity are listed in the CAP/CLSI MM09 test validation worksheet as IVD (not LDT) terminology only and cross-contamination is not addressed (Supplementary material ). Required analytical performance characteristics for IVDs in general, listed by the US Medicare, Medicaid, and Clinical Laboratory Improvement Amendments (CLIA) programs § 493.1253(c), comprise similar metrics to CAP/CLSI MM09 for IVDs, however not specified for NGS applications . In the EU, required general analytical performance characteristics provided by the IVDR (Annex I chapter II 9.1 (a)) also include a larger scope and correspond to the CAP and CLIA criteria and cover also the specified FDA recommendations for NGS (Supplementary material ). The performance metrics addressed by the evaluated publications, therefore do not meet CLIA and IVDR requirements as well. Such tests therefore are not suitable to be marketed and applied for clinical pharmacogenetic profiling in the USA or the EU. For an appropriate analytical validation of an NGS based qualitative test for pharmacogenetic profiling at least accuracy, precision, limit of detection and specificity should be assessed as recommended by the FDA guidance document “for Stakeholders and Food and Drug Administration Staff” and as analyzed in this literature search. Furthermore, we suggest that also the reportable range and the range of outcomes expected in a normal population (normal interval) should be determined as recommended additionally by the CAP and the CLIA. The most extensive performance evaluations were identified for targeted NGS panels that encompass a large variety of target genes, but without a direct focus on pharmacogenetic profiling. Here, Silver A et al. (2022) provided the best analytical validation in terms of the scope of performance criteria. The evaluation included accuracy and precision as recommended by the FDA and detection rates of targeted alleles, however did not include analytical specificity in terms of interference, cross-reactions and cross-contamination (Supplementary material ). Furthermore, the reportable range was determined, which was however not a criterion of the present literature assessment. Ramudo-Cela L et al. (2020) evaluated accuracy of a targeted NGS panel according to FDA recommendations and partly precision in terms of reproducibility as sample triplicates were applied. The NGS panels however comprise a total of 430 (Silver A et al., 2022) and 389 (Ramudo-Cela L et al., 2020) genes and too large a panel size is not recommended as it can affect the efficiency of the laboratory, due to the complexity in interpreting the outcomes and the depth of coverage required . Beaubier N et al. (2019) evaluated accuracy and LOD for different variant types as recommended by the FDA, however the intended use of the panel was to detect somatic alterations and microsatellite instability in solid tumors. It is therefore not validated for pharmacogenetic profiling although CYP2D6 and UGT1A1 were included in the panel. In the publication by Klein K et al. (2019), the lowest level of concordance with FDA recommendations was found with respect to the evaluation of performance criteria for a targeted NGS panel. Minor allele frequencies were addressed, however without determining detection limits and accuracy was evaluated as overall accuracy and not as recommended by the FDA via PPA, NPA and PPV. No further criteria were evaluated. Most of the publications assessed, include profiling of the highly polymorphic gene CYP2D6 which is also a challenging locus to analyze due to complex structural variation. A more comprehensive analysis of complex genetic information to assess the pharmacogenetic profile associated with the drug metabolism of a patient can be provided by WES and WGS . In our literature search two publications were identified evaluating performance of WES or WES and WGS in terms of also the according pharmacogenes of interest (Supplementary material ) . Both publications mainly focus on concordance analyses with orthogonal methods and quality metrics such as call rates or depth of coverage and not on the FDA analytical validation criteria. However, Gulilat M et al. (2019) evaluated accuracy of WES only without PPV as recommended and Yang W et al. (2016) addressed minor allele fractions only without reporting the determination of detection limits according to FDA guidance. Recommendations providing guidance on the analytical validation of WGS in clinical context were published in recent years. They indicate that traditional approaches with regard to the evaluation of performance metrics for the complete NGS based assay alone is not sufficient . Large-scale massive parallel sequencing applications such as WES and WGS do not target specific genes, sequence contexts, diseases or associated variant types per se. Therefore, here the aim of analytical validation is to evaluate and determine reliable metrics for a suitable performance with regard to the largest percentage of area analyzed to assure high quality of sequencing results overall. It is intended that variant calling over the target region is sensitive and precise and also allows for the determination of regions or bases that fail requirements for appropriate variant calls . Metrics considering aspects of genome complexity such as sequence content and the variety of different variant types should also be assessed, as variability in the accuracy of variant calling is also context-driven. Diagnostic accuracy is often evaluated in terms of sensitivity (PPA) and specificity (NPA). However, due to the expected large quantity of true negatives in WGS, precision is recommended as a more helpful metric for accuracy than specificity (NPA). In general, also for WGS, the assessment and extent of metrics as published in the non-binding guidance document for NGS by the FDA is recommended and the number of samples is suggested to be adapted to variant type or the analyzed region . In the present literature search, the only publication evaluating performance of WGS for pharmacogenetic profiling did not address analytical performance criteria including precision as recommended. Current studies evaluating NGS-based test performance for pharmacogenetic profiling mainly provide concordance assessments with orthogonal methods (Supplementary material ). Compared to targeted NGS panels analytical sensitivity for exome sequencing is lower as depth of coverage is not uniform or insufficient affecting also analytical specificity and variant calling. Thus, Sanger Sequencing is needed to complete content and identify false positive calls . For the profiling of pharmacogenes, so far, short read WGS has been proposed to be a more suitable NGS application than WES as it provides a more comprehensive testing without increasing costs significantly . With regard to further genome wide tests such as chromosomal microarrays the performance of WGS for CNV detection has been shown to be at least comparable . However, in the current literature no appropriate analytical validation of WGS for pharmacogenomic profiling has been addressed. Linderman et al. (2014) provided an evaluation of the analytical performance of WES and WGS in general using five reference samples and technical replicates, by assessing the reproducibility (intra-run, inter-run, inter-mode and inter-machine), the concordance with microarrays and Sanger Sequencing and the sensitivity. Here, for WES and WGS the focus was on sensitive and precise as possible variant calling over the target region and thereby the determination of bases for which appropriate requirements were not met. Here, the performance analysis was based on ranges recommended by The American College of Medical Genetics and Genomics . However, short read based NGS currently applied for WES or WGS can be prone to coverage bias in GC rich regions also due to a dependence on PCR for template amplification and therefore may miss hidden genetic variation in numerous genomic regions with high GC content. Therefore, disruptions of relevant genes with such GC rich regions or impairments of dosage-sensitive genes leading to dosage change may remain undetected in patients assumed to have a genetic disorder. Many structural variants and indels undetected by short read based NGS were detected by long read sequencing in recent studies . However, the low throughput in the application of platforms such as Nanopore ONT and PacBio prohibits cost-effective deep sequencing without enrichment of the target sequence of interest . Nevertheless, barcoding and multiplexing of samples has been reported to reduce per sample expenses in CYP2D6 SMRT sequencing and enable cost comparability to commercial genotyping assays. Moreover, full gene SMRT provides further CYP2D6 variant information such as e.g. the characterization of CNVs, sub allele resolution and novel tandem arrangement which may be convenient for routine clinical testing . Also in long read sequencing PCR based enrichment can introduce phasing errors and amplification bias and also leads to a loss of information on native modifications. CRISPR-Cas-based enrichment strategies such as those for Nanopore Cas9-targeted sequencing (nCATS) can prevent such bias and information loss. Still, a relatively high amount of native DNA is needed and limitations in the detection of some SNVs remain. For the performance of this application, however, it is assumed that identification of SNVs will progress with accelerating improvements of algorithms for base- and variant-calling in future and will also increase performance for mutation surveillance . Furthermore, many recent publications indicate that these evolving techniques hold considerable promise for clinical application . However, for third generation sequencing with regard to drug metabolizing enzymes only one study providing validation results and performance comparisons with currently FDA cleared or approved nucleic acid based tests was detected via our literature screening based on the keyword search. The study detected was however based on low numbers of tested samples. Still, it provides a promising outlook for third generation sequencing technologies to detect variation also in the complex CYP2D6 locus. Currently, WES and WGS are performed based on second generation sequencing techniques as the emerging techniques in third generation sequencing are characterized by a high error rate. However, third generation sequencing provides a promising utility for the analysis of structurally complex genomic regions and may become a helpful tool for the assessment of the whole genome for clinical use . Thus, they are increasingly provided by commercial service laboratories, currently however for research evaluations only . A limitation of the present assessment is that other suitable keywords such as performance or concordance were not applied in this literature search. Therefore, several publications focusing on analytical validation in terms of analytical performance may have been missed. Another limitation is that NGS based diagnostic tests applied as clinical assays are mainly validated and applied in laboratories as LDT and documented data on analytical validation is not publicly available . In 2017, the FDA published a discussion paper on the oversight of LDTs. It addressed also an increase of transparency in terms of providing publicly available information on analytical and clinical validity of LDTs. However, a final guidance was not established . Due to the new in vitro diagnostic regulation in the EU it will be mandatory for laboratories to provide appropriate performance evaluations to the regulatory authorities. The establishment of transparency in terms of publicly available information on the validity of laboratory developed and applied NGS technologies is not addressed by the new IVDR in the EU . Clinical NGS and knowledge in pharmacogenomics are evolving rapidly. However, current studies evaluating NGS-based test performance for pharmacogenetic profiling mainly provide concordance assessments with orthogonal methods without considering further relevant analytical performance characteristics recommended by the FDA or required according to the CAP and CLIA guidance and in the EU according to the IVDR. For the use of next generation sequencing applications such as WES, WGS and also third generation sequencing for pharmacogenetic profiling in the clinical setting, a more comprehensive analytical performance analysis including at least the recommended performance characteristics is needed. However, it became clear that for publication of such assessments a standard in reporting on analytical validation of NGS based tests is not in place and may be increasingly necessary to facilitate the implementation of NGS tests in clinical use. Feasibility evaluations to extract clinically important pharmacogenetic information from sequencing data obtained by diagnostic testing for other clinical conditions suggest a potential to reuse relevant information for therapy management. Therefore, regulatory guidance to establish analytical validation requirements also for such approaches may be necessary. Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 |
Early noninvasive prenatal paternity testing by targeted fetal DNA analysis | b133be97-40e3-4e96-8668-27013de04e1d | 10372148 | Forensic Medicine[mh] | In prenatal care, technical advances in the analysis of fetal DNA circulating in maternal blood (cffDNA) have enabled great progress in early noninvasive prenatal screening and diagnosis of several genetic conditions – . These advances could also contribute to the handling of difficult situations in forensic science, such as the question of paternity during early pregnancy of sexually abused women. A noninvasive test would offer the possibility of avoiding procedures associated to a certain degree of harm to the mother and the fetus (amniocentesis – and chorionic villus sampling , ), in addition to addressing the question of paternity as early as six weeks of amenorrhea. Because the presence of fetal DNA in the blood stream of the mother is due to the continuous remodeling of the placenta, with trophoblast cells undergoing apoptotic events, its analysis presents several challenges to traditional DNA profiling. CffDNA is mostly short (about 160–200 base-pairs) and it represents 5–15% of the total cell free plasma DNA – , thus generating an in vivo DNA mixture characterized by a large excess of maternal DNA (unbalanced DNA mixture) , . As the background maternal DNA interferes with the detection of fetal DNA , analytical methods need to be either highly sensitive for DNA mixture deconvolution or specific to the fetal DNA fraction. Early studies primarily focused on the use of single-nucleotide polymorphisms (SNPs) because of their short size and the possibility of scaling-up by high throughput technologies. First, SNP microarray were employed for large-scale marker analysis (High-throughput SNP array) – which compensates for the SNP’s low discrimination power. The advent of massive parallel sequencing (MPS) – offered an extremely efficient method for fetal DNA analysis from maternal plasma: the sequencing of thousands to millions of DNA molecules on a genome-wide scale, makes it possible to reveal the minor fetal genetic component against a background of highly homologous maternal DNA. Yet, several studies indicated that thousands of SNP markers are recommended to achieve a high accuracy as the loss in uninformative SNPs and false negatives due to the low fetal fraction can be high , , , , . Improved results came with the use of SNP sets selected based on population variability data (356 SNPs) and the use of Unique Molecular Identifiers (UMI) for more reliable genotyping . Finally, in an effort of reducing the number of possibly linked loci and high costs, an MPS-based assay targeting 60 microhaplotypes was also tested , . Yet, these studies show that a well-accepted method including careful data interpretation has not been developed. The evaluation of factors influencing the accuracy of the results, such as degree of marker polymorphism and number for informativeness (maximize number but lowering incidental findings), sequencing depth for a sufficient signal-to-noise ratio, thresholds of fetal fraction as well as sequencing strategy is still ongoing, including validation of the method on a large clinical sample set from early stages of pregnancy. Alternatives approaches aiming at reducing the recurrent problem of low signal-to-noise ratio propose the targeting the fetal DNA with allele-specific amplifications. These include methylation markers and more frequently compound markers such as SNP-SNPs and DIP-STRs , . DIP-STR markers can resolve extremely unbalanced two-source DNA mixtures of same-or-opposite sex donors, up to a 1:1000 minor:major DNA ratio. Positive results were obtained in targeting DNA sequences unique to the fetal DNA transmitted by the father. These sequences are biallelic deletion/insertion polymorphisms (DIP) of several nucleotides located physically very close to the STR marker, for combined analysis. The multiallelic haplotype composed of both DIP and STR alleles is analyzed by using PCR primers overlapping the deleted/inserted sequence (S-DIP, L-DIP primers) on one side and downstream the STR region on the other side (STR primer) (Fig. ). A forensic set of 10 markers was validated for casework and a larger set of 28 DIP-STRs showed an efficient detection of fetal DNA in the plasma of 48 women, whose blood was collected in advanced pregnancy . DIP-STR markers were also used to test zygosity of twin pregnancies . These results show that DIP-STR can be used as a supplementary method especially when cffDNA accounts for less than 10% of total cell free DNA. In this study, we aimed at improving the number of markers validated for prenatal paternity testing on a large number of cases all collected early in gestation. We studied a set of 47 DIP-STRs in 87 pregnancies, with blood samples collected throughout pregnancy starting from the first trimester. Here, we also provide the general framework for the paternity index calculation with an evaluation of the minimum number of informative markers necessary for determining paternity.
Type of informative markers As described above, the analysis of circulating fetal DNA is based on allele-specific amplifications of the paternally transmitted DIP-STR haplotype when its DIP allele is not shared with the maternal DNA (Fig. ). To do so, primers for cffDNA analysis are selected based on the DIP genotype of the mother. Markers informative for the genotypes of the fetus are always DIP homozygous in the mother (SS or LL) and they are analyzed using primers specific to the opposite DIP allele (L- and S-primers, respectively). The fetal allele is then targeted if a DIP-STR haplotype containing a DIP allele different from the mother is transmitted by the father. Considering all possible genotype assortments, three types of informative markers exist (Fig. ): (i) markers of type A, the father is homozygous for the other DIP allele than the one of the mother. In this case, the paternal haplotype can be targeted in cffDNA analysis; (ii) markers of type B, the father is heterozygous at the DIP locus, in this case the cffDNA can be targeted only if the paternally transmitted haplotype carries the DIP allele different from the mother; finally, (iii) markers of type C, the father is homozygous for the same DIP allele of the mother, in this case no paternal haplotype can be targeted in maternal plasma. Any DIP-STR result inconsistent with the expected genotype of the alleged father, including any positive results from markers of type C, would support the exclusion of the alleged father. In Table is reported the average number of informative marker observed in the cases presented. When using the set of 47 DIP-STRs the number of parental genotypes of type A are about five and the number of either type B or C is 11–12, with a total of about 28 informative markers available for each family to determine the paternity. Performance of DIP-STR specific amplification of cffDNA Singleplex genotyping of markers of type A was done on selected DIP-STRs with the aim of testing their performance in the first trimester across cases. The consumption of cffDNA was optimized to enable the testing of several markers per case and each marker across several cases. Moreover, to further control for PCR specificity, the priority was given to the genotyping of cases with fathers heterozygous for the STR. Samples from the second and third trimester were used for confirmation. The electropherograms of few representative samples are illustrated in Fig. in addition to examples of positive results from the longest DIP-STRs of our collection (Fig. ). We didn’t observe a difference between standard EDTA and Streck Cell-Free DNA BCT ® collection tubes in the success rate of cffDNA amplification. All cases of amplification failure of fetal DNA clustered in the first trimester, with 10 false negative cases out of 164 fetal allele interrogated (Table ). This false negative rate is of about 6% in the first trimester and close to zero in samples from second and third trimester. Eight markers longer than 300 bp were not included in this estimate, as well as the marker rs71070706-rs147416097 that is not sensitive enough and should be eliminated for clinical applications. The markers associated to these weak and negative results are all characterized by low PCR efficiency as previously reported, Refs. , , these are L-rs55886629-rs56078928, L-rs59055342-rs71557834, L-rs61437086-rs140235473, S-rs71725104-rs10656000, S-rs138331044-rs113027169, S- and L-rs145299629-rs200177067. No false positive amplification of maternal alleles were observed except for few markers L-rs56821990-rs200925554, S- and L-rs3216342-rs10639027, S-rs61437086-rs140235473, S-rs140762-rs139631506, S-rs552898832-rs58403232, which produce extra peaks 100/200 bp far from the expected signal of constant size, each time an excess of DNA is used. Paternity determination After the initial screening of the 47 DIP genotypes of the parents, the complete transmitted DIP-STR haplotype was determined only for markers of type A. Therefore, the PI calculated for each case is based on genotypes of type A. As expected, CPI values increase with the number of informative markers. At least seven markers are necessary to reach a CPI higher than 1,000, which is the value accepted in most country for the verbal conclusion about practically proven paternity. As expected, one or two markers of type A are never sufficient to demonstrate paternity (Fig. ). Markers of type C, as markers of type A, are extremely useful for the question of paternity, especially for a quick and clear exclusion of paternity. Markers of type C for the alleged father can be of type A or of type B for the biological father, the associated probabilities of such genotype assortment are (s 4 l 2 ) + (l 4 s 2 ) + (s 4 2sl) + (l 4 2sl). It should be considered that when the biological father is heterozygous, his transmitted allele can be informative or not. It follows that, considering half of the cases where the biological father is DIP heterozygous, the genotypes potentially showing allelic inconsistencies have the following probabilities: (s 4 l 2 ) + (l 4 s 2 ) + (s 4 sl) + (l 4 sl). This probability is roughly 6.25%. In these cases, positive fetal DNA results are obtained which are not expected based on the alleged father’s genotype. In addition to these cases, markers of type C for the biological father can be of type A or B for the alleged father. They similarly contribute to paternity exclusion: (s 4 l 2 ) + (l 4 s 2 ) + (s 4 sl) + (l 4 sl). Under these genotypes, the expected fetal DNA results is not observed. This means that, considering balanced S and L allele frequencies, 12.5% of the markers show a genetic inconsistency in cases of non-paternity. It should be noted that, when considering the whole DIP-STR haplotype information, many more markers (about 28 when using a set of 47 markers) provide data for determining paternity (Table ).
As described above, the analysis of circulating fetal DNA is based on allele-specific amplifications of the paternally transmitted DIP-STR haplotype when its DIP allele is not shared with the maternal DNA (Fig. ). To do so, primers for cffDNA analysis are selected based on the DIP genotype of the mother. Markers informative for the genotypes of the fetus are always DIP homozygous in the mother (SS or LL) and they are analyzed using primers specific to the opposite DIP allele (L- and S-primers, respectively). The fetal allele is then targeted if a DIP-STR haplotype containing a DIP allele different from the mother is transmitted by the father. Considering all possible genotype assortments, three types of informative markers exist (Fig. ): (i) markers of type A, the father is homozygous for the other DIP allele than the one of the mother. In this case, the paternal haplotype can be targeted in cffDNA analysis; (ii) markers of type B, the father is heterozygous at the DIP locus, in this case the cffDNA can be targeted only if the paternally transmitted haplotype carries the DIP allele different from the mother; finally, (iii) markers of type C, the father is homozygous for the same DIP allele of the mother, in this case no paternal haplotype can be targeted in maternal plasma. Any DIP-STR result inconsistent with the expected genotype of the alleged father, including any positive results from markers of type C, would support the exclusion of the alleged father. In Table is reported the average number of informative marker observed in the cases presented. When using the set of 47 DIP-STRs the number of parental genotypes of type A are about five and the number of either type B or C is 11–12, with a total of about 28 informative markers available for each family to determine the paternity.
Singleplex genotyping of markers of type A was done on selected DIP-STRs with the aim of testing their performance in the first trimester across cases. The consumption of cffDNA was optimized to enable the testing of several markers per case and each marker across several cases. Moreover, to further control for PCR specificity, the priority was given to the genotyping of cases with fathers heterozygous for the STR. Samples from the second and third trimester were used for confirmation. The electropherograms of few representative samples are illustrated in Fig. in addition to examples of positive results from the longest DIP-STRs of our collection (Fig. ). We didn’t observe a difference between standard EDTA and Streck Cell-Free DNA BCT ® collection tubes in the success rate of cffDNA amplification. All cases of amplification failure of fetal DNA clustered in the first trimester, with 10 false negative cases out of 164 fetal allele interrogated (Table ). This false negative rate is of about 6% in the first trimester and close to zero in samples from second and third trimester. Eight markers longer than 300 bp were not included in this estimate, as well as the marker rs71070706-rs147416097 that is not sensitive enough and should be eliminated for clinical applications. The markers associated to these weak and negative results are all characterized by low PCR efficiency as previously reported, Refs. , , these are L-rs55886629-rs56078928, L-rs59055342-rs71557834, L-rs61437086-rs140235473, S-rs71725104-rs10656000, S-rs138331044-rs113027169, S- and L-rs145299629-rs200177067. No false positive amplification of maternal alleles were observed except for few markers L-rs56821990-rs200925554, S- and L-rs3216342-rs10639027, S-rs61437086-rs140235473, S-rs140762-rs139631506, S-rs552898832-rs58403232, which produce extra peaks 100/200 bp far from the expected signal of constant size, each time an excess of DNA is used.
After the initial screening of the 47 DIP genotypes of the parents, the complete transmitted DIP-STR haplotype was determined only for markers of type A. Therefore, the PI calculated for each case is based on genotypes of type A. As expected, CPI values increase with the number of informative markers. At least seven markers are necessary to reach a CPI higher than 1,000, which is the value accepted in most country for the verbal conclusion about practically proven paternity. As expected, one or two markers of type A are never sufficient to demonstrate paternity (Fig. ). Markers of type C, as markers of type A, are extremely useful for the question of paternity, especially for a quick and clear exclusion of paternity. Markers of type C for the alleged father can be of type A or of type B for the biological father, the associated probabilities of such genotype assortment are (s 4 l 2 ) + (l 4 s 2 ) + (s 4 2sl) + (l 4 2sl). It should be considered that when the biological father is heterozygous, his transmitted allele can be informative or not. It follows that, considering half of the cases where the biological father is DIP heterozygous, the genotypes potentially showing allelic inconsistencies have the following probabilities: (s 4 l 2 ) + (l 4 s 2 ) + (s 4 sl) + (l 4 sl). This probability is roughly 6.25%. In these cases, positive fetal DNA results are obtained which are not expected based on the alleged father’s genotype. In addition to these cases, markers of type C for the biological father can be of type A or B for the alleged father. They similarly contribute to paternity exclusion: (s 4 l 2 ) + (l 4 s 2 ) + (s 4 sl) + (l 4 sl). Under these genotypes, the expected fetal DNA results is not observed. This means that, considering balanced S and L allele frequencies, 12.5% of the markers show a genetic inconsistency in cases of non-paternity. It should be noted that, when considering the whole DIP-STR haplotype information, many more markers (about 28 when using a set of 47 markers) provide data for determining paternity (Table ).
This study demonstrates that DIP-STR markers can be used to detect paternal alleles in cffDNA from seven weeks into pregnancy and that a small set of markers (about 40) is sufficient to address the question of paternity. Our results are based on a large number of cases, namely 87, all collected during the first trimester with additional samples per case collected in the second and/or third trimester for results confirmation. This enabled us to have for each marker at least one informative case and possibly several. It should be noted, that cffDNA accounts for approximately 5–20% of the total cffDNA, with an upward trend evident through pregnancy. Previous studies indicate the fetal fraction in the 1 st trimester as low as 3–4% in 30% of the cases , . It is therefore essential to test the feasibility and accuracy of the method in real cases and early during pregnancy. Recent reports of similar studies either didn’t include samples from the first trimester , , or included about 10 cases – . This doesn’t allow to have all markers informative in at least one case and therefore evaluate marker specific false positive/negative rate which may vary. In our study, we tested all DIP-STR markers available, including long ones (> 300 bp) and less specific/sensitive to produce empirical data for marker exclusion. The failure rate of each marker in the first trimester was estimated on a selection of cases showing informative genotypes of type A, with parents opposite homozygous for the DIP alleles. Here, the false negative rate reached 6%. Interestingly, markers longer than 320 bp to 386 bp worked in 8/22 PCR assays, yet we didn’t include them in the estimate of the false negative rate. The partial success with much longer markers is probably due to the allele-specificity of the method which enables the use of the much reduced fetal DNA fraction that is longer than the average apoptotic DNA fragments. If necessary, long DIP-STR can be used for cffDNA detection if results are interpreted using the related increased drop-out probability. Few observed false positive signals were limited to extra peaks far from the expected allele size that could be easily excluded with the GeneMapper ® software’s bin set system for scoring the data. We didn’t observe a difference between standard EDTA and Streck Cell-Free DNA BCT ® collection tubes in the quality and/or quantity of cffDNA. This may be due to the fact that the Streck Cell-Free DNA BCT ® collection tubes were still stored at 4 °C immediately after collection and sample processing was never delayed longer than four hours. Finally, the comparison of false negative rates of DIP-STRs to other markers was limited by the reduced number of studies including a sufficient number of first trimester cases. When these cases are included, often results are pooled for the whole sample collection including all time points. Several studies by Ou’s group employing the MPS technology reported false negative rates of 20–50% , , . The lowest reported values are around 15% when using microhaplotypes and fetal fractions estimated at less than 5% . With the DIP-STR markers, the molecular approach employed for DNA mixture deconvolution assures a fetal specificity that is superior to any currently proposed solution. Interestingly, the minor DNA specificity achieved by combining two types of polymorphisms also contributes to the enhanced informativeness of the markers. In previous studies, the use of SNPs associated with MPS showed that a larger number of SNPs compared with STRs, needed to be interrogated to compensate for the lower discrimination power. About 50 SNPs, having allele frequencies between 0.2 and 0.8, are equivalent to the use of 12 STRs in postnatal tests . In terms of variability, DIP biallelic polymorphisms are expected to perform as SNPs, yet the haplotypes from combining STR variants are associated with much smaller allele frequencies and higher discrimination power. Moreover, MPS SNP genotyping without a method for targeting the minor DNA, generates noisy data resulting from low fetal allele concentrations, against which the analysis of more SNPs is required to allow for filtering low-quality data. Deep sequencing results require an effective analysis algorithm which can indeed be developed, yet it is still not clear whether a routine high-throughput service laboratory would have a standard analysis pipeline that doesn’t need to be adjusted depending on the efficiency of each experiment, read coverage and accuracy of the estimated fetal fraction. It is worth emphasizing that our method currently based on PCR-CE, is compatible with a procedure that includes screening of informative markers by DIP multiplex genotyping of the reference DNA (mother and alleged father) followed by singleplex genotyping of selected informative markers. To make use of all the genetic data and avoid the risk of insufficient markers for paternity determination it is necessary to develop two large multiplex for S- and L-allele specific genotyping. Depending on the number of markers selected, the MPS technology may be solution of choice. In this case, the signal to background noise ratio would still be high because of the target enrichment of fetal DNA. Moreover, the results would include all informative and uninformative markers to produce high-confidence fetal genotype calls. The associated Bayesian framework would provide an unsupervised approach convenient for also testing several alleged fathers with one set of data. However, it should be noted that as few as seven informative markers with parents DIP homozygous for different alleles (markers of type A), were sufficient in our cases to determine paternity. Moreover, at least 12.5% of the genotyped markers are expected to show clear cases of paternity exclusion (unexpected positive or negative fetal DNA amplifications) for an unrelated alleged father. This means that many cases of prenatal paternity testing can be handled with a quick and cost effective solution that uses the technology available in all forensic genetics laboratories. Certainly, markers of type B and the whole DIP-STR haplotype information should still be considered to increase the weight of the evidence. In conclusion, the current study provides several elements supporting the further use of DIP-STR markers for non-invasive prenatal paternity testing. These are: comprehensive marker set, first trimester validation of all markers, values of false positive and negative rates and a statistical framework for interpretation. This work provides the basis for the forensic development of a standardized prenatal paternity test. Further, deep sequencing based multiplexing is recommended for the improvement of the testing efficiency.
Sample collection Inclusion criteria for enrolled couples were singleton pregnancies with known paternity. Maternal blood samples (10 ml) were drawn longitudinally from 87 women at 7–13 weeks of amenorrhea (first trimester) (Table ). For each case, additional blood samples were collected during either the second trimester (14–26 weeks) and/or the third trimester (27–40 weeks). Venous blood samples were drawn into EDTA blood collection tubes and Streck Cell-Free DNA BCT ® tubes (Streck, USA). Plasma was separated from the blood cells via double centrifugation (1,600 g for 10 min, tube transfer, and centrifugation at 18,000 g for 10 min) within 2–4 h of blood being drawn. Aliquots of 1 ml were stored at − 20 °C until further processing . DNA samples from both parents of the developing baby were collected by buccal swab. The current study was approved by the Centre Hospitalier Universitaire Vaudois and Université de Lausanne institutional review board, research protocol number (2019-01601 CER-VD). Written informed consent was obtained from all participants. All methods were performed in accordance with the relevant guidelines and regulations. DNA extraction Cell free circulating DNA was extracted in duplicate from 2 ml of plasma by using the QIAamp Circulating Nucleic Acid Kit (Qiagen AG, Basel, Switzerland) according to the manufacturer’s protocol. Absorbed DNA was eluted with 60 µl of provided elution buffer. The synthetic DNA RT-SPCY-T02 (Eurogentec, Angers, France) was added to the plasma to function as positive control for circulating DNA extraction. According to the manufacturer’s protocol, 2 ul of a tenfold diluted RT-SPCY-T02 was added to 2 ml of plasma. Reference buccal samples were extracted using the QIAamp DNA Mini Kit (Qiagen AG, Basel, Switzerland) according to the manufacturer’s protocol and eluted in 100 µl final volume. Both genomic and circulating DNA samples were stored at − 20 °C. DNA was extracted using the QIAamp DNA Mini kit (Qiagen AG Switzerland) according to the manufacturer’s guidelines and quantified using the kit QuantiFiler Trio on a QuantStudio™ 5 System (Life Technologies Europe, Zug, Switzerland). The commercial DNAs CEPH 1347-02, Control DNA 007 (Life Technologies Europe, Zug, Switzerland), 2800 M Control DNA (Promega, Dübendorf, Switzerland) were genotyped as a reference controls for allele designation. DIP-STR genetic markers The DIP-STR markers genotyped for this study include 24 markers previously published , , and 23 newly developed (Table ). PCR reactions for the markers were performed as previously published DIP-STR genotyping protocols , , , using 10 ul of cffDNA. S- and L-DIP-STR specific amplifications were done in singleplex according to published protocols , , , . For cffDNA amplifications PCR conditions are modified to increase sensitivity with 36 cycles and twice the amount of PCR primers. To identify informative markers for plasma DNA analysis, reference DNA samples from the mothers and the fathers were first genotyped for 47 DIP markers using seven multiplex reactions as described in Supplementary Table and in Moriot et al. 2019 . PCR fragments were separated by capillary electrophoresis after adding 1 μl PCR amplicon to 8.5 μl deionized formamide HI-DI (Life Technologies Europe, Zug, Switzerland) and to 0.5 μl 600 LIZ size standard (Life Technologies Europe, Zug, Switzerland). Capillary electrophoresis was performed using an ABI PRISM 3500 xl Genetic Analyzer (Life Technologies Europe, Zug, Switzerland) according to the manufacturer's instruction and analyzed using the GeneMapper ® ID v3.2.1 software (Life Technologies Europe, Zug, Switzerland), with a minimum peak height threshold of 50 Relative Fluorescence Unit (RFU). The commercial DNA CEPH 1347–02 (Life Technologies Europe, Zug, Switzerland) was used as positive control of amplification and internal standard for allele designations. For standard PCR reactions (28–30 cycles) 0.5 ng of commercial reference DNA was used, for all PCR reactions with increased number of cycles (34–36) 0.0125 ng of commercial reference DNA was used. Probability of paternity The paternity index (PI) was calculated as the ratio of likelihood values of two hypotheses (H0: the test man is the biological father of the child; H1: the test man is unrelated) based on 158 European allele frequency . H0 is equal to 1 if the alleged father is homozygous for the observed DIP-STR haplotype shared with the child, and it is 0.5 when the alleged father is heterozygous for the observed DIP-STR. H1 corresponds to the frequency of all the homozygous and heterozygous individuals of the observed DIP-STR in the populations. H0/H1 = 1/(the frequency of the observed DIP-STR) if the alleged father is homozygous for the DIP-STR and 0.5/(the frequency of the observed DIP-STR) if the alleged father is heterozygous for the DIP-STR). The CPI is the product of the PI of unlinked loci. All marker combinations used for each CPI calculation included unlinked markers either located on different chromosomes or chromosomal arms. Those located on the same chromosomal region were more distant than 6 Mb, on average at about 40 Mb distance, with the exception of two cases at 1.5 Mb and three cases at 0.5 Mb which tested negative for allelic associations in Europe . According to the Swiss national technical specification for parentage testing, inclusion of parenthood is noted when the CPI is greater than 369 which corresponds to a log 10 (CPI) of 2.57. Ethics approval The current study was approved by the Centre Hospitalier Universitaire Vaudois and Université de Lausanne institutional review board, research protocol number (2019–01,601 CER-VD). Consent to participate Each blood sample used was freely donated under conditions of informed consent to participate.
Inclusion criteria for enrolled couples were singleton pregnancies with known paternity. Maternal blood samples (10 ml) were drawn longitudinally from 87 women at 7–13 weeks of amenorrhea (first trimester) (Table ). For each case, additional blood samples were collected during either the second trimester (14–26 weeks) and/or the third trimester (27–40 weeks). Venous blood samples were drawn into EDTA blood collection tubes and Streck Cell-Free DNA BCT ® tubes (Streck, USA). Plasma was separated from the blood cells via double centrifugation (1,600 g for 10 min, tube transfer, and centrifugation at 18,000 g for 10 min) within 2–4 h of blood being drawn. Aliquots of 1 ml were stored at − 20 °C until further processing . DNA samples from both parents of the developing baby were collected by buccal swab. The current study was approved by the Centre Hospitalier Universitaire Vaudois and Université de Lausanne institutional review board, research protocol number (2019-01601 CER-VD). Written informed consent was obtained from all participants. All methods were performed in accordance with the relevant guidelines and regulations.
Cell free circulating DNA was extracted in duplicate from 2 ml of plasma by using the QIAamp Circulating Nucleic Acid Kit (Qiagen AG, Basel, Switzerland) according to the manufacturer’s protocol. Absorbed DNA was eluted with 60 µl of provided elution buffer. The synthetic DNA RT-SPCY-T02 (Eurogentec, Angers, France) was added to the plasma to function as positive control for circulating DNA extraction. According to the manufacturer’s protocol, 2 ul of a tenfold diluted RT-SPCY-T02 was added to 2 ml of plasma. Reference buccal samples were extracted using the QIAamp DNA Mini Kit (Qiagen AG, Basel, Switzerland) according to the manufacturer’s protocol and eluted in 100 µl final volume. Both genomic and circulating DNA samples were stored at − 20 °C. DNA was extracted using the QIAamp DNA Mini kit (Qiagen AG Switzerland) according to the manufacturer’s guidelines and quantified using the kit QuantiFiler Trio on a QuantStudio™ 5 System (Life Technologies Europe, Zug, Switzerland). The commercial DNAs CEPH 1347-02, Control DNA 007 (Life Technologies Europe, Zug, Switzerland), 2800 M Control DNA (Promega, Dübendorf, Switzerland) were genotyped as a reference controls for allele designation.
The DIP-STR markers genotyped for this study include 24 markers previously published , , and 23 newly developed (Table ). PCR reactions for the markers were performed as previously published DIP-STR genotyping protocols , , , using 10 ul of cffDNA. S- and L-DIP-STR specific amplifications were done in singleplex according to published protocols , , , . For cffDNA amplifications PCR conditions are modified to increase sensitivity with 36 cycles and twice the amount of PCR primers. To identify informative markers for plasma DNA analysis, reference DNA samples from the mothers and the fathers were first genotyped for 47 DIP markers using seven multiplex reactions as described in Supplementary Table and in Moriot et al. 2019 . PCR fragments were separated by capillary electrophoresis after adding 1 μl PCR amplicon to 8.5 μl deionized formamide HI-DI (Life Technologies Europe, Zug, Switzerland) and to 0.5 μl 600 LIZ size standard (Life Technologies Europe, Zug, Switzerland). Capillary electrophoresis was performed using an ABI PRISM 3500 xl Genetic Analyzer (Life Technologies Europe, Zug, Switzerland) according to the manufacturer's instruction and analyzed using the GeneMapper ® ID v3.2.1 software (Life Technologies Europe, Zug, Switzerland), with a minimum peak height threshold of 50 Relative Fluorescence Unit (RFU). The commercial DNA CEPH 1347–02 (Life Technologies Europe, Zug, Switzerland) was used as positive control of amplification and internal standard for allele designations. For standard PCR reactions (28–30 cycles) 0.5 ng of commercial reference DNA was used, for all PCR reactions with increased number of cycles (34–36) 0.0125 ng of commercial reference DNA was used.
The paternity index (PI) was calculated as the ratio of likelihood values of two hypotheses (H0: the test man is the biological father of the child; H1: the test man is unrelated) based on 158 European allele frequency . H0 is equal to 1 if the alleged father is homozygous for the observed DIP-STR haplotype shared with the child, and it is 0.5 when the alleged father is heterozygous for the observed DIP-STR. H1 corresponds to the frequency of all the homozygous and heterozygous individuals of the observed DIP-STR in the populations. H0/H1 = 1/(the frequency of the observed DIP-STR) if the alleged father is homozygous for the DIP-STR and 0.5/(the frequency of the observed DIP-STR) if the alleged father is heterozygous for the DIP-STR). The CPI is the product of the PI of unlinked loci. All marker combinations used for each CPI calculation included unlinked markers either located on different chromosomes or chromosomal arms. Those located on the same chromosomal region were more distant than 6 Mb, on average at about 40 Mb distance, with the exception of two cases at 1.5 Mb and three cases at 0.5 Mb which tested negative for allelic associations in Europe . According to the Swiss national technical specification for parentage testing, inclusion of parenthood is noted when the CPI is greater than 369 which corresponds to a log 10 (CPI) of 2.57.
The current study was approved by the Centre Hospitalier Universitaire Vaudois and Université de Lausanne institutional review board, research protocol number (2019–01,601 CER-VD).
Each blood sample used was freely donated under conditions of informed consent to participate.
Supplementary Table S1.
|
Impact of random variation in albuminuria and estimated glomerular filtration rate on patient enrolment and duration of clinical trials in nephrology | 19b81a42-ae43-40b0-ad9d-ee9eedee1470 | 9306498 | Internal Medicine[mh] | INTRODUCTION Chronic kidney disease (CKD) is present in approximately 700 million people around the world and is associated with substantial morbidity and mortality. Despite the high prevalence, there are few proven effective therapies to slow progressive kidney function loss. End‐stage kidney disease (ESKD) is used as a clinical endpoint in clinical trials of CKD progression. However, ESKD is a late manifestation of CKD which requires large trials of long duration to assess drug efficacy and safety. Therefore, clinical trials typically enroll patients with high‐risk CKD in order for sufficient endpoints to occur within the clinical trial. Low estimated glomerular filtration rate (eGFR) and high albuminuria are risk markers of CKD , , and are commonly used in clinical trials to enrich the population with participants more likely to progress to ESKD. However, both albuminuria and eGFR (serum creatinine) show a substantial within‐patient day‐to‐day variation. , , , , The high intraindividual variation in albuminuria and eGFR contributes to high screen failure rates in clinical trials. Screen failures, commonly defined as individuals who undergo screening but are not enrolled in a clinical trial, cause a waste of effort and time for participants and investigators. Novel strategies to reduce screen failures in order to improve efficiency of clinical trial conduct in nephrology would be very valuable. The day‐to‐day variation in albuminuria and eGFR can be attributed to a combination of progression of underlying disease, day‐to‐day biological variation due to changes in exercise, diet or hydration status and medication adherence, and measurement variation. This random variation is unlikely to impact the patients' risk with regard to kidney outcomes. We hypothesized that the use of less stringent albuminuria and eGFR inclusion criteria, in participants who met the inclusion criteria of a trial based on pre‐screening values prior to the clinical trial, may reduce screen failure rates, without decreasing the event rate and statistical power of a clinical trial. The aim of this study, therefore, was to determine whether a screening approach with less stringent albuminuria and eGFR thresholds would decrease the screen failure rate without adversely impacting on overall power and study duration of clinical trials.
METHODS 2.1 Patients and study design We performed a post‐hoc analysis of the ALTITUDE trial. The ALTITUDE trial was a randomized, double‐blind, placebo‐controlled trial that included 8561 participants with type 2 diabetes at high risk of kidney or cardiovascular outcomes. The inclusion criteria of this study were useful for our analysis as it was enriched for albuminuria, but also allowed participants with lower levels of albuminuria to be randomized based on the presence of other risk factors. The study design and principal results of the ALTITUDE trial have been published elsewhere. , For our analysis we selected participants at the first study visit (pre‐screening), who were randomly assigned to placebo treatment. We selected patients with a urinary albumin creatinine ratio (UACR) >300 mg/g and an eGFR between 30 mL/min/1.73 m 2 and 60 mL/min/1.73 m 2 . We then used more flexible cut‐offs stepwise at the next visit at 3 months (qualifying visit; Figure ). We used three different strategies for inclusion at the qualifying visit: firstly, stepwise lowering of UACR cut‐offs at the qualifying visit (eg, ≥300 mg/g [base scenario], ≥210 mg/g, ≥150 mg/g, ≥30 mg/g); secondly, stepwise increasing of eGFR cut‐offs at the qualifying visit (eg, 30‐60 mL/min/1.73 m 2 [base scenario], 30‐66 mL/min/1.73 m 2 , 30‐75 mL/min/1.73 m 2 , 30‐90 mL/min/1.73 m 2 , ≥30 mL/min/1.73 m 2 ), and lastly, a combination of lowering UACR and increasing eGFR inclusion criteria at the qualifying visit (eg, UACR ≥300 mg/g and eGFR 30‐60 mL/min/1.73 m 2 [base scenario]; UACR ≥270 mg/g and eGFR 30‐66 mL/min/1.73 m 2 ; UACR ≥225 mg/g and eGFR 30‐75 mL/min/1.73 m 2 ; UACR ≥150 mg/g and eGFR 30‐90 mL/min/1.73 m 2 ; UACR ≥ 0 mg/g and eGFR ≥30). See also Figure for the design of this study. 2.2 Measurements At each visit three consecutive first‐morning‐void urine samples were collected for measurement of urinary albumin and urinary creatinine to compute the UACR. UACR and serum creatinine were measured at a central laboratory at Week 0 (pre‐screening visit) and Week 12 (qualifying visit). The Modification of Diet in Renal Disease formula was used to calculate the eGFR. 2.3 Outcomes The endpoints used for this study were composite cardiovascular and kidney events, as originally defined in the ALTITUDE trial. , The cardiovascular endpoint was a composite of the first occurrence of any of the following: cardiac death; resuscitated cardiac arrest; nonfatal myocardial infarction; nonfatal stroke; and unplanned hospitalization for heart failure. The kidney endpoint was a composite endpoint consisting of a sustained doubling of serum creatinine, ESKD, or death due to kidney‐related cause. The composite kidney‐cardiovascular endpoint was a combination of the individual composite cardiovascular and kidney endpoints. 2.4 Statistical analysis and simulations Baseline characteristics are presented as mean ± SD or median (interquartile range) for variables with a nonparametric distribution. Categorical baseline characteristics are shown as proportions. Baseline characteristics were those recorded at the pre‐screening visit and are shown for the total eligible population and stratified by albuminuria (UACR ≥300 mg/g and <300 mg/g) at the qualifying visit. Differences in baseline characteristics were tested with unpaired t ‐tests, Mann‐Whitney U ‐tests or χ 2 tests as appropriate. To determine the within‐individual variability over time in UACR and eGFR we calculated the coefficient of variation in the placebo arm of the ALTITUDE trial using the UACR and eGFR values collected at pre‐screening, Month 3 and Month 6. For each scenario the number of eligible participants, the number of renal and cardiovascular events and the event rates were determined. Event rates were calculated as events per 100 patient‐years. We then used these event rates to calculate the duration of a future clinical trial using statistical simulations. In these simulations we designed a clinical trial with enrolment of 5220 participants assuming a 24‐month inclusion period and 36‐month follow‐up period. Under these conditions, a total of 961 endpoints provided 90% power to detect a 20% relative risk reduction. We calculated the duration of enrolment and total duration of the clinical trial to accrue 961 endpoints assuming three scenarios with screen failure rates of 40%, 50% or 60%. Using nonlinear modelling the chance of reaching an event was calculated iteratively for each patient using the scale and shape parameter of Weibull fit of the survival models created from each of the UACR and eGFR screening scenarios. Statistical analyses were performed using STATA 15SE (StataCorp LLC, College Station, Texas) and R (version 3.4). A two‐sided P value <0.05 was considered to indicate statistical significance.
Patients and study design We performed a post‐hoc analysis of the ALTITUDE trial. The ALTITUDE trial was a randomized, double‐blind, placebo‐controlled trial that included 8561 participants with type 2 diabetes at high risk of kidney or cardiovascular outcomes. The inclusion criteria of this study were useful for our analysis as it was enriched for albuminuria, but also allowed participants with lower levels of albuminuria to be randomized based on the presence of other risk factors. The study design and principal results of the ALTITUDE trial have been published elsewhere. , For our analysis we selected participants at the first study visit (pre‐screening), who were randomly assigned to placebo treatment. We selected patients with a urinary albumin creatinine ratio (UACR) >300 mg/g and an eGFR between 30 mL/min/1.73 m 2 and 60 mL/min/1.73 m 2 . We then used more flexible cut‐offs stepwise at the next visit at 3 months (qualifying visit; Figure ). We used three different strategies for inclusion at the qualifying visit: firstly, stepwise lowering of UACR cut‐offs at the qualifying visit (eg, ≥300 mg/g [base scenario], ≥210 mg/g, ≥150 mg/g, ≥30 mg/g); secondly, stepwise increasing of eGFR cut‐offs at the qualifying visit (eg, 30‐60 mL/min/1.73 m 2 [base scenario], 30‐66 mL/min/1.73 m 2 , 30‐75 mL/min/1.73 m 2 , 30‐90 mL/min/1.73 m 2 , ≥30 mL/min/1.73 m 2 ), and lastly, a combination of lowering UACR and increasing eGFR inclusion criteria at the qualifying visit (eg, UACR ≥300 mg/g and eGFR 30‐60 mL/min/1.73 m 2 [base scenario]; UACR ≥270 mg/g and eGFR 30‐66 mL/min/1.73 m 2 ; UACR ≥225 mg/g and eGFR 30‐75 mL/min/1.73 m 2 ; UACR ≥150 mg/g and eGFR 30‐90 mL/min/1.73 m 2 ; UACR ≥ 0 mg/g and eGFR ≥30). See also Figure for the design of this study.
Measurements At each visit three consecutive first‐morning‐void urine samples were collected for measurement of urinary albumin and urinary creatinine to compute the UACR. UACR and serum creatinine were measured at a central laboratory at Week 0 (pre‐screening visit) and Week 12 (qualifying visit). The Modification of Diet in Renal Disease formula was used to calculate the eGFR.
Outcomes The endpoints used for this study were composite cardiovascular and kidney events, as originally defined in the ALTITUDE trial. , The cardiovascular endpoint was a composite of the first occurrence of any of the following: cardiac death; resuscitated cardiac arrest; nonfatal myocardial infarction; nonfatal stroke; and unplanned hospitalization for heart failure. The kidney endpoint was a composite endpoint consisting of a sustained doubling of serum creatinine, ESKD, or death due to kidney‐related cause. The composite kidney‐cardiovascular endpoint was a combination of the individual composite cardiovascular and kidney endpoints.
Statistical analysis and simulations Baseline characteristics are presented as mean ± SD or median (interquartile range) for variables with a nonparametric distribution. Categorical baseline characteristics are shown as proportions. Baseline characteristics were those recorded at the pre‐screening visit and are shown for the total eligible population and stratified by albuminuria (UACR ≥300 mg/g and <300 mg/g) at the qualifying visit. Differences in baseline characteristics were tested with unpaired t ‐tests, Mann‐Whitney U ‐tests or χ 2 tests as appropriate. To determine the within‐individual variability over time in UACR and eGFR we calculated the coefficient of variation in the placebo arm of the ALTITUDE trial using the UACR and eGFR values collected at pre‐screening, Month 3 and Month 6. For each scenario the number of eligible participants, the number of renal and cardiovascular events and the event rates were determined. Event rates were calculated as events per 100 patient‐years. We then used these event rates to calculate the duration of a future clinical trial using statistical simulations. In these simulations we designed a clinical trial with enrolment of 5220 participants assuming a 24‐month inclusion period and 36‐month follow‐up period. Under these conditions, a total of 961 endpoints provided 90% power to detect a 20% relative risk reduction. We calculated the duration of enrolment and total duration of the clinical trial to accrue 961 endpoints assuming three scenarios with screen failure rates of 40%, 50% or 60%. Using nonlinear modelling the chance of reaching an event was calculated iteratively for each patient using the scale and shape parameter of Weibull fit of the survival models created from each of the UACR and eGFR screening scenarios. Statistical analyses were performed using STATA 15SE (StataCorp LLC, College Station, Texas) and R (version 3.4). A two‐sided P value <0.05 was considered to indicate statistical significance.
RESULTS 3.1 Baseline characteristics Of the 8561 participants included in the ALTITUDE trial, 995 participants were assigned to placebo and had albuminuria of >300 mg/g and eGFR ≥30 mL/min/1.73 m 2 and <60 mL/min/1.73 m 2 at the pre‐screening visit (Table ). These participants were eligible for the present analysis. Baseline characteristics from the pre‐screening visit are shown in Table . The base scenario was defined by participants who had an UACR >300 mg/g at the pre‐screening and qualifying visit (Week 12) and consisted of 848 participants (85.2%; median UACR 1239 mg/g; median eGFR 44 mL/min/1.73 m 2 ). A total of 147 participants (14.8%) had a UACR >300 mg/g at the pre‐screening visit and a UACR <300 mg/g at the qualifying visit. These participants were thus excluded in the base scenario but would become potentially eligible when more flexible inclusion scenarios were used. Baseline characteristics recorded at pre‐screening of the 147 participants are shown in Table . Participants with a UACR >300 mg/g at the qualifying visit had a higher median UACR at pre‐screening compared to patients with a UACR <300 mg/g. The other baseline characteristics were not statistically different in either of the groups (Table ). During follow‐up, the median within‐individual variation over time in UACR was 31.8% (25th‐75th percentile 19.3‐51.9) and the within‐individual variation in eGFR was 9.0% (25th‐75th percentile 5.6‐13.7). 3.2 Effect of more flexible inclusion criteria on renal and cardiovascular events and event rates Lowering the UACR qualification threshold increased the number of eligible participants who would otherwise fail the screening (Table ). For example, when applying a UACR criterion at the qualifying visit of 210 mg/g (30% decrease from 300 mg/g) 75 additional participants (51% of all screen failures) qualified (Table ). Lowering the UACR inclusion criterion at the qualifying visit increased the total number of eligible patients and the number of cardiovascular and renal events, indicating that participants who would otherwise be screen failures contribute cardiovascular and renal events. By lowering the UACR inclusion criterion only a modest decrease in average renal and cardiovascular event rate was observed due to the inclusion of participants with lower UACR values (Table ). For example, using a UACR criterion of 210 mg/g at the qualifying visit resulted in an increase in the number of eligible patients from 848 to 923, and an increase in renal events from 117 events to 122 events. The event rate showed a moderate decrease from 5.6 (4.6‐6.7) events per 100 patient‐years to 5.3 (4.4‐6.4) events per 100 patient‐years (Table ). We also tested whether increasing eGFR thresholds would influence the number of renal and cardiovascular events and event rates. Relaxing the eGFR criterion resulted in an increase in the number of eligible participants (Table ) and a decrease in the number of screen failures, while it only resulted in a modest, decrease in average renal and cardiovascular event rates (Table ). Similar results were observed when relaxing both eGFR and UACR thresholds in that the number of screen failures decreased without considerably affecting the renal or cardiovascular event rates (Table ). 3.3 Effect of lowering UACR , eGFR or UACR / eGFR criteria on trial duration We performed simulations to test whether relaxing inclusion criteria by decreasing UACR thresholds and/or increasing eGFR thresholds, resulted in a change in trial duration. We used the observed event rates in the ALTITUDE trial and the increase in the number of eligible participants when using less stringent UACR or eGFR inclusion criteria as inputs for statistical simulation (Figure 2). Reducing the UACR threshold for inclusion did not result in an increase in trial duration for the renal and cardiovascular endpoint. This was true for the three scenarios assuming a 40%, 50% or 60% screen failure rate (Figure ). Similarly, when using less stringent eGFR inclusion criteria (Figure ) or a combination of less stringent UACR and eGFR criteria (Figure ), no increase in trial duration was observed. The results were not different when we modelled a clinical trial using a composite cardiovascular and renal endpoint (Figure ).
Baseline characteristics Of the 8561 participants included in the ALTITUDE trial, 995 participants were assigned to placebo and had albuminuria of >300 mg/g and eGFR ≥30 mL/min/1.73 m 2 and <60 mL/min/1.73 m 2 at the pre‐screening visit (Table ). These participants were eligible for the present analysis. Baseline characteristics from the pre‐screening visit are shown in Table . The base scenario was defined by participants who had an UACR >300 mg/g at the pre‐screening and qualifying visit (Week 12) and consisted of 848 participants (85.2%; median UACR 1239 mg/g; median eGFR 44 mL/min/1.73 m 2 ). A total of 147 participants (14.8%) had a UACR >300 mg/g at the pre‐screening visit and a UACR <300 mg/g at the qualifying visit. These participants were thus excluded in the base scenario but would become potentially eligible when more flexible inclusion scenarios were used. Baseline characteristics recorded at pre‐screening of the 147 participants are shown in Table . Participants with a UACR >300 mg/g at the qualifying visit had a higher median UACR at pre‐screening compared to patients with a UACR <300 mg/g. The other baseline characteristics were not statistically different in either of the groups (Table ). During follow‐up, the median within‐individual variation over time in UACR was 31.8% (25th‐75th percentile 19.3‐51.9) and the within‐individual variation in eGFR was 9.0% (25th‐75th percentile 5.6‐13.7).
Effect of more flexible inclusion criteria on renal and cardiovascular events and event rates Lowering the UACR qualification threshold increased the number of eligible participants who would otherwise fail the screening (Table ). For example, when applying a UACR criterion at the qualifying visit of 210 mg/g (30% decrease from 300 mg/g) 75 additional participants (51% of all screen failures) qualified (Table ). Lowering the UACR inclusion criterion at the qualifying visit increased the total number of eligible patients and the number of cardiovascular and renal events, indicating that participants who would otherwise be screen failures contribute cardiovascular and renal events. By lowering the UACR inclusion criterion only a modest decrease in average renal and cardiovascular event rate was observed due to the inclusion of participants with lower UACR values (Table ). For example, using a UACR criterion of 210 mg/g at the qualifying visit resulted in an increase in the number of eligible patients from 848 to 923, and an increase in renal events from 117 events to 122 events. The event rate showed a moderate decrease from 5.6 (4.6‐6.7) events per 100 patient‐years to 5.3 (4.4‐6.4) events per 100 patient‐years (Table ). We also tested whether increasing eGFR thresholds would influence the number of renal and cardiovascular events and event rates. Relaxing the eGFR criterion resulted in an increase in the number of eligible participants (Table ) and a decrease in the number of screen failures, while it only resulted in a modest, decrease in average renal and cardiovascular event rates (Table ). Similar results were observed when relaxing both eGFR and UACR thresholds in that the number of screen failures decreased without considerably affecting the renal or cardiovascular event rates (Table ).
Effect of lowering UACR , eGFR or UACR / eGFR criteria on trial duration We performed simulations to test whether relaxing inclusion criteria by decreasing UACR thresholds and/or increasing eGFR thresholds, resulted in a change in trial duration. We used the observed event rates in the ALTITUDE trial and the increase in the number of eligible participants when using less stringent UACR or eGFR inclusion criteria as inputs for statistical simulation (Figure 2). Reducing the UACR threshold for inclusion did not result in an increase in trial duration for the renal and cardiovascular endpoint. This was true for the three scenarios assuming a 40%, 50% or 60% screen failure rate (Figure ). Similarly, when using less stringent eGFR inclusion criteria (Figure ) or a combination of less stringent UACR and eGFR criteria (Figure ), no increase in trial duration was observed. The results were not different when we modelled a clinical trial using a composite cardiovascular and renal endpoint (Figure ).
DISCUSSION To improve clinical trials for progression of CKD we assessed the utility of a screening approach with more flexible albuminuria and eGFR thresholds to decrease screen failure rates, accelerate trial enrolment, and improve the feasibility of trial conduct. We observed that lowering of the albuminuria‐based and increasing of the eGFR‐based inclusion criteria, for participants who met the inclusion criteria of a trial based on pre‐screening values prior to the clinical trial, increases the number of eligible participants and decreases screen failure rates without the need to increase sample size or prolong trial duration. Our proposed approach thus simplifies enrolment of participants into a trial and increases efficiency of trial conduct. The number of clinical trials undertaken in nephrology lags behind other therapeutic areas in medicine. , Moreover, clinical trials in nephrology compared to cardiology are often smaller, shorter in duration and less frequently involve clinically meaningful endpoints. There are multiple reasons why trials in nephrology are less frequently undertaken, including lack of global clinical trial networks, higher than average adverse event rates, and endpoints being late manifestations of CKD progression, requiring large trials of long duration. In addition, screen failure rates in recent large kidney outcome trials are high, ranging between 45% and 60%. The high screen failure rates prolong clinical trial recruitment, increase costs and cause a waste of effort and time from participants and clinicians, which may temper interest in participating in future research. A large proportion of screen failures in clinical trials in nephrology are due to albuminuria and eGFR values not falling within the protocol‐specified range. , , The high screen failure rate can be in part attributed to the large day‐to‐day variability in these laboratory variables. Indeed, in the ALTITUDE trial we found a within‐individual variability over time of 31.8% for albuminuria and 9.0% for eGFR. This variability was of similar magnitude to that observed in other studies. Waikar et al , reported a variability in eGFR of 6.5% and a variability in albuminuria of 32.5%. Another study in patients with type 2 diabetes with increased albuminuria reported a variability in albuminuria of 31.8%. , Removing trial inefficiencies such as reducing screen failure rates by simplifying inclusion criteria may thus help to stimulate clinical trials in nephrology. It is important to note that simplifying trial procedures, including efforts to reduce screen failures, will reduce disappointment among participants and investigators and reduce wasted effort and time. Frequent screen failures can lead to frustration among site investigators and can decrease engagement and commitment. As site investigators can participate in multiple clinical trials, it is likely that they may choose to spend more time and effort on less complex and simpler trials. We did not capture this aspect in our simulations, which may have resulted in an underestimation of the efficiency gains and an overestimation of the trial duration when applying less stringent inclusion criteria. Previous studies have shown that higher albuminuria levels are associated with a higher risk of developing kidney and cardiovascular outcomes. Clinical trials therefore enrich populations for participants with higher albuminuria and lower eGFR levels to collect sufficient endpoints within the 3‐ to 4‐year trial duration. One would expect that lowering the albuminuria threshold for inclusion in a clinical trial would dilute the risk profile of the population and reduce kidney and cardiovascular event rates. However, in our study, event rates only modestly decreased despite inclusion of participants with lower degrees of albuminuria. The likely explanation for this is that, in contrast to other studies, we selected a cohort of participants who had high albuminuria (UACR >300 mg/g) at a previous (pre‐screening) visit. The finding that the event rates only changed modestly supports the possibility that the lower albuminuria level at qualification was indeed at least in part explained by random day‐to‐day variation. Optimizing and simplifying eligibility criteria for a clinical trial is only one of many approaches to facilitating patient enrolment and clinical trial conduct. Several other approaches are also proposed. These include approaches to increase the willingness of patients to participate in a trial and removing barriers for participation such as less frequent site visits, implementing decentralized study procedures, and developing more efficient informed consent procedures. Other strategies to promote clinical trial conduct include better use of electronic medical records to facilitate pre‐screening approaches, developing implementable study protocols, and considering novel biomarkers that assist in selecting optimal trial populations who are most likely to respond and tolerate the investigational drug. Another approach to reducing screen failures would be to re‐screen patients who have a measurement outside the screening limits. However, this requires another visit for participants and greater effort from trial sites. Relaxing albuminuria‐ and eGFR‐based inclusion criteria is a simpler approach, which is also less burdensome for patients. The disadvantage of using enrichment criteria, such as albuminuria and eGFR, for clinical trials is that it reduces the generalizability of the clinical trial. Some patients with normoalbuminuria may also progress towards kidney failure, as illustrated by studies that show that the prevalence of non‐albuminuric CKD is increasing. , , However, patients with normoalbuminuria are often excluded from kidney outcome trials. Methods to identify patients with normoalbuminuria who are at risk of progression are needed. In this respect, novel biomarkers, such as tumor necrosis factor receptor (TNFR)‐1 and TNFR‐2, may help as they have been shown to predict kidney outcomes in patients with type 2 diabetes and normoalbuminuria. , This study has some limitations. First, the duration between a pre‐screening and qualifying visit in most trials in nephrology is 2 to 4 weeks whereas in our study we used a 3‐month period, which may have led to a larger variation in albuminuria than would be observed in practice. In addition, the analyses were performed post hoc. Future trials which have implemented the proposed strategy are ongoing (NCT03819153) and will test the utility of our proposed approach prospectively. In conclusion, relaxing albuminuria‐ and eGFR‐based inclusion criteria for a clinical trial for participants who met these criteria based on pre‐screening values prior to the trial, decreases screen failure rates without prolonging trial duration. This approach may increase recruitment feasibility and enhance efficiency in the conduct of clinical trials.
Hiddo J. L. Heerspink: Conceptualization. Simke W. Waijer and Hiddo J. L. Heerspink: Data curation. Simke W. Waijer , Michele Provenzano , Skander Mulder and Hiddo J. L. Heerspink: Formal analysis. Hiddo J. L. Heerspink: Investigation. Simke W. Waijer , Michele Provenzano and Hiddo J. L. Heerspink: Writing – original draft. All authors: Writing – review and editing.
S.W. Waijer, M. Provenzano and S. Mulder have nothing to disclose. P. Rossing has received research support and personal fees from AstraZeneca and Novo Nordisk, and personal fees from Eli Lilly and Company, Bayer, Boehringer Ingelheim, Astellas Pharma Inc., Gilead, Merck, Merck Sharp and Dohme, Sanofi and Vifor Pharma. All fees were given to the Steno Diabetes Centre Copenhagen. F. Persson has served as a consultant, on advisory boards or as educator for AstraZeneca, Novo Nordisk, Boehringer Ingelheim, Sanofi, Mundipharma, MSD, Novartis, Amgen and has received research grants to institution from Novo Nordisk, Boehringer Ingelheim, Amgen and AstraZeneca. V. Perkovic has received fees for advisory boards, steering committee roles, or scientific presentations from AbbVie, Astellas, AstraZeneca, Bayer, Baxter, Bristol‐Myers Squibb, Boehringer Ingelheim, Dimerix, Durect, Eli Lilly, Gilead, GlaxoSmithKline, Janssen, Merck, Mitsubishi Tanabe, Mundipharma, Novartis, Novo Nordisk, Pfizer, PharmaLink, Relypsa, Retrophin, Sanofi, Servier, Vifor and Tricida. H. J. L. Heerspink is supported by a VIDI (917.15.306) grant from the Netherlands Organization for Scientific Research and has served as a consultant for AbbVie, Astellas, AstraZeneca, Bayer, Boehringer Ingelheim, Chinook, CSL Pharma, Dimerix Fresenius, Goldfinch, Gilead, Janssen, Merck, Mundipharma, Mitsubishi Tanabe, NovoNordisk and Travere Pharmaceuticals; and has received grant support from AstraZeneca, Boehringer Ingelheim, Janssen and NovoNordisk.
The peer review history for this article is available at https://publons.com/publon/10.1111/dom.14660 .
Figure S1 . Effects of more flexible UACR inclusion criteria, more flexible eGFR inclusion criteria and more flexible UACR/eGFR criteria on trial duration for the composite renal and cardiovascular endpoint. Click here for additional data file.
|
Pharmacist Use of a Population Management Dashboard for Safe Anticoagulant Prescribing: Evaluation of a Nationwide Implementation Effort | 9c543157-91f0-4908-be12-f9e033b25134 | 11935639 | Pathologic Processes[mh] | Use of a population health dashboard to monitor for and correct off‐label direct oral anticoagulant dosing by clinical pharmacists was associated with improved prescribing and a reduction in the composite of stroke and venous thromboembolism.
Health systems and policymakers should consider investing in anticoagulation stewardship efforts that support pharmacists in reviewing and correcting off‐label direct oral anticoagulant dosing.
This is a retrospective analysis of nationwide anticoagulant use within the VHA system. The design of the overarching project has previously been published. Several other publications describe the VHA PMT dashboard design and use in detail. , This study was reviewed and approved by the VHA Ann Arbor Institutional Review Board with a waiver of informed consent given the retrospective nature of data analysis. Briefly, the DOAC PMT dashboard was first made available to select sites in August 2016, then expanded to other sites beginning in April 2017. Starting in January 2018, the DOAC PMT dashboard was made available to all VHA sites nationally. Our prior analysis of DOAC PMT dashboard use demonstrated near universal adoption by early 2019. Data for this analysis are available from the corresponding author upon reasonable request, as constrained by VHA policy. Population and Predictor Variables Our analysis included all patients dispensed DOAC medications from a VHA outpatient pharmacy between August 2015 and December 2019 with an associated diagnosis of atrial fibrillation or venous thromboembolism (VTE). All DOAC prescriptions made during hospitalization, dispensed by non‐VHA pharmacies, or ordered by non‐VHA clinicians (eg, community health care providers) are not included in the default user view of the DOAC PMT dashboard. Any sites using the Cerner electronic health record system were excluded from the analysis. All assessments occurred at the level of the VHA site (hospital or clinic) and divided on the basis of the date at which the site achieved “moderate” or “high” DOAC PMT dashboard use, defined a priori as at least 1 login on 2 to 5 separate days of the month. Consistent with our prior analysis, sites were grouped into 4 usage start dates: those with moderate‐high usage before April 2017, between May 2017 and December 2017, between January 2018 and June 2018, and after June 2018. Outcome Definition and Assessment The Reach, Effectiveness, Adoption, Implementation, and Maintenance framework guided evaluation of the VHA DOAC PMT dashboard. This framework is a commonly used implementation science evaluation framework. A prior analysis focused on the reach, adoption, and maintenance outcomes. For this analysis, we focused on effectiveness and implementation outcomes. Effectiveness was defined in 2 ways. First, we defined the off‐label dosing of DOAC medications. This was determined according to VHA criteria for use guidelines (Table ) on the basis of the phase 3 clinical trial protocols, specifically in relation to the indication for therapy (eg, stroke prevention in atrial fibrillation, treatment of VTE) and renal function (or combination of age, body weight, and renal function for apixaban use in stroke prevention for patients with atrial fibrillation). This outcome was reported as a proportion of all DOAC prescriptions at each site, measured monthly during the study period. The second effectiveness outcome was the occurrence of clinical adverse events, which were defined in 2 groups. The first is bleeding. The second is a composite of ischemic stroke and VTE. Clinical outcomes were collected using the International Classification of Diseases , Ninth Revision ( ICD‐9 ) and Tenth Revision ( ICD‐10 ). For the outcome of bleeding, we followed the methodology used by Perino et al. These were supplemented with a list of blood transfusion codes previously used by Siontis et al. We followed the approach and ICD codes used by Perino et al for VTE outcomes, including a 7‐day “blanking” period after the index VTE diagnosis when looking for recurrent VTE events. For the ischemic stroke outcome, we used ICD codes from Olivier et al ( ICD‐9 ) and Lawrence et al ( ICD‐10 ). , Additional details and the list of relevant ICD codes can be found in Tables . Clinical adverse event rates are reported per 100 patient‐years. For the outcome of implementation, we explored the percentage of patients with an off‐label DOAC dosing prescription who had a change in their prescription within 7 days. This differed slightly from our preplanned analysis as we were unable to obtain access to DOAC PMT dashboard flag‐specific data. Any change in the medication, dosage, or frequency of administration was considered a change in the prescription, to assess the implementation outcome. Post hoc, we performed 2 additional sensitivity analyses to explore the potential impact of the DOAC PMT dashboard use on the first effectiveness outcome (off‐label DOAC dosing). These analyses defined “very high” DOAC PMT dashboard use at ≥10 days and ≥15 days with at least 1 login in a month (Table ). This higher threshold value was selected after inspecting the distribution of monthly login days to determine if sites with the most frequent use demonstrated any difference in effectiveness outcomes from those with less frequent use. Data Sources Several data sources were used in this analysis. The first included data on the frequency of VHA DOAC PMT dashboard use at the provider and site levels, provided by the VHA Pharmacy Benefits Management Center for Medication Safety. Three sites were excluded from the analysis due to inaccurate DOAC PMT dashboard data. The second data source was the VHA Clinical Data Warehouse, which includes all medication prescription data and ICD diagnosis codes. The third data source was the Centers for Medicare and Medicaid Services data for veterans receiving care outside the VHA that was paid for by Medicare. Statistical Analysis To assess for changes in site‐level monthly off‐label DOAC prescribing based on the timing when each site achieved moderate‐high DOAC PMT dashboard use, we constructed a linear regression model that included an interaction between time (measured by month), the timing of when the site achieved moderate‐high DOAC PMT dashboard use (defined above), and pre–post dashboard use. The model was fit at the group‐month level with adjustment for median site population without adjustment for patient‐level characteristics given an ecological analytic approach. The slope and change in slope of DOAC PMT dashboard use across each implementation time group was estimated from the model using marginal effects estimation. All analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC) and Stata version 18 (StataCorp, College Station, TX). Statistical significance was set with a P value of <0.05.
Our analysis included all patients dispensed DOAC medications from a VHA outpatient pharmacy between August 2015 and December 2019 with an associated diagnosis of atrial fibrillation or venous thromboembolism (VTE). All DOAC prescriptions made during hospitalization, dispensed by non‐VHA pharmacies, or ordered by non‐VHA clinicians (eg, community health care providers) are not included in the default user view of the DOAC PMT dashboard. Any sites using the Cerner electronic health record system were excluded from the analysis. All assessments occurred at the level of the VHA site (hospital or clinic) and divided on the basis of the date at which the site achieved “moderate” or “high” DOAC PMT dashboard use, defined a priori as at least 1 login on 2 to 5 separate days of the month. Consistent with our prior analysis, sites were grouped into 4 usage start dates: those with moderate‐high usage before April 2017, between May 2017 and December 2017, between January 2018 and June 2018, and after June 2018.
The Reach, Effectiveness, Adoption, Implementation, and Maintenance framework guided evaluation of the VHA DOAC PMT dashboard. This framework is a commonly used implementation science evaluation framework. A prior analysis focused on the reach, adoption, and maintenance outcomes. For this analysis, we focused on effectiveness and implementation outcomes. Effectiveness was defined in 2 ways. First, we defined the off‐label dosing of DOAC medications. This was determined according to VHA criteria for use guidelines (Table ) on the basis of the phase 3 clinical trial protocols, specifically in relation to the indication for therapy (eg, stroke prevention in atrial fibrillation, treatment of VTE) and renal function (or combination of age, body weight, and renal function for apixaban use in stroke prevention for patients with atrial fibrillation). This outcome was reported as a proportion of all DOAC prescriptions at each site, measured monthly during the study period. The second effectiveness outcome was the occurrence of clinical adverse events, which were defined in 2 groups. The first is bleeding. The second is a composite of ischemic stroke and VTE. Clinical outcomes were collected using the International Classification of Diseases , Ninth Revision ( ICD‐9 ) and Tenth Revision ( ICD‐10 ). For the outcome of bleeding, we followed the methodology used by Perino et al. These were supplemented with a list of blood transfusion codes previously used by Siontis et al. We followed the approach and ICD codes used by Perino et al for VTE outcomes, including a 7‐day “blanking” period after the index VTE diagnosis when looking for recurrent VTE events. For the ischemic stroke outcome, we used ICD codes from Olivier et al ( ICD‐9 ) and Lawrence et al ( ICD‐10 ). , Additional details and the list of relevant ICD codes can be found in Tables . Clinical adverse event rates are reported per 100 patient‐years. For the outcome of implementation, we explored the percentage of patients with an off‐label DOAC dosing prescription who had a change in their prescription within 7 days. This differed slightly from our preplanned analysis as we were unable to obtain access to DOAC PMT dashboard flag‐specific data. Any change in the medication, dosage, or frequency of administration was considered a change in the prescription, to assess the implementation outcome. Post hoc, we performed 2 additional sensitivity analyses to explore the potential impact of the DOAC PMT dashboard use on the first effectiveness outcome (off‐label DOAC dosing). These analyses defined “very high” DOAC PMT dashboard use at ≥10 days and ≥15 days with at least 1 login in a month (Table ). This higher threshold value was selected after inspecting the distribution of monthly login days to determine if sites with the most frequent use demonstrated any difference in effectiveness outcomes from those with less frequent use.
Several data sources were used in this analysis. The first included data on the frequency of VHA DOAC PMT dashboard use at the provider and site levels, provided by the VHA Pharmacy Benefits Management Center for Medication Safety. Three sites were excluded from the analysis due to inaccurate DOAC PMT dashboard data. The second data source was the VHA Clinical Data Warehouse, which includes all medication prescription data and ICD diagnosis codes. The third data source was the Centers for Medicare and Medicaid Services data for veterans receiving care outside the VHA that was paid for by Medicare.
To assess for changes in site‐level monthly off‐label DOAC prescribing based on the timing when each site achieved moderate‐high DOAC PMT dashboard use, we constructed a linear regression model that included an interaction between time (measured by month), the timing of when the site achieved moderate‐high DOAC PMT dashboard use (defined above), and pre–post dashboard use. The model was fit at the group‐month level with adjustment for median site population without adjustment for patient‐level characteristics given an ecological analytic approach. The slope and change in slope of DOAC PMT dashboard use across each implementation time group was estimated from the model using marginal effects estimation. All analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC) and Stata version 18 (StataCorp, College Station, TX). Statistical significance was set with a P value of <0.05.
During the study period, 128 652 patients were prescribed DOAC therapy. Demographics, medication use, and comorbidities are summarized in the . When broken down by the timing of initial moderate‐high DOAC PMT dashboard use, the number of sites per implementation group ranges from 8 to 56. These sites averaged between 731 and 1483 DOAC‐treated patients per site. Data on sites achieving very high DOAC PMT dashboard use by timing are shown in Table . Effectiveness Analysis Across 123 VHA sites, the mean site‐level percentage of patients with off‐label DOAC prescribing each month ranged between 8.6% in January 2016 and 6.8% in December 2019. Before achieving moderate‐high DOAC PMT dashboard use, the mean (SD) off‐label DOAC dosing prescription was 7.9% (0.6%), 8.0% (0.4%), 8.1% (0.4%), and 10.2% (0.9%) for the groups that first achieved moderate‐high DOAC PMT dashboard use on/before April 2017, between May and December 2017, between January and June 2018, or on/after July 2018, respectively. Mean rates of off‐label DOAC dosing following implementation were 6.9% (0.7%), 7.5% (0.4%), 7.2% (0.3%), and 10.2% (0.4%), respectively. Across all sites, the mean off‐label monthly DOAC dosing prescription rate declined from 8.7% (1.2%) to 7.6% (1.2%) from before and after implementation. Three important themes emerge when analyzing off‐label DOAC use by the site‐level timing of initial moderate‐high DOAC PMT dashboard use (Figure ; Figures ). First, sites that were the latest to initiate DOAC PMT dashboard use (June 2018 or after) had a higher rate of off‐label DOAC dosing prescriptions than sites with earlier moderate‐high DOAC PMT dashboard use ( P <0.001 versus all other groups). Second, all 4 groups had either a slight increase or no significant trend in off‐label DOAC dosing prescriptions before adopting the DOAC PMT dashboard. Third, sites with the earliest moderate‐high level of DOAC PMT dashboard use (on or before April 2017), had a notable decline in off‐label DOAC dosing prescriptions beginning in September 2017 that persisted during the remainder of the study period. In the linear regression model, all 4 groups had a statistically significant decline in the trend line of off‐label DOAC dosing prescriptions from before to after achieving moderate‐high DOAC PMT dashboard use (Table ). In the sensitivity analyses, a deline in the trend line of off‐label DOAC prescribing was also seen for very high DOAC PMT dashboard monthly use, but staitiscal significance was limited by lower power in some subgroups (Table ). Mean (SD) rates of bleeding overall were 3.0 (0.5) per 100 patient‐years before and 2.9 (0.4) per 100 patient‐years after implementation. For each of the 4 groups, the mean pre‐ and postimplementation rates per 100 patient‐years were 3.7 (0.6) and 3.1 (0.3) for the group adopting on/before April 2017, 2.9 (0.3) and 2.8 (0.3) for May–December 2017, 2.8 (0.3) and 2.6 (0.3) for January to June 2018, and 3.0 (0.5) and 3.3 (0.4) for on/after July 2018. Over the study period, the group who adopted the DOAC PMT dashboard at a moderate‐high level at the earliest time period (April 2017 or earlier) had nonsignificant change in the downward trend of bleeding events following adoption at a moderate‐high level (change of slope, 0.04 [95% CI, –0.02 to 0.10]; P =0.22). However, the figure demonstrates a continued reduction in the monthly percentage of patients experiencing bleeding events following adoption of the DOAC PMT dashboard at a moderate‐high rate. The second group (May to December 2017) had a nonsignificantly significant decline in the slope of bleeding events (change in slope, −0.03 [95% CI, –0.07 to 0.004]; P =0.08), while the third (January to June 2018) and fourth groups (July 2018 or later) also had nonsignificant and smaller declines in the slope (change in slope, −0.01 [95% CI, –0.06 to 0.05];, P =0.84; and change in slope, −0.01 [95% CI, –0.07 to 0.05]; P =0.76, respectively). This is graphically represented in Figure . Mean (SD) composite rates of VTE and stroke overall were 2.7 (0.2) per 100 patient‐years before and 2.2 (0.3) per 100 patient‐years after implementation. For each of the 4 groups, the mean pre‐ and postimplementation rates per 100 patient‐years were 2.7 (0.2) and 2.2 (0.3) for the group adopting on/before April 2017, 2.7 (0.1) and 2.1 (0.2) for May to December 2017, 2.5 (0.1) and 2.0 (0.2) for January to June 2018, and 3.0 (0.2) and 2.6 (0.2) for on/after July 2018. Over the study period, all 4 groups had a reduction in the composite rate of VTE or stroke (Figure ). The change was smallest for the earliest adopting group (change in slope, −0.01 [95% CI, –0.03 to –0.01]; P =0.021) and largest for the latest adopting group (change in slope, −0.43 [95% CI, –0.05 to –0.03]; P <0.001). The second and third adopting groups had similar changes (change in slope, −0.03 [95% CI, –0.04 to –0.02]; P <0.001; and change in slope, −0.02 [95% CI, –0.04 to –0.01]; P <0.001, respectively). In sensitivity analyses using very high DOAC PMT dashboard usage thresholds, the trends in both bleeding and stroke/VTE outcomes are largely similar to those in the primary analysis (Tables ; Figures ). Implementation Analysis To assess the fidelity in the DOAC PMT dashboard use at the site level, we assessed the percentage of patients with an off‐label DOAC prescription that had a change to their prescription within 7 days. As shown in Figure , there was no meaningful difference in the rate of off‐label DOAC use based on the site‐level timing of DOAC PMT dashboard adoption. In regression analysis, there was no statistically significant change in the rate of flags resolved within 7 days from before to after DOAC PMT dashboard adoption (Table ).
Across 123 VHA sites, the mean site‐level percentage of patients with off‐label DOAC prescribing each month ranged between 8.6% in January 2016 and 6.8% in December 2019. Before achieving moderate‐high DOAC PMT dashboard use, the mean (SD) off‐label DOAC dosing prescription was 7.9% (0.6%), 8.0% (0.4%), 8.1% (0.4%), and 10.2% (0.9%) for the groups that first achieved moderate‐high DOAC PMT dashboard use on/before April 2017, between May and December 2017, between January and June 2018, or on/after July 2018, respectively. Mean rates of off‐label DOAC dosing following implementation were 6.9% (0.7%), 7.5% (0.4%), 7.2% (0.3%), and 10.2% (0.4%), respectively. Across all sites, the mean off‐label monthly DOAC dosing prescription rate declined from 8.7% (1.2%) to 7.6% (1.2%) from before and after implementation. Three important themes emerge when analyzing off‐label DOAC use by the site‐level timing of initial moderate‐high DOAC PMT dashboard use (Figure ; Figures ). First, sites that were the latest to initiate DOAC PMT dashboard use (June 2018 or after) had a higher rate of off‐label DOAC dosing prescriptions than sites with earlier moderate‐high DOAC PMT dashboard use ( P <0.001 versus all other groups). Second, all 4 groups had either a slight increase or no significant trend in off‐label DOAC dosing prescriptions before adopting the DOAC PMT dashboard. Third, sites with the earliest moderate‐high level of DOAC PMT dashboard use (on or before April 2017), had a notable decline in off‐label DOAC dosing prescriptions beginning in September 2017 that persisted during the remainder of the study period. In the linear regression model, all 4 groups had a statistically significant decline in the trend line of off‐label DOAC dosing prescriptions from before to after achieving moderate‐high DOAC PMT dashboard use (Table ). In the sensitivity analyses, a deline in the trend line of off‐label DOAC prescribing was also seen for very high DOAC PMT dashboard monthly use, but staitiscal significance was limited by lower power in some subgroups (Table ). Mean (SD) rates of bleeding overall were 3.0 (0.5) per 100 patient‐years before and 2.9 (0.4) per 100 patient‐years after implementation. For each of the 4 groups, the mean pre‐ and postimplementation rates per 100 patient‐years were 3.7 (0.6) and 3.1 (0.3) for the group adopting on/before April 2017, 2.9 (0.3) and 2.8 (0.3) for May–December 2017, 2.8 (0.3) and 2.6 (0.3) for January to June 2018, and 3.0 (0.5) and 3.3 (0.4) for on/after July 2018. Over the study period, the group who adopted the DOAC PMT dashboard at a moderate‐high level at the earliest time period (April 2017 or earlier) had nonsignificant change in the downward trend of bleeding events following adoption at a moderate‐high level (change of slope, 0.04 [95% CI, –0.02 to 0.10]; P =0.22). However, the figure demonstrates a continued reduction in the monthly percentage of patients experiencing bleeding events following adoption of the DOAC PMT dashboard at a moderate‐high rate. The second group (May to December 2017) had a nonsignificantly significant decline in the slope of bleeding events (change in slope, −0.03 [95% CI, –0.07 to 0.004]; P =0.08), while the third (January to June 2018) and fourth groups (July 2018 or later) also had nonsignificant and smaller declines in the slope (change in slope, −0.01 [95% CI, –0.06 to 0.05];, P =0.84; and change in slope, −0.01 [95% CI, –0.07 to 0.05]; P =0.76, respectively). This is graphically represented in Figure . Mean (SD) composite rates of VTE and stroke overall were 2.7 (0.2) per 100 patient‐years before and 2.2 (0.3) per 100 patient‐years after implementation. For each of the 4 groups, the mean pre‐ and postimplementation rates per 100 patient‐years were 2.7 (0.2) and 2.2 (0.3) for the group adopting on/before April 2017, 2.7 (0.1) and 2.1 (0.2) for May to December 2017, 2.5 (0.1) and 2.0 (0.2) for January to June 2018, and 3.0 (0.2) and 2.6 (0.2) for on/after July 2018. Over the study period, all 4 groups had a reduction in the composite rate of VTE or stroke (Figure ). The change was smallest for the earliest adopting group (change in slope, −0.01 [95% CI, –0.03 to –0.01]; P =0.021) and largest for the latest adopting group (change in slope, −0.43 [95% CI, –0.05 to –0.03]; P <0.001). The second and third adopting groups had similar changes (change in slope, −0.03 [95% CI, –0.04 to –0.02]; P <0.001; and change in slope, −0.02 [95% CI, –0.04 to –0.01]; P <0.001, respectively). In sensitivity analyses using very high DOAC PMT dashboard usage thresholds, the trends in both bleeding and stroke/VTE outcomes are largely similar to those in the primary analysis (Tables ; Figures ).
To assess the fidelity in the DOAC PMT dashboard use at the site level, we assessed the percentage of patients with an off‐label DOAC prescription that had a change to their prescription within 7 days. As shown in Figure , there was no meaningful difference in the rate of off‐label DOAC use based on the site‐level timing of DOAC PMT dashboard adoption. In regression analysis, there was no statistically significant change in the rate of flags resolved within 7 days from before to after DOAC PMT dashboard adoption (Table ).
In this nationwide analysis of the VHA system, sites with earlier moderate‐high pharmacist use of the DOAC PMT dashboard had lower rates of off‐label DOAC dosing prescriptions. Furthermore, sites that used the DOAC PMT dashboard at a moderate‐high level on or before June 2018 demonstrated a statistically significant decline in the overall percentage of off‐label DOAC dosing prescriptions, while sites with moderate‐high dashboard use on or after July 2018 or later did not. Rates of stroke or VTE declined in all groups following the adoption of the DOAC PMT dashboard, while rates of bleeding declined significantly for those in the second group of moderate‐high use adoption. This analysis provides evidence that most of the DOAC PMT dashboard's 3 main purposes were likely successful and strongly support this approach to antithrombotic stewardship. The first purpose is to support VHA anticoagulation pharmacists who are charged with overseeing the management of patients prescribed DOAC medications. Before the implementation of the DOAC PMT dashboard and guided by national VHA guidance initially published shortly after DOACs entered the US market, most VHA centers had adopted policies and practices that provided close monitoring of all DOAC prescriptions in the acute care and ambulatory settings (VHA Directive 1033, July 2015). Examples of front‐end anticoagulation stewardship tools put into place across VHA include prior authorization drug requests, clinical decision support tools, and formal consultation to anticoagulation services. Many VHA sites also employed regular follow‐up protocols to ensure patient adherence to the DOAC medication, assessment for any DOAC‐related side effects, and a review of the medical record with DOAC dose adjustment when necessary (eg, change in renal function). As such, the quality of DOAC care across the VHA system nationally was high, even before the DOAC PMT dashboard was implemented. Our data demonstrated this finding with a low rate of off‐label DOAC prescribing before the DOAC PMT dashboard was developed in August 2016. In fact, most VHA sites had an off‐label DOAC dosing rate of ≈8% during that period, less than half the 20% rates reported in other health systems across several other analyses. , Despite this, it is notable that moderate‐high use of the DOAC PMT dashboard before July 2018 was still associated with a decline in off‐label DOAC dosing use within this high‐functioning health care delivery system (Figure ). The second purpose of the DOAC PMT dashboard is to reduce DOAC‐associated adverse events by improving on‐label dosing. Prior analyses have found associations between off‐label DOAC dosing and increased rates of stroke/systemic embolism, major bleeding, hospitalization, and all‐cause death. , However, this association does not prove that “fixing” off‐label dosing would result in patients experiencing fewer adverse events. Rather, it may be a marker of underlying risk for adverse events. For instance, in one prior analysis, patients receiving off‐label underdosing of DOACs experienced higher rates of major bleeding than those receiving on‐label dosing (2.9% versus 0.4%). Our analysis was able to demonstrate an association between DOAC PMT dashboard adoption and decreased rates of both bleeding and thromboembolic events. Of note, thrombotic events declined for all sites following moderate‐high DOAC PMT dashboard adoption, while sites that adopted the DOAC PMT dashboard at a moderate‐high level between May and December 2017 (the second group) also had a statistically significant decline in bleeding. This provides critical evidence, albeit not prospective or randomized, that supports the linkage between correcting off‐label DOAC prescribing and an association with fewer adverse clinical events. The third purpose of the VHA DOAC PMT dashboard is to improve the efficiency of pharmacists who oversee ambulatory prescribing of DOAC medications. Prior VHA analyses of pharmacist work found a dramatic improvement in the percentage of cases reviewed that received an intervention (55% versus 20%) and markedly reduced time to intervention (16 versus 64 minutes) with the DOAC PMT dashboard compared with traditional pharmacist case review. Similar findings were seen in an analysis of the Kaiser Permanente health system, where more than 10 times as many patients (21 891 versus 2089) could be monitored using a population health dashboard compared with traditional pharmacist management. One potential explanation for the lack of bleeding outcome benefit for all groups assessed in this analysis may be the way in which the effectiveness analysis was prespecified. We categorized VHA centers according to the monthly frequency of DOAC PMT dashboard use. It is possible that our threshold values were not appropriate. However, the sensitivity analysis using higher dashboard use cut points was consistent with the primary analysis for bleeding. It is possible that monthly DOAC PMT dashboard use metrics may not be the best way to assess and categorize clinical use of the dashboard by pharmacists. Measures of actual clinical interventions, which were unavailable for our analysis, may provide a different assessment of anticoagulation pharmacists' work and clinical impact. However, a prior analysis of 40 VHA centers comparing those with the highest and lowest quintile of DOAC PMT dashboard access found a statistically significant reduction in “questionable DOAC dosing” with PMT dashboard use versus nonuse. There may also be issues of statistical power for the sites with later adoption and insufficient postimplementation data available. Similarly, our prespecified implementation outcome of change in prescription within 7 days may also be a less‐than‐ideal measure, as it does not account for other impacts that the pharmacists have on safe DOAC prescribing (eg, renal and liver function testing, changes to interacting medications). Given that anticoagulant medications are currently the leading cause of adverse drug events in US emergency departments, , other analyses using different metrics for DOAC PMT dashboard effectiveness and implementation are warranted. Our analysis of the VHA DOAC PMT dashboard is characterized by several strengths. These include the use of a large population, a prespecified analytic plan, and an established implementation science evaluation framework. There are also several important limitations to consider. First, while the population was large and drawn from across the entire United States, there were important differences between the veteran population and the general US population. Notably, this population is mostly men, receiving care in an urban setting, and with a larger proportion receiving dabigatran than many non‐VHA populations. While the overall proportion of DOAC‐treated patients using dabigatran is not nearly as large as in the Kaiser Permanente analysis, it is still significantly higher than most other DOAC‐treated populations in the United States. , Second, pharmacists in the VHA have unique prescribing and drug management authority that may not be available in non‐VHA health systems. Third, despite the large sample size, this nonrandomized study invariably is limited by issues of confounding, which must be considered. Fourth, as noted above, the VHA system was already delivering high‐quality DOAC care, which may limit the ability to detect a difference in implementation of effectiveness outcomes following adoption of a PMT dashboard‐facilitated clinical management strategy. Fifth, while this analysis focuses on off‐label DOAC use and associated clinical adverse events, we were unable to assess for other elements of antithrombotic stewardship facilitated by the DOAC PMT dashboard, including reductions in concurrent aspirin use, initiation of proton pump inhibitor therapy to reduce gastrointestinal bleeding risk, and interventions to ensure timely medication fill at the pharmacy. Sixth, the DOAC PMT dashboard does not display medications prescribed by non‐VHA clinicians or filled at non‐VHA pharmacies in the default configuration. This gap may impact the overall assessment of effectiveness from the perspective of the entire VHA patient population. Seventh, our analysis was conducted at a population level and did not adjust for patient‐specific factors at each site. However, the use of an interrupted time series–like analysis allows for sites to be their own control when comparing effectiveness outcomes before and after they achieved moderate to high DOAC PMT dashboard use. Eighth, the DOAC PMT dashboard includes flags for items other than off‐label prescribing, including concurrent nonsteroidal anti‐inflammatory drug use and delayed medication fill (concern about adherence). It is possible that pharmacist efforts addressing these issues were larger drivers of the clinical outcome improvements seen than the pharmacist effort addressing off‐label DOAC prescribing. Finally, as a retrospective study, we can comment only on association and not causation. Nonetheless, this analysis demonstrates the impact of a population management approach to safe and effective DOAC prescribing. A similar approach could be used for other medications that require close monitoring, such as antiarrhythmic medications (eg, screening for thyroid and liver function with amiodarone), immunosuppressive medications, and other rheumatologic agents. In fact, within the VHA, this approach is being used in cardiology, endocrinology, gastroenterology, geriatrics, mental health, neurology, oncology, and transplant medicine. In summary, this nationwide analysis of the VHA DOAC PMT dashboard found an association between moderate‐high use and key effectiveness outcomes, showing lower rates of off‐label DOAC prescribing and reduced adverse clinical events. Future studies should explore different analytic approaches, compare VHA and non‐VHA care models, and aim to better characterize the clinical efficiencies with a population health approach to DOAC management associated with both off‐label DOAC prescribing and other elements of antithrombotic stewardship.
This project was funded by the Agency for Healthcare Research and Quality through grant R18HS026874.
Dr Barnes reports grant funding from Boston Scientific and consulting for Pfizer, Bristol‐Myers Squibb, Janssen, Bayer, AstraZeneca, Sanofi, Anthos, Abbott Vascular, and Boston Scientific; and Board of Directors, Anticoagulation Forum. A.L. Allen reports speaking, AstraZeneca; and Board of Directors, Anticoagulation Forum. The remaining authors have no disclosures to report.
Data S1. Supporting Information.
|
Epidemiology of Acute Heart Failure in Critically Ill Patients With COVID-19: An Analysis From the Critical Care Cardiology Trials Network | 863dacd5-19e6-4b3a-9fba-2340adf1b064 | 8762923 | Internal Medicine[mh] | We analyzed consecutive admissions to ICUs of patients with COVID-19 from March 2020 to December 2020 across 6 academic medical centers in the United States using data from the Critical Care Cardiology Trials Network. Participating centers entered comprehensive clinical data into a central case-report form for patients with primary diagnoses of COVID-19 who had been admitted to all ICUs at their institutions. All patients admitted to the ICUs with cardiogenic shock (CS) (either classic or vasodilatory) or with acute HF without CS were classified as having an acute HF syndrome and were compared to patients without acute HF. CS was defined by sustained hemodynamic impairment (systolic blood pressure < 90 mmHg) and evidence of end-organ hypoperfusion due to low cardiac output. The distinction between classic and vasodilatory CS was based on high vs low systemic vascular resistance by using either invasive hemodynamic or clinical assessment. Classification of acute HF without CS was based on clinician assessment using local diagnostic standards and the entirety of the clinical record. Admissions for acute HF were further classified as de novo vs acute-on-chronic presentations based on the absence or presence of a prior diagnosis of HF, respectively. The protocol and waiver of informed consent were approved by the Institutional Review Board at Mass General Brigham and at each center. Baseline patient characteristics, presenting clinical features and ICU resource use were summarized according to presenting HF categories. Categorical variables are presented as counts and percentages, and continuous variables are presented as medians with 25th–75th percentiles. Differences between groups were evaluated using the Pearson χ 2 test for categorical variables and the Wilcoxon rank sum test for continuous variables.
Among 901 admissions to an ICU due to COVID-19, 80 (8.9%) had acute HF, including 18 (2.0%) with classic CS and 37 (4.1%) with vasodilatory CS. In our cohort, patients critically ill with COVID-19 and with acute HF had a median age of 64 (25th–75th percentile, 55–76) years and were predominantly male (70.0%). More than half were de novo presentations of HF (n = 45). Compared to patients critically ill due to COVID-19 but without acute HF, those with acute HF were more likely to have prior HF (43.8% vs 8.8%; P < 0.001), coronary artery disease (26.3% vs 9.5%; P < 0.001), atrial fibrillation (27.5% vs 8.8%; P < 0.001), or chronic kidney disease (32.5% vs 14.6%) ( P < 0.001) ( ). These comorbidities were more common in acute-on-chronic HF than in de novo HF ( ). Presentations with acute HF were most commonly due to left ventricular-predominant failure. Among patients with acute HF who had available presenting data for left ventricular ejection fraction (n = 67), 65.6% had left ventricular systolic dysfunction (LVEF < 50%), which was more common in patients with de novo (74.3%) vs acute-on-chronic HF (56.3%; P = 0.03) ( ). Of patients with acute HF, 16% had concurrent acute coronary syndromes ( ). Pulmonary vascular disease (eg, pulmonary hypertension, pulmonary embolism) was identified as a contributor in a minority of patients with biventricular (n = 5; 31.3%) and isolated right ventricular failure (n = 4; 25.0%). Acute myocarditis was not strictly defined or captured in this dataset. As compared to those without acute HF, patients with acute HF had significantly higher circulating biomarkers of myocardial injury (median baseline cardiac troponin (cTn): 3.2x [1.6x–8.7x] vs 1.0x [0.4x–2.6x], the 99th percentile upper reference limit [URL]; median peak cTn 12.7x [4.1x–53.3x] vs 2.1x [0.7x-7.0x] 99th percentile URL; P < 0.001 for both) and hemodynamic stress (median baseline N-terminal pro-B-type natriuretic peptide [NT-proBNP]: 2391 [976– 357] vs 381 [114– 459] pg/mL; median peak NT-proBNP: 5146 [ 319–23,446] vs 742 [186– 510] pg/mL; P < 0.001 for both) ( ). Although peak NT-proBNP concentrations were similar in de novo and acute-on-chronic HF (median 4518 [ 230–23,446] vs 5589 [ 505–23,977] pg/mL; P = 0.39), cTn was significantly higher in patients with de novo vs acute-on-chronic HF (median peak cTn 21.6x [7.4x–71.0x] vs 5.9x [2.1x–26.2x] 99th percentile URL; P = 0.004) ( ). This pattern was consistent in a sensitivity analysis excluding patients with acute coronary syndrome or cardiac arrest prior to ICU admission (median peak cTn 16.9x [7.3x–29.2x] vs 5.2x [2.1x–13.0x] 99th percentile URL; P = 0.019). In contrast to the distinct patterns observed with cardiovascular biomarkers, patients critically ill with COVID-19 with and without acute HF had similarly elevated biomarkers of systemic inflammation—median peak high-sensitivity C-reactive protein 176 (43–280) vs 123 (22–257) mg/L ( P = 0.14); median interleukin-6 (IL-6) 72 (54–304) vs 91 (30–297) pg/mL ( P = 0.98); and median ferritin 1480 (575–3,522) vs 1375 (652–2798) mg/L ( P = 0.60) ( ). However, patients with de novo HF tended to have more inflamation than those with acute-on-chronic HF ( ). Patients who are critically ill due to COVID-19 and have acute HF had modestly higher indices of disease severity as compared to those without acute HF (median Sequential Organ Failure Assessment score 8 [ –10] vs 6 [ –9]; P = 0.025), but similar patterns of ICU resource use, including mechanical ventilation ( P = 0.22) and acute renal replacement therapy ( P = 0.26). The median ICU length-of-stay among ICU survivors was similar in patients with and without acute HF (10.4 [2.9–17.9] vs 8.0 [3.6–18.2] days; P = 0.98) ( ). Patients critically ill with COVID-19 and with acute HF were more likely than patients without acute HF to experience cardiac arrest either before or during ICU admission (30.0% vs 10.8%; P < 0.001). In-hospital mortality was moderately higher in patients with vs without acute HF (43.8% vs 32.4%; P = 0.040). Patients with acute HF were more likely to have a cardiovascular (eg, acute myocardial infarction, HF, stroke, arrhythmia) mode of death (45.7% vs 16.5%; P < 0.001) and less likely to have a respiratory mode (54.3% vs 71.8%; P = 0.034) ( ).
Prior HF is an important prognostic indicator in COVID-19. , Our analysis extends this observation by demonstrating that pre-existing HF is also an important risk factor for the development of severe acute HF syndromes in patients critically ill with COVID-19. At the same time, more than half of admissions to ICUs for acute HF occurred in patients without prior diagnoses of HF, highlighting the clinically important risk of de novo myocardial dysfunction and HF in this population. In a single-center analysis of hospitalized (critically ill and noncritically ill) patients with COVID-19, 37 were identified as having de novo HF, 8 of whom had no prior cardiovascular disease or known risk factors. The point prevalence of de novo HF in our cohort was > 8-fold higher than that observed in that study (5.0% vs 0.6%), probably related to the higher overall risk of our exclusively ICU-based population. Nevertheless, we also observed that many patients with de novo HF had no known prior cardiovascular disease or risk factors. Collectively, these findings underscore the importance of recognizing this subset of patients and investigating the mechanisms of myocardial injury so we can tailor acute and chronic therapies and future preventive interventions. The biomarker profiles observed in our study also offer potentially important and clinically relevant insights. Both cTn and natriuretic peptide concentrations were strongly associated with acute HF presentation in critically ill patients with COVID-19; however, cTn was particularly elevated in de novo compared with acute-on-chronic HF, suggesting more acute myocardial injury in this group. Notably, although patients with COVID-19 in ICUs and with acute HF had elevated inflammatory markers, the degree of inflammation was comparable to those without acute HF, suggesting that the hyperinflammatory phenotype may not distinguish presentation with acute HF. Whether these biomarker patterns reflect the underlying mechanisms driving acute HF syndromes in critically ill patients with COVID-19 warrants further investigation (eg, correlation with cardiac MRI, endomyocardial biopsy). Finally, although mortality rates were high in patients critically ill with COVID-19, both with and without acute HF, those with acute HF had higher risks of cardiac arrest and of dying from a cardiovascular cause, which may have implications for optimal triage of these patients (eg, to cardiac ICUs). It is important to note that mortality estimates from our study period may be higher than contemporary estimates due to subsequent adoption of effective therapies (eg, corticosteroids). In conclusion, acute HF is an important complication in patients critically ill with COVID-19, occurring in approximately 1 in every 11 such patients. Although the risk of acute HF is higher in patients with prior HF, > 50% of acute HF syndromes in patients critically ill with COVID-19 are de novo presentations of HF. Among critically ill COVID-19 patients, presentation with acute HF is characterized more by elevations in biomarkers of myocardial injury and hemodynamic stress than by elevations in biomarkers of inflammation, and myocardial injury appears to be a particularly distinguishing feature of patients with de novo HF.
The present analysis was supported by Evergrande COVID-19 Response Fund Award from the Massachusetts Consortium on Pathogen Readiness (DDB, DAM, EAB). DDB is supported by Harvard Catalyst KL2/Catalyst Medical Research Investigator Training (National Center for Advancing Translational Sciences grant UL 1TR002541). SS reports receiving personal fees from Abiomed outside the submitted work. ASV is supported by the National Heart, Lung, and Blood Institute T32 postdoctoral training grant T32HL007604 and the Daniel Pierce Family Fellowship in Advanced Heart Disease. All other authors report no disclosures relevant to the contents of this paper.
|
Effectiveness of a comic book intervention on mental health literacy among adolescents and youth in Burkina Faso: a randomized controlled trial protocol | 35ff471c-06d1-4f34-8ea2-17ddbe2fe3d0 | 11715181 | Health Literacy[mh] | It is estimated that one in seven adolescents experiences a mental disorder globally, yet this vulnerable group is largely undiagnosed and untreated . The World Health Organization categorizes adolescents from the ages of 10 to 19 years of age and youth as 15 to 24 . This mental health crisis for adolescents and youth is especially critical in low-and middle-income countries (LMICs), where the high demand for quality mental health services is often met by low supply, as seen in Sub-Saharan Africa (SSA) . Burkina Faso, one of seven SSA countries ranked in the lower 5th percentile out of the Human Development Index , exemplifies this issue. While Burkina Faso has over 20 million inhabitants, there are only eleven psychiatrists, eighty-six nurses specialized in mental health, five psychologists, and ten neurologists who are practicing in the public healthcare systems . A national survey from 2015 indicated that 41.5% of Burkina Faso’s population currently suffers from at least one mental disorder, with depression being the most common . The country also has a very young population, with almost half (45.3%) under 15 years old . Approximately one in four adolescents aged 10 to 19 years (26.4%) in Burkina Faso reported experiencing depressive symptoms . Another study surveyed Burkina Faso youth aged 12 to 20 years on the prevalence of self-injurious thoughts and behaviors and found that 8% of 12–13-year-olds reported life was not worth living, with this percentage increasing to 20% among 18–20-year-olds . Burkina Faso’s 2020–2024 Mental Health Strategic Plan identifies several priority areas, including conducting mental health research (Priority Area 6) and focusing on specific groups such as adolescents (Priority Area 13). Advancing research on adolescent mental health can focus on promotion, prevention, and/or treatment. The WHO recommends increasing mental health knowledge and understanding as a strategy to prevent and promote mental health issues . Increasing mental health literacy (MHL) in this population represents an essential intervention pathway to understand mental health conditions, their symptoms, causes, and available treatments. The concept of MHL is defined as “understanding how to obtain and maintain positive mental health; understanding mental disorders and their treatments; decreasing stigma related to mental disorders, and enhancing help-seeking efficacy” . Individuals with high MHL are better equipped to make informed decisions about mental health treatment and use effective coping strategies . While adolescents and youth broadly have lower rates of MHL , this problem is more pronounced in LMICs . Most of the studies that target MHL take place in high-income countries and depend on the availability of trained mental health professionals, such as therapists, psychiatrists, and social workers, to deliver the interventions . This furthers the need for MHL interventions delivered by non-specialist health providers (NSHPs) such as teachers, lay health workers, and community health workers . In resource-constrained settings it can also be important to intervene efficiently. “Light-touch” interventions are low-cost and minimally invasive interventions . Light-touch interventions are typically aimed at improving psychological well-being for the subclinical population . Such interventions have been demonstrated to improve mental health in low-income settings . Many government organizations create brochures to educate the public on the basics of mental health education (see https://www.nimh.nih.gov/health/publications ). The use of brochures to increase health literacy is a widespread practice, and there have been mixed results regarding whether a brochure is more effective than a more visually engaging delivery mechanism, such as an illustrated brochure, brochure with photographs, or video . Comic books offer a low-cost, engaging approach to health education, accessible to varying literacy levels by providing pictures that provide visual explanations of the text . A comic's “fun factor” of a comic draws in individuals who would otherwise be reluctant to read text-only narratives . There is some evidence that comic books are especially effective for an adolescent population in communicating health knowledge . However, to the best of our knowledge, there have been no randomized control trials testing the effectiveness of using a comic book to improve MHL for adolescents and youth in LMICs. This study aims to evaluate whether an illustrated comic book or a brochure-like flyer (text-only) that describes the mental disorders of anxiety and depression as well as information around help-seeking and coping, is more effective at increasing mental health literacy for certain age groups. We hypothesize that both the comic book and the flyer will increase mental health literacy for all ages compared to a control group that receives no intervention. However, we hypothesize that the comic book will be more effective in increasing mental health literacy with younger adolescents (ages 10–14) compared to older adolescents and youth (ages 15–24) due to the engaging nature of illustrations for young readers. The primary outcome will be mental health literacy and secondary outcomes will include anxiety, depression, and intentions to cope. We will also explore how the information in the comic book and flyer might impact anxiety and depression scores as previous studies have shown that at times, individuals are negatively impacted after consuming information about depression . The primary and secondary outcomes will be assessed directly following the implementation of the intervention for the comic book and flyer groups.
Participants and recruitment This randomized controlled trial study is part of the research grant DASH for the design and evaluation of adolescent health interventions and policies in Sub-Saharan Africa (SSA). DASH is a network comprising public health research and training institutions from seven SSA countries (Burkina Faso, Ethiopia, Ghana, Nigeria, South Africa, Tanzania, Uganda), as well as Germany and the United States. It aims to address critical research gaps related to the need for interventions and policies and their design, evaluation, and transportability across three domains: nutrition and physical activity, sexual and reproductive health, and mental health and violence ( https://dash-rhissa.org/ ). This trial is integrated into the initial round of longitudinal data collection of a cohort of adolescents and young adults for each country, which will hereafter be referred to as the ‘DASH cohort study’. A Health and Demographic Surveillance System (HDSS) operates in each of the study communities, meaning that the DASH cohort study can use existing sampling frames of the full population to sample and recruit participants. Within the DASH cohort study, we will sample and recruit 2,007 adolescents and youth aged between 10 and 24 years (with equivalent proportions from the age categories 10–14; 15–19; and 20–24 years) in the Burkina Faso study community. Where the initial sample is not sufficient to reach 2,007 participants due to refusal to participate or incorrect sampling frames, we will draw further individuals from the existing sampling frame, until 2,007 participants are enrolled. These individuals will be followed up annually for four waves of data collection. This sample of 2,007 will also serve as the study sample for this trial.
This randomized controlled trial study is part of the research grant DASH for the design and evaluation of adolescent health interventions and policies in Sub-Saharan Africa (SSA). DASH is a network comprising public health research and training institutions from seven SSA countries (Burkina Faso, Ethiopia, Ghana, Nigeria, South Africa, Tanzania, Uganda), as well as Germany and the United States. It aims to address critical research gaps related to the need for interventions and policies and their design, evaluation, and transportability across three domains: nutrition and physical activity, sexual and reproductive health, and mental health and violence ( https://dash-rhissa.org/ ). This trial is integrated into the initial round of longitudinal data collection of a cohort of adolescents and young adults for each country, which will hereafter be referred to as the ‘DASH cohort study’. A Health and Demographic Surveillance System (HDSS) operates in each of the study communities, meaning that the DASH cohort study can use existing sampling frames of the full population to sample and recruit participants. Within the DASH cohort study, we will sample and recruit 2,007 adolescents and youth aged between 10 and 24 years (with equivalent proportions from the age categories 10–14; 15–19; and 20–24 years) in the Burkina Faso study community. Where the initial sample is not sufficient to reach 2,007 participants due to refusal to participate or incorrect sampling frames, we will draw further individuals from the existing sampling frame, until 2,007 participants are enrolled. These individuals will be followed up annually for four waves of data collection. This sample of 2,007 will also serve as the study sample for this trial.
To estimate the sample size and conduct the power analysis, we used the R package “Pwr” and the statistical software G*Power . We calculated the minimum detectable effect size (MDE) for primary and secondary outcomes, which represents the smallest effect size that would be sufficient to detect statistical significance, based on a predetermined level of significance ( α ), sample size ( N ), and statistical power (1— β ). We began by computing the power for t-tests of means to achieve a target power of 0.80. This process was conducted separately for each of the two waves in our study, which include three treatment arms and three age subgroups. In our calculations, we used the guideline α = 1 − (1 − 0.05) ^(1/√ h ) , where h was computed based on the number of treatment arms and outcomes. The overall sample size was set at N = 2007. For Wave 1, we allocated the sample size as N/3 per treatment arm. Anticipating a 10% attrition rate for follow-up measurements in Wave 2, the adjusted sample size was calculated as N = n = (sample/3) * (1—attrition). For the subgroup analysis in Wave 1, the sample size per age subgroup was assumed as N = sample/9, (i.e., N divided by 3 treatments arms * 3 age subgroups). Similarly, the adjusted sample size for Wave 2 subgroup analysis, the adjusted sample size was n = (sample/9) * (1—attrition). We used R Power to input the calculated MDEs and illustrate how each subgroup sample size corresponds to power levels (See Table ). There is a lack of studies investigating the impact of comic book intervention on mental health literacy in Africa. To ensure that the calculated effect estimate is something that we could reasonably expect, we compare our calculated effect sizes with those from a related study in a non-mental health domain. Shin and colleagues (2022) investigated the effectiveness of a comic book intervention in East Africa on knowledge about human papillomavirus (HPV) types . They reported that the mean percentage of correctly answered questions about HPV pre-intervention was 44%, which increased to 82.9% post-intervention. The study included a sample size of n = 64 for the pre-test and n = 72 for the post-test assessments. Cohen's d effect size was calculated using the following formulas. [12pt]{minimal}
$$d=$$ d = M e a n 1 - M e a n 2 PooledStandardDeviation [12pt]{minimal}
$$PooledSD=^{2}1+(n2-1)*{SD}^{2}2}{n1+n2-2}}$$ P o o l e d S D = n 1 - 1 ∗ SD 2 1 + n 2 - 1 ∗ SD 2 2 n 1 + n 2 - 2 This resulted in Cohen’s d = 3.58, which indicates a large effect size. In addition, there was a previous study on age differences in mental health literacy that used Cohen’s h values, a measure similar to Cohen’s d but applicable to differences between proportions or probabilities. A difference with h = 0.2 is considered “small”, h = 0.5 indicates a “medium” difference, and h = 0.8 represents a “large” difference. Their findings revealed that adults aged 70 years exhibited lower accuracy in identifying depression symptoms compared to those aged 18–24 years (Cohen’s h = 0.64), 25–39 years ( h = 0.61), 40–54 years ( h = 0.53), and 55–69 years ( h = 0.50). Based on this, we assume a medium effect size for the impact of age on mental health literacy. In summary, our sample size is likely adequate for detecting both large and medium effect sizes, as observed in previous studies.
In this randomized controlled trial, each participant will be randomly assigned (1:1:1) to receive either a comic book (Intervention 1, 15 min long) a flyer (Intervention 2, 15 min long), or no intervention (Control). The randomization process will be stratified for the age category (i.e., 10–14 years; 15–19; 20–24) to ensure that within each category, a similar number is allocated to each arm (see Fig. ). The allocation sequence will be generated by colleagues at the Technical University of Munich, who are not further involved in the conduct of the trial. They will use statistical computing software to define a random allocation sequence for all potentially eligible participants included in the pre-defined sample roster. As part of the DASH cohort study, which this trial is nested, all participants will take part in a survey interview of approximately 60–90 min. Each interview will begin with the data collector providing the consent form. If the participant consents, they will be asked survey questions regarding mental health literacy, anxiety, and depression before the intervention is administered, which will serve as baseline measures. For participants assigned to the comic book intervention, the data collector will provide the printed comic book to the participants, reading the text aloud. For participants assigned to the flyer intervention, the process will be the same where the data collector will provide the printed flyer, reading the text out loud. Participants in the control group will be provided with no additional instructions or interventions. Participants will then be asked other survey questions regarding sexual and reproductive health providing a brief break between the intervention implementation and the post-intervention survey. All participants will be surveyed on mental health literacy, anxiety, depression, and intentions to cope as a post measure.
Comic book The original comic book titled “Let’s Talk About It” is a 28-page Graphic Guide to Mental Health that was originally co-created by the Cartoon Studies Lab for the Ohio State Department of Health (USA), specifically designed for middle and high school students ( https://www.cartoonstudies.org/css-studio/cartooningprojects/mentalhealth/ ). As a light-touch intervention, we chose two pages from the comic book that specifically highlight the two most prevalent mental disorders in this population: Anxiety and Depression (see Fig. ). These pages primarily convey information about mental health, understanding mental disorders, guidance on seeking help, and addressing stigma through storytelling with an illustrated cast of rabbit characters. Each page briefly explains how biological processes in the brain result in anxiety and depression and how a disorder differs from feeling stressed or feeling blue. It also covers the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) based on symptomology for each disorder. Notably, we adapted the rabbit character by using a darker shade of brown fur to reflect a representative skin color that is prevalent within the cultural context. Also, we collaborated closely with the local mental health expert practicing in Burkina Faso, to translate and validate the English text into French, ensuring it was culturally sensitive and appropriate. Furthermore, we integrated local information platforms at the bottom of the page to provide participants with additional resources for further learning. Flyer The flyer intervention is designed to resemble the commonly used health information material in public health initiatives. The flyer contains text identical to that of the comic book. However, where text within illustrations could be confusing without accompanying visuals, we adjusted the wording slightly to maintain clarity. For example, in the comic book's text accompanying the first picture for Anxiety states: “Anxiety is an alarm from our brain’s fight, flight or freeze response. Hide. Fight. Run,” which may not fully convey its meaning without the illustration. The flyer version reads: “Anxiety is an alarm from our brain’s fight, flight or freeze response. It can look like fighting, freezing, or running away.” (See Fig. ). This adaptation ensures that the information remains accessible and understandable without accompanying visuals, maintaining the goal of the flyer as an educational tool while nevertheless remaining very close to the comic book text.
The original comic book titled “Let’s Talk About It” is a 28-page Graphic Guide to Mental Health that was originally co-created by the Cartoon Studies Lab for the Ohio State Department of Health (USA), specifically designed for middle and high school students ( https://www.cartoonstudies.org/css-studio/cartooningprojects/mentalhealth/ ). As a light-touch intervention, we chose two pages from the comic book that specifically highlight the two most prevalent mental disorders in this population: Anxiety and Depression (see Fig. ). These pages primarily convey information about mental health, understanding mental disorders, guidance on seeking help, and addressing stigma through storytelling with an illustrated cast of rabbit characters. Each page briefly explains how biological processes in the brain result in anxiety and depression and how a disorder differs from feeling stressed or feeling blue. It also covers the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) based on symptomology for each disorder. Notably, we adapted the rabbit character by using a darker shade of brown fur to reflect a representative skin color that is prevalent within the cultural context. Also, we collaborated closely with the local mental health expert practicing in Burkina Faso, to translate and validate the English text into French, ensuring it was culturally sensitive and appropriate. Furthermore, we integrated local information platforms at the bottom of the page to provide participants with additional resources for further learning.
The flyer intervention is designed to resemble the commonly used health information material in public health initiatives. The flyer contains text identical to that of the comic book. However, where text within illustrations could be confusing without accompanying visuals, we adjusted the wording slightly to maintain clarity. For example, in the comic book's text accompanying the first picture for Anxiety states: “Anxiety is an alarm from our brain’s fight, flight or freeze response. Hide. Fight. Run,” which may not fully convey its meaning without the illustration. The flyer version reads: “Anxiety is an alarm from our brain’s fight, flight or freeze response. It can look like fighting, freezing, or running away.” (See Fig. ). This adaptation ensures that the information remains accessible and understandable without accompanying visuals, maintaining the goal of the flyer as an educational tool while nevertheless remaining very close to the comic book text.
Our primary outcome will be the sum score on a revised version of a Universal Mental Health Literacy Scale for Adolescents (UMHL-A) , measured both before and after the intervention in wave 1 and a year later in wave 2. This scale assesses mental health literacy across domains such as mental health knowledge, mental disorder knowledge, help-seeking behavior, and stigma. The revised UMHL-A consists of 10 items, each rated on a 5-point Likert scale (1–5 points), yielding a total score ranging from of 10 to 50. Higher scores indicate a greater level of mental health literacy. Our secondary outcomes will include the total scores on the 2-item Patient Health Questionnaire (PHQ-2) , the 2-item Generalized Anxiety Disorder (GAD-2) , and the 2-item Intentions to Cope scale , all measured post-intervention. Outcome assessment will happen as part of waves 1 and 2 of the DASH cohort study; as part of the DASH cohort study, an attempt will be made to identify and follow-up all participants from wave 1 in wave 2. Outcomes will be assessed during the survey interview conducted by a data collector. All data collectors will be trained in administering the full survey, which should ensure outcomes are collected for all participants. All scales used in our study will be translated into French and reviewed by local mental health experts for cultural sensitivity. To assess reliability, we will calculate Cronbach’s alpha within our sample . First level of data entry and storage will be on a tablet before forms can be uploaded online. At this level only data collectors involved in the interview and/or data manager will have access to the data collected. Data collectors are responsible for privacy and confidentiality of the data on the device. Second level of data storage will be on the online server. Application developers will be responsible for setting up the server for online data storage. Server setup will include password protected access and different levels of access to ensure only authorized individuals can access the data online and only access their allowable sections of the application and/or data sources. Data cleaning will first be done on the tablet before the forms are uploaded online. Data collectors will be responsible for checking sections of the questionnaire for errors. The second level of data cleaning will be done by data managers after downloading the data from the online server. At this step, data will be checked for any duplicates (using household identifiers) and any failed logical checks. Any data errors resulting from how the questions are setup and/or asked will be communicated to program managers immediately.
We will assess participants’ sociodemographic characteristics, including age, gender, education levels, ethnicity, religion, household characteristics, and their parents’ highest level of education and occupation with outcome measures. To assess the effectiveness of an intervention on mental health literacy (MHL), we will build a multiple regression model. We will also assess each sub-scale component of MHL (knowledge about mental health, knowledge about mental health, help-seeking, and stigma). Here, the independent variable is “3 arms”, representing different intervention groups, and the outcome will be “MHL after the intervention.” We will control for baseline MHL scores by including them as a covariate, following the recommendation by Senn (2006) to adjust for potential baseline differences . For the analysis, we will employ dummy coding for the three trial arms to compare each intervention arm against the control arm. In addition, covariates such as depressive and anxiety symptoms, intentions to cope, age, gender, and education levels will be included in the model. Secondary outcomes will be examined using three separate linear regression models, with the trial arm” treated as a factor and the control arm serving as the reference category for each secondary outcome. To ensure the validity of our multiple regression analysis, we will calculate variance inflation factor values to check for multicollinearity among the independent variables. We will conduct statistical analyses using R version 4.1.0 ( www.r-project.org ). The final cleaned, de-identified and locked dataset of the trial will be accessible by the DASH network partners. Public access to the data will be made available upon review of the request and approval by each institute’s Principal Investigators.
This study will contribute valuable information on light-touch interventions for improving mental health outcomes for adolescents and youth in LMICs. We will use an efficient, low cost, scalable, and validated intervention to enhance MHL among adolescents and youth in Burkina Faso. Our findings will add knowledge on the effectiveness of using a comic book to increase knowledge about mental health and mental disorders, facilitate help-seeking behaviors, reduce stigma, and thereby enhance overall MHL. In addition, we will determine if comic books are more effective for specific age groups of adolescents and youth compared to a text-only flyer. This work will advance Burkina Faso’s 2020–2024 Mental Health Strategic Plan and employ strategies suggested by the WHO Mental Health Action Plan 2013—2020. Increasing MHL has the potential to improve mental health outcomes as low levels of MHL have been linked to adverse mental health outcomes . As randomized control trial studies on the effectiveness of comic books in impacting mental health literacy are few, this study will also add high quality research to future researchers who are considering this method of delivery. Due to the nature of this light-touch intervention, its long-term effectiveness may be limited. In addition, the exposure to the light-touch intervention would benefit from repeated exposure over a period of time. By harnessing the potential of the comic book approach and incorporating lessons learned from this study, we aim to create a comprehensive comic intervention, including a digital format, in the future. Moreover, we will investigate the effect of our comic intervention on other countries part of the DASH study, including Ethiopia, Ghana, Nigeria, Uganda, South Africa, and Tanzania. Our study will provide valuable insights into innovative and engaging ways of communicating mental health information to adolescents and youth through NSHP delivery agents.
Supplementary Material 1.
|
Comparing healthcare quality: A common framework for both ordinal and cardinal data with an application to primary care variation in England | 8faee2a8-d0ff-464f-b29a-382fc35b23d7 | 9804671 | Family Medicine[mh] | INTRODUCTION Comparing the quality of healthcare providers and measuring the degree of variation in quality are major policy concerns in many countries (Busse et al., ), with patients in England commonly said to face a “postcode lottery” in which their choice of healthcare provider and hence the quality of care they can expect to receive is determined by where they live. Making quality comparisons between healthcare providers or geographical areas is a routine exercise based on quantitative indicators of structure, process and outcome quality (Mainz, ), with the degree of variation captured using summary statistics such as the extremal quotient, coefficient of variation and systematic component of variation (Ibáñez et al., ). However, these summary statistics are only appropriate for quality indicators measured on a cardinal scale, such as staff to patient ratios, proportions of patients receiving indicated treatment and risk‐adjusted mortality rates. Nowadays, cardinal quality indicators are increasingly being supplemented by multicategory response information from patient experience surveys in which, importantly, respondents are typically asked to assess their quality of care by choosing between one of several ranked categories (e.g., very poor, poor, OK, good, very good). For example, England initiated a national patient survey program in 2001 (DeCourcy et al., ), with surveys now regularly conducted of patient experience in a range of primary and secondary care settings (NHS England, ). A critical limitation of this patient‐reported data for the summary evaluation of both the performance of individual healthcare providers and the variation between them is its qualitative or ordinal nature. In particular, the mean is not well defined for polytomous categorical response data, which in turn severely restricts the choice of dispersion measures. A common workaround has been to impose some numerical, perhaps latent, scale on the ordinal data, but this chosen scale is essentially arbitrary and different scales can yield substantially different results. For example, the resultant ranking of healthcare providers by mean quality levels will not in general be robust to simple monotonic transformations of the chosen scale (cf. Bond & Lang, ) and this non‐robustness problem extends to measures of variation that are a function of the mean (Allison & Forster, ). Another popular option is to collapse the number of categories to yield a binary 0/1 indicator that is amenable to analysis in terms of the proportion of patients reporting good (as opposed to not good) care (see e.g., Bruyneel et al., ). However, the choice of cutoff is again arbitrary yet impactful, and information is also inevitably discarded in the process. Neither of these standard approaches is therefore entirely satisfactory despite their widespread use in practice. The main contributions of this paper are twofold. First, we propose a quality assessment framework that is directly applicable to ordinal as well as cardinal quality indicators, without the need to first convert ordinal indicators into a cardinal scale such as the proportion meeting a binary threshold. For this purpose, we build directly on the methods used by Allanson to assess regional variation in ordinal indicators of health on the basis of the statistical preference criterion (De Schuymer et al., ; Montes et al., ), showing how this approach can be applied to cardinal as well as ordinal indicators and providing a novel application to the context of healthcare performance evaluation motivated by the notion that patients face a postcode lottery in healthcare provision. Specifically, we make use of information about the care quality profiles or distributions of all healthcare providers serving some population of interest to provide intelligible measures of both the comparative quality of each provider and the variation in quality between them. The comparative quality of a provider is defined as the difference in the chances that the quality of care received by a randomly chosen patient treated by that provider will be better rather than worse than that received by a randomly chosen patient from the population as a whole. The measure of variation is equal to the average absolute difference in the chances that the quality of care received by patients will be better rather than worse as a result of being treated by one provider rather than another, leading us to call it the “lottery” index. This index will take a minimum value of zero if all quality profiles are identical such that there is no difference in the chances that a randomly chosen patient treated by one provider will receive better rather than worse care than one treated by another. Conversely, it will take a maximum value of one if the quality of care provided by any one provider is certain to be either strictly better or strictly worse than that provided by any other, which will only be the case for non‐overlapping quality profiles. The intuition and mathematics behind our measures are set out in detail in the assessment framework section below. Second, we show how our assessment framework can generate useful new insights into the performance of healthcare systems by applying it to three different practice‐level indicators of the quality of primary care services in England—categorical response data from the annual GP Patient Survey (GPPS), ordinal inspection ratings from the Care Quality Commission (CQC), and cardinal measures of process quality from the Quality and Outcomes Framework (QOF)—all of which are published in searchable online databases to help inform patients' choices. Primary care services in England are delivered through general practices (“practices” hereafter) with the average practice responsible for the care of about 7000 adult patients. All practices are a member of one of nearly 200 Clinical Commissioning Groups (CCGs), which are responsible for the planning and commissioning of health care services for their local populations. We therefore examine variation in quality both between practices and between CCGs, where our analysis is most likely to be of interest to healthcare managers and policymakers responsible for the delivery of services at the population level rather than to individual patients looking to choose a practice that will meet their own personal care needs and preferences. The CQC, the independent regulator of health and social care service providers, reported wide variation between practices in the mean number of full‐time equivalent general practitioners (GPs) per head of registered population in 2018/19, with the geographical concentration of poor quality care, as shown by inspection ratings, making it difficult for people living in some areas to access good care (CQC , pp. 19, 20). NHS England and Ipsos MORI ( , p. 10) report considerable variation across individual CCGs in the proportion of patients describing their practice as either fairly or very good in the 2019 GPPS, ranging from 69.1% to 92.1%. Patients were given the right to choose their practice in 2015, with the aim of improving the quality of access to GP services, although practices are not bound to accept patients living outside their catchment area. Santos et al. investigate patients' choice of family doctor and show that individuals are more likely to choose practices with higher standards of care as measured by their total QOF score across all achievement indicators, trading off practice quality against distance. Policy concern about variation in the quality of healthcare services relates specifically to that part of the variation not warranted by differences in patient need or preferences. Accordingly, measures of healthcare performance are often standardized with the aim of identifying this unwarranted variation by controlling for the effects of differences in patient characteristics not under the control of providers such as age, sex, ethnicity, health and deprivation (see e.g., ; Public Health England, ). To investigate the impact of standardization on variation in primary care quality at the CCG level we report results based on both raw and indirectly standardized practice quality profiles, where the latter are what would be expected if quality outcomes conditional upon socio‐demographic characteristics were the same in each practice as in England as a whole. The main empirical analysis is based on data from the 2019 English GPPS questionnaire, which was sent out to more than 2 million people asking for feedback on their experiences. Practice‐level experience data, weighted by age and gender to resemble the population of eligible patients within each practice, are reported for nearly 7000 practices across 195 CCGs. We make use of the data on the proportions of patients in each practice reporting their overall experience as very poor, fairly poor, neither good nor poor, fairly good, and very good to explore the variation in primary care quality both between practices within each CCG and between CCGs in England. We also investigate the variation in primary care quality between CCGs using the CQC overall rating and total QOF score for each practice to see if these indicators provide ordinally equivalent information to the GPPS on some common latent “primary care quality” characteristic. Analysis of the CQC and QOF data is restricted to the CCG level because the practice‐level quality profiles for these indicators consist simply of a single overall rating or score. Comparative quality indices are calculated for all practices and CCGs using the GPPS data, and for all CCGs using the CQC and QOF data. The remainder of the paper is organized as follows. The next section introduces the conceptual framework, motivating the definitions of the comparative quality and lottery indices and outlining the indirect standardization procedure. Section discusses the various sources of data on practice quality which are employed in the empirical study, with the results presented in Section . The final section provides a discussion of the findings and concludes.
ASSESSMENT FRAMEWORK The basic building block of our assessment framework is the comparative evaluation of the quality profiles of pairs of healthcare providers (i.e., practices or CCGs) based on information about the care quality profile or distribution of each healthcare provider. We start with a simple numerical example to provide the intuition behind the approach, before turning to the general mathematical formulation and properties of the comparative quality and lottery indices. Finally, we outline the indirect standardization procedure. 2.1 Assessing pairwise quality differences Figure provides an example in which the quality profiles for two practices, A and B, are given as the proportion of patients in each practice who report their care as either “poor”, “OK”, or “good”—a three‐valued ordinal scale. We first note that neither conversion to a numerical scale nor dichotomization of the categories leads to a robust ranking of the quality profiles of the two practices. With numerical scaling, the mean quality of the two practices will be the same if the response options are assumed to be evenly spaced, being equal to 2.1 if the categories are scored 1, 2 and 3. But A has the higher mean if the distance between good and OK is greater than between OK and poor, whereas B has the higher mean if the opposite is the case. With dichotomization, A has the higher proportion of patients reporting quality as good (rather than OK or poor) but a lower proportion reporting quality as either good or OK (rather than poor). It follows that neither approach can provide a robust basis for an analysis of the variation in quality between practices. The calculation of the lottery index may be thought of in terms of the outcome of a lottery in which the patient has an equal chance of being assigned to A or B with the quality level for each practice determined by a random draw from the quality profile for that practice. The patient “wins” or “loses” depending on whether they are assigned to the practice with the higher or lower randomly chosen quality level, and will be indifferent to the lottery outcome if the quality levels delivered by the two practices are the same. Patients have a (15 + 15 + 3) = 33% chance of “winning” if assigned to A, a (4 + 16 + 12) = 32% chance of “winning” if assigned to B and will be indifferent to the lottery outcome in the remaining (20 + 3 + 12) = 35% of draws. Hence the difference in “winning” chances of (33 − 32) = 1% provides a measure of the degree to which the profile of A is superior to that of B. We proceed to calculate the lottery index as the absolute value of this difference, where this is equal by definition to the absolute difference in the chances that a patient randomly assigned to one practice will receive better rather than worse care than if assigned to the other. More generally, consider some population in which each individual is a patient of one (and only one) of a set of K ≥ 2 healthcare providers, such that the patient list of each provider is independent of that of any other. Let P Q k ≥ Q k ′ = P Q k > Q k ′ + P Q k = Q k ′ be the probability that the quality of care received by a randomly chosen patient with provider k ∈ K is at least as good as—that is, strictly better than or the same as—that received by a randomly chosen patient with provider k ′ ∈ K . Following Allanson , the pairwise quality difference is defined as the difference in chances that the quality of care received by a randomly chosen patient with provider k′ is (strictly) better rather than worse than that received by one with provider k : (1) Δ k k ′ = − Δ k ′ k = P Q k ′ ≥ Q k − P Q k ≥ Q k ′ = P Q k ′ > Q k − P Q k > Q k ′ ; ∀ k , k ′ ∈ K Δ k k ′ will take a value of zero if the quality profiles of the two providers are equivalent, although this does not necessarily imply that they are identical; a maximum value of one when the worst quality of care provided by provider k′ is strictly better than the best quality provided by provider k ; and a minimum value of minus one when the opposite is the case. The normative significance of the pairwise quality difference derives from the use of the statistical preference criterion (De Schuymer et al., ) for the comparative evaluation of quality profiles. According to this criterion one profile is better than another if the patient receiving the (strictly) higher quality care of any randomly chosen pair of patients is more likely to be registered with the first rather than the second provider. The criterion is more general and powerful than first‐order stochastic or rank dominance (De Baets & De Meyer, ), which is commonly employed to compare ordinal distributions but can lead to incomplete orderings (see e.g., Gutacker & Street, ). Statistical preference will always say whether one quality profile is better, worse or equivalent to another, whereas rank dominance often leaves things undefined—neither better nor worse, but not equivalent either. Thus, A and B in the numerical example are not comparable by rank dominance since the proportion of patients who receive poor care is lower in B but the proportion receiving no better than OK care is lower in A. Moreover, statistical preference is not only able to rank all quality profiles but also provides a “graded” comparison of them (De Baets & De Meyer, ), with the pairwise difference in winning chances offering a readily intelligible measure of the degree to which one profile is better or worse than another. 2.2 The comparative quality index In the absence of an external standard, a summary measure of comparative quality for each provider can be obtained by calculating a pairwise index for it relative to some common benchmark patient quality profile, such as that of the whole population (Allanson, ). The comparative quality index : (2) Δ k = ∑ k ′ = 1 K p k ′ P Q k > Q k ′ − P Q k ′ > Q k = ∑ k ′ = 1 K p k ′ Δ k ′ k ; ∀ k ∈ K offers a summary measure of the quality of provider k compared to all K providers, where p k ′ is the proportion of total registrations with provider k′ . The index may be used to generate a complete ranking of providers by quality but will generally be more informative than a simple measure of “league table” position. Δ k can take values in the closed interval from –(1 – p k ) to +(1 – p k ), since Δ kk = 0 by definition, with the sign of the index indicating whether the care quality of provider k is better or worse than the benchmark and its magnitude indicating the degree of any separation between the two profiles. By construction, Δ k takes a weighted average of zero across all providers, that is, ∑ k p k Δ k = 0. 2.3 The lottery index The lottery index provides a measure of the variation in quality between providers in terms of the average absolute value of the pairwise quality differences over all pairs of providers (Allanson, ). Specifically, the index is defined as the normalized average absolute value: (3) L = ∑ k = 1 K ∑ k ′ = 1 K p k p k ′ | Δ k k ′ | / 1 − ∑ k = 1 K p k 2 where the normalization factor 1 − ∑ k p k 2 implies that L may be interpreted as the patient‐weighted mean absolute difference in the chances that quality will be better rather than worse as a result of being cared for by one provider rather than another. The interpretation in terms of the average absolute difference in the chances of winning rather than losing over all distinct pairwise lotteries follows directly from the definition of the pairwise index | Δ k k ′ | . Alternatively, the index may be interpreted as a measure of the potential value to patients of exercising the right to choose their healthcare provider rather than it being determined by the accident of where they live. This follows since | Δ k k ′ | in Equation may also be written as: (4) | Δ k k ′ | = 2 max P Q k ′ > Q k , P Q k > Q k ′ − P Q k ′ > Q k + P Q k > Q k ′ ; ∀ k , k ′ ∈ K So L may also be interpreted as twice the mean increase in the probability that patient care will be better than it would otherwise have been if patients chose the provider with the better quality profile of any pair of providers rather than being randomly assigned to one of them. A third interpretation is in terms of the degree of “postcode discrimination” faced by patients on the basis of where they live due to the variation in care quality across providers. Specifically, L may be interpreted as a summary measure of discrimination between pairs of providers given that Δ k k ′ is formally equivalent to the Le Breton et al. first‐order discrimination index Δ 1 if provider k′ has the better profile of the two providers. L will take a minimum value of zero if and only if the comparative quality—but not necessarily the quality profiles—of all providers is the same and a maximum value of one if there is complete separation of the patient lists for each provider into disjoint strata in the population quality profile. The index is sensitive to any change in the quality of care received by any patient unless the change is over some quality range occupied exclusively by others cared for by the same provider as the patient. For binary 0/1 quality indicators (e.g., good or bad), L is simply the weighted average of the absolute pairwise differences in the proportion of patients receiving good care. But, as shown by the example, it can also be calculated for ordinal measures with three or more categories without the need for dichotomization. Given independent patient lists, the simplest way to compute L for an ordinal quality indicator is to calculate the pairwise indices using the approach employed in the numerical example and then take the weighted average over all pairs. A more computationally efficient approach if there are more than three health categories makes use of the relation Δ k k ′ = 1 − 2 P Q k > Q k ′ + 0.5 P Q k = Q k ′ in the first step. Supporting Information : Appendix 1 provides Stata code to compute values of Δ k k ′ , Δ k and L for a set of healthcare providers from ordinal quality data. For cardinal indicators, the pairwise indices can be calculated exactly from the relation Δ k k ′ = G b / G B if practice k′ has the higher mean quality of the two providers (Monti & Santori, ), where G b is the conventional between‐group Gini coefficient (Pyatt, ) and G b and is the variant proposed in Yitzhaki and Lerman . Alternatively, L may be approximated to any required degree of accuracy by rounding the data and then treating the resultant discretized variable like any other ordinal indicator. For both types of indicator, the measures are calculated directly from the data not from predicted or simulated quality profiles (see e.g., Gutacker & Street, ). 2.4 Standardization of practice quality profiles Previous studies have revealed systematic differences in how patients from different socio‐demographic groups evaluate the quality of primary care services (see e.g., Lyratzopoulos et al., ; Paddison et al., ). Individual response data from the GPPS could in principle be used to estimate directly standardized quality profiles calculated on the basis that all practices had the same socio‐demographic composition as the whole population. However, the sample size of the GPPS is not large enough to provide reliable estimates of group‐specific quality profiles at the practice level and the approach is in any case inapplicable to the practice‐wide CQC ratings and QOF scores. Instead we employ an indirect standardization procedure based on the estimation of a distribution regression model (Chernozhukov et al., ) for each quality indicator to predict the practice quality profiles that would be expected if quality outcomes conditional upon socio‐demographic characteristics were the same in each practice as in England as a whole. Specifically, the proportion of the patients of a practice expected to experience a quality level no better than q ( q = 1, … Q − 1 of Q discrete quality levels) is given by the prediction from a binary choice model in which the dependent variable takes a value equal to the proportion of patients reporting experience no better than q .
Assessing pairwise quality differences Figure provides an example in which the quality profiles for two practices, A and B, are given as the proportion of patients in each practice who report their care as either “poor”, “OK”, or “good”—a three‐valued ordinal scale. We first note that neither conversion to a numerical scale nor dichotomization of the categories leads to a robust ranking of the quality profiles of the two practices. With numerical scaling, the mean quality of the two practices will be the same if the response options are assumed to be evenly spaced, being equal to 2.1 if the categories are scored 1, 2 and 3. But A has the higher mean if the distance between good and OK is greater than between OK and poor, whereas B has the higher mean if the opposite is the case. With dichotomization, A has the higher proportion of patients reporting quality as good (rather than OK or poor) but a lower proportion reporting quality as either good or OK (rather than poor). It follows that neither approach can provide a robust basis for an analysis of the variation in quality between practices. The calculation of the lottery index may be thought of in terms of the outcome of a lottery in which the patient has an equal chance of being assigned to A or B with the quality level for each practice determined by a random draw from the quality profile for that practice. The patient “wins” or “loses” depending on whether they are assigned to the practice with the higher or lower randomly chosen quality level, and will be indifferent to the lottery outcome if the quality levels delivered by the two practices are the same. Patients have a (15 + 15 + 3) = 33% chance of “winning” if assigned to A, a (4 + 16 + 12) = 32% chance of “winning” if assigned to B and will be indifferent to the lottery outcome in the remaining (20 + 3 + 12) = 35% of draws. Hence the difference in “winning” chances of (33 − 32) = 1% provides a measure of the degree to which the profile of A is superior to that of B. We proceed to calculate the lottery index as the absolute value of this difference, where this is equal by definition to the absolute difference in the chances that a patient randomly assigned to one practice will receive better rather than worse care than if assigned to the other. More generally, consider some population in which each individual is a patient of one (and only one) of a set of K ≥ 2 healthcare providers, such that the patient list of each provider is independent of that of any other. Let P Q k ≥ Q k ′ = P Q k > Q k ′ + P Q k = Q k ′ be the probability that the quality of care received by a randomly chosen patient with provider k ∈ K is at least as good as—that is, strictly better than or the same as—that received by a randomly chosen patient with provider k ′ ∈ K . Following Allanson , the pairwise quality difference is defined as the difference in chances that the quality of care received by a randomly chosen patient with provider k′ is (strictly) better rather than worse than that received by one with provider k : (1) Δ k k ′ = − Δ k ′ k = P Q k ′ ≥ Q k − P Q k ≥ Q k ′ = P Q k ′ > Q k − P Q k > Q k ′ ; ∀ k , k ′ ∈ K Δ k k ′ will take a value of zero if the quality profiles of the two providers are equivalent, although this does not necessarily imply that they are identical; a maximum value of one when the worst quality of care provided by provider k′ is strictly better than the best quality provided by provider k ; and a minimum value of minus one when the opposite is the case. The normative significance of the pairwise quality difference derives from the use of the statistical preference criterion (De Schuymer et al., ) for the comparative evaluation of quality profiles. According to this criterion one profile is better than another if the patient receiving the (strictly) higher quality care of any randomly chosen pair of patients is more likely to be registered with the first rather than the second provider. The criterion is more general and powerful than first‐order stochastic or rank dominance (De Baets & De Meyer, ), which is commonly employed to compare ordinal distributions but can lead to incomplete orderings (see e.g., Gutacker & Street, ). Statistical preference will always say whether one quality profile is better, worse or equivalent to another, whereas rank dominance often leaves things undefined—neither better nor worse, but not equivalent either. Thus, A and B in the numerical example are not comparable by rank dominance since the proportion of patients who receive poor care is lower in B but the proportion receiving no better than OK care is lower in A. Moreover, statistical preference is not only able to rank all quality profiles but also provides a “graded” comparison of them (De Baets & De Meyer, ), with the pairwise difference in winning chances offering a readily intelligible measure of the degree to which one profile is better or worse than another.
The comparative quality index In the absence of an external standard, a summary measure of comparative quality for each provider can be obtained by calculating a pairwise index for it relative to some common benchmark patient quality profile, such as that of the whole population (Allanson, ). The comparative quality index : (2) Δ k = ∑ k ′ = 1 K p k ′ P Q k > Q k ′ − P Q k ′ > Q k = ∑ k ′ = 1 K p k ′ Δ k ′ k ; ∀ k ∈ K offers a summary measure of the quality of provider k compared to all K providers, where p k ′ is the proportion of total registrations with provider k′ . The index may be used to generate a complete ranking of providers by quality but will generally be more informative than a simple measure of “league table” position. Δ k can take values in the closed interval from –(1 – p k ) to +(1 – p k ), since Δ kk = 0 by definition, with the sign of the index indicating whether the care quality of provider k is better or worse than the benchmark and its magnitude indicating the degree of any separation between the two profiles. By construction, Δ k takes a weighted average of zero across all providers, that is, ∑ k p k Δ k = 0.
The lottery index The lottery index provides a measure of the variation in quality between providers in terms of the average absolute value of the pairwise quality differences over all pairs of providers (Allanson, ). Specifically, the index is defined as the normalized average absolute value: (3) L = ∑ k = 1 K ∑ k ′ = 1 K p k p k ′ | Δ k k ′ | / 1 − ∑ k = 1 K p k 2 where the normalization factor 1 − ∑ k p k 2 implies that L may be interpreted as the patient‐weighted mean absolute difference in the chances that quality will be better rather than worse as a result of being cared for by one provider rather than another. The interpretation in terms of the average absolute difference in the chances of winning rather than losing over all distinct pairwise lotteries follows directly from the definition of the pairwise index | Δ k k ′ | . Alternatively, the index may be interpreted as a measure of the potential value to patients of exercising the right to choose their healthcare provider rather than it being determined by the accident of where they live. This follows since | Δ k k ′ | in Equation may also be written as: (4) | Δ k k ′ | = 2 max P Q k ′ > Q k , P Q k > Q k ′ − P Q k ′ > Q k + P Q k > Q k ′ ; ∀ k , k ′ ∈ K So L may also be interpreted as twice the mean increase in the probability that patient care will be better than it would otherwise have been if patients chose the provider with the better quality profile of any pair of providers rather than being randomly assigned to one of them. A third interpretation is in terms of the degree of “postcode discrimination” faced by patients on the basis of where they live due to the variation in care quality across providers. Specifically, L may be interpreted as a summary measure of discrimination between pairs of providers given that Δ k k ′ is formally equivalent to the Le Breton et al. first‐order discrimination index Δ 1 if provider k′ has the better profile of the two providers. L will take a minimum value of zero if and only if the comparative quality—but not necessarily the quality profiles—of all providers is the same and a maximum value of one if there is complete separation of the patient lists for each provider into disjoint strata in the population quality profile. The index is sensitive to any change in the quality of care received by any patient unless the change is over some quality range occupied exclusively by others cared for by the same provider as the patient. For binary 0/1 quality indicators (e.g., good or bad), L is simply the weighted average of the absolute pairwise differences in the proportion of patients receiving good care. But, as shown by the example, it can also be calculated for ordinal measures with three or more categories without the need for dichotomization. Given independent patient lists, the simplest way to compute L for an ordinal quality indicator is to calculate the pairwise indices using the approach employed in the numerical example and then take the weighted average over all pairs. A more computationally efficient approach if there are more than three health categories makes use of the relation Δ k k ′ = 1 − 2 P Q k > Q k ′ + 0.5 P Q k = Q k ′ in the first step. Supporting Information : Appendix 1 provides Stata code to compute values of Δ k k ′ , Δ k and L for a set of healthcare providers from ordinal quality data. For cardinal indicators, the pairwise indices can be calculated exactly from the relation Δ k k ′ = G b / G B if practice k′ has the higher mean quality of the two providers (Monti & Santori, ), where G b is the conventional between‐group Gini coefficient (Pyatt, ) and G b and is the variant proposed in Yitzhaki and Lerman . Alternatively, L may be approximated to any required degree of accuracy by rounding the data and then treating the resultant discretized variable like any other ordinal indicator. For both types of indicator, the measures are calculated directly from the data not from predicted or simulated quality profiles (see e.g., Gutacker & Street, ).
Standardization of practice quality profiles Previous studies have revealed systematic differences in how patients from different socio‐demographic groups evaluate the quality of primary care services (see e.g., Lyratzopoulos et al., ; Paddison et al., ). Individual response data from the GPPS could in principle be used to estimate directly standardized quality profiles calculated on the basis that all practices had the same socio‐demographic composition as the whole population. However, the sample size of the GPPS is not large enough to provide reliable estimates of group‐specific quality profiles at the practice level and the approach is in any case inapplicable to the practice‐wide CQC ratings and QOF scores. Instead we employ an indirect standardization procedure based on the estimation of a distribution regression model (Chernozhukov et al., ) for each quality indicator to predict the practice quality profiles that would be expected if quality outcomes conditional upon socio‐demographic characteristics were the same in each practice as in England as a whole. Specifically, the proportion of the patients of a practice expected to experience a quality level no better than q ( q = 1, … Q − 1 of Q discrete quality levels) is given by the prediction from a binary choice model in which the dependent variable takes a value equal to the proportion of patients reporting experience no better than q .
DATA AND METHODS 3.1 Data Patient experience data for 6926 practices were obtained from the 2019 results of the annual GPPS (NHS England, ). The survey asked patients about a range of issues associated with using the services offered by their practice, including how they would describe their overall experience using a 5‐category semantic differential scale, as well as various questions about their own personal circumstances. The specific question was: “Overall, how would you describe your experience of your GP practice?”, with response categories: “Very good”, “Fairly good”, “Neither good nor poor”, “Fairly poor”, “Very poor”. Postal questionnaires were sent out in January 2019 to 2.33 million adult patients in England of whom 770512 in 6999 practices completed the survey representing a response rate of 33.1% (Ipsos MORI, ). All practices listed on NHS Digital as having eligible patients were included in the survey apart from an unspecified number that chose to opt out as they felt it was inappropriate to their patient population. Patients were eligible for inclusion in the survey if they had a valid NHS number, had been registered with a practice continuously for at least 6 months before being selected, and were 16 years of age or over. The sample was based on a proportionately stratified, unclustered design, with the sample size for each practice selected to ensure that confidence intervals were as consistent as possible between practices. Practice‐level data are published on a weighted basis to ensure that the results are more representative of the population of adult patients registered with each practice by correcting for the sampling design and to reduce the impact of non‐response bias. No overall experience data are available for 73 practices due to the suppression of data for questions answered by fewer than 10 people to protect confidentiality. Inspection ratings data for 6670 practices was obtained from the January 2019 CQC Care Directory (CQC, ). The Care Directory is updated monthly and includes the latest published ratings of all practices that have been subject to inspection in England, which in January 2019 dated back as far as November 2014. Practices are given an overall rating for the “whole population” of service users on a 4‐category semantic differential scale following a visit by an inspection team and taking account of the views of both patients and staff. The overall rating is based on a detailed assessment of the quality of care across six patient subgroups in terms of whether the service is safe, effective, caring, responsive to people's needs and well‐led. The CQC uses a risk‐based approach to target inspections in which practices rated “Inadequate” and “Requiring improvement” are required to make changes and subject to re‐inspection within six and 12 months respectively, while those rated “Good” or “Outstanding” are not liable to re‐inspection unless there is monitoring evidence of quality change (CQC, ). The most recent rating was used for practices with multiple ratings based on different inspection dates. The rating for the main branch of a practice was used where ratings were available for more than one location. QOF scores for 6854 practices with achievement data were obtained from the QOF 2018‐19 results (NHS Digital, ). The QOF is a voluntary, annual incentive payment scheme for all practices in England that rewards practices for the provision of “quality care”, with 95.1% of practices participating in the reporting year from April 1, 2018 to March 31, 2019. The QOF provides an indication of overall practice achievement through a points system, with points awarded against a range of 77 clinical care and public health indicators based, for example, on the proportion of patients on specified disease registers who receive defined interventions. The headline measure of practice achievement published by NHS England is percentage attainment of the maximum 559 QOF points available, but an alternative measure is also provided which takes account of instances where practices cannot achieve points because they have no patients pertinent to an indicator. We use the publicly reported scores and refrain from making an adjustment by adding “exception reported” patients back into the population denominator, which typically provides a less favorable measure of performance. QOF percentage attainment data are rounded to 1 decimal place to calculate the indirectly standardized quality profiles. 3.2 Methods The main analysis of patient experience data was based on the full GPPS sample of 6926 practices. A sub‐set of 6427 matched practices with valid GPPS, CQC and QOF data was used to generate comparable CCG‐level results for all three practice quality indicators. All sample practices belonged to one of 195 CCGs, with the number per CCG varying between 10 and 169, and a mean of 35.5. Practice weights based on the number of registered patients aged 16 years old and over in December 2018 (NHS Digital, ) were used to construct CCG quality profiles as weighted averages of sample practice profiles and, after adjusting for missing practices within each CCG, to ensure the national representativeness of results at the CCG level. Practice‐level comparative quality and within‐CCG lottery indices were calculated using the GPPS practice quality profiles, and CCG‐level comparative quality and between‐CCG lottery indices using the CCG quality profiles for all three indicators. Analysis of the CQC and QOF data was restricted to the CCG level because the practice‐level quality profiles for these indicators are degenerate distributions, consisting simply of an overall rating or score. This does not prevent the application of the measurement framework at the practice level but it does limit the informational value of the resultant indices. In particular, for continuous quality indicators the comparative quality indices of practices will simply be given by their rank in the population‐level quality distribution less half, while the between‐practice lottery index will equal one in the absence of ties. We report both total and indirectly standardized indices. For the estimation of indirectly standardized quality profiles, distribution regression models for each practice quality measure were specified as a function of sex, age, ethnicity, health and deprivation, where the models allow for main effects only given the nature of the socio‐demographic data. The composition of each practice patient list in terms of sex (female, male), age group (16–24, 25–34, 35–44, 45–54, 65–74, 75–84, 85+), ethnicity (White, Asian, Black, Mixed, Other) and health status (presence of long‐term condition) are separately reported in the GPPS data. One missing health status value was replaced by the CCG mean for the practice. Deprivation was measured by the 2019 Index of Multiple Deprivation score for the Lower Super Output Area in which the practice was located (Ministry of Housing, Communities & Local Government, ). The specifications also include a set of intercept dummy variables to allow for separate impacts on practice quality attributable to CCGs themselves. Predictions for each practice were based on the socio‐demographic characteristics of the practice and CCG shares of the English patient population to avoid adjusting for factors over which CCGs may have influence. In our base case analysis we employ a linear probability distribution regression model (LPDRM) for convenience but, as a robustness check, also calculate indirectly standardized profiles using a generalized linear distribution regression model (GLDRM) with a probit link function and a binomial distribution with the parameter n set equal to the number of survey responses in a practice for the GGPS data, and to one for the CQC and QOF data. Estimated counterfactual cumulative proportions were censored where necessary to lie in the unit interval, with the resultant set of predictions scaled to match the sample mean. Finally, bootstrap standard errors were obtained for all comparative quality and lottery indices by the resampling of practices within each CCG to reflect the organizational structure. All analysis was conducted using Stata version 15.1.
Data Patient experience data for 6926 practices were obtained from the 2019 results of the annual GPPS (NHS England, ). The survey asked patients about a range of issues associated with using the services offered by their practice, including how they would describe their overall experience using a 5‐category semantic differential scale, as well as various questions about their own personal circumstances. The specific question was: “Overall, how would you describe your experience of your GP practice?”, with response categories: “Very good”, “Fairly good”, “Neither good nor poor”, “Fairly poor”, “Very poor”. Postal questionnaires were sent out in January 2019 to 2.33 million adult patients in England of whom 770512 in 6999 practices completed the survey representing a response rate of 33.1% (Ipsos MORI, ). All practices listed on NHS Digital as having eligible patients were included in the survey apart from an unspecified number that chose to opt out as they felt it was inappropriate to their patient population. Patients were eligible for inclusion in the survey if they had a valid NHS number, had been registered with a practice continuously for at least 6 months before being selected, and were 16 years of age or over. The sample was based on a proportionately stratified, unclustered design, with the sample size for each practice selected to ensure that confidence intervals were as consistent as possible between practices. Practice‐level data are published on a weighted basis to ensure that the results are more representative of the population of adult patients registered with each practice by correcting for the sampling design and to reduce the impact of non‐response bias. No overall experience data are available for 73 practices due to the suppression of data for questions answered by fewer than 10 people to protect confidentiality. Inspection ratings data for 6670 practices was obtained from the January 2019 CQC Care Directory (CQC, ). The Care Directory is updated monthly and includes the latest published ratings of all practices that have been subject to inspection in England, which in January 2019 dated back as far as November 2014. Practices are given an overall rating for the “whole population” of service users on a 4‐category semantic differential scale following a visit by an inspection team and taking account of the views of both patients and staff. The overall rating is based on a detailed assessment of the quality of care across six patient subgroups in terms of whether the service is safe, effective, caring, responsive to people's needs and well‐led. The CQC uses a risk‐based approach to target inspections in which practices rated “Inadequate” and “Requiring improvement” are required to make changes and subject to re‐inspection within six and 12 months respectively, while those rated “Good” or “Outstanding” are not liable to re‐inspection unless there is monitoring evidence of quality change (CQC, ). The most recent rating was used for practices with multiple ratings based on different inspection dates. The rating for the main branch of a practice was used where ratings were available for more than one location. QOF scores for 6854 practices with achievement data were obtained from the QOF 2018‐19 results (NHS Digital, ). The QOF is a voluntary, annual incentive payment scheme for all practices in England that rewards practices for the provision of “quality care”, with 95.1% of practices participating in the reporting year from April 1, 2018 to March 31, 2019. The QOF provides an indication of overall practice achievement through a points system, with points awarded against a range of 77 clinical care and public health indicators based, for example, on the proportion of patients on specified disease registers who receive defined interventions. The headline measure of practice achievement published by NHS England is percentage attainment of the maximum 559 QOF points available, but an alternative measure is also provided which takes account of instances where practices cannot achieve points because they have no patients pertinent to an indicator. We use the publicly reported scores and refrain from making an adjustment by adding “exception reported” patients back into the population denominator, which typically provides a less favorable measure of performance. QOF percentage attainment data are rounded to 1 decimal place to calculate the indirectly standardized quality profiles.
Methods The main analysis of patient experience data was based on the full GPPS sample of 6926 practices. A sub‐set of 6427 matched practices with valid GPPS, CQC and QOF data was used to generate comparable CCG‐level results for all three practice quality indicators. All sample practices belonged to one of 195 CCGs, with the number per CCG varying between 10 and 169, and a mean of 35.5. Practice weights based on the number of registered patients aged 16 years old and over in December 2018 (NHS Digital, ) were used to construct CCG quality profiles as weighted averages of sample practice profiles and, after adjusting for missing practices within each CCG, to ensure the national representativeness of results at the CCG level. Practice‐level comparative quality and within‐CCG lottery indices were calculated using the GPPS practice quality profiles, and CCG‐level comparative quality and between‐CCG lottery indices using the CCG quality profiles for all three indicators. Analysis of the CQC and QOF data was restricted to the CCG level because the practice‐level quality profiles for these indicators are degenerate distributions, consisting simply of an overall rating or score. This does not prevent the application of the measurement framework at the practice level but it does limit the informational value of the resultant indices. In particular, for continuous quality indicators the comparative quality indices of practices will simply be given by their rank in the population‐level quality distribution less half, while the between‐practice lottery index will equal one in the absence of ties. We report both total and indirectly standardized indices. For the estimation of indirectly standardized quality profiles, distribution regression models for each practice quality measure were specified as a function of sex, age, ethnicity, health and deprivation, where the models allow for main effects only given the nature of the socio‐demographic data. The composition of each practice patient list in terms of sex (female, male), age group (16–24, 25–34, 35–44, 45–54, 65–74, 75–84, 85+), ethnicity (White, Asian, Black, Mixed, Other) and health status (presence of long‐term condition) are separately reported in the GPPS data. One missing health status value was replaced by the CCG mean for the practice. Deprivation was measured by the 2019 Index of Multiple Deprivation score for the Lower Super Output Area in which the practice was located (Ministry of Housing, Communities & Local Government, ). The specifications also include a set of intercept dummy variables to allow for separate impacts on practice quality attributable to CCGs themselves. Predictions for each practice were based on the socio‐demographic characteristics of the practice and CCG shares of the English patient population to avoid adjusting for factors over which CCGs may have influence. In our base case analysis we employ a linear probability distribution regression model (LPDRM) for convenience but, as a robustness check, also calculate indirectly standardized profiles using a generalized linear distribution regression model (GLDRM) with a probit link function and a binomial distribution with the parameter n set equal to the number of survey responses in a practice for the GGPS data, and to one for the CQC and QOF data. Estimated counterfactual cumulative proportions were censored where necessary to lie in the unit interval, with the resultant set of predictions scaled to match the sample mean. Finally, bootstrap standard errors were obtained for all comparative quality and lottery indices by the resampling of practices within each CCG to reflect the organizational structure. All analysis was conducted using Stata version 15.1.
RESULTS We first present results based on the full sample of practices with GPPS patient experience data, looking in turn at the indices calculated using the practice and CCG‐level quality profiles. We subsequently compare the indices calculated using the GPPS, CQC and QOF CCG‐level quality profiles constructed from the matched sample of practices with valid data for all three indicators. 4.1 GPPS patient experience This section reports results based on the full GPPS sample of 6926 practices. The proportions of adult patients in England reporting their overall experience of their practice as very poor, fairly poor, neither good nor poor, fairly good and very good were 2.1%, 4.4%, 10.6%, 37.8% and 45.1% respectively. Scoring these responses 1–5, practice quality was 4.19 on average with a standard deviation of 0.30 across all practices. It might thus appear that the variation in reported experience between practices was low relative to the mean, but the coefficient of variation can be made arbitrarily large or small through the choice of alternative scoring schemes. For example, if the responses were scored instead from −2 to +2, with 0 providing a natural measure of neither good nor poor, then the coefficient of variation would be 24.9% not 7.1%. Some other approach is therefore required to meaningfully assess the degree of variation in reported experience. Figure shows the distribution of practice‐level comparative quality index values, which have a patient‐weighted mean of zero by construction. The variation in comparative quality across individual practices is considerable, ranging from a 0.520 or 52.0 percentage point (pp) higher chance that a patient from the best practice would have reported a better rather than worse experience than one from anywhere in England to a 61.4pp lower chance for the worst practice. The standard deviation of the comparative quality index is 16.8pp, with within‐CCG differences accounting for 83.0% of the variance in practice‐level comparative quality and only 17.0% due to between‐CCG differences. Thus there was much more variation between practices within each CCG than between CCGs, where the former is of more relevance for the exercise of patient choice given the evidence that patients are only willing to travel a limited distance to access better quality GP services (Santos et al., ). Responses to the patient experience question are commonly collapsed into a dichotomous variable for presentational purpose by combining very poor/fairly poor/neither good nor poor into one category and fairly good/very good into the other (see e.g., NHS England and Ipsos MORI, ). However, the use of this binary quality indicator leads to a marked reduction in the ability to discriminate between “average” and “good” practices, while continuing to capture the extent to which “bad” practices offer poorer quality care. Thus, a patient from the best practice is now estimated to have had only a 17.1pp higher chance of reporting a better rather than worse experience than one from anywhere in England, whereas a patient from the worst practice would have had a 50.7pp lower chance. Overall, dichotomization leads to a substantial underestimate of the variation in the quality of care between practices with the standard deviation of the comparative quality index falling to 9.8pp as a result. The first row ([a] Full sample GPPS) of results in Table reports an average 17.8pp absolute difference in the chances that patient experience was better rather than worse as a result of being registered with one practice rather than another within the same CCG. Thus, on average, it was 8.9pp (=17.8/2) more likely that patient experience would have been better than it would otherwise have been as a result of being able to choose the better of any pair of practices within a CCG rather than being randomly assigned to one of them. Figure maps the variation in quality between practices within individual CCGs, ranging from a 9.0pp absolute difference in patients' chances of reporting a better rather than worse experience as a result of being registered with one practice rather than another in the most homogeneous CCG to a 30.3pp difference in the least. The expected value of the within‐CCG lottery index is not a function of the number of practices within a CCG although, unsurprisingly, the conditional variance is decreasing in the number of practices. Moreover, differences between the socio‐demographic composition of practices within individual CCGs account for relatively little of the total variation in practice quality within CCGs, with predicted within‐CCG variation highest in the more heterogeneous and segregated metropolitan areas based on the distribution regression estimates in Supporting Information , Tables and . Table reports that the within‐CCG lottery index based on the LPDRM indirectly standardized profile was 5.9pp rather than 17.8pp, leaving a residual or “unexplained” 11.9pp average absolute difference in the chances that reported patient experience would have been better rather than worse as a result of being registered with one practice rather than another within the same CCG. Figure shows that there was also considerable variation in the comparative quality levels of CCGs, ranging from a 18.3pp higher chance that a patient from the best CCG would have reported better rather than worse experience than one from anywhere in England to a 17.4pp lower chance for the worst CCG. Dichotomization again leads to a reduction in measured variation, particularly between “average” and “good” CCGs: the range in chances shrinks to 9.3pp higher for the best CCG to 12.6pp lower for the worst, that is to the difference in the proportion of patients reporting their experience as either fairly or very good between the best and worst performing CCGs (cf. NHS England and Ipsos MORI, , p. 10). Dichotomization also leads to some re‐ranking of CCGs in terms of performance, with the Kendall's rank correlation coefficient τ a between the two rankings implying that the full and dichotomized measures are 86.5pp (95% CI, 0.837–0.893) more likely to agree than differ over which of any pair of CCGs had the strictly better quality profile (cf. Newson, ). Figure maps the comparative quality of CCGs and shows that patient experience tends to be worse in CCGs located in metropolitan regions and surrounding areas than in the more rural “shire” counties. This geographical pattern is strongly associated with socio‐demographic differences between CCGs, with the LPDRM estimates in Supporting Information : Table implying that patient experience would be expected to have been worse in CCGs containing higher proportions of patients of prime working age (25–54 year olds), in the Asian ethnic group, with long‐term health conditions and living in more deprived areas. The between‐CCG lottery index would have been 5.2pp rather than 7.9pp if the only source of variation in practice quality was differences in the socio‐demographic composition of patient lists, leaving an unexplained or residual 2.7pp absolute difference in the chances that patient experience was better rather than worse as a result of being registered with one CCG rather than another. Dichotomization leads to a loss of contrast between better and worse performing CCGs but no fundamental change in the geographical pattern. 4.2 Comparative analysis of three practice quality measures This section reports results based on the matched sample of 6427 practices with valid GPPS, CQC and QOF quality data. The left‐hand plot in Figure and first row ([b] Common sample) of Table present results based on the GPPS CCG‐level quality profiles, which are virtually the same as those discussed above for the full GPPS sample. We compare these results to those obtained with the CQC and QOF indicators. The proportions of the patient population in England registered at a practice with a latest CQC rating of inadequate, requires improvement, good and outstanding were 0.7%, 2.9%, 90.4% and 6.0% respectively. This profile is likely to exaggerate the quality of GP services in January 2019 to the extent that the CQC inspection regime was quicker at picking up improvement in poorly rated practices than deterioration in highly rated ones. But limiting the analysis to practices that have been recently inspected may lead to the opposite problem as poorer quality practices were targeted for re‐inspection. Figure plots CCG comparative quality indices based on the latest CQC inspection ratings of all practices, ranging from a 74.7pp higher chance that a patient from the best CCG would have been in a practice with a higher rather than lower rating than one from anywhere in England to a 22.8pp lower chance for the worst CCG. However the best CCG—comprising a few, mostly outstanding practices—is an extreme outlier and the between‐CCG lottery index of 9.9pp reported in Table is not that much higher than that of the GPPS measure despite the much larger range. The association between the ranking of CCGs by CQC inspection rating and GPPS patient experience is positive but weak. Kendall's τ a is only 0.305 (95% CI, 0.203–0.406), implying that there was only a 30.5pp higher chance that the two measures would agree rather than differ over which of any pair of CCGs had the strictly better quality profile. The null hypothesis that τ a is equal to 1, which would be the value if the two measures produced identical rankings of CCGs, can be rejected decisively implying that GPPS patient experience and CQC inspection rating data do not provide alternative sources of ordinally equivalent information on some common latent “primary care quality” characteristic. Unlike for the GPPS measure, very little of the variation in inspection ratings between CCGs can be accounted for by practice‐level differences in socio‐demographic composition. The distribution regression results are given in Supporting Information Tables and , with Table reporting an LPDRM indirectly standardized lottery index of 0.0184 that is only 18.6% of the raw value. Levels of QOF achievement were very high with 14.5% of patients registered in practices achieving the maximum score of 559 QOF points, mean percentage achievement of 96.9pp (541.6 points), and standard deviations of 5.4pp (30.2 points) and 3.0pp (16.7 points) at the practice and CCG levels respectively. The right hand plot of CCG comparative quality indices in Figure is based on QOF scores, ranging from a 66.2pp higher chance that a patient from the best CCG would have been in a practice with a higher rather than lower QOF score than one from anywhere in England to a 93.3pp lower chance for the worst CCG. Table reports a between‐CCG lottery index of 0.2931 based on percentage achievement of the maximum score, with the alternative measure of percentage achievement of points available to the practice yielding the same result to 4 significant figures. Lottery indices for the separate clinical, public health and public health additional services domains are somewhat lower, but all are above 0.2 despite more than half of practices achieving the maximum score in the latter two domains. These considerably higher estimates of the variation in care quality compared to both the GPPS and CQC indices cannot simply be dismissed as an artifact of the cardinality of QOF scores: collapsing the total QOF score into a 5‐category variable with population proportions for England as a whole identical to those for the GPPS measure only reduces the index value to 0.2577. Rather they would appear to reflect the relatively high degree of variation in QOF scores between CCGs as compared to within CCGs, with the between‐CCG standard deviation of 3.0pp reported above similar in magnitude to a weighted‐average within‐CCG standard deviation of practice quality of 3.8pp: between‐CCG differences accounted for as much as 30.4% of the overall variance in practice‐level total QOF scores. The associations between the ranking of CCGs by QOF achievement and by the other two quality indicators are both weakly positive, with Kendall's τ a equal to 0.333 (95% CI, 0.236–0.431) for GPPS patient experience and 0.267 (95% CI, 0.160–0.374) for CQC inspection ratings. Only 30.7% of variation (0.0899/0.2925) in QOF achievement between CCGs was accounted for by differences in the socio‐demographic composition of practice lists, with the GLDRM yielding a somewhat higher estimate of the proportion of “explained” variation in this case. Illustrative distribution regression model results for QOF achievement are presented in Supporting Information Tables and . By way of summary, Figure maps the comparative quality indices by CCG quintile for the three alternative practice quality indicators. The maps share some similar features, which is to be expected given the positive association between the corresponding comparative quality indices. In particular, all show a concentration of CCGs with poorer levels of primary care quality in the London area. Nevertheless, the prevailing impression is of pervasive differences in the ranking of individual CCGs across the three measures, with nine CCGs in the top quintile on one measure and the bottom on another.
GPPS patient experience This section reports results based on the full GPPS sample of 6926 practices. The proportions of adult patients in England reporting their overall experience of their practice as very poor, fairly poor, neither good nor poor, fairly good and very good were 2.1%, 4.4%, 10.6%, 37.8% and 45.1% respectively. Scoring these responses 1–5, practice quality was 4.19 on average with a standard deviation of 0.30 across all practices. It might thus appear that the variation in reported experience between practices was low relative to the mean, but the coefficient of variation can be made arbitrarily large or small through the choice of alternative scoring schemes. For example, if the responses were scored instead from −2 to +2, with 0 providing a natural measure of neither good nor poor, then the coefficient of variation would be 24.9% not 7.1%. Some other approach is therefore required to meaningfully assess the degree of variation in reported experience. Figure shows the distribution of practice‐level comparative quality index values, which have a patient‐weighted mean of zero by construction. The variation in comparative quality across individual practices is considerable, ranging from a 0.520 or 52.0 percentage point (pp) higher chance that a patient from the best practice would have reported a better rather than worse experience than one from anywhere in England to a 61.4pp lower chance for the worst practice. The standard deviation of the comparative quality index is 16.8pp, with within‐CCG differences accounting for 83.0% of the variance in practice‐level comparative quality and only 17.0% due to between‐CCG differences. Thus there was much more variation between practices within each CCG than between CCGs, where the former is of more relevance for the exercise of patient choice given the evidence that patients are only willing to travel a limited distance to access better quality GP services (Santos et al., ). Responses to the patient experience question are commonly collapsed into a dichotomous variable for presentational purpose by combining very poor/fairly poor/neither good nor poor into one category and fairly good/very good into the other (see e.g., NHS England and Ipsos MORI, ). However, the use of this binary quality indicator leads to a marked reduction in the ability to discriminate between “average” and “good” practices, while continuing to capture the extent to which “bad” practices offer poorer quality care. Thus, a patient from the best practice is now estimated to have had only a 17.1pp higher chance of reporting a better rather than worse experience than one from anywhere in England, whereas a patient from the worst practice would have had a 50.7pp lower chance. Overall, dichotomization leads to a substantial underestimate of the variation in the quality of care between practices with the standard deviation of the comparative quality index falling to 9.8pp as a result. The first row ([a] Full sample GPPS) of results in Table reports an average 17.8pp absolute difference in the chances that patient experience was better rather than worse as a result of being registered with one practice rather than another within the same CCG. Thus, on average, it was 8.9pp (=17.8/2) more likely that patient experience would have been better than it would otherwise have been as a result of being able to choose the better of any pair of practices within a CCG rather than being randomly assigned to one of them. Figure maps the variation in quality between practices within individual CCGs, ranging from a 9.0pp absolute difference in patients' chances of reporting a better rather than worse experience as a result of being registered with one practice rather than another in the most homogeneous CCG to a 30.3pp difference in the least. The expected value of the within‐CCG lottery index is not a function of the number of practices within a CCG although, unsurprisingly, the conditional variance is decreasing in the number of practices. Moreover, differences between the socio‐demographic composition of practices within individual CCGs account for relatively little of the total variation in practice quality within CCGs, with predicted within‐CCG variation highest in the more heterogeneous and segregated metropolitan areas based on the distribution regression estimates in Supporting Information , Tables and . Table reports that the within‐CCG lottery index based on the LPDRM indirectly standardized profile was 5.9pp rather than 17.8pp, leaving a residual or “unexplained” 11.9pp average absolute difference in the chances that reported patient experience would have been better rather than worse as a result of being registered with one practice rather than another within the same CCG. Figure shows that there was also considerable variation in the comparative quality levels of CCGs, ranging from a 18.3pp higher chance that a patient from the best CCG would have reported better rather than worse experience than one from anywhere in England to a 17.4pp lower chance for the worst CCG. Dichotomization again leads to a reduction in measured variation, particularly between “average” and “good” CCGs: the range in chances shrinks to 9.3pp higher for the best CCG to 12.6pp lower for the worst, that is to the difference in the proportion of patients reporting their experience as either fairly or very good between the best and worst performing CCGs (cf. NHS England and Ipsos MORI, , p. 10). Dichotomization also leads to some re‐ranking of CCGs in terms of performance, with the Kendall's rank correlation coefficient τ a between the two rankings implying that the full and dichotomized measures are 86.5pp (95% CI, 0.837–0.893) more likely to agree than differ over which of any pair of CCGs had the strictly better quality profile (cf. Newson, ). Figure maps the comparative quality of CCGs and shows that patient experience tends to be worse in CCGs located in metropolitan regions and surrounding areas than in the more rural “shire” counties. This geographical pattern is strongly associated with socio‐demographic differences between CCGs, with the LPDRM estimates in Supporting Information : Table implying that patient experience would be expected to have been worse in CCGs containing higher proportions of patients of prime working age (25–54 year olds), in the Asian ethnic group, with long‐term health conditions and living in more deprived areas. The between‐CCG lottery index would have been 5.2pp rather than 7.9pp if the only source of variation in practice quality was differences in the socio‐demographic composition of patient lists, leaving an unexplained or residual 2.7pp absolute difference in the chances that patient experience was better rather than worse as a result of being registered with one CCG rather than another. Dichotomization leads to a loss of contrast between better and worse performing CCGs but no fundamental change in the geographical pattern.
Comparative analysis of three practice quality measures This section reports results based on the matched sample of 6427 practices with valid GPPS, CQC and QOF quality data. The left‐hand plot in Figure and first row ([b] Common sample) of Table present results based on the GPPS CCG‐level quality profiles, which are virtually the same as those discussed above for the full GPPS sample. We compare these results to those obtained with the CQC and QOF indicators. The proportions of the patient population in England registered at a practice with a latest CQC rating of inadequate, requires improvement, good and outstanding were 0.7%, 2.9%, 90.4% and 6.0% respectively. This profile is likely to exaggerate the quality of GP services in January 2019 to the extent that the CQC inspection regime was quicker at picking up improvement in poorly rated practices than deterioration in highly rated ones. But limiting the analysis to practices that have been recently inspected may lead to the opposite problem as poorer quality practices were targeted for re‐inspection. Figure plots CCG comparative quality indices based on the latest CQC inspection ratings of all practices, ranging from a 74.7pp higher chance that a patient from the best CCG would have been in a practice with a higher rather than lower rating than one from anywhere in England to a 22.8pp lower chance for the worst CCG. However the best CCG—comprising a few, mostly outstanding practices—is an extreme outlier and the between‐CCG lottery index of 9.9pp reported in Table is not that much higher than that of the GPPS measure despite the much larger range. The association between the ranking of CCGs by CQC inspection rating and GPPS patient experience is positive but weak. Kendall's τ a is only 0.305 (95% CI, 0.203–0.406), implying that there was only a 30.5pp higher chance that the two measures would agree rather than differ over which of any pair of CCGs had the strictly better quality profile. The null hypothesis that τ a is equal to 1, which would be the value if the two measures produced identical rankings of CCGs, can be rejected decisively implying that GPPS patient experience and CQC inspection rating data do not provide alternative sources of ordinally equivalent information on some common latent “primary care quality” characteristic. Unlike for the GPPS measure, very little of the variation in inspection ratings between CCGs can be accounted for by practice‐level differences in socio‐demographic composition. The distribution regression results are given in Supporting Information Tables and , with Table reporting an LPDRM indirectly standardized lottery index of 0.0184 that is only 18.6% of the raw value. Levels of QOF achievement were very high with 14.5% of patients registered in practices achieving the maximum score of 559 QOF points, mean percentage achievement of 96.9pp (541.6 points), and standard deviations of 5.4pp (30.2 points) and 3.0pp (16.7 points) at the practice and CCG levels respectively. The right hand plot of CCG comparative quality indices in Figure is based on QOF scores, ranging from a 66.2pp higher chance that a patient from the best CCG would have been in a practice with a higher rather than lower QOF score than one from anywhere in England to a 93.3pp lower chance for the worst CCG. Table reports a between‐CCG lottery index of 0.2931 based on percentage achievement of the maximum score, with the alternative measure of percentage achievement of points available to the practice yielding the same result to 4 significant figures. Lottery indices for the separate clinical, public health and public health additional services domains are somewhat lower, but all are above 0.2 despite more than half of practices achieving the maximum score in the latter two domains. These considerably higher estimates of the variation in care quality compared to both the GPPS and CQC indices cannot simply be dismissed as an artifact of the cardinality of QOF scores: collapsing the total QOF score into a 5‐category variable with population proportions for England as a whole identical to those for the GPPS measure only reduces the index value to 0.2577. Rather they would appear to reflect the relatively high degree of variation in QOF scores between CCGs as compared to within CCGs, with the between‐CCG standard deviation of 3.0pp reported above similar in magnitude to a weighted‐average within‐CCG standard deviation of practice quality of 3.8pp: between‐CCG differences accounted for as much as 30.4% of the overall variance in practice‐level total QOF scores. The associations between the ranking of CCGs by QOF achievement and by the other two quality indicators are both weakly positive, with Kendall's τ a equal to 0.333 (95% CI, 0.236–0.431) for GPPS patient experience and 0.267 (95% CI, 0.160–0.374) for CQC inspection ratings. Only 30.7% of variation (0.0899/0.2925) in QOF achievement between CCGs was accounted for by differences in the socio‐demographic composition of practice lists, with the GLDRM yielding a somewhat higher estimate of the proportion of “explained” variation in this case. Illustrative distribution regression model results for QOF achievement are presented in Supporting Information Tables and . By way of summary, Figure maps the comparative quality indices by CCG quintile for the three alternative practice quality indicators. The maps share some similar features, which is to be expected given the positive association between the corresponding comparative quality indices. In particular, all show a concentration of CCGs with poorer levels of primary care quality in the London area. Nevertheless, the prevailing impression is of pervasive differences in the ranking of individual CCGs across the three measures, with nine CCGs in the top quintile on one measure and the bottom on another.
DISCUSSION Evidence on the quality of healthcare services is increasingly being provided by multicategory response information from patient experience surveys, supplementing the routine collection of standard cardinal quality indicators. This paper proposes an assessment framework that is directly applicable to both ordinal and cardinal quality indicators, providing intelligible measures of both the comparative quality of each member of a set of healthcare providers serving some population and the variation in quality between them. Our approach is motivated by the concept of statistical preference whereby one healthcare provider is judged to be better than another if the patient receiving the (strictly) higher quality care of any randomly chosen pair of patients is more likely to be registered with the first rather than the second provider. Unlike first order stochastic dominance, statistical preference will provide a graded comparison of all possible pairs of care quality profiles. The resultant measures are sensitive to the full distribution of quality scores for each provider, not just the mean nor the proportion meeting some binary quality threshold. The GPPS offers a large‐scale, annual survey of patients' experience in virtually all practices in England, with practice‐level multicategory response data made publicly available in a timely fashion. We find significant variation in primary care quality levels both between practices within individual CCGs and between CCGs in 2019, with the right to choose between any two practices within a CCG leading on average to an 8.9pp higher chance that patient experience would be better than it would otherwise have been under random assignment. Dichotomization leads to a reduction in measured variation, with the loss of contrast most marked between “average” and “good” providers. Practice‐level information on primary care quality is also available in the form of ordinal CQC inspection ratings and cardinal QOF achievement scores, which are generated for regulatory and performance incentive purposes respectively. We show that neither provide an alternative source of ordinally equivalent information on some common latent “primary care quality” variable to either the GPPS or each other. Allen et al. have previously found a similar lack of agreement between routine performance indicators, measuring patient satisfaction and the management of chronic conditions, and CQC inspection ratings based on the limited ability of the former to predict the latter. Additionally, the measured level of between‐CCG variation is much higher using QOF scores than with the other two quality indicators. Why this is the case is unclear though we do demonstrate that it is not due to the cardinality of the QOF indicator by showing that the value of the lottery index is relatively insensitive to the grouping of QOF scores. Elimination of the postcode lottery in GP patient experience would provide a measurable, policy‐relevant objective to the extent that such variation was due to factors within the control of the National Health Service. In particular, attainment of the goal would not require that all individual patients could expect to receive the same quality of care, which is surely unrealistic, but rather that their experience was equally likely to be better rather than worse as a result of being registered with one practice or CCG rather than another to the extent that this was achievable. A limitation of our study is that the indirect standardization procedure is based on practice‐level rather than individual patient data, which prevents the specification of interaction terms between socio‐demographic characteristics and runs the risk of ecological bias. Nevertheless our findings are consistent with those from other studies (see e.g., Lyratzopoulos et al., ; Paddison et al., ) in showing that patient experience tended to be worse in practices located in more deprived areas with higher proportions of patients of prime working age, in the Asian ethnic group and in poorer health. These variations may be due to differences in reporting behavior between different patient subgroups and/or systematic disparities in the actual standard of care provided to them, with our finding that socio‐demographic characteristics explained much smaller but still significant proportions of between‐CCG variation in both CQC ratings and QOF scores suggesting that both factors were of importance (see e.g., Lyratzopoulos et al., , Burt at al., ; Fisher et al., ; for further evidence on this point). In conclusion, the proposed approach provides a general framework to assess variation between healthcare providers or geographical areas making full use of the information provided by the ordinal quality indicators that are now routinely available. Further studies are required both to elicit healthcare service decision makers' views on the utility of our proposed new performance metric and to explore whether our empirical findings are more generally characteristic of the scale of healthcare variation in other clinical settings and countries. Finally, we note that the statistical preference criterion can also be used to compare care quality profiles that are not independent of each other. In particular it would be of interest to evaluate changes in healthcare provider quality over time taking account of the temporal dependence of patient experience, with the impact of the COVID‐19 epidemic on GP practice quality an obvious topic for investigation.
No conflicts of interest exist.
The paper raises no ethical issues.
Supporting Information S1 Click here for additional data file.
|
Distribution of | 389a5241-d534-415c-8f8e-ba72c7810e86 | 11610360 | Microbiology[mh] | Clostridioides difficile is amongst the most common intestinal bacterial pathogens worldwide. C. difficile infection (CDI) is associated with disturbed gut microbiota and presents from mild diarrhoea to colitis or pseudomembranous colitis . C. difficile spores are ubiquitous and transmission routes include humans, animals, and the environment . Colonized animals, regardless of the symptoms, can be a multiplying host and source, from which spores can be transmitted to humans via direct contact, faecal contamination of meat, and contamination of vegetables via manure or irrigation. Meat products and root vegetables are often found to be contaminated by C. difficile spores. Animals and food are therefore potentially important in overall C. difficile transmissions . Two main typing approaches are currently used. C. difficile strains are distributed into ribotypes by PCR-based ribotyping, whereas genome-based sequence types (STs) divide strains into five main clades (I to V) and four cryptic C. difficile clades (C-I to C-IV) . Clades C-I to C-IV are divergent branches in the overall C. difficile species structure and according to average nucleotide identity (ANI) values should not be regarded as C. difficile species . PCR ribotyping (RT) and whole-genome sequencing (WGS) have detected overlap between ribo- and genotypes shared between humans, environment, animals or food ; however, One Health-related studies covering more than two sources/reservoirs in combination with extensive geographic coverage are rare . To date, One Health studies with WGS data are mainly performed in a single country, on low numbers of strains and mostly on reservoir pairs (humans/animals; humans/environment) , and large multi-country studies focus either on a single sequence type (ST) or RT or a single source only (humans, ; animals, ; food, ). Evidence that C. difficile ribotypes differ in prevalence in different transmission routes is building. Genomic analysis, on the level of a single hospital, revealed that certain ribotypes showed high (RT027) medium (RT001, RT106) or low (RT078/126, RT014/020) probability of hospital transmission, based on the degree of genomic similarity with other strains of the same ribotype in the same hospital . This correlates well with 2 distribution patterns observed in a comparative genomic analysis of a collection of strains isolated from humans from 19 European countries ; within-country clustering likely associated with intra-hospital transmissions (RT027, RT001/072, RT176, RT018) and cross-country transmissions (RT078, RT014, RT020, RT002, RT015) . In this study, we present a unique collection containing strains concurrently sampled from 4 sources (humans in hospital and community settings, animals, and food) in 13 European countries. This collection provides the opportunity to compare the genomic relatedness of contemporary strains from various sources and countries, enabling the detection of associations between ribotypes and sources, and potential ribotype association with transmission routes. Strain collection Collection of samples from human and food sources (potatoes) was described previously . Thirteen countries were included in the study, no samples from humans were collected from Austria and no samples from animals were collected from Belgium (Supplementary Figure S1). Specifically, for the human samples, these were residual diagnostic diarrhoeal faecal samples collected across 12 European countries during the COMBACTE-CDI point-prevalence study in 2018 . All the human origin C. difficile isolates from the COMBACTE-CDI point-prevalence study (i.e. samples that yielded a positive C. difficile culture, and regardless of toxigenic status) were analysed in this study for PCR ribotyping, toxinotyping and whole-genome sequencing (WGS). For animal origin strains, we contacted authors of veterinary C. difficile -related publications or veterinarians through the COMBACTE-CDI network. Contact persons were asked for either available contemporary C. difficile strains or samples (faeces or rectal swab) from piglets. Data on animal age were collected if available. If contemporary strains or samples were not available, contacts were asked to submit strains isolated from pigs from any time period and/or C. difficile strains isolated from other animals from any time period. Isolation dates together with additional details are available in Supplementary Table S1. Detailed isolation procedures are described in Supplementary Doc S1. PCR ribotyping and toxinotyping DNA extraction and capillary-based PCR ribotyping were performed as described for strains isolated from humans in the COMBACTE-CDI study . Strains from potatoes or from animal faecal samples were checked for clonality with crude PCR ribotyping before inclusion in capillary-based PCR ribotyping . Toxinotyping was performed as described in Rupnik . Whole-genome sequencing (WGS) Residual DNA extracts from capillary-based PCR ribotyping were used for library preparations using the Nextera XT DNA Library Prep kit (Illumina). Sequencing was performed with MiSeq v3 600-cycles kit (MiSeq platform) or with NextSeq 550 High-Output v 2.5 300-cycles kit (Illumina NextSeq 500 platform). EPISEQ®CS cloud-based software application (bioMérieux) was used for quality control checks of the FASTQ files (Supplementary Doc S1). Sequence analysis with EPISEQ®CS application De novo assembly from WGS data, contamination check, conventional MLST prediction and whole-genome Multilocus Sequence Typing (wgMLST)-based analysis were performed with the bioMérieux EPISEQ®CS cloud-based application v1.2.0. The wgMLST scheme for C. difficile in EPISEQ®CS has been created using 259 reference genomes and contains a total of 8745 C. difficile loci (including 1999 core loci and 7 MLST loci). With similarity set at 95% (isolates considered possibly related), the EPISEQ®CS dendrogram was edited using Microsoft Publisher to add metadata. Calculated distances (number of pairwise allelic differences for detected wgMLST loci) were downloaded from the application and presented as a table with metadata. Prim’s algorithm (Jarník–Prim–Dijkstra) was applied to the distance matrix for constructing the minimum spanning tree (MST) graph. The MST graph displayed in the EPISEQ®CS interactive window panel was edited by dragging nodes with the mouse cursor for visualization without overlapping branches and edited for annotations using source code and vector graphics editors (Notepad++, Inkscape v1.2.1). Sequence analysis with BioNumerics A whole-genome single nucleotide variant analysis (wgSNV) was performed with BioNumerics software v7.5 (bioMérieux). Raw reads were mapped to the reference C. difficile genomes. The following genomes were used as a reference: CD630, clade 1/PCR ribotype 012 (NCBI accession # AM180355.1); DSM 102859, clade 3/PCR ribotype 023 (NZ_CP020378.1); M120, clade 5/PCR ribotype 078 (NZ_CP068555.1) and R20291, clade 2/PCR ribotype 027 (NZ_CP029423.1). The SNP calling was performed using the “strict SNP filtering,” removing all SNPs with at least one unreliable base (i.e. N), ambiguous base (non-ATCG) or gap and all non-informative SNPs. SNPs were called if they had at least 5× coverage, once in both directions and the minimum distance of 12 bp between SNPs. Genomic analysis of variant strains from divergent clades Genomes were assembled de novo with SPAdes v3.13.1, and Kraken2 v2.1.2 was used to assign taxonomy and check for contamination. Concatenated MLST alleles were extracted from C. difficile genomes, from 10 divergent isolates and an additional 33 representatives of the diversity of the C. difficile population, from 8 clades (1–5 and cryptic clades C-I to C-III), to assess the distribution of divergent isolates within the C. difficile population. Neighbour-joining tree, based on MAFFT-aligned concatenated MLST loci, was constructed using MEGA 11. C. difficile isolates with negative PCR results for PaLoc genes and 115-bp region were screened for the presence of tcd A and tcd B genes with BLAST search (coverage and identity thresholds of 40% and 60%, respectively). Extracted tcd A sequences were aligned with MUSCLE and maximum likelihood trees were constructed in MEGA11. A BLAST search was used also to identify the presence of binary toxin locus (CDTLoc) and divergent CDT genes carried on bacteriophage phiSemix9P1 . Collection of samples from human and food sources (potatoes) was described previously . Thirteen countries were included in the study, no samples from humans were collected from Austria and no samples from animals were collected from Belgium (Supplementary Figure S1). Specifically, for the human samples, these were residual diagnostic diarrhoeal faecal samples collected across 12 European countries during the COMBACTE-CDI point-prevalence study in 2018 . All the human origin C. difficile isolates from the COMBACTE-CDI point-prevalence study (i.e. samples that yielded a positive C. difficile culture, and regardless of toxigenic status) were analysed in this study for PCR ribotyping, toxinotyping and whole-genome sequencing (WGS). For animal origin strains, we contacted authors of veterinary C. difficile -related publications or veterinarians through the COMBACTE-CDI network. Contact persons were asked for either available contemporary C. difficile strains or samples (faeces or rectal swab) from piglets. Data on animal age were collected if available. If contemporary strains or samples were not available, contacts were asked to submit strains isolated from pigs from any time period and/or C. difficile strains isolated from other animals from any time period. Isolation dates together with additional details are available in Supplementary Table S1. Detailed isolation procedures are described in Supplementary Doc S1. PCR ribotyping and toxinotyping DNA extraction and capillary-based PCR ribotyping were performed as described for strains isolated from humans in the COMBACTE-CDI study . Strains from potatoes or from animal faecal samples were checked for clonality with crude PCR ribotyping before inclusion in capillary-based PCR ribotyping . Toxinotyping was performed as described in Rupnik . Whole-genome sequencing (WGS) Residual DNA extracts from capillary-based PCR ribotyping were used for library preparations using the Nextera XT DNA Library Prep kit (Illumina). Sequencing was performed with MiSeq v3 600-cycles kit (MiSeq platform) or with NextSeq 550 High-Output v 2.5 300-cycles kit (Illumina NextSeq 500 platform). EPISEQ®CS cloud-based software application (bioMérieux) was used for quality control checks of the FASTQ files (Supplementary Doc S1). Sequence analysis with EPISEQ®CS application De novo assembly from WGS data, contamination check, conventional MLST prediction and whole-genome Multilocus Sequence Typing (wgMLST)-based analysis were performed with the bioMérieux EPISEQ®CS cloud-based application v1.2.0. The wgMLST scheme for C. difficile in EPISEQ®CS has been created using 259 reference genomes and contains a total of 8745 C. difficile loci (including 1999 core loci and 7 MLST loci). With similarity set at 95% (isolates considered possibly related), the EPISEQ®CS dendrogram was edited using Microsoft Publisher to add metadata. Calculated distances (number of pairwise allelic differences for detected wgMLST loci) were downloaded from the application and presented as a table with metadata. Prim’s algorithm (Jarník–Prim–Dijkstra) was applied to the distance matrix for constructing the minimum spanning tree (MST) graph. The MST graph displayed in the EPISEQ®CS interactive window panel was edited by dragging nodes with the mouse cursor for visualization without overlapping branches and edited for annotations using source code and vector graphics editors (Notepad++, Inkscape v1.2.1). Sequence analysis with BioNumerics A whole-genome single nucleotide variant analysis (wgSNV) was performed with BioNumerics software v7.5 (bioMérieux). Raw reads were mapped to the reference C. difficile genomes. The following genomes were used as a reference: CD630, clade 1/PCR ribotype 012 (NCBI accession # AM180355.1); DSM 102859, clade 3/PCR ribotype 023 (NZ_CP020378.1); M120, clade 5/PCR ribotype 078 (NZ_CP068555.1) and R20291, clade 2/PCR ribotype 027 (NZ_CP029423.1). The SNP calling was performed using the “strict SNP filtering,” removing all SNPs with at least one unreliable base (i.e. N), ambiguous base (non-ATCG) or gap and all non-informative SNPs. SNPs were called if they had at least 5× coverage, once in both directions and the minimum distance of 12 bp between SNPs. Genomic analysis of variant strains from divergent clades Genomes were assembled de novo with SPAdes v3.13.1, and Kraken2 v2.1.2 was used to assign taxonomy and check for contamination. Concatenated MLST alleles were extracted from C. difficile genomes, from 10 divergent isolates and an additional 33 representatives of the diversity of the C. difficile population, from 8 clades (1–5 and cryptic clades C-I to C-III), to assess the distribution of divergent isolates within the C. difficile population. Neighbour-joining tree, based on MAFFT-aligned concatenated MLST loci, was constructed using MEGA 11. C. difficile isolates with negative PCR results for PaLoc genes and 115-bp region were screened for the presence of tcd A and tcd B genes with BLAST search (coverage and identity thresholds of 40% and 60%, respectively). Extracted tcd A sequences were aligned with MUSCLE and maximum likelihood trees were constructed in MEGA11. A BLAST search was used also to identify the presence of binary toxin locus (CDTLoc) and divergent CDT genes carried on bacteriophage phiSemix9P1 . DNA extraction and capillary-based PCR ribotyping were performed as described for strains isolated from humans in the COMBACTE-CDI study . Strains from potatoes or from animal faecal samples were checked for clonality with crude PCR ribotyping before inclusion in capillary-based PCR ribotyping . Toxinotyping was performed as described in Rupnik . Residual DNA extracts from capillary-based PCR ribotyping were used for library preparations using the Nextera XT DNA Library Prep kit (Illumina). Sequencing was performed with MiSeq v3 600-cycles kit (MiSeq platform) or with NextSeq 550 High-Output v 2.5 300-cycles kit (Illumina NextSeq 500 platform). EPISEQ®CS cloud-based software application (bioMérieux) was used for quality control checks of the FASTQ files (Supplementary Doc S1). De novo assembly from WGS data, contamination check, conventional MLST prediction and whole-genome Multilocus Sequence Typing (wgMLST)-based analysis were performed with the bioMérieux EPISEQ®CS cloud-based application v1.2.0. The wgMLST scheme for C. difficile in EPISEQ®CS has been created using 259 reference genomes and contains a total of 8745 C. difficile loci (including 1999 core loci and 7 MLST loci). With similarity set at 95% (isolates considered possibly related), the EPISEQ®CS dendrogram was edited using Microsoft Publisher to add metadata. Calculated distances (number of pairwise allelic differences for detected wgMLST loci) were downloaded from the application and presented as a table with metadata. Prim’s algorithm (Jarník–Prim–Dijkstra) was applied to the distance matrix for constructing the minimum spanning tree (MST) graph. The MST graph displayed in the EPISEQ®CS interactive window panel was edited by dragging nodes with the mouse cursor for visualization without overlapping branches and edited for annotations using source code and vector graphics editors (Notepad++, Inkscape v1.2.1). A whole-genome single nucleotide variant analysis (wgSNV) was performed with BioNumerics software v7.5 (bioMérieux). Raw reads were mapped to the reference C. difficile genomes. The following genomes were used as a reference: CD630, clade 1/PCR ribotype 012 (NCBI accession # AM180355.1); DSM 102859, clade 3/PCR ribotype 023 (NZ_CP020378.1); M120, clade 5/PCR ribotype 078 (NZ_CP068555.1) and R20291, clade 2/PCR ribotype 027 (NZ_CP029423.1). The SNP calling was performed using the “strict SNP filtering,” removing all SNPs with at least one unreliable base (i.e. N), ambiguous base (non-ATCG) or gap and all non-informative SNPs. SNPs were called if they had at least 5× coverage, once in both directions and the minimum distance of 12 bp between SNPs. Genomes were assembled de novo with SPAdes v3.13.1, and Kraken2 v2.1.2 was used to assign taxonomy and check for contamination. Concatenated MLST alleles were extracted from C. difficile genomes, from 10 divergent isolates and an additional 33 representatives of the diversity of the C. difficile population, from 8 clades (1–5 and cryptic clades C-I to C-III), to assess the distribution of divergent isolates within the C. difficile population. Neighbour-joining tree, based on MAFFT-aligned concatenated MLST loci, was constructed using MEGA 11. C. difficile isolates with negative PCR results for PaLoc genes and 115-bp region were screened for the presence of tcd A and tcd B genes with BLAST search (coverage and identity thresholds of 40% and 60%, respectively). Extracted tcd A sequences were aligned with MUSCLE and maximum likelihood trees were constructed in MEGA11. A BLAST search was used also to identify the presence of binary toxin locus (CDTLoc) and divergent CDT genes carried on bacteriophage phiSemix9P1 . Collection of contemporary European strains from humans, animal, and food sources Altogether 13 European countries participated in this study (Supplementary Figure S1). The initial strain set contained 1043 isolates from humans (isolates from the COMBACTE-CDI point-prevalence study, i.e. hospital and community; n = 280), food (potatoes; n = 504) and animals ( n = 259). All the human C. difficile isolates from the COMBACTE-CDI point-prevalence study were included, whereas strains from animal and food sources were checked for clonality. In the animal strain collection, clonality was defined as the same PCR ribotype in faecal samples from the same farm. In food strains, clonality was defined as the same PCR ribotype in the same potato sample. After the removal of clonal isolates from the animal and food collection, the final 441 strains were included in the comparison between countries and across the sources based on their ribotype and whole-genome sequence (humans n = 280, potato n = 65, animal n = 96, Supplementary Table S1). The median number of strains per country was 34 (range 6–73) (Supplementary Figure S1). Spain, France, Italy, Poland, and Romania were well represented with human, animal and food origin strains. Sweden and Slovakia contributed lower numbers of strains (8–19) but from all three sources. Poland had the highest number of food origin strains ( n = 10). France and Spain had the most animal/pig origin strains ( n = 33, n = 10, respectively). No samples from animals were obtained from Greece, Ireland, Belgium and the UK. MLST sequence types correlate well with EPISEQ®CS wgMLST clusters All 441 sequenced genomes were distributed into 83 STs (Supplementary Table S1). Of these, 9 STs had more than 10 genomes, with ST11 (80 genomes) and ST1 (54 genomes) being the most common ( and ; Supplementary Table S1). With the EPISEQ®CS application, the wgMLST clustering based on 95% similarity was performed on a panel of 430 isolates from all 441 sequenced isolates included in ribotype/source analysis (Supplementary Figure S1). This panel of 430 sequences excluded sequenced food origin isolates previously shown to be identical ( n = 6, RT912, ), or that did not pass the EPISEQ®CS C. difficile identification tag ( n = 5). Of 430 genomes, 385 (90%) were distributed in 53 clusters (with ≥2 genomes) and 45 genomes were singletons (Supplementary Figure S2). Individual clusters contained from 2 to 71 strains. Eight of 53 clusters included 2 closely related STs. All other EPISEQ®CS clusters included a single ST (Supplementary Figure S2; Supplementary Table S1). Some STs could be further differentiated with EPISEQ®CS into two or more clusters. Most diverse in this respect were ST2 (28 strains, 4 clusters), ST3 (20 strains, 5 clusters) and ST11 (80 strains, 4 clusters). Individual ST could include from one to eight PCR ribotypes (Supplementary Table S1). The greatest diversity was detected in ST11 (eight RTs; 078/126 and related RTs), ST14, ST2 (both with RT 014/020 and related RTs) and ST6 (RT005 and related RTs). While most STs clustered together, at least one (ST3) showed three unrelated clusters for isolates with RT 009, RT 001, and RT484, respectively ( ; pink shadowed squares). Individual RTs were not homogeneous regarding the STs and could include from one to four STs. Most diverse were RT014 and RT020 (Supplementary Table S2). Sequence types and relationship between isolates from different sources An overview of the relationship between isolates ( n = 430) from different sources, STs and RTs is summarized in . The intermingling of diverse PCR ribotypes and sources was observed in small and large clusters ( ; Supplementary Figure S2). All top 5 STs (ST11, ST1, ST2, ST6, ST3) represented by at least 20 isolates, with more than one PCR ribotype per ST, include strains originating from all source combinations. The only highly represented ST that did not include strains from various sources was ST26 ( n = 16 human isolates only, linked mainly to RT039). Two of the largest ST groups and selected small ST groups are shown in more detail in . ST11 with RT078 and RT126 and a small sub-cluster with RT045 showed the shortest allelic distances between isolates from diverse sources (source: animal n = 50, food n = 10, humans in hospital n = 13, humans in the community n = 7). The ST1 cluster is comprised of two related sub-clusters (allele difference = 315), RT181 only had human hospital and community origin isolates, and mixed clusters RT027, RT176, RT198, and RT016 had human and two food origin isolates. ST6, and particularly RT005 was well represented among animal origin isolates but was also found in human and food origin isolates. ST2 included predominately isolates from humans and some from animal and food sources. wgSNP analysis did not always confirm close genetic relatedness within clusters Clusters identified using the wgMLST approach showed varying degrees of genetic relatedness after further detailed wgSNP genomic analysis. Within individual clusters, between 0 and 864 SNPs (median 48) were found between pairs of isolates. For the RT181 cluster, isolates showed a close genetic relationship (0–10 SNPs). All isolates in this cluster were from Romania, except for one isolate from the UK. The RT027 cluster, consisting of 23 isolates (21 human and 2 food isolates) exhibited a close genetic relationship but with slightly higher genetic heterogeneity compared to RT181. Within this cluster, there were 2–68 SNPs found between isolates (both food isolates differ from the closest human isolate by >10 SNPs), and sub-clustering of strains based on the country of origin was observed . Low genetic heterogeneity was confirmed for ST11 (RTs 078, 126 and 193). All 71 isolates in this cluster had no more than 30 SNPs difference. Sub-clusters of isolates from human and animal or food sources in ST11 were from different countries. The wgSNP analysis did identify strains from different sources with relatively low numbers of SNP differences. Specifically, among RT045 strains, there were strains isolated from both humans and pigs collected in Poland that had less than 10 SNPs difference. A similar pattern was seen for some isolates from pigs from both Poland and Slovakia . Similarly, among RT078 and RT126 strains, sub-clusters with less than 10 SNPs were identified. For RT078, there was a sub-cluster consisting of six strains, including two isolated from humans in France and Spain, three from pigs and one from a pigeon from Spain. Notably, two animal origin strains from this cluster were isolated in 2010 and 2017 . For RT126, there was a cluster of 6 strains (<10 SNPs), including 2 isolated from humans from Italy and Spain, 3 from pigs from France, and 1 from a potato from Romania . Ribotype distribution across countries and reservoirs The 441 strains were distributed across 106 PCR ribotypes (Supplementary Table S2). Of those, 57 had a single representative strain, 23 had 2 or 3 representatives and 26 had 4 or more representatives . PCR ribotypes 078, 126, 014, 181 and 027 were the most identified in our collection and are present in several countries. The only exception was RT181, with 26 strains from only 2 countries and mainly from hospitals, potentially indicating an outbreak situation. Diversity within each source was high ( (A) and ). Strains isolated from humans in the hospital were distributed into 66 PCR ribotypes, those from humans in the community into 41, from animals in 30 and from food in 36 PCR ribotypes. Each of the 4 sources had from 7 to 34 non-shared PCR ribotypes. The most common PCR ribotypes for each source collectively are shown in (B). Of 106 PCR ribotypes, 38 were positioned in intersections of different sources ( (A)). Ten PCR ribotypes were shared between all 4 sources, 9 PCR ribotypes between 3 sources and 18 PCR ribotypes between 2 sources ( , Supplementary Table S2). Any combination of two or three sources had mostly from one to four common PCR ribotypes and the highest number of shared PCR ribotypes ( n = 26) was observed between isolates from humans in the hospital and community settings. Of these 26 ribotypes, 9 were only found in humans and shared between hospital and community patients. The only combination without any shared PCR ribotypes was Food/Animal/Community . Divergent C. difficile lineages were detected in food and exceptionally in human strains Neighbour-joining phylogenetic tree on concatenated MLST loci demonstrated that 10 isolates belonged to one of the 3 cryptic clades (C-I to C-III) (Supplementary Figure S3, ). Five isolates were highly divergent also on the wgMLST dendrogram (Supplementary Figure S2) and another five were not analysed with EPISEQ®CS as these were not recognized as C. difficile . Four of these 10 cryptic isolates, all from food, carried divergent toxin A gene (Supplementary Figure S3, ). The single human isolate did not show any variant forms of toxin A or B. None of these divergent isolates had CDT genes or variant CDT genes. Altogether 13 European countries participated in this study (Supplementary Figure S1). The initial strain set contained 1043 isolates from humans (isolates from the COMBACTE-CDI point-prevalence study, i.e. hospital and community; n = 280), food (potatoes; n = 504) and animals ( n = 259). All the human C. difficile isolates from the COMBACTE-CDI point-prevalence study were included, whereas strains from animal and food sources were checked for clonality. In the animal strain collection, clonality was defined as the same PCR ribotype in faecal samples from the same farm. In food strains, clonality was defined as the same PCR ribotype in the same potato sample. After the removal of clonal isolates from the animal and food collection, the final 441 strains were included in the comparison between countries and across the sources based on their ribotype and whole-genome sequence (humans n = 280, potato n = 65, animal n = 96, Supplementary Table S1). The median number of strains per country was 34 (range 6–73) (Supplementary Figure S1). Spain, France, Italy, Poland, and Romania were well represented with human, animal and food origin strains. Sweden and Slovakia contributed lower numbers of strains (8–19) but from all three sources. Poland had the highest number of food origin strains ( n = 10). France and Spain had the most animal/pig origin strains ( n = 33, n = 10, respectively). No samples from animals were obtained from Greece, Ireland, Belgium and the UK. All 441 sequenced genomes were distributed into 83 STs (Supplementary Table S1). Of these, 9 STs had more than 10 genomes, with ST11 (80 genomes) and ST1 (54 genomes) being the most common ( and ; Supplementary Table S1). With the EPISEQ®CS application, the wgMLST clustering based on 95% similarity was performed on a panel of 430 isolates from all 441 sequenced isolates included in ribotype/source analysis (Supplementary Figure S1). This panel of 430 sequences excluded sequenced food origin isolates previously shown to be identical ( n = 6, RT912, ), or that did not pass the EPISEQ®CS C. difficile identification tag ( n = 5). Of 430 genomes, 385 (90%) were distributed in 53 clusters (with ≥2 genomes) and 45 genomes were singletons (Supplementary Figure S2). Individual clusters contained from 2 to 71 strains. Eight of 53 clusters included 2 closely related STs. All other EPISEQ®CS clusters included a single ST (Supplementary Figure S2; Supplementary Table S1). Some STs could be further differentiated with EPISEQ®CS into two or more clusters. Most diverse in this respect were ST2 (28 strains, 4 clusters), ST3 (20 strains, 5 clusters) and ST11 (80 strains, 4 clusters). Individual ST could include from one to eight PCR ribotypes (Supplementary Table S1). The greatest diversity was detected in ST11 (eight RTs; 078/126 and related RTs), ST14, ST2 (both with RT 014/020 and related RTs) and ST6 (RT005 and related RTs). While most STs clustered together, at least one (ST3) showed three unrelated clusters for isolates with RT 009, RT 001, and RT484, respectively ( ; pink shadowed squares). Individual RTs were not homogeneous regarding the STs and could include from one to four STs. Most diverse were RT014 and RT020 (Supplementary Table S2). An overview of the relationship between isolates ( n = 430) from different sources, STs and RTs is summarized in . The intermingling of diverse PCR ribotypes and sources was observed in small and large clusters ( ; Supplementary Figure S2). All top 5 STs (ST11, ST1, ST2, ST6, ST3) represented by at least 20 isolates, with more than one PCR ribotype per ST, include strains originating from all source combinations. The only highly represented ST that did not include strains from various sources was ST26 ( n = 16 human isolates only, linked mainly to RT039). Two of the largest ST groups and selected small ST groups are shown in more detail in . ST11 with RT078 and RT126 and a small sub-cluster with RT045 showed the shortest allelic distances between isolates from diverse sources (source: animal n = 50, food n = 10, humans in hospital n = 13, humans in the community n = 7). The ST1 cluster is comprised of two related sub-clusters (allele difference = 315), RT181 only had human hospital and community origin isolates, and mixed clusters RT027, RT176, RT198, and RT016 had human and two food origin isolates. ST6, and particularly RT005 was well represented among animal origin isolates but was also found in human and food origin isolates. ST2 included predominately isolates from humans and some from animal and food sources. Clusters identified using the wgMLST approach showed varying degrees of genetic relatedness after further detailed wgSNP genomic analysis. Within individual clusters, between 0 and 864 SNPs (median 48) were found between pairs of isolates. For the RT181 cluster, isolates showed a close genetic relationship (0–10 SNPs). All isolates in this cluster were from Romania, except for one isolate from the UK. The RT027 cluster, consisting of 23 isolates (21 human and 2 food isolates) exhibited a close genetic relationship but with slightly higher genetic heterogeneity compared to RT181. Within this cluster, there were 2–68 SNPs found between isolates (both food isolates differ from the closest human isolate by >10 SNPs), and sub-clustering of strains based on the country of origin was observed . Low genetic heterogeneity was confirmed for ST11 (RTs 078, 126 and 193). All 71 isolates in this cluster had no more than 30 SNPs difference. Sub-clusters of isolates from human and animal or food sources in ST11 were from different countries. The wgSNP analysis did identify strains from different sources with relatively low numbers of SNP differences. Specifically, among RT045 strains, there were strains isolated from both humans and pigs collected in Poland that had less than 10 SNPs difference. A similar pattern was seen for some isolates from pigs from both Poland and Slovakia . Similarly, among RT078 and RT126 strains, sub-clusters with less than 10 SNPs were identified. For RT078, there was a sub-cluster consisting of six strains, including two isolated from humans in France and Spain, three from pigs and one from a pigeon from Spain. Notably, two animal origin strains from this cluster were isolated in 2010 and 2017 . For RT126, there was a cluster of 6 strains (<10 SNPs), including 2 isolated from humans from Italy and Spain, 3 from pigs from France, and 1 from a potato from Romania . The 441 strains were distributed across 106 PCR ribotypes (Supplementary Table S2). Of those, 57 had a single representative strain, 23 had 2 or 3 representatives and 26 had 4 or more representatives . PCR ribotypes 078, 126, 014, 181 and 027 were the most identified in our collection and are present in several countries. The only exception was RT181, with 26 strains from only 2 countries and mainly from hospitals, potentially indicating an outbreak situation. Diversity within each source was high ( (A) and ). Strains isolated from humans in the hospital were distributed into 66 PCR ribotypes, those from humans in the community into 41, from animals in 30 and from food in 36 PCR ribotypes. Each of the 4 sources had from 7 to 34 non-shared PCR ribotypes. The most common PCR ribotypes for each source collectively are shown in (B). Of 106 PCR ribotypes, 38 were positioned in intersections of different sources ( (A)). Ten PCR ribotypes were shared between all 4 sources, 9 PCR ribotypes between 3 sources and 18 PCR ribotypes between 2 sources ( , Supplementary Table S2). Any combination of two or three sources had mostly from one to four common PCR ribotypes and the highest number of shared PCR ribotypes ( n = 26) was observed between isolates from humans in the hospital and community settings. Of these 26 ribotypes, 9 were only found in humans and shared between hospital and community patients. The only combination without any shared PCR ribotypes was Food/Animal/Community . C. difficile lineages were detected in food and exceptionally in human strains Neighbour-joining phylogenetic tree on concatenated MLST loci demonstrated that 10 isolates belonged to one of the 3 cryptic clades (C-I to C-III) (Supplementary Figure S3, ). Five isolates were highly divergent also on the wgMLST dendrogram (Supplementary Figure S2) and another five were not analysed with EPISEQ®CS as these were not recognized as C. difficile . Four of these 10 cryptic isolates, all from food, carried divergent toxin A gene (Supplementary Figure S3, ). The single human isolate did not show any variant forms of toxin A or B. None of these divergent isolates had CDT genes or variant CDT genes. This study is the first large-scale multi-country comparison of largely contemporaneous C. difficile strains from different sources including humans in hospital and community settings, animals (predominantly pigs) and food (potato). With 83 identified STs and 106 identified PCR ribotypes, the collection shows high overall and reservoir/source diversity of C. difficile across Europe. Genotypes from all main five clades and three cryptic clades were represented. The top represented genotypes were from clades 5, 2 and 1. The majority of the most common STs and PCR ribotypes were detected in several countries and across all four sources. The three exceptions are RT181 (ST1) prevailing in a single country, RT027 (ST1) mainly found in the hospital setting and rarely in food, and RT039 (ST26) found only in humans. The RT181 cluster included highly related strains and was composed mainly of strains isolated from hospitals in Romania where C. difficile is the main nosocomial pathogen . An outbreak with RT181 was also previously described in Greece . The four top PCR ribotypes (078, 126, 014, 027) found in this study, are also among the most common globally . Based on pan-European studies, subsequent ribotype waves were observed in Europe over the last two decades. RT001 and RT014 were the most common in 2005 and 2008, respectively , while RT027 largely predominated in Eastern European countries in 2012/2013 . In our strain collection from the year 2018, the RT027 was found only in 4 of 13 countries, instead RT078 (ST11) was most common in our strain collection. Of note, in contrast to three previous studies with strains of human origin only, our study also includes strains from animals, contributing significantly to high RT078 representation. Here we have confirmed the within-country and non-country-based clustering described previously for some ribotypes . Additionally, our results suggest that food can contribute to within- and between-country spread. Frequent interchanges between humans, animals and food with high numbers of sequence types and ribotypes are supported by the large number of genomic clusters seen within strains from two or more sources. The highest number of genomic clusters from mixed sources was between human isolates from hospital and community settings (26 shared PCR ribotypes). This is consistent with evidence that community infections contribute to overall disease burden, enriching the hospital reservoir with C. difficile from community sources . Our data also support the role of food and animals in transmission, as most well-represented STs/PCR ribotypes included strains from human, animal and food sources. We found a comparable number of ribotypes shared between animals and humans ( n = 9) and between food and humans ( n = 6). Some well-established RTs from the human population were also found here in food; RT027 is well well-known hypervirulent C. difficile type and RT023 is an emerging type in the human population . Three further ribotypes often found in various animal hosts, meat and meat products, were also present in humans in our collection; RT009, RT045, and RT046. RT009 was reported in cats and dogs and RT046 in pigs . RT045 seem to have the widest range and was reported from cats, horses, cattle, pigs, meat and seafood . The genomic relatedness observed on allele-based wgMLST was further analysed with cgSNP analysis. Here only a handful of strains were shown to be closely genomically related. Similar to previous studies, ST11 (with RT078, RT126 and relatives) showed the lowest genomic diversity among sources (human, different animal species, food) . In addition, related strains, isolated eight years apart, and strain pairs from diverse geographical locations, were detected in ST11. Almost two-thirds of all detected sequence types/ribotypes were associated with a single source; i.e. humans, animals or food only. Most of those ( n = 34) were specific to the hospital setting, which is probably due to the many strains from this source (197/441; 44.7%). Among those RTs found only in the community setting the most described is RT017/ST37. This TcdA-negative, TcdB-positive ribotype is predominant in Asia , but its prevalence in European countries has markedly decreased since 2005 . RT017 is more likely associated with younger individuals and this fits well with community association in this study, where 80% of community cases were <65 years old . Among ribotypes detected in our study from animals only, RT033/ST11 is the best documented within the literature. It displays a truncated version of toxin coding PaLoc and is hence TcdA and TcdB negative, but CDT positive. It is typically associated with a bovine reservoir and rarely reported in human CDI . Most ribotypes detected only from food are representatives of cryptic C. difficile clades. Divergent PCR ribotypes in our collection belonged to C-I to C-III clades and to known or new STs. Some were even not recognized by the EPISEQ®CS application (developed for clinical samples) as being C. difficile , which is in line with the genomic delineation of cryptic clades outside of C. difficile species . Such divergent lineages are commonly found in rural soils . This is of importance as representatives from divergent PCR ribotypes can occasionally produce toxins and are emerging in human infections . When divergent strains produce toxins, these are mono-toxigenic with either tcdA or tcdB gene only ; the toxin genes are also divergent and not recognized by diagnostic molecular tests . From 10 divergent strains in our study, only 4 harboured tcd A gene and as expected the similarity with ordinary tcd A gene was low. We identified one non-toxigenic strain from a human source in Spain from divergent PCR ribotypes. Indeed, human infections with divergent strains in Spain and the Netherlands have been previously described . One of the limitations of this study is that it was not possible to collect samples from animal and food sources from all countries. The sampling strategy for the animal and food sources was determined by the availability of such samples in each country for this study. Similarly, isolates derived from human samples were those identified as part of the COMBACTE-CDI point-prevalence study in 2018, therefore this study did not aim to compare a similar number of C. difficile isolates per country, but rather to use the existing isolates from the COMBACTE-CDI project to provide further genomic information pertinent to CDI researchers. However, the unequal distribution of isolates from different sources from different countries is an additional limitation of the study and might contribute to bias. Further potential limitation is the inclusion of only non-animal-based foods. The addition of animal-based foods would further enhance the overall picture, especially in the intersection of Animal/Food/Community. In summary, we have shown a substantial C. difficile diversity across human, animal and potato-associated strains, with considerable overlap of STs and ribotypes found in almost every combination of these four sources. Some globally common STs and PCR ribotypes are found in all four sources, while others were found to overlap between humans and animals or between food and humans. Some were preferentially found only from a single source. This study is registered under the ClinicalTrials.gov Identifier NCT03503474. Ethical approval was received from participating countries and from the University of Leeds for the overarching study (IRAS244784). The planning conduct and reporting of this study was in line with the 2013 Declaration of Helsinki. Informed consent was not required for the use of anonymised residual diagnostic material and data. Suppl Fig S1 to S3 EMI.pdf Supplementary Table S2 EMI_19 10 2024.xlsx Supplementary Table S1 EMI.xlsx Supplementary Doc S1.docx Supplementary Table S3 EMI.docx |
REFEEDING SYNDROME IN A PATIENT WITH AN OBSTRUCTIVE PANCREATIC CANCER: AN EMERGING COMPLICATION OF ARTIFICIAL NUTRITION IN THE GASTROENTEROLOGY WARD | 4e274ee1-da51-49ce-926f-f1cfeecb0073 | 8735258 | Internal Medicine[mh] | Refeeding syndrome (RS) is a life-threatening condition first described in severe malnourished prisoners of the Second World War . This syndrome is defined as electrolyte and fluid shifts associated with metabolic abnormalities developed during nutritional support. RS hallmark is hypophosphatemia, but also includes hypomagnesemia, hypokalemia, vitamin deficiencies, abnormal glucose metabolism and fluid retention. Prolonged fasting is the most important risk factor and RS may be precipitated by oral, enteral or parenteral nutrition , . The authors describe a case of RS in the gastroenterology ward exemplifying the importance of recognizing this underreported condition in patients with digestive pathology under nutritional therapy.
An 82 year-old female was admitted due to recurrent vomiting during 10 days. Her past medical history included cerebrovascular disease, diabetes and hypertension. On hospital admission she was febrile, dehydrated and presented low body mass index (20.8 kg/m 2 ). Initial evaluation revealed acute kidney injury (creatinine 4.1 mg/dl), hypokalaemia (K + 3.2 mg/dl), hyperphosphatemia (Pi 5.4 mg/dl) and normal serum sodium and magnesium. Fluid and electrolyte replacement were immediately started but vomiting persisted. Nasogastric intubation revealed stasis (1800 cc/24h). Upper gastrointestinal endoscopy detected lumen narrowing in second/third duodenum parts. CT scan identified a heterogeneous mass in the pancreatic head causing Wirsung duct dilation and duodenal compression ( ). Surgical resection was ruled out considering the advanced age and poor performance status. Palliative care with gastroduodenal self-expandable metallic stent (SEMS) placement was scheduled. Since the patient displayed protein-energy malnutrition and the stent could not be placed immediately to resume oral feeding, total parenteral nutrition (TPN) was instituted through a central line, after correction of hypokalemia and preventive supplementation with intravenous phosphorus. Our gastroenterology department has a protocol for TPN gradual onset, starting with 25% of energy needs with progressive increase during the first week. However, TPN was started during the weekend, and the protocol was not applied due to an institutional mistake and reduced monitoring. Thus, 25 kcal/kg were administered during the first 24 h, which corresponded to 100% of energy needs. After the first day of TPN the patient developed depressed level of consciousness, myoclonus, sinus tachycardia, polypnea, pleural effusion and peripheral edema. Hypophosphatemia (Pi 0.6 mg/dl), worsening of hypokalemia (K + 2.7 mg/dl), hypomagnesemia (Mg 2+ 0.9mg/dl) and hypernatremia (Na + 162 mg/dl) were detected in blood tests. RS diagnosis was assumed and the TPN prescription mistake was promptly identified. TPN was immediately stopped, thiamine supplementation and intensive intravenous hydration with electrolyte replacement and monitoring were implemented. Three days later, hydroelectrolytic balance was achieved and neurologic status recovered with complete resolution of clinical manifestations. Gastroduodenal SEMS was placed and oral feeding resumed. Refeeding was performed using an oral diet to supply initially 10% of the energy needs, reaching 100% at day 7, with no additional complications.
Several disorders in the gastroenterology ward may induce significant weight-loss and fluid/electrolyte imbalance, including obstructive tumors causing dysphagia and vomiting. Nevertheless, RS remains poorly recognized with unknown incidence and no well-established diagnostic criteria , . Metabolic changes, in particular hypophosphatemia and hypokalaemia, may be life-threatening. Low serum phosphorus and potassium can induce severe cardiorespiratory events such as heart failure, arrhythmias and respiratory muscle weakness, and neurological abnormalities like paresthesia, myoclonus and seizures . To manage RS, most authors use the evidence described by National Institute for Health and Clinical Excellence (NICE) guidelines . Actually, our patient presented high risk for RS given the low intake for at least 10 days and baseline low serum potassium 2, . Nutritional support should have been started, according with the protocol, with a maximum of 10 kcal/kg/day and increased slowly to achieve the total needs in 4-7 days. Close monitoring of electrolytes prior to begin TPN refeeding and during the first 10 days is of utmost importance. Although the impact of RS in patient outcome, hospitalization length and mortality are not established, some studies in critically ill showed increased mortality and longer admission when it develops and our experience with percutaneous endoscopic gastrostomy PEG-fed patients established increased mortality of hypophosphatemic patients , , . This highlights the importance of RS awareness and the need of training for physicians who prescribe nutritional support and pharmacists who play an active role in selecting and preparing TPN bags . The rules of avoiding initiating TPN by untrained staff and not to start during the weekend, when patients are not so closely monitored are highly advisable. The development of institutional protocols and multidisciplinary teams dedicated to nutritional support should be mandatory.
|
Wall shear stress modulates metabolic pathways in endothelial cells | 11175c64-92c0-498f-92da-169e6e875297 | 11753319 | Biochemistry[mh] | Endothelial cells (ECs) constitute the inner lining of all blood vessels, orchestrating the transport of nutrients to the underlying tissues and coordinating the formation of new blood vessels, a process termed angiogenesis (Baeyens et al., ). Displaying remarkable adaptability, ECs can transition from a highly proliferative and migratory state during angiogenesis into a quiescent, resting state in normal conduit blood vessels (Ballermann et al., ). Ongoing research on ECs predominantly centres around static in vitro cultures (Barry et al., ). Nevertheless, blood vessels, and consequently ECs, are constantly exposed to various hemodynamic forces, including hydrostatic pressure, cyclic stretch, and fluid shear stress induced by pulsatile blood pressure and flow (Binek et al., ). Experimental studies indicate shear stress values ranging from 1 to 6 dyn/cm 2 in the venous system and from 10 to 70 dyn/cm 2 in arteries (Binek et al., ). Laminar physiological shear stress is considered crucial in maintaining the quiescent state of endothelial cells (Chistiakov et al., ). For example, the transcription factor Krüppel-like factor 2 (KLF2), induced by shear stress, coordinates a network of genes promoting the quiescent phenotype (Chistiakov et al., ; Chiu & Chien, ). A uniform, high flow promotes vascular health by preserving ECs in a quiescent state, characterized by low cell turnover, low permeability to macromolecules, and minimal inflammatory response (Ballermann et al., ; Choi & Helmke, ). Conversely, disturbed flow, marked by low velocity and/or non-uniform direction, may foster atherosclerosis by exacerbating EC injury, pro-inflammatory activation, and increased permeability to lipoproteins and other molecules. (Ballermann et al., ; Choi & Helmke, ) The molecular mechanisms underlying these responses are only partially understood but recent studies indicate that flow alters the expression of hundreds of coding and non-coding RNAs. (Choi & Park, ; Choi et al., ) Metabolism emerged as a pivotal mechanism enabling cells to rapidly respond to dynamic changes, crucial for controlling endothelial cell phenotype (Cruys et al., ). Angiogenic ECs heavily rely on glycolysis for migration and proliferation (Davies et al., ). The enzyme PFKFB3, a key glycolysis regulator in ECs, promotes angiogenic sprouting (Craemer et al., ). Biomechanical signals from shear stress, mediated by the upregulation of KLF2 and KLF4, curtail EC glycolysis activity and mitochondrial content by repressing PFKFB3 expression (Chiu & Chien, ). These insights highlight the significance of biomechanical stimuli in maintaining the resting quiescent state and metabolism of ECs. Considering the pivotal role of physical parameters in EC physiological functioning and its critical hypothesized importance of EC metabolism, we hypothesized that shear stress triggers diverse metabolic responses beyond glycolysis downregulation. Our multi-omics study delves into how wall shear stress (WSS) regulates EC metabolism in vitro. In conclusion, our inhibitor results reveal that the repression of the enzyme GDH under shear stress induces profound local cytoskeletal remodelling, disrupting cell–cell contacts and impairing the overall integrity of endothelial tissues in vitro. Meanwhile, blocking GLS activity exhibits a milder impact on EC fitness.
Cell culture Human umbilical cord endothelial cells (HUVECs) from pooled donors were obtained from PromoCell (C-12208) and cultured in 0,2% gelatin-coated dishes in EGM2 medium supplemented with growth factors (PromoCell, C-22111), 100 units/ml penicillin and 100 μg/ml streptomycin (Gibco). For all experiments, HUVECs were used between passages 4 to 6, the medium was refreshed every 48 h or 72 h, and regularly tested for mycoplasma. Shear Stress experiments HUVECs were seeded at a density of 500.000 cells per slide in growth medium on gelatin-coated SuperFrost Excell object slides (Thermo Scientific Menzel) and were grown to confluency for 48 h. Slides were placed in a custom-designed parallel plate flow chamber within a tissue culture incubator. Wall shear stress was applied (15 dynes/cm2) for 6 to 24 h using a peristaltic pump (Masterflex LS 7550–30). For static controls, HUVECs were seeded on slides, cultured in the same incubator in the absence of shear stress, and collected at the same time. RNAseq Each slide was sheared independently in an individual chamber for 24 h and lysate from one slide was used per sample. RNA was extracted using RNEasy Kits (Qiagen) according to the manufacturer’s instructions. RNA sequencing was performed by the VIB Nucleomics Core (KU Leuven, Belgium). Briefly, RNA concentration and purity were determined spectrophotometrically using the Nanodrop ND-1000 (Nanodrop Technologies) and RNA integrity was assessed using a Bioanalyzer 2100 (Agilent). Per sample, an amount of 250 ng of total RNA was used as input. Using the Illumina TruSeq® Stranded mRNA Sample Prep Kit (protocol version: # 1,000,000,040,498 v00 October 2017) poly-A containing mRNA molecules were purified from the total RNA input using poly-T oligo-attached magnetic beads. In a reverse transcription reaction using random primers, RNA was converted into first strand cDNA and subsequently converted into double-stranded cDNA in a second strand cDNA synthesis reaction using DNA PolymeraseI and RNAse H. The cDNA fragments were extended with a single 'A' base to the 3' ends of the blunt-ended cDNA fragments after which multiple indexing adapters were ligated introducing different barcodes for each sample. Finally, enrichment PCR was carried out to enrich those DNA fragments that have adapter molecules on both ends and to amplify the amount of DNA in the library. Sequence libraries of each sample were equimolarly pooled and sequenced on Illumina NovaSeq 6000 (S1 100, 100 bp single- end reads (100–8-8–0)) at the VIB Nucleomics Core. Raw sequencing data are available in Gene Expression Omnibus under accession number GSE202220. Analysis of the bulk RNA sequencing data was performed in BIOMEX (Bock et al., ). Gene set enrichment analysis (GSEA) was performed using the MSigDB KEGG gene list. GSEA scores were calculated for sets with a minimum of 10 detected genes and only the significantly up-and downregulated genes (adjusted p < 0.05) were used. Individual genes shown in heatmaps were selected out of relevant KEGG gene lists where after they were filtered based on significant adjusted p-value and endothelial function. (Choi et al., ) RNA extraction, cDNA synthesis, and RT-qPCR Cultured cells were collected in RLT buffer and RNA was isolated using RNeasy Mini Kit (Qiagen), according to the manufacturer's protocol. Reverse transcription to cDNA was performed using SuperScriptTM III First-Strand Synthesis System (Invitrogen, 18,080,051) according to the manufacturer’s instructions. RT-qPCR amplification was performed using in‐house‐designed primers (IDT) and probes in the QuantStudio 12 K Flex Real-Time PCR system (Applied Biosystems). HPRT was used as housekeeping gene. Flow cytometry (FACS) analysis For detection of proliferation and cell cycle, HUVECs were cultured in medium containing 10 μM of EdU under WSS or in static cultures for 5 h before collection with trypsin. Cells were fixed in 4% PFA for 15 min at 4 oC and EdU + cells were detected using Click-iTTM Plus EdU Alexa FluorTM 647 Flow Cytometry Assay Kit (Invitrogen, C10635) according to the manufacturer’s instructions. Cells were subsequently treated with RNAse A (Qiagen, 19,101, 1:1000) for 15 min and stained with propidium iodide (Sigma-Aldrich, P4864, 0,5 μg/ml) for another 15 min and immediately analyzed with a Fortessa LSR-II (BD Bioscience). 13C Tracer Experiments and Metabolite Levels HUVECs were cultured in M199 (11,150,059, Gibco) supplemented with 10% dialyzed FBS and 5 mM [U-13C]-D-glucose (CLM-1396, Cambridge Isotope Laboratories) or 0.68 mM [U-13C]-D- glutamine (CLM-1822, Cambridge Isotope Laboratories) under WSS or in static cultures for 24 h. Metabolites were extracted with 600 μl of extraction buffer (80% Methanol, 2uM d27myristic acid). For analysis on condition medium, 990 μl of extraction buffer (80% Methanol, 2uM d27myristic acid) was added to 10 μl of sample. After overnight storage at − 80 °C, samples were centrifuged at 13,000 g for 20 min at 4 °C and the supernatant was transferred to a new vial for MS analysis. 10 μl of each sample was loaded into a Dionex UltiMate 3000 LC System (Thermo Scientific Bremen, Germany) equipped with a C-18 column (Acquity UPLC -HSS T3 1.8 μm; 2.1 × 150 mm, Waters) coupled to a Q Exactive Orbitrap mass spectrometer (Thermo Scientific) operating in negative ion mode. A step gradient was carried out using solvent A (10 mM TBA and 15 mM acetic acid) and solvent B (100% methanol). The gradient started with 5% of solvent B and 95% solvent A and remained at 5% B until 2 min post-injection. A linear gradient to 37%B was carried out until 7 min and increased to 41% until 14 min. Between 14 and 26 min, the gradient increased to 95% of B and remained at 95% B for 4 min. At 30 min the gradient returned to 5%B. The chromatography was stopped at 40 min. The flow was kept constant at 0.25 mL/min at the column and was placed at 40 °C throughout the analysis. The MS operated in full scan mode (m/z range: [70,0000–1050,0000]) using a spray voltage of 4.80 kV, the capillary temperature of 300 °C, sheath gas at 40.0, and auxiliary gas at 10.0. The AGC target was set at 3.0E + 006 using a resolution of 140,000, with a maximum IT fill time of 512 ms. Data collection was performed using the Xcalibur software (Thermo Fisher Scientific) while data analysis was performed with the software Xcalibur software (Thermo Fisher Scientific) and El Maven-Polly (Elucidata). Raw abundance values were normalized by protein content. The fractional contribution (FC), which is the fractional percentage of each of the labeled isotopologue out of the total amount of metabolite, was obtained by using the software El Maven-Polly (Elucidata), which is mostly based on the formula FC = (Σni = 0 i × mi)/(n × Σni = 0 mi), where n is the number of C atoms in the metabolite, i denotes the isotopologues, and m indicates the abundance of an isotopologue. Metabolomic inhibition assay HUVECs were cultured under WSS or in static cultures in medium M199 (11,150,059, Gibco) supplemented with 10% FBS and 1 mM 2-deoxyglucose (D3179, Sigma) or 10 μM Epigallocatechin gallate (E4143, Sigma) or 1 μM CB-839 (HY-12248, MedChemExpress) for 8 h. Slides were washed with PBS and used for protein extraction or immunocytochemistry. Protein extraction and western blot Whole cell protein extraction from HUVECs was performed using RIPA buffer (Thermo Scientific, 89,900) supplemented with Complete Mini protease inhibitor (Roche, 11,836,145,001) and PhosphoSTOP Phosphatase Inhibitor (Roche, 04906837001). Protein concentration was measured by the bicinchoninic acid (BCA) protein assay (Thermo Scientific, 23,225) according to the manufacturer’s instructions. Protein samples (50 μg) were denatured by adding a reducing loading buffer and incubating at 95 °C for 5 min. Proteins were resolved by SDS-PAGE and transferred to nitrocellulose membrane. Membranes were incubated overnight at 4 oC with the following primary antibodies diluted in TBST containing 5% bovine serum albumin: Glutamate Dehydrogenase (Abcam, 89,967, 1:500), Glutaminase (Abcam, 93,434, 1:500), beta Actin (Abcam, 8227, 1:2000) and Tubulin-HRP (Abcam, ab21058, 1:2000). The day after, membranes were incubated for 1 h at RT with appropriate HRP-conjugated secondary antibodies (Cell Signaling, 7074, 1:2000). Signal was visualized by Novex ECL Chemiluminescent Substrate (Invitrogen, WP20005) or SuperSignalTM West Femto Maximum Sensitivity Substrate (Thermo Scientific, 34,096) and acquired by a LAS 4000 CCD camera with ImageQuant software (GE Healthcare). Densitometric quantifications of bands were done with ImageJ software. Immunocytochemistry All methods for histology and immunostainings have been previously described (Dekker et al., ; Doddaballapur et al., ). Briefly, cells were fixed with 4% PFA for 15 min at room temperature and processed for immunocytochemistry. In general, PBS-Triton 0.5%-2%BSA was used to permeabilize the cells. Primary antibodies were incubated with the cells overnight, and secondary antibodies were incubated for 2 h. On tissues, immunostainings were performed using the following primary antibodies: anti-VE-cadherin (AF1002, R&D Systems). Sections were then incubated with the appropriate fluorescently conjugated secondary antibody Alexa 568 (Thermo Fisher Scientific). Actin was counterstained 11 with Alexa-488 Fluor conjugated phalloidin (Thermo Fisher Scientific). Nuclei were counterstained with DAPI (Thermo Fisher Scientific). Statistical analysis Data entry and all analyses were performed in a blinded fashion. All statistical analyses were performed using GraphPad Prism 9. Statistical significance was calculated by a two-tailed unpaired t-test on two experimental conditions or two-way ANOVA when more than two experimental groups were compared. Detection of mathematical outliers was performed using the Grubbs’ test in GraphPad. All data are represented as mean ± standard error of the mean (SEM). P-values are indicated as: p < 0.05 , p < 0.01 , p < 0.001 and p < 0.0001 . n values represent the number of independent experiments performed (in vitro data) and the number of animals per condition (in vivo).
Human umbilical cord endothelial cells (HUVECs) from pooled donors were obtained from PromoCell (C-12208) and cultured in 0,2% gelatin-coated dishes in EGM2 medium supplemented with growth factors (PromoCell, C-22111), 100 units/ml penicillin and 100 μg/ml streptomycin (Gibco). For all experiments, HUVECs were used between passages 4 to 6, the medium was refreshed every 48 h or 72 h, and regularly tested for mycoplasma.
HUVECs were seeded at a density of 500.000 cells per slide in growth medium on gelatin-coated SuperFrost Excell object slides (Thermo Scientific Menzel) and were grown to confluency for 48 h. Slides were placed in a custom-designed parallel plate flow chamber within a tissue culture incubator. Wall shear stress was applied (15 dynes/cm2) for 6 to 24 h using a peristaltic pump (Masterflex LS 7550–30). For static controls, HUVECs were seeded on slides, cultured in the same incubator in the absence of shear stress, and collected at the same time.
Each slide was sheared independently in an individual chamber for 24 h and lysate from one slide was used per sample. RNA was extracted using RNEasy Kits (Qiagen) according to the manufacturer’s instructions. RNA sequencing was performed by the VIB Nucleomics Core (KU Leuven, Belgium). Briefly, RNA concentration and purity were determined spectrophotometrically using the Nanodrop ND-1000 (Nanodrop Technologies) and RNA integrity was assessed using a Bioanalyzer 2100 (Agilent). Per sample, an amount of 250 ng of total RNA was used as input. Using the Illumina TruSeq® Stranded mRNA Sample Prep Kit (protocol version: # 1,000,000,040,498 v00 October 2017) poly-A containing mRNA molecules were purified from the total RNA input using poly-T oligo-attached magnetic beads. In a reverse transcription reaction using random primers, RNA was converted into first strand cDNA and subsequently converted into double-stranded cDNA in a second strand cDNA synthesis reaction using DNA PolymeraseI and RNAse H. The cDNA fragments were extended with a single 'A' base to the 3' ends of the blunt-ended cDNA fragments after which multiple indexing adapters were ligated introducing different barcodes for each sample. Finally, enrichment PCR was carried out to enrich those DNA fragments that have adapter molecules on both ends and to amplify the amount of DNA in the library. Sequence libraries of each sample were equimolarly pooled and sequenced on Illumina NovaSeq 6000 (S1 100, 100 bp single- end reads (100–8-8–0)) at the VIB Nucleomics Core. Raw sequencing data are available in Gene Expression Omnibus under accession number GSE202220. Analysis of the bulk RNA sequencing data was performed in BIOMEX (Bock et al., ). Gene set enrichment analysis (GSEA) was performed using the MSigDB KEGG gene list. GSEA scores were calculated for sets with a minimum of 10 detected genes and only the significantly up-and downregulated genes (adjusted p < 0.05) were used. Individual genes shown in heatmaps were selected out of relevant KEGG gene lists where after they were filtered based on significant adjusted p-value and endothelial function. (Choi et al., )
Cultured cells were collected in RLT buffer and RNA was isolated using RNeasy Mini Kit (Qiagen), according to the manufacturer's protocol. Reverse transcription to cDNA was performed using SuperScriptTM III First-Strand Synthesis System (Invitrogen, 18,080,051) according to the manufacturer’s instructions. RT-qPCR amplification was performed using in‐house‐designed primers (IDT) and probes in the QuantStudio 12 K Flex Real-Time PCR system (Applied Biosystems). HPRT was used as housekeeping gene.
For detection of proliferation and cell cycle, HUVECs were cultured in medium containing 10 μM of EdU under WSS or in static cultures for 5 h before collection with trypsin. Cells were fixed in 4% PFA for 15 min at 4 oC and EdU + cells were detected using Click-iTTM Plus EdU Alexa FluorTM 647 Flow Cytometry Assay Kit (Invitrogen, C10635) according to the manufacturer’s instructions. Cells were subsequently treated with RNAse A (Qiagen, 19,101, 1:1000) for 15 min and stained with propidium iodide (Sigma-Aldrich, P4864, 0,5 μg/ml) for another 15 min and immediately analyzed with a Fortessa LSR-II (BD Bioscience).
HUVECs were cultured in M199 (11,150,059, Gibco) supplemented with 10% dialyzed FBS and 5 mM [U-13C]-D-glucose (CLM-1396, Cambridge Isotope Laboratories) or 0.68 mM [U-13C]-D- glutamine (CLM-1822, Cambridge Isotope Laboratories) under WSS or in static cultures for 24 h. Metabolites were extracted with 600 μl of extraction buffer (80% Methanol, 2uM d27myristic acid). For analysis on condition medium, 990 μl of extraction buffer (80% Methanol, 2uM d27myristic acid) was added to 10 μl of sample. After overnight storage at − 80 °C, samples were centrifuged at 13,000 g for 20 min at 4 °C and the supernatant was transferred to a new vial for MS analysis. 10 μl of each sample was loaded into a Dionex UltiMate 3000 LC System (Thermo Scientific Bremen, Germany) equipped with a C-18 column (Acquity UPLC -HSS T3 1.8 μm; 2.1 × 150 mm, Waters) coupled to a Q Exactive Orbitrap mass spectrometer (Thermo Scientific) operating in negative ion mode. A step gradient was carried out using solvent A (10 mM TBA and 15 mM acetic acid) and solvent B (100% methanol). The gradient started with 5% of solvent B and 95% solvent A and remained at 5% B until 2 min post-injection. A linear gradient to 37%B was carried out until 7 min and increased to 41% until 14 min. Between 14 and 26 min, the gradient increased to 95% of B and remained at 95% B for 4 min. At 30 min the gradient returned to 5%B. The chromatography was stopped at 40 min. The flow was kept constant at 0.25 mL/min at the column and was placed at 40 °C throughout the analysis. The MS operated in full scan mode (m/z range: [70,0000–1050,0000]) using a spray voltage of 4.80 kV, the capillary temperature of 300 °C, sheath gas at 40.0, and auxiliary gas at 10.0. The AGC target was set at 3.0E + 006 using a resolution of 140,000, with a maximum IT fill time of 512 ms. Data collection was performed using the Xcalibur software (Thermo Fisher Scientific) while data analysis was performed with the software Xcalibur software (Thermo Fisher Scientific) and El Maven-Polly (Elucidata). Raw abundance values were normalized by protein content. The fractional contribution (FC), which is the fractional percentage of each of the labeled isotopologue out of the total amount of metabolite, was obtained by using the software El Maven-Polly (Elucidata), which is mostly based on the formula FC = (Σni = 0 i × mi)/(n × Σni = 0 mi), where n is the number of C atoms in the metabolite, i denotes the isotopologues, and m indicates the abundance of an isotopologue.
HUVECs were cultured under WSS or in static cultures in medium M199 (11,150,059, Gibco) supplemented with 10% FBS and 1 mM 2-deoxyglucose (D3179, Sigma) or 10 μM Epigallocatechin gallate (E4143, Sigma) or 1 μM CB-839 (HY-12248, MedChemExpress) for 8 h. Slides were washed with PBS and used for protein extraction or immunocytochemistry.
Whole cell protein extraction from HUVECs was performed using RIPA buffer (Thermo Scientific, 89,900) supplemented with Complete Mini protease inhibitor (Roche, 11,836,145,001) and PhosphoSTOP Phosphatase Inhibitor (Roche, 04906837001). Protein concentration was measured by the bicinchoninic acid (BCA) protein assay (Thermo Scientific, 23,225) according to the manufacturer’s instructions. Protein samples (50 μg) were denatured by adding a reducing loading buffer and incubating at 95 °C for 5 min. Proteins were resolved by SDS-PAGE and transferred to nitrocellulose membrane. Membranes were incubated overnight at 4 oC with the following primary antibodies diluted in TBST containing 5% bovine serum albumin: Glutamate Dehydrogenase (Abcam, 89,967, 1:500), Glutaminase (Abcam, 93,434, 1:500), beta Actin (Abcam, 8227, 1:2000) and Tubulin-HRP (Abcam, ab21058, 1:2000). The day after, membranes were incubated for 1 h at RT with appropriate HRP-conjugated secondary antibodies (Cell Signaling, 7074, 1:2000). Signal was visualized by Novex ECL Chemiluminescent Substrate (Invitrogen, WP20005) or SuperSignalTM West Femto Maximum Sensitivity Substrate (Thermo Scientific, 34,096) and acquired by a LAS 4000 CCD camera with ImageQuant software (GE Healthcare). Densitometric quantifications of bands were done with ImageJ software.
All methods for histology and immunostainings have been previously described (Dekker et al., ; Doddaballapur et al., ). Briefly, cells were fixed with 4% PFA for 15 min at room temperature and processed for immunocytochemistry. In general, PBS-Triton 0.5%-2%BSA was used to permeabilize the cells. Primary antibodies were incubated with the cells overnight, and secondary antibodies were incubated for 2 h. On tissues, immunostainings were performed using the following primary antibodies: anti-VE-cadherin (AF1002, R&D Systems). Sections were then incubated with the appropriate fluorescently conjugated secondary antibody Alexa 568 (Thermo Fisher Scientific). Actin was counterstained 11 with Alexa-488 Fluor conjugated phalloidin (Thermo Fisher Scientific). Nuclei were counterstained with DAPI (Thermo Fisher Scientific).
Data entry and all analyses were performed in a blinded fashion. All statistical analyses were performed using GraphPad Prism 9. Statistical significance was calculated by a two-tailed unpaired t-test on two experimental conditions or two-way ANOVA when more than two experimental groups were compared. Detection of mathematical outliers was performed using the Grubbs’ test in GraphPad. All data are represented as mean ± standard error of the mean (SEM). P-values are indicated as: p < 0.05 , p < 0.01 , p < 0.001 and p < 0.0001 . n values represent the number of independent experiments performed (in vitro data) and the number of animals per condition (in vivo).
Wall shear stress reveals broad transcriptomic alterations in EC metabolism in vitro To investigate the effect of WSS on EC metabolism, we carried out an unbiased transcriptomic analysis to assess global changes in HUVEC subjected to static and WSS cultures for 24 h under normal growth conditions. Principal component analysis (PCA) of all genes and hierarchical clustering analysis of the highly variable genes revealed that static control and WSS-treated HUVEC grouped into distinct clusters, indeed suggesting broad transcriptomic changes (Fig. A). Differential gene expression analysis (DGEA) of our transcriptomic data showed that, in line with the literature [(Binek et al., )], the shear stress marker KLF2 was elevated upon WSS (Fig. a), which we confirmed by real-time qPCR analysis (Fig. b), suggesting that our in vitro system can accurately mimic the in vivo response to WSS. Next, we investigated whether we could identify changes induced by WSS of transcripts involved in metabolic pathways. Gene set enrichment analysis (GSEA) (Subramanian et al., 2005) on metabolic KEGG gene sets was performed comparing static versus WSS-treated HUVEC. Our analysis showed an increased expression of genes involved in folate biosynthesis, alanine, aspartate, glutamate, glycerophospholipid, arachidonic acid, and amino/nucleotide sugar metabolism, while gene sets involved in pyrimidine metabolism, O-glycan biosynthesis, purine and drug metabolism, valine, leucine, isoleucine degradation, and glycolysis/gluconeogenesis were downregulated (Fig. B). A comparison of specific markers for glycolysis, alanine, aspartate, and glutamate metabolism revealed that the main genes involved in the glycolysis and pyruvate/lactate production (PGM1, PGM2, GPI, PFKL, PFKFB3, PKM, PFKM, TPI1, ENO1, PDHA1, LDHA, LDHB, and ACSS2) were downregulated (Fig. A), while genes involved in glutamine, asparagine, alanine, aspartate, and glutamate metabolism (GSR, G6PD, GCLC, IDH1, GOT1, GLUD1, ASS1, GFPT2, GFPT1, and ASNS) were upregulated (Fig. A). Real-time qPCR analysis further confirmed our bulk RNA sequencing data as the GLUD1, GLUL, PKM1, and PKM2 genes were upregulated (Fig. b) while HK2 was downregulated (Fig. b) upon WSS. GSEA on all KEGG gene sets was performed to associate groups of significant differentially expressed genes to biological processes [(Subramanian et al., )] not strongly related to metabolism (Fig. c). With DGEA, we then set out to obtain specific WSS-markers conserved across different biological donors. Among the top downregulated genes were GALNT15, SULT1B1, MAN1A1, RRM2, AQP1, DHRS3, and SLC46A1 (Fig. d) which have been associated with cell adhesion, DNA synthesis, replication, and repair, nucleotide metabolism and binding, glycerophospholipid and hydrolase activity and water transport. Likewise, among the top upregulated ranking genes were PTGS2, ALPL, CYP1B1, SLC7A11, ABCC3, AKR1C1, HMOX1, SLC6A4, and TXNRD1 (Fig. d) which have been linked to prostaglandin activity, amino acid and sodium/calcium transport, drug metabolism, ATPase and oxidoreductase activity, NAD cofactors biosynthesis and actin/tubulin polymerization. The meta-analysis revealed that the gene sets involved in glycolysis/gluconeogenesis were more upregulated in the static cultured ECs data set. Gene sets involved in alanine, aspartate, and glutamate metabolism were upregulated only on WSS cultured ECs data set (Fig. A). Proliferation can change metabolism and shear stress is reported to reduce proliferation [(Levesque, )]. Therefore, we set out to separate the role of WSS in promoting a pro-resting quiescent phenotype, from the effects WSS has on metabolism specifically. For this, HUVECs were cultured in the presence of EdU to confirm lower EdU incorporation upon exposure to shear stress in comparison to the static cultured HUVECs (Fig. e). Then, to exclude that proliferation/quiescent signaling was promoting the metabolic alterations seen in our transcriptomic analysis, we performed a meta- analysis of contact-inhibited versus proliferating ECs (GSE89174) [(Kalucka et al., )] and static versus WSS cultured ECs to rank differentially regulated gene sets (Fig. B). No genes were excluded or overlapped. While our analysis highlights significant metabolic changes in ECs under wall WSS, it is important to note that the quiescence states induced by WSS and contact inhibition may represent distinct biological conditions, which could influence the observed metabolic and transcriptomic profiles. In summary, we found that HUVECs subjected to WSS displayed large transcriptional alterations and upregulated genes associated not only with glutamine metabolism while downregulating genes involved in glycolysis but also in other important metabolic pathways responsible for key cellular events. Wall Shear Stress promotes glutamine anaplerosis in vitro To investigate whether WSS affects EC metabolism, we first investigated its impact on nutritional behaviour. For this, medium samples from static and WSS cultured ECs for 24 h were analysed by LC–MS and these indeed showed distinct differences in the nutritional needs upon WSS (Fig. A). Figure A shows the relative differences in metabolites abundance between the medium from static and WSS cultured ECs. Compared to the static medium, the medium from WSS showed lower pyruvate, lactate, glutamine, and asparagine levels (Fig. A). This is in contrast with glucose, which showed to be more abundant in the WSS medium (Fig. A), pointing towards a lesser reliance on glucose during WSS. Comparison of the static and WSS medium towards the basal medium allowed us to understand the nutritional preferences of the different conditions (Fig. B). Figure B reveals that upon WSS, the nutritional preference of ECs towards glucose diminishes while the preference for glutamine as a nutrient becomes apparent. The combination of higher glucose and lower lactic acid levels also indicates that the cells are less glycolytic upon exposure to WSS. For this, we plotted the correlation of the absolute concentration of glucose and lactate in each condition (Fig. C), this allowed us to roughly estimate the glycolytic rate (in the hypothetical scenario where 100% of the glucose molecules are converted into lactic acid, 1 molecule of glucose will generate 2 molecules of lactic acid). In the static setups, a clear correlation between glucose and lactate levels was found (Fig. C) indicating that about 75% of the glucose was converted into lactic acid. This finding was confirmed by 73% of fractional contribution (FC) of 13 C glucose to lactic acid from the medium (Fig. S2a). Surprisingly, no correlation was found between glucose and lactic acid in the WSS setups being in line with the decreased lactic acid levels and higher glucose levels in the medium (Fig. A–C). This corroborated the finding that glucose metabolism is less preferred by ECs undergoing WSS. Here as well, the FC of glucose to lactic acid was significantly lower in the WSS setups (Fig. S2a). In conclusion, from our nutritional medium study, we considered that the fate of glutamine and glucose were of interest for further investigation by tracer metabolomics to determine the biochemical paths these key nutrients undergo. To understand how ECs utilize glutamine and glucose under WSS in vitro versus the static EC cultures, we supplemented the medium in parallel setups with respectively [U-13C5] glutamine and [U-13C6] glucose to determine both the abundances, amount of 13 C-label incorporation (FC) and isotopologues of the key metabolites involved in the main central carbon pathways, including glycolysis, TCA-cycle, amino acids and glutamine related intermediates (Fig. A, Fig. S2a-o). Data from the tracer metabolomics setups (Fig. A) is represented as pie charts (Craemer et al., ). The size of the pie chart represents the relative change in metabolite abundances while the distinct color reveals the FC of the respective tracers. Aligning with the previous findings, Fig. D shows that, during static conditions, ECs mainly rely on glucose and a combination of glucose and glutamine for the Krebs cycle. Upon WSS, the amount of ( 13 C) labeled glucose used in glycolysis is reduced significantly while the contribution and dependency on glutamine-derived carbons for the Krebs cycle increased (Fig. A). The abundance of glycolysis intermediates (glucose 6-phosphate (G6P) and pyruvate) were found to be decreased upon WSS and this was accompanied by a decreased FC of 13 C-carbons from glucose. Consequently, while glucose is taken up by the ECs during WSS, less of its fraction becomes processed via glycolysis in comparison to static ECs. This finding was further strengthened by a significant decrease in lactate secretion in the medium (Fig. B) and a drop in the respective glycolytic m + 3 isotopologue of pyruvate and lactate (Fig. S2h-i). Hence, upon WSS, ECs become less glycolytic and rely less on glucose as the main nutrient. We also investigated the contribution of 13 C-glucose and 13 C-glutamine into the TCA intermediates. Overall, upon WSS, we found a decrease in the relative abundance of different TCA intermediates (Fig. A). The contribution of ( 13 C) carbons from glucose and glutamine to the Krebs cycle changed upon WSS (Fig. A). We observed an increase of FC by glutamine which was accompanied by a decrease of the FC of 13 C glucose carbons (Fig. A). Of interest, the increase of the m + 4 isotopologue of citrate and other TCA intermediates such as fumarate and malate by 13 C glutamine pointed towards a clockwise entry of 13 C carbons from glutamine into the TCA cycle (Fig. S2b-d) in ECs under WSS. Thus, in combination with an increase in the relative abundance of the isotopologues m + 5 and m + 4 from glutamate and asparagine respectively (Fig. S2e-f), we concluded that glutamine anaplerosis, and not glycolysis, is more prominent when endothelial cells are exposed to WSS in comparison the static cultured ECs. In conclusion, these observations indicate that during normo-physiological conditions, ECs, under WSS, favor glutamine as a carbon source for the central carbon metabolism-related pathways and reduce their glycolysis significantly. This finding adds a crucial missing piece of the puzzle of EC metabolism which was initially thought to be highly glycolytic. Glutamate dehydrogenase inhibition upon WSS impairs endothelial cell morphology in vitro To validate the importance of glutamine anaplerosis on EC fitness (the ability of ECs to maintain their morphological adaptation to flow, which includes cytoskeletal remodeling and elongation) we monitored the response of static and WSS ECs upon administration of established metabolic pathway inhibitors such as 2-deoxyglucose (2-DG) (Parniak & Kalant, ) a well-characterized glycolytic inhibitor, Epigallocatechin gallate (EGCG) (Choi & Park, ), an inhibitor of glutamate dehydrogenase (GDH) and CB-839 (Gross et al., ), a GLS1-specific inhibitor (Fig. S3a). Western blot analysis demonstrated that the exposure of ECs to WSS cultures led to the increased expression of glutaminase (GLS) and glutamate dehydrogenase (GDH) in these cultures, as expected, based on its putative mechanism of action, showing that glutamine metabolism is an essential metabolic signature of ECs undergoing WSS (Fig. S3b-c). Also, protein expression levels of beta-actin were elevated upon WSS (Fig. S3b-c) supporting previous studies (Choi & Helmke, ; Girard & Nerem, ). Immunofluorescence images of fixed cells were used to identify the morphological changes triggered by the different inhibitors both in static and WSS setups. This was accomplished by quantifying the changes in cell length, mean radius, and area between the different conditions as a readout for cellular fitness (Fig. S4a-c). Inhibition of glutamine and GDH metabolism (by CB-839 and EGCG) lowered the cellular capacity of elongation through a decrease of both cell length (Fig. S4a) and mean radius (Fig. S4b), while inhibition of glycolysis by 2DG only showed a slight reduction in the WSS setups (Fig. S4a-b). As well, the cellular area during WSS was also reduced by the treatment with EGCG and CB-839 (Fig. S4c), while 2DG, as expected, did not show any impairment (Fig. S3c). These findings show that glutamine metabolism through the anaplerosis of glutamine derived carbons into the Krebs cycle is essential for the integrity of the endothelial cells exposed to WSS. In contrast, blocking glycolysis only showed minimal effects, pointing towards a lesser dependency on glucose of ECs in normo-physiological conditions. To further evaluate cell viability and to determine whether the cell loss observed after EGCG and CB-839 treatment was caused by apoptosis, fluorescence-activated cell sorting (FACS) analysis using annexin V and propidium iodide (PI) was performed (Fig. A). The FACS analysis confirmed that greater than 90% of cells under static cultures (control) were viable and not labeled with Annexin V or PI (Fig. A). In in vitro shear stress models, higher apoptotic levels are typically observed compared to static conditions due to cell detachment and stress-induced cell death caused by shear forces, explaining a slighter higher number of apoptotic cells (18%) on the control shear stress condition in comparison with the static (Fig. A). Exposure to EGCG for 24 h significantly increased the number of apoptotic and necrotic cells by 45% and 27%, respectively. Furthermore, the treatment with CB-839 and 2DG slightly increased the number of apoptotic cells and does not influence the number of necrotic cells (Fig. A–B). These results indicate that the inhibition of glutamine uptake impairs cell viability of EC under WSS and the activity of GDH is essential to sustain the process. Mechanical signals related to the launch shear stress stimulate differential VE-cadherin and actin-mediated morphological changes that are dependent on the monolayer organization (Choi & Park, ) (Girard & Nerem, ). We tested whether the different inhibitor treatments triggered changes at focal adhesions and peripheral intercellular junctions of VE-cadherin (Barry et al., ) (Fig. C) and F-actin (Fig. D) on immunofluorescence images of fixed cells. In the untreated cell monolayers, ECs displayed the typical cobblestone morphology with intact, linear cell–cell junctions and cortical F-actin bundles at the cell periphery (Fig. C, D). However, EGCG treatment triggered the disruption of peripheral cell–cell junctions, as shown by punctate VE-cadherin staining (white arrows) (Fig. c). These changes are phenotypically similar to those induced by pro-inflammatory stimuli such as thrombin.((((Huveneers et al., ; Komarova & Malik, ; Verin et al., )))) Whereas CB-839 and 2DG treatment do not impair the morphology (Fig. C). Accordingly, analysis of F-actin through quantification of the integrated fluorescence intensity (Fig. S4d) and protein expression of b-actin (Fig. S4e-f) was shown to be decreased when ECs were treated with EGCG and CB-389 (Fig. S4d-f). Protein expression of GLS and GDH were, as expected, decreased when treated with the respective inhibitor (CB-389 and EGCG) (Fig. S4e-f). Shear stress induces substantial alterations in the cytoskeleton (e.g., actin and tubulin) but also in metabolic activity markers like GAPDH. Consequently, we acknowledge potential variability in tubulin expression, but we opted to use it as loading control as that exhibited the least variability across the experimental conditions (Fig. S3a, Fig.S4e-f). In conclusion, these inhibitor results demonstrate that when repressed, the enzyme GDH trigger severe local cytoskeletal remodeling and disrupt cell–cell contacts impairing the global integrity of endothelial tissues under WSS in vitro while blocking GLS activity has a light impairment on the ECs fitness.
To investigate the effect of WSS on EC metabolism, we carried out an unbiased transcriptomic analysis to assess global changes in HUVEC subjected to static and WSS cultures for 24 h under normal growth conditions. Principal component analysis (PCA) of all genes and hierarchical clustering analysis of the highly variable genes revealed that static control and WSS-treated HUVEC grouped into distinct clusters, indeed suggesting broad transcriptomic changes (Fig. A). Differential gene expression analysis (DGEA) of our transcriptomic data showed that, in line with the literature [(Binek et al., )], the shear stress marker KLF2 was elevated upon WSS (Fig. a), which we confirmed by real-time qPCR analysis (Fig. b), suggesting that our in vitro system can accurately mimic the in vivo response to WSS. Next, we investigated whether we could identify changes induced by WSS of transcripts involved in metabolic pathways. Gene set enrichment analysis (GSEA) (Subramanian et al., 2005) on metabolic KEGG gene sets was performed comparing static versus WSS-treated HUVEC. Our analysis showed an increased expression of genes involved in folate biosynthesis, alanine, aspartate, glutamate, glycerophospholipid, arachidonic acid, and amino/nucleotide sugar metabolism, while gene sets involved in pyrimidine metabolism, O-glycan biosynthesis, purine and drug metabolism, valine, leucine, isoleucine degradation, and glycolysis/gluconeogenesis were downregulated (Fig. B). A comparison of specific markers for glycolysis, alanine, aspartate, and glutamate metabolism revealed that the main genes involved in the glycolysis and pyruvate/lactate production (PGM1, PGM2, GPI, PFKL, PFKFB3, PKM, PFKM, TPI1, ENO1, PDHA1, LDHA, LDHB, and ACSS2) were downregulated (Fig. A), while genes involved in glutamine, asparagine, alanine, aspartate, and glutamate metabolism (GSR, G6PD, GCLC, IDH1, GOT1, GLUD1, ASS1, GFPT2, GFPT1, and ASNS) were upregulated (Fig. A). Real-time qPCR analysis further confirmed our bulk RNA sequencing data as the GLUD1, GLUL, PKM1, and PKM2 genes were upregulated (Fig. b) while HK2 was downregulated (Fig. b) upon WSS. GSEA on all KEGG gene sets was performed to associate groups of significant differentially expressed genes to biological processes [(Subramanian et al., )] not strongly related to metabolism (Fig. c). With DGEA, we then set out to obtain specific WSS-markers conserved across different biological donors. Among the top downregulated genes were GALNT15, SULT1B1, MAN1A1, RRM2, AQP1, DHRS3, and SLC46A1 (Fig. d) which have been associated with cell adhesion, DNA synthesis, replication, and repair, nucleotide metabolism and binding, glycerophospholipid and hydrolase activity and water transport. Likewise, among the top upregulated ranking genes were PTGS2, ALPL, CYP1B1, SLC7A11, ABCC3, AKR1C1, HMOX1, SLC6A4, and TXNRD1 (Fig. d) which have been linked to prostaglandin activity, amino acid and sodium/calcium transport, drug metabolism, ATPase and oxidoreductase activity, NAD cofactors biosynthesis and actin/tubulin polymerization. The meta-analysis revealed that the gene sets involved in glycolysis/gluconeogenesis were more upregulated in the static cultured ECs data set. Gene sets involved in alanine, aspartate, and glutamate metabolism were upregulated only on WSS cultured ECs data set (Fig. A). Proliferation can change metabolism and shear stress is reported to reduce proliferation [(Levesque, )]. Therefore, we set out to separate the role of WSS in promoting a pro-resting quiescent phenotype, from the effects WSS has on metabolism specifically. For this, HUVECs were cultured in the presence of EdU to confirm lower EdU incorporation upon exposure to shear stress in comparison to the static cultured HUVECs (Fig. e). Then, to exclude that proliferation/quiescent signaling was promoting the metabolic alterations seen in our transcriptomic analysis, we performed a meta- analysis of contact-inhibited versus proliferating ECs (GSE89174) [(Kalucka et al., )] and static versus WSS cultured ECs to rank differentially regulated gene sets (Fig. B). No genes were excluded or overlapped. While our analysis highlights significant metabolic changes in ECs under wall WSS, it is important to note that the quiescence states induced by WSS and contact inhibition may represent distinct biological conditions, which could influence the observed metabolic and transcriptomic profiles. In summary, we found that HUVECs subjected to WSS displayed large transcriptional alterations and upregulated genes associated not only with glutamine metabolism while downregulating genes involved in glycolysis but also in other important metabolic pathways responsible for key cellular events.
To investigate whether WSS affects EC metabolism, we first investigated its impact on nutritional behaviour. For this, medium samples from static and WSS cultured ECs for 24 h were analysed by LC–MS and these indeed showed distinct differences in the nutritional needs upon WSS (Fig. A). Figure A shows the relative differences in metabolites abundance between the medium from static and WSS cultured ECs. Compared to the static medium, the medium from WSS showed lower pyruvate, lactate, glutamine, and asparagine levels (Fig. A). This is in contrast with glucose, which showed to be more abundant in the WSS medium (Fig. A), pointing towards a lesser reliance on glucose during WSS. Comparison of the static and WSS medium towards the basal medium allowed us to understand the nutritional preferences of the different conditions (Fig. B). Figure B reveals that upon WSS, the nutritional preference of ECs towards glucose diminishes while the preference for glutamine as a nutrient becomes apparent. The combination of higher glucose and lower lactic acid levels also indicates that the cells are less glycolytic upon exposure to WSS. For this, we plotted the correlation of the absolute concentration of glucose and lactate in each condition (Fig. C), this allowed us to roughly estimate the glycolytic rate (in the hypothetical scenario where 100% of the glucose molecules are converted into lactic acid, 1 molecule of glucose will generate 2 molecules of lactic acid). In the static setups, a clear correlation between glucose and lactate levels was found (Fig. C) indicating that about 75% of the glucose was converted into lactic acid. This finding was confirmed by 73% of fractional contribution (FC) of 13 C glucose to lactic acid from the medium (Fig. S2a). Surprisingly, no correlation was found between glucose and lactic acid in the WSS setups being in line with the decreased lactic acid levels and higher glucose levels in the medium (Fig. A–C). This corroborated the finding that glucose metabolism is less preferred by ECs undergoing WSS. Here as well, the FC of glucose to lactic acid was significantly lower in the WSS setups (Fig. S2a). In conclusion, from our nutritional medium study, we considered that the fate of glutamine and glucose were of interest for further investigation by tracer metabolomics to determine the biochemical paths these key nutrients undergo. To understand how ECs utilize glutamine and glucose under WSS in vitro versus the static EC cultures, we supplemented the medium in parallel setups with respectively [U-13C5] glutamine and [U-13C6] glucose to determine both the abundances, amount of 13 C-label incorporation (FC) and isotopologues of the key metabolites involved in the main central carbon pathways, including glycolysis, TCA-cycle, amino acids and glutamine related intermediates (Fig. A, Fig. S2a-o). Data from the tracer metabolomics setups (Fig. A) is represented as pie charts (Craemer et al., ). The size of the pie chart represents the relative change in metabolite abundances while the distinct color reveals the FC of the respective tracers. Aligning with the previous findings, Fig. D shows that, during static conditions, ECs mainly rely on glucose and a combination of glucose and glutamine for the Krebs cycle. Upon WSS, the amount of ( 13 C) labeled glucose used in glycolysis is reduced significantly while the contribution and dependency on glutamine-derived carbons for the Krebs cycle increased (Fig. A). The abundance of glycolysis intermediates (glucose 6-phosphate (G6P) and pyruvate) were found to be decreased upon WSS and this was accompanied by a decreased FC of 13 C-carbons from glucose. Consequently, while glucose is taken up by the ECs during WSS, less of its fraction becomes processed via glycolysis in comparison to static ECs. This finding was further strengthened by a significant decrease in lactate secretion in the medium (Fig. B) and a drop in the respective glycolytic m + 3 isotopologue of pyruvate and lactate (Fig. S2h-i). Hence, upon WSS, ECs become less glycolytic and rely less on glucose as the main nutrient. We also investigated the contribution of 13 C-glucose and 13 C-glutamine into the TCA intermediates. Overall, upon WSS, we found a decrease in the relative abundance of different TCA intermediates (Fig. A). The contribution of ( 13 C) carbons from glucose and glutamine to the Krebs cycle changed upon WSS (Fig. A). We observed an increase of FC by glutamine which was accompanied by a decrease of the FC of 13 C glucose carbons (Fig. A). Of interest, the increase of the m + 4 isotopologue of citrate and other TCA intermediates such as fumarate and malate by 13 C glutamine pointed towards a clockwise entry of 13 C carbons from glutamine into the TCA cycle (Fig. S2b-d) in ECs under WSS. Thus, in combination with an increase in the relative abundance of the isotopologues m + 5 and m + 4 from glutamate and asparagine respectively (Fig. S2e-f), we concluded that glutamine anaplerosis, and not glycolysis, is more prominent when endothelial cells are exposed to WSS in comparison the static cultured ECs. In conclusion, these observations indicate that during normo-physiological conditions, ECs, under WSS, favor glutamine as a carbon source for the central carbon metabolism-related pathways and reduce their glycolysis significantly. This finding adds a crucial missing piece of the puzzle of EC metabolism which was initially thought to be highly glycolytic.
To validate the importance of glutamine anaplerosis on EC fitness (the ability of ECs to maintain their morphological adaptation to flow, which includes cytoskeletal remodeling and elongation) we monitored the response of static and WSS ECs upon administration of established metabolic pathway inhibitors such as 2-deoxyglucose (2-DG) (Parniak & Kalant, ) a well-characterized glycolytic inhibitor, Epigallocatechin gallate (EGCG) (Choi & Park, ), an inhibitor of glutamate dehydrogenase (GDH) and CB-839 (Gross et al., ), a GLS1-specific inhibitor (Fig. S3a). Western blot analysis demonstrated that the exposure of ECs to WSS cultures led to the increased expression of glutaminase (GLS) and glutamate dehydrogenase (GDH) in these cultures, as expected, based on its putative mechanism of action, showing that glutamine metabolism is an essential metabolic signature of ECs undergoing WSS (Fig. S3b-c). Also, protein expression levels of beta-actin were elevated upon WSS (Fig. S3b-c) supporting previous studies (Choi & Helmke, ; Girard & Nerem, ). Immunofluorescence images of fixed cells were used to identify the morphological changes triggered by the different inhibitors both in static and WSS setups. This was accomplished by quantifying the changes in cell length, mean radius, and area between the different conditions as a readout for cellular fitness (Fig. S4a-c). Inhibition of glutamine and GDH metabolism (by CB-839 and EGCG) lowered the cellular capacity of elongation through a decrease of both cell length (Fig. S4a) and mean radius (Fig. S4b), while inhibition of glycolysis by 2DG only showed a slight reduction in the WSS setups (Fig. S4a-b). As well, the cellular area during WSS was also reduced by the treatment with EGCG and CB-839 (Fig. S4c), while 2DG, as expected, did not show any impairment (Fig. S3c). These findings show that glutamine metabolism through the anaplerosis of glutamine derived carbons into the Krebs cycle is essential for the integrity of the endothelial cells exposed to WSS. In contrast, blocking glycolysis only showed minimal effects, pointing towards a lesser dependency on glucose of ECs in normo-physiological conditions. To further evaluate cell viability and to determine whether the cell loss observed after EGCG and CB-839 treatment was caused by apoptosis, fluorescence-activated cell sorting (FACS) analysis using annexin V and propidium iodide (PI) was performed (Fig. A). The FACS analysis confirmed that greater than 90% of cells under static cultures (control) were viable and not labeled with Annexin V or PI (Fig. A). In in vitro shear stress models, higher apoptotic levels are typically observed compared to static conditions due to cell detachment and stress-induced cell death caused by shear forces, explaining a slighter higher number of apoptotic cells (18%) on the control shear stress condition in comparison with the static (Fig. A). Exposure to EGCG for 24 h significantly increased the number of apoptotic and necrotic cells by 45% and 27%, respectively. Furthermore, the treatment with CB-839 and 2DG slightly increased the number of apoptotic cells and does not influence the number of necrotic cells (Fig. A–B). These results indicate that the inhibition of glutamine uptake impairs cell viability of EC under WSS and the activity of GDH is essential to sustain the process. Mechanical signals related to the launch shear stress stimulate differential VE-cadherin and actin-mediated morphological changes that are dependent on the monolayer organization (Choi & Park, ) (Girard & Nerem, ). We tested whether the different inhibitor treatments triggered changes at focal adhesions and peripheral intercellular junctions of VE-cadherin (Barry et al., ) (Fig. C) and F-actin (Fig. D) on immunofluorescence images of fixed cells. In the untreated cell monolayers, ECs displayed the typical cobblestone morphology with intact, linear cell–cell junctions and cortical F-actin bundles at the cell periphery (Fig. C, D). However, EGCG treatment triggered the disruption of peripheral cell–cell junctions, as shown by punctate VE-cadherin staining (white arrows) (Fig. c). These changes are phenotypically similar to those induced by pro-inflammatory stimuli such as thrombin.((((Huveneers et al., ; Komarova & Malik, ; Verin et al., )))) Whereas CB-839 and 2DG treatment do not impair the morphology (Fig. C). Accordingly, analysis of F-actin through quantification of the integrated fluorescence intensity (Fig. S4d) and protein expression of b-actin (Fig. S4e-f) was shown to be decreased when ECs were treated with EGCG and CB-389 (Fig. S4d-f). Protein expression of GLS and GDH were, as expected, decreased when treated with the respective inhibitor (CB-389 and EGCG) (Fig. S4e-f). Shear stress induces substantial alterations in the cytoskeleton (e.g., actin and tubulin) but also in metabolic activity markers like GAPDH. Consequently, we acknowledge potential variability in tubulin expression, but we opted to use it as loading control as that exhibited the least variability across the experimental conditions (Fig. S3a, Fig.S4e-f). In conclusion, these inhibitor results demonstrate that when repressed, the enzyme GDH trigger severe local cytoskeletal remodeling and disrupt cell–cell contacts impairing the global integrity of endothelial tissues under WSS in vitro while blocking GLS activity has a light impairment on the ECs fitness.
ECs, integral components of our vascular system, exhibit remarkable adaptability to environmental stimuli, as exemplified by their swift transition from a quiescent state to a proliferative status during angiogenesis.(Cruys et al., ) This dynamic metamorphosis is underpinned by a sophisticated utilization of metabolic pathways, which not only provide energy but also furnish essential molecules crucial for cellular functions. Not surprisingly, endothelial cell metabolism has been extensively studied (Zecchin et al., ) using in vitro systems mimicking the behavior of quiescent (Kalucka et al., ) or proliferating endothelial cells (De Bock et al. ) in health and disease (Maes et al., ) (Schoors et al., ),(Zecchin et al., ). Our research provides a crucial and missing piece of the puzzle on the normo-physiological metabolic behavior of endothelial cells in blood vessels. It highlights the importance of WSS -a physical stimulus caused by the blood flow- in the dynamic responses of endothelial cells. Prior research has suggested shear as a potential influencer of metabolic processes in various cellular contexts, including biofilms (Jones & Buie, ), epithelial cells (Miceli et al., ), and endothelial cells, more specifically: downregulation of the transcripts of glycolytic enzymes (Doddaballapur et al., ). However, direct metabolic evidence was lacking. Yet previous works have shown that glycolysis activity decreases in the presence of different shear stress stimuli ((Doddaballapur et al., ), (Han et al., ) . Our study reveals a nuanced perspective, elucidating that under physiological flow conditions, glutamine, rather than glucose, emerges as the primary energy source for quiescent ECs. This phenomenon, termed glutamine anaplerosis into the Krebs cycle, is substantiated by the heightened incorporation of 13 C carbons from glutamine into diverse Krebs intermediates in response to WSS. Our results thus support the previously hypothesized importance of mitochondria on EC metabolism (Groschner et al., ), (Kadlec et al., ), (Kluge et al., ), (Marcu et al., ). We demonstrate that glutamine anaplerosis into the Krebs cycle is indispensable for maintaining normo-physiological endothelial function. Moreover, our findings shed light on the crucial role of Glutamate Dehydrogenase (GDH) activity in this process. Suppression of GDH and Glutaminase (GLS) activity under WSS revealed a more pronounced impairment when GDH activity was inhibited, underscoring the intricate interplay of these enzymes in EC metabolism. This is explained by the fact that ECs can still synthesize glutamate via alternative pathways ((Eelen et al., ), (Huang et al., )) without impairing Krebs cycle activity. Our study demonstrates that WSS promotes glutamine anaplerosis into the Krebs cycle, underscoring ECs reliance on glutamine metabolism under shear stress. This also aligns with findings highlighting GLS1-mediated glutaminolysis as essential for EC proliferation, migration, and survival (Peyton et al., ). The metabolic landscape further evolves when WSS ceases, triggering a swift shift to glucose metabolism in ECs. This aligns with existing literature highlighting the role of glucose metabolism in supporting endothelial cells during angiogenesis (De Bock et al. ;(Schoors et al., )). Importantly, our work introduces a novel concept—the ability of ECs to rewire their metabolism in response to WSS, complementing their well-documented adaptability in cytoskeletal and signalling dynamics (Baeyens et al., ), (Girard & Nerem, ), (Roux et al., ), (Zhou et al., ). The observed decrease in intracellular TCA cycle intermediates, including glutamine, likely reflects accelerated metabolic turnover rather than suppressed metabolism. This is supported by inhibitor data (Fig. ) showing GDH inhibition disrupts VE-Cadherin adherens junctions, indicating increased metabolic flux. Under WSS, tracer metabolomics reveal enhanced glutamine-derived carbon influx into the TCA cycle, driving rapid intermediate consumption to meet biosynthetic and energetic demands. This suggests a metabolic adaptation where cells prioritize efficient processing of TCA intermediates under mechanical or physiological stress conditions (Choi et al., ). The metabolic response of ECs alters the nutritional environment of surrounding tissues and plays an important role in the absorption of nutrients from the bloodstream. This may explain the extensive crosstalk and interactions between shear stress and metabolic disturbance during the pathogenesis of atherosclerosis (disturbed shear stress environment), as strength by our transcriptomic results, showing a broad metabolic gene signature being altered upon WSS. In conformity with our findings, two important features of glucose metabolic disorders, hyperglycemia, and insulin resistance, have been described as key factors in the development of atherosclerotic plaques(((James & Allen, ; Jiang et al., ))), suggesting thus that glucose metabolism plays a role in the origin of the disorder (Fuh et al., ). The molecular mechanisms underlying these responses are only partially understood and previous studies indicate that flow alters the expression of hundreds of coding and non-coding RNAs.(Dunn et al., ), (Passerini et al., )] However, a recent study has uncovered a set of shear stress-responsive enhancing elements that will serve as a resource for future studies investigating the regulation of gene expression mediated by shear stress (Tsaryk et al., ). The importance of culture conditions as the medium composition (Lagziel et al., ) (Vande Voorde et al., )] and sample processing (Binek et al., ) was previously mentioned as key aspects to be considered when planning experiments, as they can alter cellular metabolism and cell physiology. Looking forward, our research underscores the imperative for meticulous consideration of culture conditions, medium composition, and sample processing in experimental designs. This emphasizes the need for advanced in vitro systems capable of faithfully mimicking the native vascular environment, affording precise control over biophysical stimuli. By elucidating the intricate mechanisms governing EC mechanobiology, we gain valuable insights into the etiology of diverse vascular diseases. These diseases encompass not only well-established conditions such as atherosclerosis, thrombosis, and aneurysm formation but also extend to cancer, metastasis, and diabetes. The interconnectedness of fluid shear stress with the molecular processes underlying tumor progression underscores the multifaceted nature of vascular biology and highlights the potential for targeted interventions in various pathological contexts.
Our study illuminates the critical role of WSS in shaping EC metabolic behaviour (Fig. A). Glutamine anaplerosis and GDH activity emerge as key player for EC function under normo-physiological conditions. These insights not only broaden our understanding of EC mechanobiology but also hold implications for vascular diseases, encompassing atherosclerosis, thrombosis, aneurysm formation, cancer, metastasis, and diabetes.
Below is the link to the electronic supplementary material. Supplementary file1 (DOCX 6885 KB)
|
Usual prevention in unusual settings: A scoping review of place-based health interventions in public-facing businesses | 72d4ead1-d81c-467d-908e-4f0d23aad6a1 | 11760570 | Health Promotion[mh] | An important part of public health is reaching all segments of the population where they live, work, and play . However, there are various groups in the United States that can be considered underserved, including some low-income, racial/ethnic minority, and immigrant populations. Thus, there have been increasing efforts to develop unique ways to outreach, engage, and serve these groups. Place-based interventions is one approach to reach people “where they are.” Place-based interventions can range widely, from changing physical environments and areas where groups live (e.g., creating bike lanes to facilitate exercise) to increasing access to resources in neighborhoods (e.g., introducing new farmers markets or transport facilities). A recent review examined 51 studies of locally-delivered place-based interventions across three elements of place and health: the physical, social, and economic environments . However, these interventions were mostly focused on altering the “place” of people as the intervention, which can disconnect individuals from their familiar community contexts. In contrast, delivering traditional health interventions in unconventional settings—such as salons and barbershops—maintains the familiarity and trust that these spaces offer, potentially leading to better engagement and participation from target populations. For example, a number of novel studies have developed health promotion interventions in salons and barbershops to reach African American communities, which have been summarized in a recent systematic review . However, the vast majority of health promotion activities in real-world settings and in the research literature continue to occur in traditional settings, such as primary care clinics, and clinics embedded in pharmacies, grocery stores, and community health centers. The Coronavirus Disease-2019 (COVID-19) pandemic also helped popularize these types of place-based health interventions, which were often referred to as “pop-up clinics” by offering COVID-19 vaccines across many public settings, such as groceries, malls, schools, etc. . But what types of other interventions and in which types of settings have these place-based interventions been used? Specifically, what unconventional public-facing business settings have been involved, and which communities have been served? No prior review found has examined the full scope of these types of place-based health interventions delivered at less conventional public-facing business settings across all health conditions and racial/ethnic groups. Such a review would help public health practitioners, researchers, and business owners understand what is possible in terms of implementing interventions in public-facing business settings. In particular, a review that examines not only the types of settings, but the health conditions targeted and the outcomes can further knowledge about how to reach underserved and minority populations in these settings. To address this knowledge gap, we conducted a scoping review of studies on delivering preventive health interventions in unconventional public-facing business settings, such as retail and service settings, to understand the type of interventions, the diversity of settings, the broad range of health conditions targeted, and how these interventions have been implemented across various racial and ethnic populations.
Protocol A review protocol was developed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guidelines . While this protocol was not registered, it followed a systematic approach aligned with PRISMA-ScR standards. The three major search databases used were PubMed, Google Scholar, and APA PsycNet. The search range in each database was set from 1960 to the present; however, the oldest study that met the inclusion criteria was published in 1995, and the most recent in 2023. Inclusion criteria were studies conducted in the United States, provided health interventions (e.g., education, disease screening, connection with healthcare providers, pharmacy services) to clients in one or more business settings for one or more diseases or conditions, were written in English, and were peer-reviewed. Exclusion criteria were studies that were not peer-reviewed, not published in English-written journals, or did not report study of a health intervention in a business setting. We only included peer-reviewed studies to ensure the review focused on studies which had been previously assessed and deemed acceptable for publication as a quality control measure for this review. Search method The two authors participated in all stages of the review, with one author leading study screening, selection, and data extraction and the other author assisting and double-checking this work for validity, and rerunning processes when there were discrepancies. Searches were initially conducted in PubMed using keywords and Boolean operators like “public health AND barbershop OR laundromat OR hair salon OR movie theater OR nail salon OR mechanic.” The full syntax of keywords and strategies used in these initial searches are provided in . The initial search returned a total of 2,727 unduplicated studies (PubMed 749, Google Scholar 1,930, APA PsycNet 48). Filtering and brief analysis of the titles and abstracts led to exclusion of 2,603 studies and a total of 124 studies entering the initial review process. Review process Between January and April 2024, the two authors reviewed all 124 studies. To reduce potential bias, the two authors conducted a calibration exercise, independently reviewing a subset of studies to establish consistency before proceeding with selecting studies and extracting data from the selected studies. In total, 77 studies were removed for not meeting the inclusion criteria following an initial preliminary review of the studies by title and abstract (PubMed 76, Google Scholar 1). Twelve of the excluded studies were systematic reviews, and the other 65 studies were excluded due to conduction of the studies outside of the United States, providing health interventions outside of unusual business settings, or only surveying business clients without providing additional health intervention services. After a secondary review (NM), five more studies were removed after the full text articles were assessed for eligibility due to not meeting the inclusion criteria. This resulted in a final selection of 42 studies. These 42 studies were then sorted into three categories based on targeted health conditions: 1) cardiovascular health conditions, HIV, and diabetes (20 studies), 2) cancers (13 studies), and 3) all other conditions (14 studies). Note that certain included studies may fall into more than one category due to targeting multiple diseases or conditions through its intervention. includes a CONSORT diagram summarizing the study identification and review process. Data extraction and synthesis Data extraction was conducted using an Excel-based form, which included fields for study design, business setting, health intervention, population of interest, sample size, target disease or condition, data findings, and conclusions. This Excel form served as a tracking and synthesis tool, while Zotero was used for citation management. To ensure consistency, data extraction was piloted by both reviewers on a subset of studies, following which any discrepancies were discussed and resolved. Synthesized data were categorized descriptively across intervention type, health condition, and setting, allowing for a narrative overview. To assess the quality of each study included in this review, each study was evaluated using the NIH Study Quality Assessment Tools, matched to the appropriate study type . Observational cohort and cross-sectional studies—including cross-sectional studies, longitudinal studies, descriptive studies, and cohort studies—were assessed using the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies . Pre-post studies, such as feasibility studies, program evaluations, and pilot studies, were evaluated with the Quality Assessment Tool for Before-After (Pre-Post) Studies With No Control Group . Finally, controlled intervention studies, specifically randomized controlled trials (RCTs), were assessed using the Quality Assessment of Controlled Intervention Studies . By applying these tools, we ensured that each study design underwent a rigorous and relevant quality assessment. This assessment provided insight into the methodological rigor and potential risk of bias across the included studies.
A review protocol was developed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) guidelines . While this protocol was not registered, it followed a systematic approach aligned with PRISMA-ScR standards. The three major search databases used were PubMed, Google Scholar, and APA PsycNet. The search range in each database was set from 1960 to the present; however, the oldest study that met the inclusion criteria was published in 1995, and the most recent in 2023. Inclusion criteria were studies conducted in the United States, provided health interventions (e.g., education, disease screening, connection with healthcare providers, pharmacy services) to clients in one or more business settings for one or more diseases or conditions, were written in English, and were peer-reviewed. Exclusion criteria were studies that were not peer-reviewed, not published in English-written journals, or did not report study of a health intervention in a business setting. We only included peer-reviewed studies to ensure the review focused on studies which had been previously assessed and deemed acceptable for publication as a quality control measure for this review.
The two authors participated in all stages of the review, with one author leading study screening, selection, and data extraction and the other author assisting and double-checking this work for validity, and rerunning processes when there were discrepancies. Searches were initially conducted in PubMed using keywords and Boolean operators like “public health AND barbershop OR laundromat OR hair salon OR movie theater OR nail salon OR mechanic.” The full syntax of keywords and strategies used in these initial searches are provided in . The initial search returned a total of 2,727 unduplicated studies (PubMed 749, Google Scholar 1,930, APA PsycNet 48). Filtering and brief analysis of the titles and abstracts led to exclusion of 2,603 studies and a total of 124 studies entering the initial review process.
Between January and April 2024, the two authors reviewed all 124 studies. To reduce potential bias, the two authors conducted a calibration exercise, independently reviewing a subset of studies to establish consistency before proceeding with selecting studies and extracting data from the selected studies. In total, 77 studies were removed for not meeting the inclusion criteria following an initial preliminary review of the studies by title and abstract (PubMed 76, Google Scholar 1). Twelve of the excluded studies were systematic reviews, and the other 65 studies were excluded due to conduction of the studies outside of the United States, providing health interventions outside of unusual business settings, or only surveying business clients without providing additional health intervention services. After a secondary review (NM), five more studies were removed after the full text articles were assessed for eligibility due to not meeting the inclusion criteria. This resulted in a final selection of 42 studies. These 42 studies were then sorted into three categories based on targeted health conditions: 1) cardiovascular health conditions, HIV, and diabetes (20 studies), 2) cancers (13 studies), and 3) all other conditions (14 studies). Note that certain included studies may fall into more than one category due to targeting multiple diseases or conditions through its intervention. includes a CONSORT diagram summarizing the study identification and review process.
Data extraction was conducted using an Excel-based form, which included fields for study design, business setting, health intervention, population of interest, sample size, target disease or condition, data findings, and conclusions. This Excel form served as a tracking and synthesis tool, while Zotero was used for citation management. To ensure consistency, data extraction was piloted by both reviewers on a subset of studies, following which any discrepancies were discussed and resolved. Synthesized data were categorized descriptively across intervention type, health condition, and setting, allowing for a narrative overview. To assess the quality of each study included in this review, each study was evaluated using the NIH Study Quality Assessment Tools, matched to the appropriate study type . Observational cohort and cross-sectional studies—including cross-sectional studies, longitudinal studies, descriptive studies, and cohort studies—were assessed using the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies . Pre-post studies, such as feasibility studies, program evaluations, and pilot studies, were evaluated with the Quality Assessment Tool for Before-After (Pre-Post) Studies With No Control Group . Finally, controlled intervention studies, specifically randomized controlled trials (RCTs), were assessed using the Quality Assessment of Controlled Intervention Studies . By applying these tools, we ensured that each study design underwent a rigorous and relevant quality assessment. This assessment provided insight into the methodological rigor and potential risk of bias across the included studies.
The 42 studies identified were categorized by the health condition(s) they were focused on in Tables – . contained studies providing interventions for cardiovascular health conditions, HIV, and diabetes [ – ]. contained studies providing interventions for different cancers [ , – ], and contained studies providing interventions for all other conditions not falling into the previously listed categories [ , , , – ]. Each table lists the study author(s), study design, intervention setting (e.g., barbershop, nail salon, etc.), study population, sample size, health condition targeted by intervention, geographic area of study, major data findings, and overall findings. Note that some studies provided interventions for multiple conditions with certain studies falling into multiple tables [ , , ]. Among the 42 studies, 36 studies involved health education as an intervention, 20 studies offered preventative screening as an intervention, 9 studies referred and facilitated connection to local healthcare providers and resources, and 3 studies offered pharmacy services as an intervention. Cardiovascular health conditions A total of 12 studies were identified by their provided interventions for cardiovascular health conditions. Of these, 2 studies provided interventions for peripheral artery disease, 9 for hypertension or blood pressure monitoring, and 1 for heart disease. Both studies investigating peripheral artery disease (PAD) at barbershops sampled Black men in a midwestern state, each with a sample size of 37 participants . One of these studies used a longitudinal study design , while the other was a qualitative analysis and was a sub-study of the longitudinal study . In the longitudinal study , participants completed three visits to the barbershop: a first visit initially screened participants for PAD, a second visit rescreened participants for PAD and presented a PAD education video 4–6 weeks later, and a third visit in which exit interviews and assessments were conducted. The trial ultimately diagnosed PAD in 5/31 (16.1%) of participants and overall awareness of PAD began low at the beginning of the study but significantly increased between the initial and exit visit assessments. The qualitative study was conducted at the final exit assessment of the longitudinal study and involved individual interviews to understand the perspectives of Black men in receiving barbershop-based screening and PAD education. Several common themes arose such as acknowledgement of barriers like fear, trust, and healthcare access, but participants indicated that the barbershop intervention enhanced knowledge of PAD and cardiovascular disease. A total of 9 studies investigating hypertension or blood pressure management were identified, and all 9 studies targeted these conditions in either Black men or predominantly Black populations and offered interventions at barbershops or hair salons [ , , , – ]. These studies were conducted in all five regions of the United States (Northeast, Southwest, West, Southeast, and Midwest), and study designs included program evaluations, cluster-randomized trials, exploratory studies, longitudinal studies, pilot intervention studies, cohort studies, and descriptive studies. Additionally, the sample sizes of these hypertension intervention studies ranged widely from 10 to 14,000 participants. Each of the 9 studies incorporated blood pressure screening into their interventions in different ways. For example, studies that investigated hypertension with larger sample sizes (i.e., 680–14,000) focused on taking blood pressure measurements, using self-report questions asking about recent blood pressure screenings, connecting participants to providers, or training barbers and hair stylists to serve as lay health educators [ , – ]. Overall, these larger studies found that barbershops and beauty salons served as suitable locations for blood pressure screenings and reported increased awareness of hypertension and connection to providers by participants. The other five hypertension studies utilized longitudinal or cluster-randomized trial designs and included baseline and follow-up blood pressure measurements. These studies reported that participants who were connected with pharmacy or other medical providers and provided hypertension education had notable reductions in their blood pressure [ , , – ]. Two cluster-randomized trials were of particular note due to their study designs (control and intervention groups) and results, which found that the use of barbers as lay health educators and connection with pharmacy and healthcare services led to significant reduction in mean blood pressure in the intervention groups . Two of these hypertension studies with longitudinal components utilized additional unique methods in their interventions. One study incorporated a telehealth component for follow-up care by connecting pharmacy providers with study participants recruited from barbershops online after their blood pressure had been controlled via medication . The other study was a cluster-randomized trial that directly connected some participants to hypertension specialists and observed greater reductions in systolic blood pressure than participants who were first connected to primary care providers . HIV A total of 6 studies provided interventions for HIV in this review [ – ]. Sample sizes of these studies ranged from 48–1,124 participants, and the designs of the included studies were a cluster-randomized trial, longitudinal studies, descriptive studies, and qualitative analyses. The majority of HIV intervention studies were conducted in the Northeast region of the United States, with one being conducted in the Southeastern United States . Additionally, while most of the studies primarily targeted predominantly Black populations recruited from laundromats, one study recruited participants from a variety of settings; and another study sampled diverse racial/ethnic participants . All 6 studies provided participants education about HIV and safe sex practices. One notable study was a descriptive study assessing previous HIV knowledge and providing prevention and education services to Black participants (n = 677) at corner stores, beauty supply stores, laundromats, mechanics, and barbershops . The intervention targeted Black communities heavily affected by HIV in Miami, Florida and included a survey collecting demographic information and HIV knowledge and prevention services like free condom distribution and HIV testing. Within these communities, 68.8% had never heard of PrEP (a medication highly effective at reducing the risk of getting HIV), 8% had never been tested for HIV, and 65.9% had no primary care provider . The study found that the intervention delivered at corner and food stores had the most engagement followed by laundromats, barbershops, and beauty salons. A second notable intervention study based in Delaware utilized the Popular Opinion Model (POL), a community level peer-based outreach strategy designed by the Centers for Disease Control and Prevention, to provide HIV education to over 6,000 racial/ethnic minority participants and HIV tests to 1,124 of those participants . A third notable HIV study was a cluster randomized controlled trial targeting Black men (n = 618) that compared the Shape Up! Barbers Building Better Brothers HIV risk-reduction intervention (based on the theory of planned behavior) or an attention-matched violence prevention control . The Shape Up! intervention led to a significantly increased consistent condom use in the postintervention period. These three notable studies together suggest place-based interventions can lead to decreased risky behavior associated HIV transmission (e.g. condomless sex, sex with multiple partners, etc.) and increased self-efficacy for condom use among participants [ , , ]. Diabetes A total of 5 studies provided interventions for diabetes were included in the review [ , , , , ]. Each of these studies occurred at barbershops or beauty salons, and incorporated diabetes education and prevention strategies. The designs of the 5 diabetes studies included a program evaluation, a longitudinal study, a cohort study, a focus group and qualitative research study, and a cross-sectional study. Two of the studies targeted predominantly Black populations, and two additional studies targeted Black men specifically. The remaining study included a broad sample of participants deemed “at risk” for developing diabetes. The sample sizes of the included studies ranged from 13–14,000 participants. Three studies provided a diabetes intervention in addition to interventions for other conditions [ , , ]. One study was a program evaluation targeting predominantly Black populations (n = 1,823) in barbershops and beauty salons for high blood pressure, diabetes, tobacco-use associated conditions, high cholesterol, and need for social services . Through partnership with an integrated healthcare system, local barbershops and salons in Baltimore, Maryland, and a mobile health clinic, the program screened 469 participants and connected them with free resources when necessary. Another study was a longitudinal study (n = 356) targeting people at risk for diabetes, heart disease, stroke, colon cancer, and breast cancer in laundromats and investigated perceived risk and susceptibility to these conditions; the study found 18% of participants believed their risk for diabetes was lower than it was . The third study was a cohort study that trained over 700 stylists as lay health educators and reported reaching over 14,000 clients with 60% of clients reporting they took steps to prevent or address their diabetes, hypertension, or kidney disease with a provider . The two remaining diabetes studies were a qualitative study and a cross-sectional study, each targeting Black men in barbershops . In the qualitative study, focus groups with 13 participants found that diet and exercise were recognized as ways to prevent diabetes, and people were supportive of barbershops as sites for a diabetes intervention program . The cross-sectional study sampled 290 participants and provided diagnostic hemoglobin A1C testing on site at barbershops and diabetes education based on screening results . Cancer A total of 13 studies providing cancer-related prevention, screening, education, and referral services were included in this review [ , – ]. Of the 13 studies, 6 of them provided interventions for breast cancer, 4 for prostate cancer, 2 for colorectal cancer, and 1 including both colorectal cancer and breast cancer. Seven studies provided breast cancer interventions with sample sizes ranging from 162–10,306 participants [ , – , – ]. Study designs included longitudinal studies, descriptive studies, randomized-controlled trials, cluster randomized trials, health education programs, and pilot studies. The two cluster randomized trials are noteworthy due to their inclusion of control and intervention groups in their study designs . The majority of studies occurred in beauty salons; but barbershops, churches, neighborhood health centers, laundromats, social service agencies, health fairs, and public libraries were also breast cancer intervention sites. Interventions included health education provided verbally by medical professionals and stylists trained as lay health educators; and education and connection to local resources via kiosks, magazines, store displays, and other paper materials. Regarding target populations, five of the seven interventions targeted Black women specifically, while the other two targeted participants at risk for breast cancer and all women, respectively . One notable study provided breast cancer education for Black women through touch-screen kiosks located in beauty salons, churches, health fairs, neighborhood health centers, laundromats, public libraries and social service agencies . These kiosks used an interactive computer program called Reflections of You that printed magazines for users containing tailored breast cancer education and local breast cancer resources based on participants’ answers to screening questions. These kiosks reached 4,527 participants in under 18 months and reported that 34.1% of participants over 40 had never had a mammogram before the intervention. Another descriptive study by the same lead author used the Reflections of You kiosks to identify appropriate community channels and settings for delivering evidence-based breast cancer health promotion materials . Through the 10,306 kiosks used over a four year period, the study identified laundromats were the only settings that had the highest kiosk use and highest specificity (e.g. proportion of users without health insurance, barriers to getting a mammogram, low breast cancer and mammography knowledge, etc.). Four studies provided prostate cancer interventions as two pilot studies, a descriptive study, and a non-randomized comparison study [ , , , ]. While three of the four studies targeted Black men in barbershops, one study targeted all men in barbershops, churches, industries, meal sites, car dealerships, civic organizations, and housing projects. Sample sizes of each of these studies ranged from 40–1,552 participants. Each study provided prostate cancer education and prevention materials. One notable study due to its large sample size and unique findings was a descriptive study investigating predictors of participation in free prostate cancer screenings in barbershops, churches, industries, meal sites, car dealerships, civic organizations, and housing projects . The study ultimately found that being white, having at least a high school education, being married, perceiving health benefits, and receiving a client navigator or prior education intervention were significant predictors of participation in the study’s free prostate cancer screenings. Another pilot study investigated the feasibility of training barbers to deliver customized (culturally appropriate) prostate cancer education to Black men, mostly through brochures . Through the feasibility pilot study, prostate cancer knowledge scores raised from 60% to 79%. Three studies provided colorectal cancer interventions in barbershops as a longitudinal study, a qualitative analysis, and a randomized controlled trial respectively [ , , ]. The intervention provided in each of these studies was colorectal cancer education and screening, with two studies focused on Black men and the third study focused on adults in general at risk for developing diabetes, heart disease, stroke, and breast cancer in addition to colorectal cancer. One particularly significant colorectal cancer intervention study due to its incorporation of a telehealth component and large sample size (n = 731) was the randomized controlled trial . This study aimed to test the effectiveness of a preclinical, telephone-based intervention designed to encourage and connect older Black men to colorectal cancer screening opportunities. Black male participants were recruited from barbershops initially and placed in one of three telephone intervention groups: patient navigation by a community health worker for colorectal cancer screening, motivational interviewing by a trained counselor, or both interventions. The study ultimately found that both groups of participants that received navigation by community health worker were most likely to pursue colorectal cancer screening within six months. Other conditions Fourteen studies focused on other conditions not already described above, including high cholesterol, overall physical fitness, mental health, nutrition, stroke, unintended pregnancy, violence, influenza, kidney disease, HPV, and COVID-19 [ , , , – ]. These studies included cross-sectional studies, longitudinal studies, descriptive studies, feasibility studies, a randomized controlled trial, pilot studies, cohort studies, and program evaluations. Sample sizes ranged from 20–14,000 participants, and target populations included predominantly Black populations, at risk participants for certain conditions, women across all demographic classifications, adults with children 6 months to 2 years of age, adults over 50 years of age, and citizens in the Northeast Bronx region of New York. Each study provided education, prevention strategies, and screening promotion in some capacity for its respective target condition. These studies took place in barbershops, beauty salons, nail salons, and movie theaters. One descriptive study is particularly noteworthy because of its relatively large sample size (n = 530 participants), specific target population (adults with children 6 months to 2 years of age and adults over 50 years of age), and with the intervention taking place in movie theaters . This intervention was designed to promote annual influenza vaccination by showing slides providing education about the flu and advocating for people to get their annual flu vaccine prior to presentation of upcoming movie premieres. Among moviegoers exposed to the education slides prior to the film, 24% recalled seeing the flu vaccination slides prior to the movie advertisements although some participants did not arrive to the theater before the start of the film to see the flu vaccination slides. An additional noteworthy study was a randomized controlled trial aimed at assessing the impact of an intervention conducted in barbershops on mental health and violence threat screening among Black men in Philadelphia, PA . With a sample size of 618 participants, the study found significant effects of the intervention on increasing awareness of Black manhood vulnerability. This heightened awareness contributed to a significant reduction in physical fights among participants. The study’s findings were robust, demonstrating statistically significant pathways from the intervention through both Black manhood vulnerability awareness and hypermasculinity to the outcomes studied. Another interesting study investigated the effectiveness of using beauticians to educate Black female clients about stroke warning signs and risk factors . Beauticians were trained about stroke warning signs and risk factors, and clients were asked survey questions about their stroke knowledge before and after the intervention. The study reported significant increases in client knowledge of stroke warning signs (40.7% to 50.6%) and to call 911 for stroke symptoms (86% to 94%) with this improvement sustained for five months. However, no significant increase in knowledge of the three stroke risk factors was seen before and after the intervention. One final study to note is one focused on providing COVID-19 vaccination education and resources in barbershops, hair salons, beauty salons, and faith-based organizations to Northeast Bronx citizens . Forty-five public-facing business sites across Northeast Bronx, New York participated in this COVID-19 intervention by encouraging clients to complete baseline and follow-up surveys about perceptions of COVID-19 vaccines and commitments to future vaccination, having conversations about COVID-19 and offering supporting materials, displaying posters and brochures on site encouraging vaccination, and hosting local health department staff on site. Over a span of four months, COVID-19 vaccination rates across five zip codes in Northeast Bronx were observed to increase from 5.6% to 8.7%, although causality of the intervention cannot be inferred.
A total of 12 studies were identified by their provided interventions for cardiovascular health conditions. Of these, 2 studies provided interventions for peripheral artery disease, 9 for hypertension or blood pressure monitoring, and 1 for heart disease. Both studies investigating peripheral artery disease (PAD) at barbershops sampled Black men in a midwestern state, each with a sample size of 37 participants . One of these studies used a longitudinal study design , while the other was a qualitative analysis and was a sub-study of the longitudinal study . In the longitudinal study , participants completed three visits to the barbershop: a first visit initially screened participants for PAD, a second visit rescreened participants for PAD and presented a PAD education video 4–6 weeks later, and a third visit in which exit interviews and assessments were conducted. The trial ultimately diagnosed PAD in 5/31 (16.1%) of participants and overall awareness of PAD began low at the beginning of the study but significantly increased between the initial and exit visit assessments. The qualitative study was conducted at the final exit assessment of the longitudinal study and involved individual interviews to understand the perspectives of Black men in receiving barbershop-based screening and PAD education. Several common themes arose such as acknowledgement of barriers like fear, trust, and healthcare access, but participants indicated that the barbershop intervention enhanced knowledge of PAD and cardiovascular disease. A total of 9 studies investigating hypertension or blood pressure management were identified, and all 9 studies targeted these conditions in either Black men or predominantly Black populations and offered interventions at barbershops or hair salons [ , , , – ]. These studies were conducted in all five regions of the United States (Northeast, Southwest, West, Southeast, and Midwest), and study designs included program evaluations, cluster-randomized trials, exploratory studies, longitudinal studies, pilot intervention studies, cohort studies, and descriptive studies. Additionally, the sample sizes of these hypertension intervention studies ranged widely from 10 to 14,000 participants. Each of the 9 studies incorporated blood pressure screening into their interventions in different ways. For example, studies that investigated hypertension with larger sample sizes (i.e., 680–14,000) focused on taking blood pressure measurements, using self-report questions asking about recent blood pressure screenings, connecting participants to providers, or training barbers and hair stylists to serve as lay health educators [ , – ]. Overall, these larger studies found that barbershops and beauty salons served as suitable locations for blood pressure screenings and reported increased awareness of hypertension and connection to providers by participants. The other five hypertension studies utilized longitudinal or cluster-randomized trial designs and included baseline and follow-up blood pressure measurements. These studies reported that participants who were connected with pharmacy or other medical providers and provided hypertension education had notable reductions in their blood pressure [ , , – ]. Two cluster-randomized trials were of particular note due to their study designs (control and intervention groups) and results, which found that the use of barbers as lay health educators and connection with pharmacy and healthcare services led to significant reduction in mean blood pressure in the intervention groups . Two of these hypertension studies with longitudinal components utilized additional unique methods in their interventions. One study incorporated a telehealth component for follow-up care by connecting pharmacy providers with study participants recruited from barbershops online after their blood pressure had been controlled via medication . The other study was a cluster-randomized trial that directly connected some participants to hypertension specialists and observed greater reductions in systolic blood pressure than participants who were first connected to primary care providers .
A total of 6 studies provided interventions for HIV in this review [ – ]. Sample sizes of these studies ranged from 48–1,124 participants, and the designs of the included studies were a cluster-randomized trial, longitudinal studies, descriptive studies, and qualitative analyses. The majority of HIV intervention studies were conducted in the Northeast region of the United States, with one being conducted in the Southeastern United States . Additionally, while most of the studies primarily targeted predominantly Black populations recruited from laundromats, one study recruited participants from a variety of settings; and another study sampled diverse racial/ethnic participants . All 6 studies provided participants education about HIV and safe sex practices. One notable study was a descriptive study assessing previous HIV knowledge and providing prevention and education services to Black participants (n = 677) at corner stores, beauty supply stores, laundromats, mechanics, and barbershops . The intervention targeted Black communities heavily affected by HIV in Miami, Florida and included a survey collecting demographic information and HIV knowledge and prevention services like free condom distribution and HIV testing. Within these communities, 68.8% had never heard of PrEP (a medication highly effective at reducing the risk of getting HIV), 8% had never been tested for HIV, and 65.9% had no primary care provider . The study found that the intervention delivered at corner and food stores had the most engagement followed by laundromats, barbershops, and beauty salons. A second notable intervention study based in Delaware utilized the Popular Opinion Model (POL), a community level peer-based outreach strategy designed by the Centers for Disease Control and Prevention, to provide HIV education to over 6,000 racial/ethnic minority participants and HIV tests to 1,124 of those participants . A third notable HIV study was a cluster randomized controlled trial targeting Black men (n = 618) that compared the Shape Up! Barbers Building Better Brothers HIV risk-reduction intervention (based on the theory of planned behavior) or an attention-matched violence prevention control . The Shape Up! intervention led to a significantly increased consistent condom use in the postintervention period. These three notable studies together suggest place-based interventions can lead to decreased risky behavior associated HIV transmission (e.g. condomless sex, sex with multiple partners, etc.) and increased self-efficacy for condom use among participants [ , , ].
A total of 5 studies provided interventions for diabetes were included in the review [ , , , , ]. Each of these studies occurred at barbershops or beauty salons, and incorporated diabetes education and prevention strategies. The designs of the 5 diabetes studies included a program evaluation, a longitudinal study, a cohort study, a focus group and qualitative research study, and a cross-sectional study. Two of the studies targeted predominantly Black populations, and two additional studies targeted Black men specifically. The remaining study included a broad sample of participants deemed “at risk” for developing diabetes. The sample sizes of the included studies ranged from 13–14,000 participants. Three studies provided a diabetes intervention in addition to interventions for other conditions [ , , ]. One study was a program evaluation targeting predominantly Black populations (n = 1,823) in barbershops and beauty salons for high blood pressure, diabetes, tobacco-use associated conditions, high cholesterol, and need for social services . Through partnership with an integrated healthcare system, local barbershops and salons in Baltimore, Maryland, and a mobile health clinic, the program screened 469 participants and connected them with free resources when necessary. Another study was a longitudinal study (n = 356) targeting people at risk for diabetes, heart disease, stroke, colon cancer, and breast cancer in laundromats and investigated perceived risk and susceptibility to these conditions; the study found 18% of participants believed their risk for diabetes was lower than it was . The third study was a cohort study that trained over 700 stylists as lay health educators and reported reaching over 14,000 clients with 60% of clients reporting they took steps to prevent or address their diabetes, hypertension, or kidney disease with a provider . The two remaining diabetes studies were a qualitative study and a cross-sectional study, each targeting Black men in barbershops . In the qualitative study, focus groups with 13 participants found that diet and exercise were recognized as ways to prevent diabetes, and people were supportive of barbershops as sites for a diabetes intervention program . The cross-sectional study sampled 290 participants and provided diagnostic hemoglobin A1C testing on site at barbershops and diabetes education based on screening results .
A total of 13 studies providing cancer-related prevention, screening, education, and referral services were included in this review [ , – ]. Of the 13 studies, 6 of them provided interventions for breast cancer, 4 for prostate cancer, 2 for colorectal cancer, and 1 including both colorectal cancer and breast cancer. Seven studies provided breast cancer interventions with sample sizes ranging from 162–10,306 participants [ , – , – ]. Study designs included longitudinal studies, descriptive studies, randomized-controlled trials, cluster randomized trials, health education programs, and pilot studies. The two cluster randomized trials are noteworthy due to their inclusion of control and intervention groups in their study designs . The majority of studies occurred in beauty salons; but barbershops, churches, neighborhood health centers, laundromats, social service agencies, health fairs, and public libraries were also breast cancer intervention sites. Interventions included health education provided verbally by medical professionals and stylists trained as lay health educators; and education and connection to local resources via kiosks, magazines, store displays, and other paper materials. Regarding target populations, five of the seven interventions targeted Black women specifically, while the other two targeted participants at risk for breast cancer and all women, respectively . One notable study provided breast cancer education for Black women through touch-screen kiosks located in beauty salons, churches, health fairs, neighborhood health centers, laundromats, public libraries and social service agencies . These kiosks used an interactive computer program called Reflections of You that printed magazines for users containing tailored breast cancer education and local breast cancer resources based on participants’ answers to screening questions. These kiosks reached 4,527 participants in under 18 months and reported that 34.1% of participants over 40 had never had a mammogram before the intervention. Another descriptive study by the same lead author used the Reflections of You kiosks to identify appropriate community channels and settings for delivering evidence-based breast cancer health promotion materials . Through the 10,306 kiosks used over a four year period, the study identified laundromats were the only settings that had the highest kiosk use and highest specificity (e.g. proportion of users without health insurance, barriers to getting a mammogram, low breast cancer and mammography knowledge, etc.). Four studies provided prostate cancer interventions as two pilot studies, a descriptive study, and a non-randomized comparison study [ , , , ]. While three of the four studies targeted Black men in barbershops, one study targeted all men in barbershops, churches, industries, meal sites, car dealerships, civic organizations, and housing projects. Sample sizes of each of these studies ranged from 40–1,552 participants. Each study provided prostate cancer education and prevention materials. One notable study due to its large sample size and unique findings was a descriptive study investigating predictors of participation in free prostate cancer screenings in barbershops, churches, industries, meal sites, car dealerships, civic organizations, and housing projects . The study ultimately found that being white, having at least a high school education, being married, perceiving health benefits, and receiving a client navigator or prior education intervention were significant predictors of participation in the study’s free prostate cancer screenings. Another pilot study investigated the feasibility of training barbers to deliver customized (culturally appropriate) prostate cancer education to Black men, mostly through brochures . Through the feasibility pilot study, prostate cancer knowledge scores raised from 60% to 79%. Three studies provided colorectal cancer interventions in barbershops as a longitudinal study, a qualitative analysis, and a randomized controlled trial respectively [ , , ]. The intervention provided in each of these studies was colorectal cancer education and screening, with two studies focused on Black men and the third study focused on adults in general at risk for developing diabetes, heart disease, stroke, and breast cancer in addition to colorectal cancer. One particularly significant colorectal cancer intervention study due to its incorporation of a telehealth component and large sample size (n = 731) was the randomized controlled trial . This study aimed to test the effectiveness of a preclinical, telephone-based intervention designed to encourage and connect older Black men to colorectal cancer screening opportunities. Black male participants were recruited from barbershops initially and placed in one of three telephone intervention groups: patient navigation by a community health worker for colorectal cancer screening, motivational interviewing by a trained counselor, or both interventions. The study ultimately found that both groups of participants that received navigation by community health worker were most likely to pursue colorectal cancer screening within six months.
Fourteen studies focused on other conditions not already described above, including high cholesterol, overall physical fitness, mental health, nutrition, stroke, unintended pregnancy, violence, influenza, kidney disease, HPV, and COVID-19 [ , , , – ]. These studies included cross-sectional studies, longitudinal studies, descriptive studies, feasibility studies, a randomized controlled trial, pilot studies, cohort studies, and program evaluations. Sample sizes ranged from 20–14,000 participants, and target populations included predominantly Black populations, at risk participants for certain conditions, women across all demographic classifications, adults with children 6 months to 2 years of age, adults over 50 years of age, and citizens in the Northeast Bronx region of New York. Each study provided education, prevention strategies, and screening promotion in some capacity for its respective target condition. These studies took place in barbershops, beauty salons, nail salons, and movie theaters. One descriptive study is particularly noteworthy because of its relatively large sample size (n = 530 participants), specific target population (adults with children 6 months to 2 years of age and adults over 50 years of age), and with the intervention taking place in movie theaters . This intervention was designed to promote annual influenza vaccination by showing slides providing education about the flu and advocating for people to get their annual flu vaccine prior to presentation of upcoming movie premieres. Among moviegoers exposed to the education slides prior to the film, 24% recalled seeing the flu vaccination slides prior to the movie advertisements although some participants did not arrive to the theater before the start of the film to see the flu vaccination slides. An additional noteworthy study was a randomized controlled trial aimed at assessing the impact of an intervention conducted in barbershops on mental health and violence threat screening among Black men in Philadelphia, PA . With a sample size of 618 participants, the study found significant effects of the intervention on increasing awareness of Black manhood vulnerability. This heightened awareness contributed to a significant reduction in physical fights among participants. The study’s findings were robust, demonstrating statistically significant pathways from the intervention through both Black manhood vulnerability awareness and hypermasculinity to the outcomes studied. Another interesting study investigated the effectiveness of using beauticians to educate Black female clients about stroke warning signs and risk factors . Beauticians were trained about stroke warning signs and risk factors, and clients were asked survey questions about their stroke knowledge before and after the intervention. The study reported significant increases in client knowledge of stroke warning signs (40.7% to 50.6%) and to call 911 for stroke symptoms (86% to 94%) with this improvement sustained for five months. However, no significant increase in knowledge of the three stroke risk factors was seen before and after the intervention. One final study to note is one focused on providing COVID-19 vaccination education and resources in barbershops, hair salons, beauty salons, and faith-based organizations to Northeast Bronx citizens . Forty-five public-facing business sites across Northeast Bronx, New York participated in this COVID-19 intervention by encouraging clients to complete baseline and follow-up surveys about perceptions of COVID-19 vaccines and commitments to future vaccination, having conversations about COVID-19 and offering supporting materials, displaying posters and brochures on site encouraging vaccination, and hosting local health department staff on site. Over a span of four months, COVID-19 vaccination rates across five zip codes in Northeast Bronx were observed to increase from 5.6% to 8.7%, although causality of the intervention cannot be inferred.
This unique review of place-based health interventions in public business settings found that a number of studies have been conducted on the topic in the past two decades. We reviewed 42 studies of place-based health interventions offered for various chronic health conditions and certain select business settings. The majority of the interventions offered were health education and preventative health screenings. Thirty-four (81%) of the studies focused on reaching Black populations; all studies, except for one, delivered health interventions in barbershops and beauty salons either solely or among a few other settings. The largest number of studies focused on cancer (13 studies) or cardiovascular disease (12 studies). Additionally, the specific health condition with the greatest number of controlled trials was for hypertension (3 studies). In general, studies reported that health interventions embedded in public settings were associated with positive outcomes, including increased disease awareness, improved health behaviors and disease management, and high rates of health screening and connection to healthcare services. Given that all studies targeted outreach to racial/ethnic minority populations, the findings suggest place-based interventions are an important way to reach underserved population and potentially address health disparities by providing accessible health education, screenings, and connection to services in places that they visit to purchase goods and services. However, it is essential to acknowledge that conducting health interventions in public spaces may inadvertently lead to unintended consequences, such as stigma and concerns about privacy. The presence of stigma or concerns about privacy can affect participant engagement and willingness to utilize these services, potentially undermining the effectiveness of the interventions. Thus, careful consideration of these social dynamics is crucial when designing and implementing interventions in community settings. Placements of these interventions in barbershops and beauty salons may represent familiar and trusted community settings that can enhance participant engagement. Further study is needed to expand beyond these settings to determine whether other public-facing business settings (e.g., banks, movie theatres, malls) are also effective and acceptable places for health interventions. Among the 42 studies reviewed, there was a general lack of rigor in the designed studies. Only one study received the maximum quality rating score , and no other study was within 2 point of the maximum quality rating score. Most studies were descriptive or observational one-group designs and did not include a comparison group. Although we identified 7 randomized trials, including several cluster randomized trials, these studies varied widely in terms of interventions and health conditions. As a result, we did not attempt to synthesize the results quantitatively. Instead, we summarized the findings of each study individually, highlighting their diverse approaches and outcomes. This decision was based on the heterogeneity of the studies, which made direct comparison challenging. In addition, many of these trials did not appear to be rigorously designed and may have had many threats to internal validity (e.g., confounding variables, inadequate sample size, limited follow-up or differential attrition between groups) that were not fully examined. Almost all studies relied on subjective outcome measures and did not measure objective health outcomes (e.g., service utilization, disease onset and outcome) so there is a need for further rigorous studies with objective outcomes. Together, our review concludes there is a small, growing body of studies of health interventions delivered in public-facing business settings that shows some preliminary success in reaching Black communities for a variety of health conditions, and these interventions may be a promising strategy to reach underserved populations but more rigorous and varied studies are needed to expand and deepen the evidence for these interventions to pinpoint how they are effective, who they are most effective for, and in using which interventions in what places. This review had several strengths and limitations worth noting. Given the nature of scoping reviews, we took a broad, comprehensive approach to cover a wide range of health conditions, interventions, settings, and study designs. There was wide variability in studies making it challenging to compare studies, and a meta-analysis could not be conducted to quantify a summary of outcomes. Given the range of studies, we are also limited in specificity in drawing conclusions. However, we have tried to summarize findings by health condition to organize the studies and allow researchers to focus on particular health conditions. We only included studies in the United States, and there may be various innovative place-based interventions delivered internationally in other countries that would yield new insights so that is both a limitation of our review and an opportunity for future research. Moreover, we only include published studies, and there may be a “file-drawer problem” of unpublished studies we do not include. Finally, while stakeholder consultation is recognized as beneficial in scoping reviews, no formal stakeholder consultations were conducted for this review, which we acknowledge as a limitation. Additionally, while our approach aligns with best practices for scoping reviews, the absence of protocol registration may limit transparency. We also recognize this as a potential limitation to reproducibility. These limitations notwithstanding, this review highlights unique and innovative ways to reach underserved populations in places like barbershops and beauty salons. The strongest evidence for these place-based interventions is for cardiovascular disease (especially hypertension) and cancer, but there are opportunities to study this further for various other health conditions. Together, these studies demonstrate possible collaborations between healthcare providers, researchers, and business owners with mutual goals to serve underserved communities. Finally, this review paves numerous paths for needed research in this area, including more experimental studies with objective outcomes, examination of the sustainability and scalability of these interventions, and the cost-effectiveness of interventions to support their adoption by businesses, healthcare providers, and policymakers.
S1 Table Keywords/operators/truncation used in databases for systematic review. (DOCX) S1 Checklist PRISMA 2020 checklist. (DOCX)
|
Supporting resident-centred decision-making about transitions from long-term care homes to hospital: a qualitative study protocol | a8a28b85-bd48-489e-912a-cc46af127617 | 11628961 | Patient-Centered Care[mh] | Long-term care (LTC) residents are typically nearing the final stages of their life’s journey; these stages often include many care transitions. Care transitions (called transitions henceforth) refer to transfers between different care settings (eg, hospital, LTC, home and community care). Transitions across care settings can be difficult and stressful, especially for adults nearing the end of life. Transitions near the end of life can also result in adverse health outcomes for residents. For example, a recent study found that among 555 residents who transitioned from LTC to hospital, adverse events (ie, skin tears, pressure ulcers and falls) were experienced in 37% of transitions; and 70% of these adverse events were deemed to be preventable. Furthermore, the proportion of LTC residents who die in hospital at the end of life varies markedly between countries, from 6% in Canada to 77% in Japan. Person-centred care is considered as the gold standard in LTC care in Canada and aligns with the Residents’ Bill of Rights (2021), which mandates that residents have the right to be involved in the decision-making around their care in LTC homes in Ontario. In this province, LTC is primarily available to individuals with complex medical needs or those requiring substantial assistance with activities of daily living, such as bathing, dressing and mobility. Health and social care for residents in LTC is publicly funded but requires resident copayments to cover accommodation costs. LTC homes are regulated under the under the Fixing Long-Term Care Act, 2021, which replaced the Long-Term Care Homes Act, 2007. In contrast to LTC, other care options for older adults in Ontario include retirement homes and ageing in place through community-based care. Retirement homes, which provide more independence and typically cater to older adults with fewer healthcare needs, are privately funded and not regulated as strictly as LTC. Ageing in place, supported by government-subsidised home care services, is a growing trend for older adults who require less intensive support and prefer to remain in their homes. While the Canada Health Act provides some universal healthcare coverage across provinces, the specifics of publicly funded services vary significantly. While person-centred care is the ideal, the reality of care delivery in LTC homes is often constrained by systemic challenges. These include understaffing, time constraints and heavy workloads, all of which contribute to a system that prioritises efficiency and medical needs over individualised care. As a result, care in LTC is frequently task-oriented, focusing on managing medical conditions rather than attending to individual preferences or emotional and social needs. In contrast, person-centred care shifts the focus from a system-driven, medicalised approach to one that prioritises the resident’s values, personal goals and holistic well-being. This model encourages collaboration between residents, their family or friend care partners and staff, emphasising individualised care plans that align with the resident’s identity, preferences and experiences. Person-centred care differs significantly from standard care in Ontario LTC homes, which is often shaped by systemic barriers that limit the ability of staff to fully engage with residents on a personal level. Person-centred care refers to care that encompasses the principles of respecting individuals, acknowledging their inherent human dignity, treating them as unique persons and understanding what holds significance to them in relation to their treatment and care. Moving forward, the term ‘person-centred care’ will be referred to as ‘resident-centred care’ to better align with this study’s focus on individuals residing in LTC. This term also acknowledges that family or friend care partners (ie, informal care assistants who take on the primary role of assisting an individual in managing their health) frequently play a role in care decision-making alongside the LTC residents that they support. We will use the Person-centred Practice in Long-term Care (PeoPLe) theoretical framework to guide and inform this work. The PeoPLe framework offers a resident-centred care perspective that all practices in LTC, including facilitating transitions to and from hospital, should prioritise residents’ values and preferences and respect residents’ personhood as they live with progressive disease and disability. The framework is based on five constructs: (1) prerequisites; (2) practice environment; (3) person-centred processes; (4) fundamental principles of care and (5) outcome. The prerequisite construct requires professional competency, interpersonal skills and commitment to the job. The practice environment invites shared decision-making, effective staff relationships and supportive organisational systems. The person-centred processes and fundamental principles of care require holistic approaches, authentic engagement and working with the person’s beliefs. Finally, the outcome promotes a healthful culture. We will use these constructs to help us highlight and navigate the tensions inherent in providing resident-centred care in LTC settings, where over 70% of residents have dementia and may have difficulty engaging in discussions and decision-making about their care. The framework’s explicit focus on LTC is highly relevant to our aim to develop a decision-making tool that will be useful in this setting given common structural barriers to the provision of resident-centred care in LTC, such as time constraints, heavy workloads, staffing shortages and lack of management support. Transition decisions require consideration of the benefits and harms of the transition and discussions with residents and care partners about which of these matter to them most. When an LTC resident experiences symptoms such as severe pain, respiratory distress or infection, a transfer to the hospital can provide access to diagnostic tests and treatments. However, a transfer to hospital may be misaligned with the resident’s or their family/friend care partners’ wishes, understanding of their illness trajectory and goals of care. LTC to hospital transitions can cause a variety of emotional and physical harms. For example, waiting in an emergency room, away from the resident’s home environment and familiar care team can lead to feelings of distress, anxiety, fear and lack of autonomy. It can also lead to complications such as delirium, pressure ulcers and hospital acquired infections. When decisions about transitions from LTC to hospital are resident-centred, they can improve residents’ satisfaction with care, the quality and safety of care, quality of life and well-being of residents and decrease hospital readmission rates. On the other hand, many LTC residents are transferred to hospital at the end of life and die receiving burdensome medical treatments. This process denies them the opportunity for a tranquil death in a familiar, home-like environment that aligns with the preferences of the majority of Canadians. The existing body of literature consistently underscores the insufficient support provided to LTC staff during decision-making processes associated with transitions from LTC to hospital settings. Similarly, when weighing potential risks and benefits of a transition from LTC to hospital, residents and their family/friends care partners may not fully comprehend or take into account the expected or anticipated decline in both physical and cognitive abilities that many LTC residents experience. This decline is often associated with the frailty death trajectory, which refers to a unique pattern of decline in the last 12 months of life. This trajectory for LTC residents is different from the terminal phase observed in individuals dying of cancer, for example, where there is a more clearly defined period leading up to the end of life. Notably, decisions regarding transitions from LTC to hospital (eg, transfer to hospital to insert an intravenous line to treat dehydration or feeding tube because of dysphagia—both of which are often expected in advanced frailty) as an attempt to delay or reverse a progressive decline often do not consider the natural progression of disease nor the resident’s values and goals. This may result in actions that harm residents and are stressful for family/friend care partners. An additional advantage of developing a decision support tool to be used by residents, care partners and staff when preparing for the possibility of a future transition from LTC to hospital is the parallel benefit of enhanced understanding by family and friends of the anticipated changes in the residents’ health over the course of their stay in LTC. This project will employ a participatory codesign approach, engaging residents, care partners and staff as equal collaborators. In phase 1, needs assessments, content analysis and codesign workshops will capture diverse perspectives to guide the creation of a decision-making tool for transitions between LTC homes and hospitals. Subsequent phases will involve interdisciplinary tool development, followed by pilot testing and evaluation in partnership with LTC homes. By incorporating the voices of a diverse population, including Black, Indigenous and People of Colour (BIPOC) communities, the study will ensure culturally responsive solutions, enhancing the inclusivity and applicability of the tool for all collaborators involved. Aims and objectives The aim of this study is to inform and codesign a decision-making tool that will help ensure that decisions regarding LTC to hospital transitions support residents’ autonomy and align with their preferences and priorities. Objectives Conduct a needs assessment with LTC residents, care partners and staff involved in decisions about transitions from LTC to hospital to inform the development of a decision-making tool. Undertake a content analysis of existing decision support and discussion tools to inform the codesign of our tool with a focus on cultural awareness and safety. Codesign a decision-making tool with LTC residents, care partners and staff at three LTC sites aimed at improving the decision-making experience regarding LTC to hospital transitions that is acceptable and appropriate to residents, family/friend care partners and care providers and is feasible and sustainable for long-term use in LTC settings. Research questions What are the experiences of residents, care partners and staff surrounding decision-making about transitions from LTC to hospital? What can we learn from existing decision support and discussion tools, specifically around cultural awareness and safety? What are staff, resident and care partner priorities for a decision-making tool that they feel would better inform, support and engage them before and during the decision-making process regarding transitions from LTC to hospital?
The aim of this study is to inform and codesign a decision-making tool that will help ensure that decisions regarding LTC to hospital transitions support residents’ autonomy and align with their preferences and priorities. Objectives Conduct a needs assessment with LTC residents, care partners and staff involved in decisions about transitions from LTC to hospital to inform the development of a decision-making tool. Undertake a content analysis of existing decision support and discussion tools to inform the codesign of our tool with a focus on cultural awareness and safety. Codesign a decision-making tool with LTC residents, care partners and staff at three LTC sites aimed at improving the decision-making experience regarding LTC to hospital transitions that is acceptable and appropriate to residents, family/friend care partners and care providers and is feasible and sustainable for long-term use in LTC settings. Research questions What are the experiences of residents, care partners and staff surrounding decision-making about transitions from LTC to hospital? What can we learn from existing decision support and discussion tools, specifically around cultural awareness and safety? What are staff, resident and care partner priorities for a decision-making tool that they feel would better inform, support and engage them before and during the decision-making process regarding transitions from LTC to hospital?
Conduct a needs assessment with LTC residents, care partners and staff involved in decisions about transitions from LTC to hospital to inform the development of a decision-making tool. Undertake a content analysis of existing decision support and discussion tools to inform the codesign of our tool with a focus on cultural awareness and safety. Codesign a decision-making tool with LTC residents, care partners and staff at three LTC sites aimed at improving the decision-making experience regarding LTC to hospital transitions that is acceptable and appropriate to residents, family/friend care partners and care providers and is feasible and sustainable for long-term use in LTC settings.
What are the experiences of residents, care partners and staff surrounding decision-making about transitions from LTC to hospital? What can we learn from existing decision support and discussion tools, specifically around cultural awareness and safety? What are staff, resident and care partner priorities for a decision-making tool that they feel would better inform, support and engage them before and during the decision-making process regarding transitions from LTC to hospital?
This study will be conducted in collaboration with three LTC homes in Ontario, Canada. The research team maintains continuous partnerships with these homes and is well positioned to codesign a tool with residents, care partners and staff (eg, physicians, nurses, personal support workers and administrators) to aid in decision-making about LTC to hospital transitions. Phase 1 of this project will use a three-part sequential qualitative study design that includes needs assessment interviews, content analysis of existing tools and collaborative codesign workshops. Patient and public involvement To develop the overarching study aims and design, the research team engaged with the leadership team that guides the larger study in which this project is embedded, the family and friends’ and resident councils of our partnering LTC homes and a Research Advisory Committee comprised of family/friend care partners. These discussions informed the larger 4 year study that will inform and codesign (phase 1), create (phase 2), pilot test (phase 3) and evaluate (phase 4) a decision-making tool to guide LTC to hospital transitions to be more resident-centred. We will adapt the Double Diamond method’s four stages: (1) discover, (2) define, (3) develop and (4) deliver to guide the development of our decision-support tool ( ). This participatory codesign approach treats all stakeholders as equal collaborators in the design process. This will allow for the development of a tool and indicators of success that meet the needs of residents, care partners and staff. The discover and define stages will take place during phase 1 (the period covered by this protocol) via needs assessment interviews (objective 1), content analysis of existing tools and discussion guides (objective 2) and codesign workshops (objective 3). The develop and deliver stages will comprise the tool creation by the quantitative team (comprised of patient partners, researchers, clinicians, engineers and graphic designers) within our research team (phase 2 ethics approval #M16-23-030) based on the findings from objectives 1–3, and then pilot tested (phase 3) and evaluate (phase 4) in partnership with LTC homes later in the larger project. Phase 1: objective 1 We will conduct semistructured needs assessment interviews in our three partnering LTC homes with residents (n=9); family/friend care partners, power of attorney for personal care or substitute decision-makers (n=9) and staff (n=9) with experience of decision-making regarding an LTC to hospital transition. The selection of these LTC homes is based on their existing partnerships with our research team and their diverse resident populations, which will help us capture a range of experiences and perspectives. To recruit participants, we will post recruitment posters in LTC homes, include information in newsletters, present the project at family, resident and nurse council network meetings, and work with LTC home leadership to identify interested individuals. The target of nine participants per group was determined based on previous research showing that this sample size is typically sufficient to achieve data saturation in qualitative studies involving LTC settings. We expect to reach data saturation as common themes and perspectives emerge, but we are prepared to increase the sample size if additional data are needed to ensure comprehensive coverage of the research topic. We will use a qualitative descriptive design for this substudy 1, a methodology well suited for understanding complex, real-world phenomena, such as decision-making during care transitions in LTC settings. Qualitative descriptive studies aim to provide a straightforward account of participants’ experiences and are commonly used in health research to capture the perspectives of various stakeholders. This approach aligns with our objective to explore the needs and experiences of residents, care partners and staff, allowing for a comprehensive understanding of decision-making in LTC to hospital transitions. Based on past research, we anticipate that this sample size (n=27) will be sufficient to get diverse viewpoints to address our research questions. The interview script (duration 1 hour) will be guided by the Decisional Needs Assessment Workbook by Jacobsen et al . See for the interview guide for resident participants. Sampling will be purposive and will aim to ensure maximum inclusivity and diversity of research participants through collaboration with the partnering homes’ equality, diversity and inclusion committees. As the older adult population grows in Canada, we are also witnessing wider diversity in all age categories. This is reflected in the diversity of LTC staff, with representation from BIPOC communities. Ensuring a diverse sample in our recruitment process will foster inclusivity in efforts to improve care transitions for all LTC residents, care partners and staff. Interviews will be conducted either in person or virtually by members of the research team, with informed consent obtained prior to the interviews. In cases where a participant is unable to give consent directly, a designated proxy may do so on their behalf. For residents who have a proxy, ongoing assent will be observed throughout the sessions by evaluating their verbal, behavioural or emotional responses, such as smiling or nodding, to ensure their comfort and willingness to continue participating. Interviews will include a sociodemographic questionnaire component to help us describe our sample and consider the relevance of intersecting equity considerations in our interpretation of the interview data. To strengthen the evaluation of the decision-making tool and enhance the overall robustness of the study, we will incorporate specific outcomes in phase 1 that will facilitate both quantitative and qualitative assessments in subsequent phases. Specifically, we will assess perceived adequacy of support through a structured questionnaire designed to capture gaps and satisfaction in current transition processes. Each interviewer will write a self-reflexive journal entry after each interview. Recordings of the interviews will be transcribed for analysis. Qualitative analysis software (MAXQDA) will be used to organise interview transcript data and data will be analysed using reflexive thematic analysis (RTA) according to Braun and Clarke’s approach. This robust method of analysis seeks to establish patterns of meaning by recursively engaging with the data. RTA involves a six-step process including familiarisation with the data set; coding closely to the research question; generating initial themes; developing and reviewing themes; refining themes and reporting on the themes. The robustness and rigour of the data will be assessed using Lincoln and Guba’s four-dimension criteria of credibility, dependability, confirmability and transferability. Credibility will be ensured through the prolonged engagement of participants, involving an introductory session with LTC home councils, continuous involvement in both study phases and presentation of results at the study’s completion. Dependability will be established through a comprehensive record of the data collection process and an adaptable protocol that embraces reflexivity. Confirmability will be ensured through reflexivity by having weekly team meetings and maintaining reflexive journals after interviews. Finally, transferability will be confirmed through our purposive sampling at each of the three LTC homes and the comparison of findings across research sites. Phase 1: objective 2 A content analysis of existing serious illness conversation tools and decision-making aids will be conducted guided by the PeoPLe theoretical framework. Qualitative content analysis is a method used to systematically interpret textual data by identifying patterns, themes and relationships within the material. This approach emphasises coding and categorising data to make sense of its underlying meanings, ensuring that all relevant aspects of the data are examined. In this substudy 2, we will apply directed content analysis, where pre-existing theories, such as the PeoPLe framework, guide the analysis. This method ensures a structured examination of tools, focusing on their cultural sensitivity, accessibility and effectiveness in addressing resident preferences. The review will involve a systematic search of the literature to identify tools currently in use with particular emphasis on those that prioritise cultural relevance and safety in LTC. Criteria for inclusion will encompass tools intended for facilitating discussions about serious illness or care transitions among healthcare providers, residents and care partners, with a specific focus on those designed to be culturally aware. We will develop a set of evaluative criteria based on the content analysis of existing tools and the initial feedback gathered during codesign workshops. These criteria may include aspects such as tool usability, clarity, cultural sensitivity and impact on decision-making processes. We will work with residents, care partners and staff to identify the evaluation criteria that are most meaningful to them and prioritise the measurement of these in our evaluation. This set of criteria will guide the assessment of the new decision-making tool’s effectiveness in phase 4. This approach seeks to supplement and enhance the representation of ethno-cultural diversity in our study, ensuring a more comprehensive understanding of the perspectives and preferences of residents, care partners and staff. Phase 1: objective 3 Codesign methodology emphasises collaborative, participatory processes where all stakeholders actively contribute to the development of a solution. This substudy 3 will follow a participatory action research framework, fostering a cyclical process of action, reflection and refinement. Informed by findings from objectives 1 and 2, we will facilitate two codesign workshops at each of three LTC homes (n=6) to solicit resident, family/friend care partner and staff preferences and priorities for the design of the decision-making tool. We anticipate that this number of workshops will allow us to reach data saturation as recurring themes and patterns emerge across the participant groups. However, additional workshops will be conducted if necessary to ensure comprehensive data collection. The first three workshops (duration: 2 hours each) will include residents and care partners from the three partnering LTC homes recruited through purposive sampling. This method involves deliberately selecting participants based on specific characteristics, such as their experience with transitions from LTC to hospital or their involvement in decision-making processes. A minimum of three participants from each partnering LTC home will be recruited for each codesign workshop. Based on past research, we anticipate that this sample size will be sufficient to gather diverse insight for tool development. Participants will be recruited using snowball sampling with the help of participants from objective 1. Objective 3 will be open to participants from phase 1 as well as new participants. We will obtain informed consent prior to participation in codesign workshops. The research team will share findings from objectives 1 and 2 and then facilitate a discussion of resident and care partner preferences and priorities for a tool-based solution to improving resident-centred decision-making around LTC to hospital transitions. Participants will be invited to brainstorm tool design, items included and format guided by the Ottawa Decision Support Framework ( ). This framework aims to improve decisional outcomes by supporting quality decision-making that is informed and values-based and has been instrumental in developing many patient decision aids, measures and training programmes. We will use the Ottawa Decision Support Framework to develop a series of prompts to guide the codesign workshops. This will ensure that we solicit feedback on all the decisional needs identified in the framework. The first round of codesign workshops will be recorded, transcribed, and a member of the research team will take detailed notes on participant preferences, and feedback for a decision-making tool prototype. The research staff will analyse the data from the codesign workshops to identify cross-cutting themes and preferences for a decision-making tool and consolidate a list of key priorities and design criteria based on resident and care partner values, preferences and recommendations. This list will be distributed back to workshop participants for participant validation as a method of increasing trustworthiness. The list will be revised as needed. The second round of codesign workshops will be held with key staff members involved in making decisions about LTC to hospital transitions (eg, personal support workers, nurses and physicians). Staff will be recruited through connections made in objective 1, using informational flyers that provide a description of the study and contact information for our team and snowball sampling. We will recruit three staff members at each home in different care provider roles (n=9). We will facilitate one codesign workshop at each of the three LTC homes where the researchers will share findings from objectives 1 and 2 and the first round of codesign workshops and then facilitate a discussion of staff preferences and priorities for the design, items included and format of a tool-based solution to improving resident-centred decision-making around LTC to hospital transitions. We will use the same process for informed consent, data collection and analysis as in the first round of codesign workshops. In addition, we will pilot test preliminary tool features within a subset of participating LTC homes to gather initial data on their feasibility and acceptability (phase 3). This pilot testing will involve collecting feedback on the tool’s design, functionality and user experience from both residents and staff. These preliminary findings will inform refinements and adjustments to ensure that the final tool is well aligned with the needs and preferences identified in phase 1. Timeline Phase 1 is anticipated to span approximately 6 months, from January 2024 to June 2024. Phase 2 will follow from July 2024 to December 2024. Phase 3 is scheduled from January 2025 to March 2025, with phase 4 occurring from April 2025 to June 2025. Dissemination Results from phase 1 will be used to inform the creation of the tool in phase 2 of the larger study. Phase 1 results will also be shared with residents, family/friend care partners and staff at an in-person session at each participating home. The research team will develop infographics describing phase 1 results in English and French and encourage the family, resident and nurse councils at each home to disseminate them at their council meetings. Knowledge translation activities for phase 1 findings also include presentations at conferences (eg, the Canadian Hospice and Palliative Care Association Annual conference, Ontario Long-Term Care Clinicians Annual Conference and Palliative Approach in LTC Community Practice webinars), progress reports to our partners, publications in open-access journals (eg, BMC Geriatrics) and social media posts by the researchers. Later in the larger study, the new decision-making tool will be disseminated through the Long-Term Care Community Practice, a group of professionals, patients and family/friend care partners who share best practices for palliative approaches in LTC. The tool will also be circulated via Advance Care Planning Canada, an initiative dedicated to creating a Pan-Canadian Framework for advance care planning. Ethics This project was approved by the Bruyère Health Research Ethics Board (#M16-23-036) as well as the ethics boards of each three partnering LTC homes. Informed consent will be collected from all participants prior to the start of the interviews and workshops. If a participant requires a proxy to consent on their behalf, ongoing assent will be obtained based on an assessment of how the resident expresses or indicates their preferences verbally, behaviourally or emotionally (eg, smiling, nodding, etc). This assessment will ensure that the participant is able to make a meaningful choice and has at least a minimal level of understanding. Given that 70% of residents in LTC have dementia, we anticipate that there will be involvement from this population in our research study.
To develop the overarching study aims and design, the research team engaged with the leadership team that guides the larger study in which this project is embedded, the family and friends’ and resident councils of our partnering LTC homes and a Research Advisory Committee comprised of family/friend care partners. These discussions informed the larger 4 year study that will inform and codesign (phase 1), create (phase 2), pilot test (phase 3) and evaluate (phase 4) a decision-making tool to guide LTC to hospital transitions to be more resident-centred. We will adapt the Double Diamond method’s four stages: (1) discover, (2) define, (3) develop and (4) deliver to guide the development of our decision-support tool ( ). This participatory codesign approach treats all stakeholders as equal collaborators in the design process. This will allow for the development of a tool and indicators of success that meet the needs of residents, care partners and staff. The discover and define stages will take place during phase 1 (the period covered by this protocol) via needs assessment interviews (objective 1), content analysis of existing tools and discussion guides (objective 2) and codesign workshops (objective 3). The develop and deliver stages will comprise the tool creation by the quantitative team (comprised of patient partners, researchers, clinicians, engineers and graphic designers) within our research team (phase 2 ethics approval #M16-23-030) based on the findings from objectives 1–3, and then pilot tested (phase 3) and evaluate (phase 4) in partnership with LTC homes later in the larger project.
We will conduct semistructured needs assessment interviews in our three partnering LTC homes with residents (n=9); family/friend care partners, power of attorney for personal care or substitute decision-makers (n=9) and staff (n=9) with experience of decision-making regarding an LTC to hospital transition. The selection of these LTC homes is based on their existing partnerships with our research team and their diverse resident populations, which will help us capture a range of experiences and perspectives. To recruit participants, we will post recruitment posters in LTC homes, include information in newsletters, present the project at family, resident and nurse council network meetings, and work with LTC home leadership to identify interested individuals. The target of nine participants per group was determined based on previous research showing that this sample size is typically sufficient to achieve data saturation in qualitative studies involving LTC settings. We expect to reach data saturation as common themes and perspectives emerge, but we are prepared to increase the sample size if additional data are needed to ensure comprehensive coverage of the research topic. We will use a qualitative descriptive design for this substudy 1, a methodology well suited for understanding complex, real-world phenomena, such as decision-making during care transitions in LTC settings. Qualitative descriptive studies aim to provide a straightforward account of participants’ experiences and are commonly used in health research to capture the perspectives of various stakeholders. This approach aligns with our objective to explore the needs and experiences of residents, care partners and staff, allowing for a comprehensive understanding of decision-making in LTC to hospital transitions. Based on past research, we anticipate that this sample size (n=27) will be sufficient to get diverse viewpoints to address our research questions. The interview script (duration 1 hour) will be guided by the Decisional Needs Assessment Workbook by Jacobsen et al . See for the interview guide for resident participants. Sampling will be purposive and will aim to ensure maximum inclusivity and diversity of research participants through collaboration with the partnering homes’ equality, diversity and inclusion committees. As the older adult population grows in Canada, we are also witnessing wider diversity in all age categories. This is reflected in the diversity of LTC staff, with representation from BIPOC communities. Ensuring a diverse sample in our recruitment process will foster inclusivity in efforts to improve care transitions for all LTC residents, care partners and staff. Interviews will be conducted either in person or virtually by members of the research team, with informed consent obtained prior to the interviews. In cases where a participant is unable to give consent directly, a designated proxy may do so on their behalf. For residents who have a proxy, ongoing assent will be observed throughout the sessions by evaluating their verbal, behavioural or emotional responses, such as smiling or nodding, to ensure their comfort and willingness to continue participating. Interviews will include a sociodemographic questionnaire component to help us describe our sample and consider the relevance of intersecting equity considerations in our interpretation of the interview data. To strengthen the evaluation of the decision-making tool and enhance the overall robustness of the study, we will incorporate specific outcomes in phase 1 that will facilitate both quantitative and qualitative assessments in subsequent phases. Specifically, we will assess perceived adequacy of support through a structured questionnaire designed to capture gaps and satisfaction in current transition processes. Each interviewer will write a self-reflexive journal entry after each interview. Recordings of the interviews will be transcribed for analysis. Qualitative analysis software (MAXQDA) will be used to organise interview transcript data and data will be analysed using reflexive thematic analysis (RTA) according to Braun and Clarke’s approach. This robust method of analysis seeks to establish patterns of meaning by recursively engaging with the data. RTA involves a six-step process including familiarisation with the data set; coding closely to the research question; generating initial themes; developing and reviewing themes; refining themes and reporting on the themes. The robustness and rigour of the data will be assessed using Lincoln and Guba’s four-dimension criteria of credibility, dependability, confirmability and transferability. Credibility will be ensured through the prolonged engagement of participants, involving an introductory session with LTC home councils, continuous involvement in both study phases and presentation of results at the study’s completion. Dependability will be established through a comprehensive record of the data collection process and an adaptable protocol that embraces reflexivity. Confirmability will be ensured through reflexivity by having weekly team meetings and maintaining reflexive journals after interviews. Finally, transferability will be confirmed through our purposive sampling at each of the three LTC homes and the comparison of findings across research sites.
A content analysis of existing serious illness conversation tools and decision-making aids will be conducted guided by the PeoPLe theoretical framework. Qualitative content analysis is a method used to systematically interpret textual data by identifying patterns, themes and relationships within the material. This approach emphasises coding and categorising data to make sense of its underlying meanings, ensuring that all relevant aspects of the data are examined. In this substudy 2, we will apply directed content analysis, where pre-existing theories, such as the PeoPLe framework, guide the analysis. This method ensures a structured examination of tools, focusing on their cultural sensitivity, accessibility and effectiveness in addressing resident preferences. The review will involve a systematic search of the literature to identify tools currently in use with particular emphasis on those that prioritise cultural relevance and safety in LTC. Criteria for inclusion will encompass tools intended for facilitating discussions about serious illness or care transitions among healthcare providers, residents and care partners, with a specific focus on those designed to be culturally aware. We will develop a set of evaluative criteria based on the content analysis of existing tools and the initial feedback gathered during codesign workshops. These criteria may include aspects such as tool usability, clarity, cultural sensitivity and impact on decision-making processes. We will work with residents, care partners and staff to identify the evaluation criteria that are most meaningful to them and prioritise the measurement of these in our evaluation. This set of criteria will guide the assessment of the new decision-making tool’s effectiveness in phase 4. This approach seeks to supplement and enhance the representation of ethno-cultural diversity in our study, ensuring a more comprehensive understanding of the perspectives and preferences of residents, care partners and staff.
Codesign methodology emphasises collaborative, participatory processes where all stakeholders actively contribute to the development of a solution. This substudy 3 will follow a participatory action research framework, fostering a cyclical process of action, reflection and refinement. Informed by findings from objectives 1 and 2, we will facilitate two codesign workshops at each of three LTC homes (n=6) to solicit resident, family/friend care partner and staff preferences and priorities for the design of the decision-making tool. We anticipate that this number of workshops will allow us to reach data saturation as recurring themes and patterns emerge across the participant groups. However, additional workshops will be conducted if necessary to ensure comprehensive data collection. The first three workshops (duration: 2 hours each) will include residents and care partners from the three partnering LTC homes recruited through purposive sampling. This method involves deliberately selecting participants based on specific characteristics, such as their experience with transitions from LTC to hospital or their involvement in decision-making processes. A minimum of three participants from each partnering LTC home will be recruited for each codesign workshop. Based on past research, we anticipate that this sample size will be sufficient to gather diverse insight for tool development. Participants will be recruited using snowball sampling with the help of participants from objective 1. Objective 3 will be open to participants from phase 1 as well as new participants. We will obtain informed consent prior to participation in codesign workshops. The research team will share findings from objectives 1 and 2 and then facilitate a discussion of resident and care partner preferences and priorities for a tool-based solution to improving resident-centred decision-making around LTC to hospital transitions. Participants will be invited to brainstorm tool design, items included and format guided by the Ottawa Decision Support Framework ( ). This framework aims to improve decisional outcomes by supporting quality decision-making that is informed and values-based and has been instrumental in developing many patient decision aids, measures and training programmes. We will use the Ottawa Decision Support Framework to develop a series of prompts to guide the codesign workshops. This will ensure that we solicit feedback on all the decisional needs identified in the framework. The first round of codesign workshops will be recorded, transcribed, and a member of the research team will take detailed notes on participant preferences, and feedback for a decision-making tool prototype. The research staff will analyse the data from the codesign workshops to identify cross-cutting themes and preferences for a decision-making tool and consolidate a list of key priorities and design criteria based on resident and care partner values, preferences and recommendations. This list will be distributed back to workshop participants for participant validation as a method of increasing trustworthiness. The list will be revised as needed. The second round of codesign workshops will be held with key staff members involved in making decisions about LTC to hospital transitions (eg, personal support workers, nurses and physicians). Staff will be recruited through connections made in objective 1, using informational flyers that provide a description of the study and contact information for our team and snowball sampling. We will recruit three staff members at each home in different care provider roles (n=9). We will facilitate one codesign workshop at each of the three LTC homes where the researchers will share findings from objectives 1 and 2 and the first round of codesign workshops and then facilitate a discussion of staff preferences and priorities for the design, items included and format of a tool-based solution to improving resident-centred decision-making around LTC to hospital transitions. We will use the same process for informed consent, data collection and analysis as in the first round of codesign workshops. In addition, we will pilot test preliminary tool features within a subset of participating LTC homes to gather initial data on their feasibility and acceptability (phase 3). This pilot testing will involve collecting feedback on the tool’s design, functionality and user experience from both residents and staff. These preliminary findings will inform refinements and adjustments to ensure that the final tool is well aligned with the needs and preferences identified in phase 1. Timeline Phase 1 is anticipated to span approximately 6 months, from January 2024 to June 2024. Phase 2 will follow from July 2024 to December 2024. Phase 3 is scheduled from January 2025 to March 2025, with phase 4 occurring from April 2025 to June 2025.
Phase 1 is anticipated to span approximately 6 months, from January 2024 to June 2024. Phase 2 will follow from July 2024 to December 2024. Phase 3 is scheduled from January 2025 to March 2025, with phase 4 occurring from April 2025 to June 2025.
Results from phase 1 will be used to inform the creation of the tool in phase 2 of the larger study. Phase 1 results will also be shared with residents, family/friend care partners and staff at an in-person session at each participating home. The research team will develop infographics describing phase 1 results in English and French and encourage the family, resident and nurse councils at each home to disseminate them at their council meetings. Knowledge translation activities for phase 1 findings also include presentations at conferences (eg, the Canadian Hospice and Palliative Care Association Annual conference, Ontario Long-Term Care Clinicians Annual Conference and Palliative Approach in LTC Community Practice webinars), progress reports to our partners, publications in open-access journals (eg, BMC Geriatrics) and social media posts by the researchers. Later in the larger study, the new decision-making tool will be disseminated through the Long-Term Care Community Practice, a group of professionals, patients and family/friend care partners who share best practices for palliative approaches in LTC. The tool will also be circulated via Advance Care Planning Canada, an initiative dedicated to creating a Pan-Canadian Framework for advance care planning.
This project was approved by the Bruyère Health Research Ethics Board (#M16-23-036) as well as the ethics boards of each three partnering LTC homes. Informed consent will be collected from all participants prior to the start of the interviews and workshops. If a participant requires a proxy to consent on their behalf, ongoing assent will be obtained based on an assessment of how the resident expresses or indicates their preferences verbally, behaviourally or emotionally (eg, smiling, nodding, etc). This assessment will ensure that the participant is able to make a meaningful choice and has at least a minimal level of understanding. Given that 70% of residents in LTC have dementia, we anticipate that there will be involvement from this population in our research study.
10.1136/bmjopen-2024-086748 online supplemental file 1
|
Using artificial intelligence reading label system in diabetic retinopathy grading training of junior ophthalmology residents and medical students | f38b1262-0d96-4a1a-88be-e6ec7ed0866f | 8994224 | Ophthalmology[mh] | This article evaluates the efficiency of using an artificial intelligence reading label system in the diabetic retinopathy grading training of junior ophthalmology resident doctors and medical students. Through reading training, the kappa score of the DR grading was elevated. It showed that the artificial intelligence reading label system was a valuable tool in training resident doctors and medical students in doing diabetic retinopathy grading.
Diabetic retinopathy (DR) is the most common microvascular complication of diabetes and the leading cause of irreversible vision loss in working-age adults . The prevalence of diabetes in China is estimated to be around 10–11% , Thus, China has the largest population of diabetes in the world, creating a high burden of DR. Early diagnosis and treatment of DR can cause timely medical intervention, thus preventing progression of the disease and avoiding the occurrence of severe visual impairment . Therefore, it is crucial to accurately screen and grade the disease. According to the White Paper on Eye Health in China, there are about 44,800 ophthalmologists in China . Among these, qualified specialists in fundus diseases are in a severe short supply. An effective DR screening programme should ensure that screeners and graders are systematically trained and qualified to read DR photos; the duration of this training process usually takes several months. For example, in the case of the UK Gloucestershire Retinal Education Group DR screening project, the total course duration was 40 weeks . If certain methods can reduce the time required for training, it will significantly improve the efficiency of DR screening and will be beneficial for DR prevention and control. In recent years, because of the rapid development of artificial intelligence (AI) techniques, AI techniques based on machine learning play a significant role in DR screening, which acquires high sensitivity and specificity through the learning of a large number of fundus photo training data sets . But the fundus photo training data sets needed manual annotation by qualified specialists, and the AI reading results also needed to be confirmed by retina experts. Thus, to train junior ophthalmologists in DR reading and AI data set annotation, DR reading training is vital for ophthalmology residency training. The purpose of this study is to evaluate the efficiency of using an artificial intelligence reading label system in the diabetic retinopathy grading training of junior ophthalmology resident doctors and medical students.
Reading methods A total of 520 fundus photographs centered on the macular region were included in this study. Photographs were randomly divided into 8 groups, with 65 images for each group. The severity of diabetic retinopathy was graded based on the international clinical diabetic retinopathy severity scale . Photographs of no DR, mild non-proliferative DR (NPDR), moderate NPDR, severe NPDR and proliferative DR (PDR) were included in each group. Three senior consultants made the diagnosis gold standard for each image. Participants were randomly recruited from all first-year ophthalmology residents and medical students entering clinical studies who were interested in this training. Thirteen junior ophthalmology residents and medical students participated in the training. Six of them were first-year residents of the ophthalmology residency training programme at Peking Union Medical College Hospital (PUMCH). Seven of them were medical students at Peking Union Medical College (PUMC). Thirteen participants performed DR reading using the AI reading label system, made DR grading, and labelled the classic lesions of each image. Reading training was performed for 8 rounds with 65 images per round. After each round’s labelling, the participants were gathered to study the diagnosis gold standard. Each round lasted for 1 week, and the whole process lasted for 8 weeks. The sensitivity and specificity according to the diagnosis glod standard were summarized after each round. Grading methods Fundus photographs were divided into 5 levels according to the DR severity degrees. No DR, mild NPDR, moderate NPDR, severe NPDR, or PDR were labelled as degrees 0, 1, 2, 3, or 4, respectively. Degree 0 is defined as ‘without DR’ and degrees 1 to 4 are defined as ‘with DR’. Degrees 0 and 1 are defined as ‘non-referral DR’, while degrees 2 to 4 are defined as ‘referral DR’. Degrees 0 to 2 are defined as ‘non-severe DR’, while degrees 3 and 4 are defined as ‘severe DR’. Introduction to the Reading label system The reading label system was originally developed for manual grading and annotation in training the AI deep learning model. It was a web-based annotation system and provided adaptively enhanced versions of the original images for reference. The readers logged in with their accounts, and the system loaded a certain number of images randomly. After reading the photos and marking the main abnormal lesions, the readers chose a grade, and the system automatically compared the results with the gold standards to calculate sensitivity and specificity. Statistical methods Diagnosis results were collected according to the diagnosis golden gold standard and analyzed statistically using SPSS 25 (IBM, NY, USA). Three diagnosis classifications were set as with/without DR, referral/non-referral DR, and severe/non-severe DR. We calculated the sensitivity and specificity of each classification. Sensitivity was calculated as the number of correctly diagnosed positive examples divided by the total number of positive examples. The specificity was calculated as the number of correctly diagnosed negative examples divided by the total number of negative examples. The harmonic mean of the sensitivity and specificity of each classification was calculated. The kappa score was calculated by combining the harmonic means of the three classifications. Kappa scores of 0.61 to 0.80 were determined to be of significant consistency, while kappa scores above 0.80 were determined to be highly consistent. The discrepancy in kappa scores before and after training was compared to evaluate the effect of DR reading training.
A total of 520 fundus photographs centered on the macular region were included in this study. Photographs were randomly divided into 8 groups, with 65 images for each group. The severity of diabetic retinopathy was graded based on the international clinical diabetic retinopathy severity scale . Photographs of no DR, mild non-proliferative DR (NPDR), moderate NPDR, severe NPDR and proliferative DR (PDR) were included in each group. Three senior consultants made the diagnosis gold standard for each image. Participants were randomly recruited from all first-year ophthalmology residents and medical students entering clinical studies who were interested in this training. Thirteen junior ophthalmology residents and medical students participated in the training. Six of them were first-year residents of the ophthalmology residency training programme at Peking Union Medical College Hospital (PUMCH). Seven of them were medical students at Peking Union Medical College (PUMC). Thirteen participants performed DR reading using the AI reading label system, made DR grading, and labelled the classic lesions of each image. Reading training was performed for 8 rounds with 65 images per round. After each round’s labelling, the participants were gathered to study the diagnosis gold standard. Each round lasted for 1 week, and the whole process lasted for 8 weeks. The sensitivity and specificity according to the diagnosis glod standard were summarized after each round.
Fundus photographs were divided into 5 levels according to the DR severity degrees. No DR, mild NPDR, moderate NPDR, severe NPDR, or PDR were labelled as degrees 0, 1, 2, 3, or 4, respectively. Degree 0 is defined as ‘without DR’ and degrees 1 to 4 are defined as ‘with DR’. Degrees 0 and 1 are defined as ‘non-referral DR’, while degrees 2 to 4 are defined as ‘referral DR’. Degrees 0 to 2 are defined as ‘non-severe DR’, while degrees 3 and 4 are defined as ‘severe DR’.
The reading label system was originally developed for manual grading and annotation in training the AI deep learning model. It was a web-based annotation system and provided adaptively enhanced versions of the original images for reference. The readers logged in with their accounts, and the system loaded a certain number of images randomly. After reading the photos and marking the main abnormal lesions, the readers chose a grade, and the system automatically compared the results with the gold standards to calculate sensitivity and specificity.
Diagnosis results were collected according to the diagnosis golden gold standard and analyzed statistically using SPSS 25 (IBM, NY, USA). Three diagnosis classifications were set as with/without DR, referral/non-referral DR, and severe/non-severe DR. We calculated the sensitivity and specificity of each classification. Sensitivity was calculated as the number of correctly diagnosed positive examples divided by the total number of positive examples. The specificity was calculated as the number of correctly diagnosed negative examples divided by the total number of negative examples. The harmonic mean of the sensitivity and specificity of each classification was calculated. The kappa score was calculated by combining the harmonic means of the three classifications. Kappa scores of 0.61 to 0.80 were determined to be of significant consistency, while kappa scores above 0.80 were determined to be highly consistent. The discrepancy in kappa scores before and after training was compared to evaluate the effect of DR reading training.
Training results for all the participants Thirteen participants were randomly recruited from all first-year ophthalmology residents and medical students entering clinical studies, including three men and ten women. The average age of the participants was 25.54 ± 2.96 years. In the DR reading training, the average harmonic means of each diagnosis classification and the average kappa scores are shown in Table . Through the eight rounds of reading, the average kappa score was elevated from 0.67 to 0.81. The average kappa score of the first 4 rounds was 0.77, which means significant consistency. The average kappa score of the latter 4 rounds was elevated to 0.81, which signifies highly consistent. There has been an escalating trend in diagnostic accuracy. The growth curve of reading training is shown in Fig. . The harmonic mean of with/without DR was elevated from 0.55 to 0.73, and the harmonic mean of referral/non-referral DR was elevated from 0.76 to 0.81. The harmonic mean of severe/non-severe DR was elevated from 0.75 to 0.85. Training results for each group The 13 participants were divided into two groups. Group 1 consisted of junior ophthalmology residents who had basic knowledge of ophthalmology. Group 2 consisted of medical students who did not have any ophthalmology knowledge base. The average kappa score of each group was calculated separately. As shown in Table , after eight rounds of reading, the average kappa score of Group 1 was elevated from 0.71 to 0.76. The average kappa score of Group 2 was elevated from 0.63 to 0.84. Figures and show the growth curves according to the kappa scores of the two groups.
Thirteen participants were randomly recruited from all first-year ophthalmology residents and medical students entering clinical studies, including three men and ten women. The average age of the participants was 25.54 ± 2.96 years. In the DR reading training, the average harmonic means of each diagnosis classification and the average kappa scores are shown in Table . Through the eight rounds of reading, the average kappa score was elevated from 0.67 to 0.81. The average kappa score of the first 4 rounds was 0.77, which means significant consistency. The average kappa score of the latter 4 rounds was elevated to 0.81, which signifies highly consistent. There has been an escalating trend in diagnostic accuracy. The growth curve of reading training is shown in Fig. . The harmonic mean of with/without DR was elevated from 0.55 to 0.73, and the harmonic mean of referral/non-referral DR was elevated from 0.76 to 0.81. The harmonic mean of severe/non-severe DR was elevated from 0.75 to 0.85.
The 13 participants were divided into two groups. Group 1 consisted of junior ophthalmology residents who had basic knowledge of ophthalmology. Group 2 consisted of medical students who did not have any ophthalmology knowledge base. The average kappa score of each group was calculated separately. As shown in Table , after eight rounds of reading, the average kappa score of Group 1 was elevated from 0.71 to 0.76. The average kappa score of Group 2 was elevated from 0.63 to 0.84. Figures and show the growth curves according to the kappa scores of the two groups.
In recent years, AI technology based on classic machine learning (ML) or deep learning (DL) has been widely used in a variety of fundus disease screenings, including DR. Gulshan et al, who used the deep learning algorithm for the screening of DR and obtained extremely high sensitivity and specificity . Takahashi et al. used a modified deep learning algorithm model for the screening and grading of DR, which can obtain grading results similar to those of ophthalmologists . However, even if the application of AI technology in DR screening and grading has achieved very high accuracy, the final results can only be used as a diagnostic reference. Training junior ophthalmologists to grow rapidly and perform DR reading accurately is still an important part of ophthalmologist training. If junior ophthalmologists can master the DR reading method through centralized training quickly, it is not only conducive to the growth of ophthalmologists but also reserves the strength of physicians for labelling AI training sets. Therefore, it is of great significance to find an efficient DR reading training method. There is no previous discussion on the standard method of DR reading training, and there is no literature exploring the use of an AI reading label system for reading training and learning. In this study, the AI reading label system was used for the DR reading training of junior ophthalmology residents and medical students. In this DR reading training, after 8 rounds of reading, the average kappa score of 13 participants increased from 0.67 in the first reading to 0.81 in the eighth reading. The average kappa score of the first four rounds was 0.77, indicating significant agreement, and the average kappa score of the last four rounds was 0.81, indicating that after training, the overall reading accuracy of the participants was significantly improved. The kappa score did not linearly increased each time, which may be because the difficulty level could not be completely consistent with the picture loaded at each time, resulting in the bias of the results. In our previous studies, during the DR reading training of the AI dataset, we calculated the overall kappa scores for doctors of different seniorities. Seventeen attendings and six consultants in the fundus speciality read 20,503 fundus photographs, and the overall kappa scores were 0.67 for attendings and 0.71 for consultants . In our training, the overall kappa score was elevated from 0.67 to 0.81, with a higher score than the attendings and consultants, despite the trainees’ lower levels of professional training. At the same time, the trainees were also divided into two groups for statistics. The first group consisted of junior ophthalmology residents with a certain basic knowledge of ophthalmology who also attended ophthalmology courses and participated in the clinical work of ophthalmology. The second group consisted of medical students who had not learned the basic knowledge of ophthalmology before the start of reading training and who had not participated in the courses and clinical work of ophthalmology after the start of reading. The initial kappa score of the two groups reflected the difference in the knowledge base of the two groups of readers, with an initial kappa score of 0.71 in Group 1 and 0.62 in Group 2, reflecting that the accuracy of the basic reading was higher in Group 1 than in Group 2. As the training progressed, the difference between the two groups gradually narrowed, and the kappa scores increased to 0.76 in Group 1 and 0.84 in Group 2 for the eighth reading, with a more significant increase in the medical student group. The average kappa score of the first four rounds was 0.77, and the average kappa score of the last four rounds was 0.81 in Group 1, 0.71 in the first four rounds and 0.82 in the last four rounds in Group 2, which also reflected that the gap in reading accuracy between the two groups was reduced. After reading training, even medical students without an ophthalmological knowledge base could be familiar with the law of DR reading and achieve a certain diagnostic accuracy. The results of the harmonic mean of with/without DR, referral/non-referral DR, and severe/non-severe DR showed that the harmonic mean of with/without DR was the lowest, and the harmonic mean of referral/non-referral DR and severe/non-severe DR was relatively higher, which may be because it may not be precise for the presence or absence of microhemangioma based on fundus colour photography alone. The small microhemangioma in the picture may be confused with poor-quality artefacts at the time of photography, leading to incorrect conclusions. This also suggests that for the reading training, we should be cautious in selecting the fundus photographs used for the training, try to select the pictures with good quality, and eliminate the possible confounding factors caused by the poor quality of picture shooting. This study also has some limitations. Since the original application of the reading label system used in the training was to train the AI deep learning model, which is not used for the physicians’ reading training, the system cannot immediately give the correct grading answer after labelling and needs to uniformly conduct the retrospective learning of picture grading after each label, which has an effect on the reading learning efficiency. Also, the number of people included in the training was small, and there may be some errors in the statistical mean. To make this system more conducive to reading training, the AI reading result prompt function can be added, and the gold standard is given after each round of picture labelling for comparison, which can increase training efficiency and strengthen the effect of reading training. The residents and medical students who participated in the training had different backgrounds in previous medical education and therefore had different foundations prior to training, which may have influenced the training outcomes to some extent. To reduce the bias caused by this factor, the participants’ basic ophthalmic knowledge needs to be examined, and they should be divided into different groups according to the results before training. In conclusion, the use of an artificial intelligence DR reading label system can effectively improve the DR reading level of junior ophthalmologists and can achieve a certain reading accuracy in a short time with hundreds of images, which is a feasible reading training method.
|
Proteo-Transcriptomic Analysis of the Venom Gland of the Cone Snail | acb8ee92-80de-48ed-9767-032285a62b0b | 11946857 | Biochemistry[mh] | Cone snails are marine mollusks which belong to the large family Conidae. This family groups about a thousand species of gastropod mollusks which possess a venom gland and modified radular teeth . These carnivorous predators feed on fish, mollusks, or worms . They produce and use a complex venom mixture intended to paralyze small prey or deter potential predators for defensive purposes . Conus snail venoms constitute a rich source of new drug leads due to the modulatory actions of their components towards key physiological receptors such as ion channels, G-protein coupled receptors (GPCRs), and other proteins . The small and highly structured peptides called conopeptides (with one or without disulfide bond) or conotoxins (disulfide-rich) are mainly responsible for the observed biological activity of cone snail venoms, although some small molecules and proteins can also contribute to the envenomation process . Most Conus species produce highly variable venoms, the origin of which remains unclear due to the interference of multiple factors, such as age, environment, feeding habit, envenomation strategy, etc. . In fact, intraspecific venom variation is known to occur not only within species but also within a single specimen, and even between consecutive venom injections [ , , , ]. For instance, some cones have demonstrated the ability to purposefully select different assortments of conotoxins following predatory or defense stimuli. The deadly fish-hunter, Gastridium geographus , has displayed chemical and biological differences in predatory and defense venom . This observation was directly correlated with the uneven distribution of toxins across the venom gland. For example, previous studies on G. geographus and Pionoconus striatus have revealed a distinct separation between the nature of conotoxins found in the proximal section of the venom gland, which is closer to the venom bulb, and the distal section of the venom gland, which is nearer to the proboscis, where the venom stings occur . Although common, the diversification of cone venoms is an unpredictable phenomenon that occurs at different levels. Therefore, integrating analytical techniques is essential to accelerate the identification of distinct predatory and defensive conotoxin repertoires . For this purpose, multiple omics approaches, or “venomics”, are combined to provide sequence and structural information. First, venom profiling is performed to record a fingerprint of the injected venom and compare it to the dissected venoms. Next, venom gland transcriptomics allows the recovery of all toxin sequences and their classification into gene superfamilies . Proteomics then permits us to identify the conopeptides in each venom sample to establish correlations between the venom gland and the injected venoms. To date, very few studies have delved into venom gland heterogeneity. Among the cones that have been characterized through venomics are piscivorous species such as Gastridium geographus, Pionoconus striatus , and Pionoconus consors , and vermivorous species such as Vituliconus planorbis and Rhizoconus vexillum . Molluscivorous species remain understudied. In this work, we present the whole venom characterization of the mollusk-hunting cone Cylinder canonicus ( ) . The shells of cones in the Cylinder clade present a “tented” pattern made of white triangles over a brown-to-yellow background. C. canonicus resembles the closely related species C. textile , and to a lesser extent C. episcopatus . Such cone species naturally live in coral reef environments throughout the Indo-Pacific oceans. Within the same clade are also found well-known cones such as Cylinder victoriae and Cylinder ammiralis , for which evidence of distinct predatory and defensive venoms was reported but possible correlation with specific sections of the duct was not investigated . To decipher the origins of the injected venom, we implemented a venomics strategy to the study of the venom of C. canonicus by combining transcriptomics and proteomics analyses. Information on the spatial distribution of conotoxins along the venom gland will provide further clues to help understand the complex venom–ecology relationships in cones.
2.1. Venom Profiling: LC-MS Traces of Predatory-Evoked and Dissected Venoms Overall, predatory-milked venoms collected from the three individuals of C. canonicus displayed similar profiles of intermediate complexity ( ). However, a detailed investigation of the detected components revealed notable differences. Indeed, mass profiling showed important variations, with a total of 159 unique masses detected ( ), but only 12 (8%) being shared between the three individuals, which include 839.3, 2114.8, 2158.8, 1024.4, 1453.5, 1454.5, 2995.2, 2913.1, 2940.0 Da, 2740.1, 3041.2, and 3265.3. The major late eluting peaks (between 23.7–25.0 min) include 2913.1, 2927.1, and 2740.1 Da peptides, and their combined area comprises more than a third of the total summed peak areas (30%). Interestingly, one specimen (MV1) injected only two (2913.1, 2927.1) out of these three major peptides. To determine the origin of the predation-evoked venom, extracts of the venom gland were also analyzed using LC-MS. Clear compartmentalization was observed, with distally- (D and DC) and proximally-dissected venoms (P and PC) presenting drastically different profiles ( A,B). For instance, peaks detected between 23.0–25.0 min in the distal sections were the most intense ions as they represent 40% relative to the total peak area (1024.4 Da, 2913.1 Da, and 2740.1 Da), while being minor in the proximal sections (3%). In contrast, proximal sections displayed major peaks early in the chromatograms, especially at 8.3 min (3139.9 Da), 8.6 min (2940.0 Da), 9.2 min (3122.9 Da), 15.0 min (2046.5 Da), and 15.5 min (1209.4 Da). These ions correspond to 30% of the total area under the curve ( B) compared to being present at only 5% in distal sections. When compared to the predation-evoked venom LC-MS trace, it appears that the distal traces are the most similar. In particular, the late eluting components (second half of the chromatogram) show a good correlation in terms of the intensity of the major ions. In the first half of the chromatogram, however, contributions from the proximal parts are also apparent, although not prominent. In total, 239 unique masses were detected throughout the venom gland sections, of which 46 (~20%) were common to proximal, distal, and predation-evoked venoms ( C). Overall, dissected venoms were more complex than milked venoms, especially the proximal venoms. In an attempt to gain further information on the contribution of each section of the duct to the composition of the predatory venom, mass profiling was performed on each venom sample by MALDI-MS analyses. 2.2. Venom Profiling: MALDI-TOF-MS of Predatory-Evoked and Dissected Venoms Matrix-assisted Laser Desorption/Ionization mass spectrometry (MALDI-MS) was performed in parallel to LC-MS to acquire a mass profile or a fingerprint of the venom samples . The predatory milked venoms of the three C. canonicus specimens, as well as the distal and proximal dissected venoms, revealed great diversity through MALDI-MS ( ). A total of 410 masses in total were detected, of which only 17 (4%) were shared between all samples. Major peaks include peaks at 1453.5, 1739.1, 2093.1, 2112.9, 2990.0, and 2910.0. From the fingerprint profiles ( ), the contribution of each venom duct section to the composition of the predatory venom does not appear as straightforward as the LC-MS results showed. Indeed, although some ions show some clear concordance between distal and predatory venom samples (i.e., 2112.67 Da), other ions are common specifically to proximal and predatory venom samples (i.e., 1739.05 Da). Mass spectrometry is an efficient strategy to evaluate venom complexity; however, it does not provide information on the identity of the toxins. Primary sequence information is necessary to more precisely decipher the nature of the conotoxins used in predation by C. canonicus . This identification will be accomplished through the coupling of proteomic and transcriptomic data. 2.3. Transcriptomics: mRNA Transcripts Expression in the Whole Venom Gland of Cylinder canonicus In total, 98 full conopeptide precursors were extracted from the 252,104 transcripts constituting the entire transcriptome of C. canonicus . Precursors were distributed into 22 gene superfamilies, of which the most diversified were M, O1, O2, and conkunitzins in the venom gland tissues ( ). These major gene superfamilies accounted for more than 50% of all conotoxin precursors. Overall, 14 paralogs were classified into the superfamily M (14%), and 12 paralogs for each of the O1, O2, and conkunitzins gene superfamilies (12%). Other gene superfamilies were annotated with lesser paralogs such as the T gene superfamily (8 paralogs), Con-ikot-ikots (7 paralogs), and P gene superfamily (5 paralogs). Finally, minor gene superfamilies often comprise less than four paralogs, including A, B, D, F, H, I2, I3, S, and U, as well as coninsulins, elevenins, conopressins-conophysins, conorfamides, and prohormones. Sequence alignments of all conopeptide precursors according to gene superfamilies may be found in the ( , ). 2.4. Proteomics: Characterization of the Predatory-Evoked Venoms and Dissected Venom Duct Sections LC-MS and MALDI-MS revealed some levels of diversity in predatory venoms and across the venom gland of C. canonicus specimens. Therefore, prior to performing a bottom-up proteomics approach, all three predatory venom samples were pooled (to minimize the intraspecific variations), as well as the D/DC and P/PC sections together (as LC-MS revealed similar composition). Thus, all three venom samples (predatory, D, and P sections) were analyzed by LC-MS/MS, and the resulting fragmentation spectra were processed with bioinformatic tools to match sequences in the database (corresponding to the whole venom gland transcriptome of C. canonicus ). Consequently, a total of 466 unique transcriptomic sequences were matched and identified from all venom samples. Out of the 466 sequences, more than 85% were found in the distal section of the venom gland (399 sequences), compared to 45% in the proximal section (209 sequences) and 12% in the pooled predatory venoms (56 sequences) ( A). As expected from the LC-MS profiles, distal and proximal venom gland sections both display a higher complexity than predatory-pooled milking venoms. Each transcriptomic sequence that was identified was then run through a similarity-based protein BLAST search to identify protein functions or the gene superfamily for the conotoxins. In total, 354 proteins were non-related to the venom function, also qualified as housekeeping proteins (76%), 84 proteins were venom-related proteins (18%), and finally, 25 sequences (5%) were identified as conotoxins precursors ( B, , ). Housekeeping proteins generally include proteins involved in cell maintenance, constantly required by the cell, such as GAPDH, ribosomal proteins, Ras protein, histones, and elongation factor. They were logically detected in dissected venom gland sections ( , ). However, venom-related proteins were found in both milked and dissected venoms. They include proteins that participate in venom production and delivery such as protein disulfide isomerase, carboxypeptidase E, von Willebrand factor A-like, or transmembrane protease serine 2-like ( , ). An extensive list of the entire proteome with annotations may be found in the Supplementary Information ( , ). About 5% of the entire bottom-up proteomic data were matched to 25 sequences of conotoxins identified in the milked and dissected venoms. They were distributed across nine gene superfamilies, including O1, T, O2, M, I1, E, S, P, and A, in order of abundance ( C). Gene superfamilies with the highest amounts of paralog identified were O1 (six peptides), T (five peptides), M, and O2 (four peptides). Fifteen of these sequences had already been identified by the previous transcriptomics analysis. They correspond to one A-conotoxin (Can01), four M-conotoxins (Can21, Can22, and two paralogs of Can28), three O1-conotoxins (Can33-Can35), three O2-conotoxins (Can45, Can50, and Can53), one P-conotoxin (Can55), one S-conotoxin (Can61), and three T-conotoxins (three paralogs of Can62) ( ). Remarkably, proteomics analyses have allowed us to uncover 10 additional conotoxins that were not present in our annotated transcriptome: four O1- (Can102, Can103, Can104, and Can105), two T- (Can107, Can108), two I1- (Can100 and Can101), one O2- (Can106), and one E-conotoxin (Can99) ( ). The main reason for this discrepancy lies in our stringent criteria for the selection of our transcriptomic sequences: full precursors containing the signal peptide, starting with a methionine and ending with a stop codon. Indeed, among the 10 “missed” sequences, 9 were incomplete precursors (either missing the N- or C-terminal part and therefore rejected in our bioinformatic analysis). Yet, this result is highly significant given that 40% of the proteomically validated conotoxin sequences were not initially listed in our retrieved precursors, showing the importance of combining both transcriptomics and proteomics for the most complete venom coverage. For instance, the conotoxin sequence of O1-Can102 (full contig sequence is 176 amino acids long) demonstrated the largest amount of peptide-spectrum matches (PSM) within predatory and distal venoms ( A,B), and remained also major in the proximal venoms. In addition to the validation of transcriptomic sequences, proteomics uncovers a complementary layer of diversity in the form of post-translational modifications (PTMs) and other modifications involved during the gene-RNA-protein translation processes. PTMs and other modifications that are involved in venom production have been assumed to play an important role in improving peptide activity and/or stability . The most common PTMs identified in the proteomics of all samples, outside of the carbamidomethylation (due to sample preparation), include hydroxylation and carboxylation ( C). Furthermore, deamidation of asparagine (Asn) and glutamine residues (Gln) was also frequently detected ( C). Additionally, a diversity of cysteine frameworks is observed in the conotoxins identified through proteo-transcriptomics ( , , ). First, the VI/VII cysteine framework (C-C-CC-C-C) was the most represented in all conotoxins, especially in the gene superfamily O (O1, O2, and O3). This arrangement was also identified in a few M- and U-conotoxins. The framework IX was found in a few O1-, conkunitzins, and P gene superfamilies. Additionally, a few frameworks were only encountered in specific gene superfamilies such as framework III (CC-C-C-CC) in M-conotoxins, framework V (CC-CC) only in T-conotoxins, framework VIII (C-C-C-C-C-C-C-C-C-C) in S-conotoxins, framework IX (C-C-C-C-C-C) in P-conotoxins, or even framework XI (C-C-CC-CC-C-C) only in I-conotoxins (I2, I3). Many other conotoxins displayed no cysteine or the classification is unknown. A few other sequences, especially those that were identified through proteomics, were missing a fragment of the mature peptide, making identification difficult, such as the conotoxins I1-Can101, O1-Can103 and O1-Can104 ( ) ( , ).
Overall, predatory-milked venoms collected from the three individuals of C. canonicus displayed similar profiles of intermediate complexity ( ). However, a detailed investigation of the detected components revealed notable differences. Indeed, mass profiling showed important variations, with a total of 159 unique masses detected ( ), but only 12 (8%) being shared between the three individuals, which include 839.3, 2114.8, 2158.8, 1024.4, 1453.5, 1454.5, 2995.2, 2913.1, 2940.0 Da, 2740.1, 3041.2, and 3265.3. The major late eluting peaks (between 23.7–25.0 min) include 2913.1, 2927.1, and 2740.1 Da peptides, and their combined area comprises more than a third of the total summed peak areas (30%). Interestingly, one specimen (MV1) injected only two (2913.1, 2927.1) out of these three major peptides. To determine the origin of the predation-evoked venom, extracts of the venom gland were also analyzed using LC-MS. Clear compartmentalization was observed, with distally- (D and DC) and proximally-dissected venoms (P and PC) presenting drastically different profiles ( A,B). For instance, peaks detected between 23.0–25.0 min in the distal sections were the most intense ions as they represent 40% relative to the total peak area (1024.4 Da, 2913.1 Da, and 2740.1 Da), while being minor in the proximal sections (3%). In contrast, proximal sections displayed major peaks early in the chromatograms, especially at 8.3 min (3139.9 Da), 8.6 min (2940.0 Da), 9.2 min (3122.9 Da), 15.0 min (2046.5 Da), and 15.5 min (1209.4 Da). These ions correspond to 30% of the total area under the curve ( B) compared to being present at only 5% in distal sections. When compared to the predation-evoked venom LC-MS trace, it appears that the distal traces are the most similar. In particular, the late eluting components (second half of the chromatogram) show a good correlation in terms of the intensity of the major ions. In the first half of the chromatogram, however, contributions from the proximal parts are also apparent, although not prominent. In total, 239 unique masses were detected throughout the venom gland sections, of which 46 (~20%) were common to proximal, distal, and predation-evoked venoms ( C). Overall, dissected venoms were more complex than milked venoms, especially the proximal venoms. In an attempt to gain further information on the contribution of each section of the duct to the composition of the predatory venom, mass profiling was performed on each venom sample by MALDI-MS analyses.
Matrix-assisted Laser Desorption/Ionization mass spectrometry (MALDI-MS) was performed in parallel to LC-MS to acquire a mass profile or a fingerprint of the venom samples . The predatory milked venoms of the three C. canonicus specimens, as well as the distal and proximal dissected venoms, revealed great diversity through MALDI-MS ( ). A total of 410 masses in total were detected, of which only 17 (4%) were shared between all samples. Major peaks include peaks at 1453.5, 1739.1, 2093.1, 2112.9, 2990.0, and 2910.0. From the fingerprint profiles ( ), the contribution of each venom duct section to the composition of the predatory venom does not appear as straightforward as the LC-MS results showed. Indeed, although some ions show some clear concordance between distal and predatory venom samples (i.e., 2112.67 Da), other ions are common specifically to proximal and predatory venom samples (i.e., 1739.05 Da). Mass spectrometry is an efficient strategy to evaluate venom complexity; however, it does not provide information on the identity of the toxins. Primary sequence information is necessary to more precisely decipher the nature of the conotoxins used in predation by C. canonicus . This identification will be accomplished through the coupling of proteomic and transcriptomic data.
In total, 98 full conopeptide precursors were extracted from the 252,104 transcripts constituting the entire transcriptome of C. canonicus . Precursors were distributed into 22 gene superfamilies, of which the most diversified were M, O1, O2, and conkunitzins in the venom gland tissues ( ). These major gene superfamilies accounted for more than 50% of all conotoxin precursors. Overall, 14 paralogs were classified into the superfamily M (14%), and 12 paralogs for each of the O1, O2, and conkunitzins gene superfamilies (12%). Other gene superfamilies were annotated with lesser paralogs such as the T gene superfamily (8 paralogs), Con-ikot-ikots (7 paralogs), and P gene superfamily (5 paralogs). Finally, minor gene superfamilies often comprise less than four paralogs, including A, B, D, F, H, I2, I3, S, and U, as well as coninsulins, elevenins, conopressins-conophysins, conorfamides, and prohormones. Sequence alignments of all conopeptide precursors according to gene superfamilies may be found in the ( , ).
LC-MS and MALDI-MS revealed some levels of diversity in predatory venoms and across the venom gland of C. canonicus specimens. Therefore, prior to performing a bottom-up proteomics approach, all three predatory venom samples were pooled (to minimize the intraspecific variations), as well as the D/DC and P/PC sections together (as LC-MS revealed similar composition). Thus, all three venom samples (predatory, D, and P sections) were analyzed by LC-MS/MS, and the resulting fragmentation spectra were processed with bioinformatic tools to match sequences in the database (corresponding to the whole venom gland transcriptome of C. canonicus ). Consequently, a total of 466 unique transcriptomic sequences were matched and identified from all venom samples. Out of the 466 sequences, more than 85% were found in the distal section of the venom gland (399 sequences), compared to 45% in the proximal section (209 sequences) and 12% in the pooled predatory venoms (56 sequences) ( A). As expected from the LC-MS profiles, distal and proximal venom gland sections both display a higher complexity than predatory-pooled milking venoms. Each transcriptomic sequence that was identified was then run through a similarity-based protein BLAST search to identify protein functions or the gene superfamily for the conotoxins. In total, 354 proteins were non-related to the venom function, also qualified as housekeeping proteins (76%), 84 proteins were venom-related proteins (18%), and finally, 25 sequences (5%) were identified as conotoxins precursors ( B, , ). Housekeeping proteins generally include proteins involved in cell maintenance, constantly required by the cell, such as GAPDH, ribosomal proteins, Ras protein, histones, and elongation factor. They were logically detected in dissected venom gland sections ( , ). However, venom-related proteins were found in both milked and dissected venoms. They include proteins that participate in venom production and delivery such as protein disulfide isomerase, carboxypeptidase E, von Willebrand factor A-like, or transmembrane protease serine 2-like ( , ). An extensive list of the entire proteome with annotations may be found in the Supplementary Information ( , ). About 5% of the entire bottom-up proteomic data were matched to 25 sequences of conotoxins identified in the milked and dissected venoms. They were distributed across nine gene superfamilies, including O1, T, O2, M, I1, E, S, P, and A, in order of abundance ( C). Gene superfamilies with the highest amounts of paralog identified were O1 (six peptides), T (five peptides), M, and O2 (four peptides). Fifteen of these sequences had already been identified by the previous transcriptomics analysis. They correspond to one A-conotoxin (Can01), four M-conotoxins (Can21, Can22, and two paralogs of Can28), three O1-conotoxins (Can33-Can35), three O2-conotoxins (Can45, Can50, and Can53), one P-conotoxin (Can55), one S-conotoxin (Can61), and three T-conotoxins (three paralogs of Can62) ( ). Remarkably, proteomics analyses have allowed us to uncover 10 additional conotoxins that were not present in our annotated transcriptome: four O1- (Can102, Can103, Can104, and Can105), two T- (Can107, Can108), two I1- (Can100 and Can101), one O2- (Can106), and one E-conotoxin (Can99) ( ). The main reason for this discrepancy lies in our stringent criteria for the selection of our transcriptomic sequences: full precursors containing the signal peptide, starting with a methionine and ending with a stop codon. Indeed, among the 10 “missed” sequences, 9 were incomplete precursors (either missing the N- or C-terminal part and therefore rejected in our bioinformatic analysis). Yet, this result is highly significant given that 40% of the proteomically validated conotoxin sequences were not initially listed in our retrieved precursors, showing the importance of combining both transcriptomics and proteomics for the most complete venom coverage. For instance, the conotoxin sequence of O1-Can102 (full contig sequence is 176 amino acids long) demonstrated the largest amount of peptide-spectrum matches (PSM) within predatory and distal venoms ( A,B), and remained also major in the proximal venoms. In addition to the validation of transcriptomic sequences, proteomics uncovers a complementary layer of diversity in the form of post-translational modifications (PTMs) and other modifications involved during the gene-RNA-protein translation processes. PTMs and other modifications that are involved in venom production have been assumed to play an important role in improving peptide activity and/or stability . The most common PTMs identified in the proteomics of all samples, outside of the carbamidomethylation (due to sample preparation), include hydroxylation and carboxylation ( C). Furthermore, deamidation of asparagine (Asn) and glutamine residues (Gln) was also frequently detected ( C). Additionally, a diversity of cysteine frameworks is observed in the conotoxins identified through proteo-transcriptomics ( , , ). First, the VI/VII cysteine framework (C-C-CC-C-C) was the most represented in all conotoxins, especially in the gene superfamily O (O1, O2, and O3). This arrangement was also identified in a few M- and U-conotoxins. The framework IX was found in a few O1-, conkunitzins, and P gene superfamilies. Additionally, a few frameworks were only encountered in specific gene superfamilies such as framework III (CC-C-C-CC) in M-conotoxins, framework V (CC-CC) only in T-conotoxins, framework VIII (C-C-C-C-C-C-C-C-C-C) in S-conotoxins, framework IX (C-C-C-C-C-C) in P-conotoxins, or even framework XI (C-C-CC-CC-C-C) only in I-conotoxins (I2, I3). Many other conotoxins displayed no cysteine or the classification is unknown. A few other sequences, especially those that were identified through proteomics, were missing a fragment of the mature peptide, making identification difficult, such as the conotoxins I1-Can101, O1-Can103 and O1-Can104 ( ) ( , ).
3.1. Characterization of Cylinder Canonicus’ Venom Through a Venomics Approach In this study, the venom composition of the molluscivorous C. canonicus was extensively uncovered through a venomics approach. The composition of the predatory venom and venom gland sections was analyzed by mass spectrometry, transcriptomics, and proteomics. First, mass spectrometry-based analyses demonstrated the complex and diverse nature of C. canonicus venom. Both MS techniques have allowed us to detect around 400 different masses, with a total of 399 masses through LC-ESI-MS and 410 through MALDI-MS. LC-ESI-MS chromatograms revealed comparable profiles for predatory venoms collected from different specimens of C. canonicus ( ). Although the general profile is preserved, we noted different levels of complexity between the specimens. These results are consistent with previous works, which have demonstrated intraspecific variation between specimens of the same Conus species, such as C. consors , C. striatus, C. catus , and C. geographus . In particular, the predatory venom of specimen 4 was more complex than that of specimens 2 and 1. Yet, the major components (i.e., 2158.8, 2913.1, 2927.2, and 2740.1 Da) were conserved among all three specimens. This suggests that this species selects a very precise cocktail of toxins for prey capture. 3.2. Effect of the Instrumentation Chosen for Venom Profiling Differences in instrumentation and chemistry between LC-ESI-MS and MALDI-TOF MS allow the profiling of different venom components. Indeed, previous venomics studies have revealed the complementarity of integrating different approaches for venom profiling . Overall, the high diversity between predatory milked and dissected venoms was confirmed again by MALDI-MS. Ion detection performed in reflectron mode allows increased resolution of peaks; however, the sensitivity is inversely affected . Contrary to LC-MS analyses, sample mixtures are not separated prior to MALDI-MS. This is associated with the ion suppression effect, caused by different parameters such as the nature of the matrix. As a result, major components might be detected to the disadvantage of minor components, which become undistinguishable from noise . For this reason, LC-MS results were more revealing in the context of the predatory venom’s origin in the venom gland. 3.3. Compartmentalization of the Venom Gland vs. Predatory Venom The profiling of the venom gland of C. canonicus demonstrates a compartmentalization of venom content across the duct. Indeed, two separate profiles are observed between the distal and proximal sections of the venom duct. This compartmentalization has been suggested to be the underlying mechanism behind the evolved ability of some species to inject different predatory and defensive venoms, as proposed in previous studies [ , , ]. Overall, the separation between each section of the gland seems consistent with the possibility of two distinct venom compositions arising from distal and proximal parts. Yet, some components show a continuous distribution across the duct but with varying intensities. When LC-MS traces are compared, the origin of the predatory venom appears to be largely contributed by the distal part. However, closer investigation shows that the predatory venom contains a blend of mass peaks from the entire venom gland. If the intensity of the ions could be considered, then the resemblance between distal and predatory venoms would become more apparent. Therefore, the comparison between the venom gland and the milked venoms should be assessed relative to the amount of each component. From our results, the venom injected for prey capture appears selected in the venom gland portion that is closer to the proboscis (distal section). To better understand the prey capture strategy of C. canonicus , the biological and structural characterization of these predatory conotoxins on prey tissues will be the focus of further studies. 3.4. Venom Characterization Through Transcriptomics and Proteomics Combined proteo-transcriptomics successfully elucidated 550 protein sequences, including 354 housekeeping proteins, 84 venom-related proteins, and 108 conotoxin precursors. “Housekeeping” proteins are considered to be ubiquitous, i.e., present in every cell tissue. They include proteins required for cell maintenance such as RNA translation (ribosomal proteins), degradation (ubiquitin chains), folding (chaperones), cellular structure (actin-like), or transport (vesicle-trafficking). They are expressed consistently, in a measured amount to allow proper cell functioning . Therefore, their high expression in the venom gland over the milked venom is coherent. Unlike “housekeeping” proteins, venom-related proteins regroup various types of proteins that are involved in venomous functions. These proteins were detected both in the venom gland but also in the predatory venom of C. canonicus . They include various enzymes and proteins that allow the venom gland to produce and mature venom precursors (conotoxin/conopeptide precursors) but also respond to signals that initiate venom release and delivery into prey . We identified for example an angiotensin-converting enzyme-like protein (ACE), a von Willebrand factor A-like protein (VWF), other metalloproteinases, carboxypeptidases, protein disulfide isomerases (PDI), puromycin-sensitive aminopeptidase-like (metal-lopeptidase involved in protein degradation) , polyubiquitin-like proteins (in-volved in degradation and recycling of proteins). For instance, the ACE is known to catalyze the production of angiotensin II, involved in blood pressure regulation, but its implication in predation is unknown . The VWF-like protein was also previously identified in another cone of the same subclade, C. ammiralis, but its function has not established yet. It is known to interfere with platelet aggregation ( ). 3.5. Identification of the Major Constituents of the Predatory-Evoked Venom Through our combined proteo-transcriptomic study, 16 conotoxin sequences were identified in the predatory venom of C. canonicus ( , ). Most represented by far is the superfamily O1, with five conotoxin sequences detected. This is highly relevant to the prey capture strategy, given that related peptides found in the venom of C. textile were found to produce potent effects in mollusks. In particular, King-Kong, TxIA, or TxVIA (the same peptide identified by several groups, hence different names) was found to be the most potent mollusk-specific toxin in C. textile venom, and it resembles the sequence Can033 . Indeed, the most intense ion detected in the LC-MS of the predatory venom of C. canonicus corresponds to the mature conotoxin of sequence Can33. The next most intense ion is Can102, which also belongs to the same family. Based on the strong contractive effect on limpet foot muscles found for TxVIA, we can expect that similar activity of Can33 and Can102 would rapidly lead to the incapacitation of the prey . Three O2 sequences (Can45-Can50-Can53) were also identified, with one (Can50) resembling Gla(3)-TxVI, another peptide identified in C. textile venom. Unfortunately, the biological activity of the O2 superfamily of conotoxins on prey-specific tissues is unknown; hence, we cannot hypothesize their possible role in prey capture. Similarly, no specific role can be attributed to the 3 T superfamily conotoxins found in the predatory venom of C. canonicus , as biological characterization is also lacking for this family, as well as for the single M, A, S, P, and E conotoxins found. However, the presence of these conotoxins strongly suggests an essential ecological role in predation, and further physiological and pharmacological characterization of mollusk tissues/receptors is warranted.
In this study, the venom composition of the molluscivorous C. canonicus was extensively uncovered through a venomics approach. The composition of the predatory venom and venom gland sections was analyzed by mass spectrometry, transcriptomics, and proteomics. First, mass spectrometry-based analyses demonstrated the complex and diverse nature of C. canonicus venom. Both MS techniques have allowed us to detect around 400 different masses, with a total of 399 masses through LC-ESI-MS and 410 through MALDI-MS. LC-ESI-MS chromatograms revealed comparable profiles for predatory venoms collected from different specimens of C. canonicus ( ). Although the general profile is preserved, we noted different levels of complexity between the specimens. These results are consistent with previous works, which have demonstrated intraspecific variation between specimens of the same Conus species, such as C. consors , C. striatus, C. catus , and C. geographus . In particular, the predatory venom of specimen 4 was more complex than that of specimens 2 and 1. Yet, the major components (i.e., 2158.8, 2913.1, 2927.2, and 2740.1 Da) were conserved among all three specimens. This suggests that this species selects a very precise cocktail of toxins for prey capture.
Differences in instrumentation and chemistry between LC-ESI-MS and MALDI-TOF MS allow the profiling of different venom components. Indeed, previous venomics studies have revealed the complementarity of integrating different approaches for venom profiling . Overall, the high diversity between predatory milked and dissected venoms was confirmed again by MALDI-MS. Ion detection performed in reflectron mode allows increased resolution of peaks; however, the sensitivity is inversely affected . Contrary to LC-MS analyses, sample mixtures are not separated prior to MALDI-MS. This is associated with the ion suppression effect, caused by different parameters such as the nature of the matrix. As a result, major components might be detected to the disadvantage of minor components, which become undistinguishable from noise . For this reason, LC-MS results were more revealing in the context of the predatory venom’s origin in the venom gland.
The profiling of the venom gland of C. canonicus demonstrates a compartmentalization of venom content across the duct. Indeed, two separate profiles are observed between the distal and proximal sections of the venom duct. This compartmentalization has been suggested to be the underlying mechanism behind the evolved ability of some species to inject different predatory and defensive venoms, as proposed in previous studies [ , , ]. Overall, the separation between each section of the gland seems consistent with the possibility of two distinct venom compositions arising from distal and proximal parts. Yet, some components show a continuous distribution across the duct but with varying intensities. When LC-MS traces are compared, the origin of the predatory venom appears to be largely contributed by the distal part. However, closer investigation shows that the predatory venom contains a blend of mass peaks from the entire venom gland. If the intensity of the ions could be considered, then the resemblance between distal and predatory venoms would become more apparent. Therefore, the comparison between the venom gland and the milked venoms should be assessed relative to the amount of each component. From our results, the venom injected for prey capture appears selected in the venom gland portion that is closer to the proboscis (distal section). To better understand the prey capture strategy of C. canonicus , the biological and structural characterization of these predatory conotoxins on prey tissues will be the focus of further studies.
Combined proteo-transcriptomics successfully elucidated 550 protein sequences, including 354 housekeeping proteins, 84 venom-related proteins, and 108 conotoxin precursors. “Housekeeping” proteins are considered to be ubiquitous, i.e., present in every cell tissue. They include proteins required for cell maintenance such as RNA translation (ribosomal proteins), degradation (ubiquitin chains), folding (chaperones), cellular structure (actin-like), or transport (vesicle-trafficking). They are expressed consistently, in a measured amount to allow proper cell functioning . Therefore, their high expression in the venom gland over the milked venom is coherent. Unlike “housekeeping” proteins, venom-related proteins regroup various types of proteins that are involved in venomous functions. These proteins were detected both in the venom gland but also in the predatory venom of C. canonicus . They include various enzymes and proteins that allow the venom gland to produce and mature venom precursors (conotoxin/conopeptide precursors) but also respond to signals that initiate venom release and delivery into prey . We identified for example an angiotensin-converting enzyme-like protein (ACE), a von Willebrand factor A-like protein (VWF), other metalloproteinases, carboxypeptidases, protein disulfide isomerases (PDI), puromycin-sensitive aminopeptidase-like (metal-lopeptidase involved in protein degradation) , polyubiquitin-like proteins (in-volved in degradation and recycling of proteins). For instance, the ACE is known to catalyze the production of angiotensin II, involved in blood pressure regulation, but its implication in predation is unknown . The VWF-like protein was also previously identified in another cone of the same subclade, C. ammiralis, but its function has not established yet. It is known to interfere with platelet aggregation ( ).
Through our combined proteo-transcriptomic study, 16 conotoxin sequences were identified in the predatory venom of C. canonicus ( , ). Most represented by far is the superfamily O1, with five conotoxin sequences detected. This is highly relevant to the prey capture strategy, given that related peptides found in the venom of C. textile were found to produce potent effects in mollusks. In particular, King-Kong, TxIA, or TxVIA (the same peptide identified by several groups, hence different names) was found to be the most potent mollusk-specific toxin in C. textile venom, and it resembles the sequence Can033 . Indeed, the most intense ion detected in the LC-MS of the predatory venom of C. canonicus corresponds to the mature conotoxin of sequence Can33. The next most intense ion is Can102, which also belongs to the same family. Based on the strong contractive effect on limpet foot muscles found for TxVIA, we can expect that similar activity of Can33 and Can102 would rapidly lead to the incapacitation of the prey . Three O2 sequences (Can45-Can50-Can53) were also identified, with one (Can50) resembling Gla(3)-TxVI, another peptide identified in C. textile venom. Unfortunately, the biological activity of the O2 superfamily of conotoxins on prey-specific tissues is unknown; hence, we cannot hypothesize their possible role in prey capture. Similarly, no specific role can be attributed to the 3 T superfamily conotoxins found in the predatory venom of C. canonicus , as biological characterization is also lacking for this family, as well as for the single M, A, S, P, and E conotoxins found. However, the presence of these conotoxins strongly suggests an essential ecological role in predation, and further physiological and pharmacological characterization of mollusk tissues/receptors is warranted.
To conclude, this work presents a comprehensive characterization of the venom produced by Cylinder canonicus through proteo-transcriptomics analyses. We identify 108 conotoxins belonging to 24 gene superfamilies, especially M, O1, O2, conkunitzin, and T. In correlation with this, we have characterized some of the major components of the predatory venoms as being O1-conotoxins. Our results also suggest that C. canonicus selects the conotoxins to be injected into their predatory venom, mostly in the distal part of the venom gland. We also demonstrate compartmentalization of the venom gland, which correlates to previous works, and the ability to produce two types of venom (predatory and defensive). This is a valuable contribution as it provides new transcriptome and proteome data that can be useful for a better understanding of cone snail venom ecology. In further work, we will attempt to obtain defensive venom to complete the study and better understand the purpose of the compartmentalization of the venom gland.
5.1. Cone Snail Collection and Venom Extraction Specimens of Cylinder canonicus were collected in February 2023 from the barrier reef surrounding the Island of Mayotte in the Mozambique channel under national and local permits/authorizations. Cone snails were then acclimated and maintained in marine aquariums in the laboratory. Specimens of C. canonicus were fed weekly or fortnightly locally collected gastropod mollusks of the family Nassariidae ( Tritia sp. ). Milked venom was obtained from three different C. canonicus specimens using previously described methods and venom gland extracts (sections) and tissues (RNA) were retrieved from two other specimens. Predatory-evoked milking venoms were obtained by luring cones with the live mollusk prey. Once the cone extends its yellow proboscis out, we intercept the sting with a collecting Eppendorf tube, lined with a fine slice of another prey’s foot over a parafilm. Milking venoms were translucid with a small white precipitate at the bottom of the tube. A whole venom gland was used for mRNA extraction, and a second venom gland was dissected into four sections (distal, distal-central, proximal-central, and proximal) before extracting the venom content for mass spectrometry and proteomic analyses. Finally, predatory-evoked venoms were milked from three specimens of C. canonicus (MV1, MV2 and MV4). Predatory behavior was initiated by facing a Nassariidae mollusk to each C. canonicus specimen and intercepting the injection with an Eppendorf tube covered with parafilm and a thin layer of prey mollusk foot. Milked samples were lyophilized and stored at −20 °C before use. 5.2. RNA Extraction and Sequencing One Cylinder canonicus specimen was dissected on ice to extract the venom gland in the form of a long and thin duct. The venom duct was placed in 1 mL of TRIzol reagent (Invitrogen, Carlsbad, CA, USA) in an Eppendorf tube and the purified total ribonucleic acid (RNA) was extracted by following the manufacturer’s instructions . Next, mRNA was purified from the retrieved total RNA using the “Oligotex mRNA Mini Kit” (Qiagen, Valencia, CA, USA). Finally, mRNA extracts were submitted to Montpellier GenomiX (MGX, BioCampus Montpellier, CNRS) for sequencing. Complementary DNA (cDNA) libraries were constructed and sequenced using a high-throughput Illumina sequencer (Illumina Inc., San Diego, CA, USA) with a TruSeq Stranded mRNA Sample Prep Kit and a TruSeq Library indices following the manufacturer’s instructions. Paired-end sequencing (2 × 150 bp) yielded several millions of short-read sequences after filtering poor-quality reads. cDNA short raw reads were controlled for quality using FastQC v.0.11.9 ( http://www.bioinformatics.babraham.ac.uk/projects/fastqc ), trimmed, then assembled into longer contigs using Trinity v2.13.2. The assembled contigs were then translated into amino acids (six reading frames) in silico . 5.3. Transcriptome Annotation Conotoxin sequences were then retrieved using, in parallel, a manual and automatic approach. First, a locally-built transcriptomic platform (VenUM) was used to manually extract conopeptide precursors from the venom gland transcriptome of Cylinder canonicus . Registered conopeptide precursors found on ConoServer ( www.conoserver.org ) from species of the same clade, such as C. victoriae and C. textile , were used as queries to search against the transcriptome of C. canonicus to extract and annotate conopeptide precursors. Additionally, the ConoPrec tool from ConoServer was employed to identify cysteine frameworks, cleavage sites, and mature sequences . For further validation of conopeptide superfamilies and other protein annotations, protein BLAST searches were conducted using BLAST+ v2.16.0+ and the non-redundant protein sequences and Swiss-Prot r2024_03 databases . To classify gene superfamilies automatically, the bioinformatic tool ConoDictor v2.4.1 was utilized ( https://github.com/koualab/conodictor ) . This tool enables sequence classification and predicts superfamily assignments . Finally, multiple conopeptide sequence alignments were generated using the open-source software Jalview v.2.11.4.0 ( www.jalview.org ) and ClustalW v2.1 with default parameters ( www.clustal.org/clustal2/ ) . 5.4. Mass Spectrometry (MS) LC-ESI-MS. Crude lyophilized venoms were resuspended in water and concentrations were evaluated by a NanoPhotometer ® (Implen GmbH, Munich, Germany). Electrospray ionization mass spectrometry coupled with liquid chromatography (LC-ESI-MS) experiments were performed following previous protocols . Separation by liquid chromatography was performed on an Acquity H-Class ultrahigh-performance liquid chromatography (UPLC) system. Around 20 µg of each venom sample were injected onto a Kinetex C 18 100 Å column (2.1 mm × 150 mm, 3 μm) (Phenomenex, Torrance, CA, USA) fitted with a precolumn for separation at 0.4 mL/min in a 0–80% gradient of a solvent B during 60 min (A: water (H 2 O) + 0.1% formic acid (FA); B: acetonitrile (ACN) + 0.1% formic acid). ESI-MS analyses were performed on a Waters Synapt G2-S (Waters Corporation, Milford, MA, USA) equipped with a time of flight (TOF) detection cellule and was used in positive mode across a masse range of 1000–3000 Dalton (Da) and scan time of 1 s. Each sample was infused in the ionization source at a flow rate of 5 µL/min. Then, ionization was performed with a capillary tension of 3 kV. Temperatures were set at 120 °C for the source and 350 °C for the desolvation gas. Gas flows were regulated at 850 L/h for the desolvation gas and at 6 L/h for the nebulizing gas. Total ion current (TIC) chromatograms and molecular masses were processed with Mass Lynx software (version 4.1, Waters, Corp., Milford, MA, USA). MALDI-TOF-MS. MALDI-TOF mass spectra were acquired using a RapifleX ® MALDI mass spectrometer (Bruker Daltonics, Billerica, MA, USA). All samples were diluted at 0.10 mg/mL. In parallel, a solution of α-cyano-4-hydroxycinnamic acid (HCCA, ACROS Organics, USA) matrix was saturated at 10 mg/mL by dissolving 10.0 mg of HCCA in 1 mL of 30% ACN in aqueous 0.1% TFA (70:30:0.1, ACN/H 2 O/TFA). Samples were spotted on an MTP 384 polished steel plate (Bruker Daltonics, Billerica, MA, USA) according to the dried-droplet spotting technique which consists of mixing 1 µL of the sample with 1 µL of HCCA matrix (α-Cyano-4-hydroxycinnamic acid) and spotting 1 µL of each sample in two different spots. External calibration was performed using a mixture of standard peptides (Bradykinin 757.399 Da, Angiotensin II 1046.542 Da, Angiotensin I 1296.685 Da, Substance P 1347.735 Da, Bombesin 1619.822 Da, ACTH clip 2093.086 Da, ACTH clip 2465.198 Da, Somatostatin 3147.471 Da). Mass spectrometry spectra were acquired in positive ion mode at a frequency of 5000 Hz, by applying 6000 shots/per sample spot (4 × 1500 shots) with a laser power of 30%. Ion detection was performed on a time of flight (TOF) detector in reflectron mode with a mass range of 500–5000 Da. FlexControl 3.0 software was used for data acquisition and FlexAnalysis 4.0 (Bruker Corporation, Billerica, MA, USA) for data treatment. 5.5. Proteomics The milked predatory venom of three specimens of Cylinder canonicus , as well as dissected venoms of distal (D and DC) and proximal (PC and P) sections of the venom gland were prepared prior to proteomic analysis as described . A total of 50 µg of pooled milked (all 3 predatory venoms), distal (D + DC), and proximal (PC + P) venoms were first diluted in 89 µL of triethylammonium bicarbonate (TEAB) 100 mM under stirring during 30 min at room temperature. Disulfide bonds were reduced with dithiothreitol (DTT) at 10 mM for 30 min at 60 °C and alkylated with iodoacetamide (IAA) at 50 mM before 30 min incubation in the dark at room temperature. The venom samples were then enzymatically digested with 1.5 µg Trypsin (Gold, Promega, Madison, WI, USA) and incubated overnight at 30 °C. Samples were desalted on OMICS C 18 tips (OMIX Tips C 18 reverse-phase resin, Agilent Technologies Inc., Santa Clara, CA, USA) for purification and concentration of peptide/protein content, then dehydrated in a vacuum centrifuge. Venom samples were analyzed in nano-flow liquid chromatography coupled to tandem mass spectrometry (Nano-LC-MS/MS). Samples were resuspended in 20 μL of buffer A (0.1% formic acid), and 1 μL was loaded onto an analytical reversed-phase column (250 × 75 mm, Acclaim Pepmap 100 C 18 , Thermo Fisher Scientific). Samples then were separated with an Ultimate 3000 RSLC system (Thermo Fisher Scientific, Waltham, MA, USA) coupled to a Q Exactive HF-X instrument (Thermo Fisher Scientific, Waltham, MA, USA) via a nano-electrospray source (nanoESI), using a 123 min gradient of 6–40% of buffer B (80% ACN, 0.1% formic acid) and a flow rate of 300 nL/min . Raw data were loaded into PEAKS ® Studio 8 software (Bioinformatics Solutions Inc., Waterloo, Canada). Precursors and fragment ions were identified with a monoisotopic mass tolerance of 0.1 Da. Carbamidomethylation was set as the potential post-translation modification to be present. The full venom gland transcriptome of Cylinder canonicus was used as a database for peptide/protein sequencing. After spectra processing and protein identification, sequences were filtered by applying an FDR (False Discovery Rate) of 1%, at least 2 unique peptides were set to confirm a protein, and a de novo only ALC (average local confidence) score of 80%. 5.6. Identification of Conotoxin in Venom Samples Proteomics and transcriptomics conotoxin results were further exploited to identify conotoxin detected in LC-MS of predatory venoms. ConoServer tools ConoPrec and ConoMass were used for mass prediction ( http://www.conoserver.org ) . First, the FASTA file of all conotoxin precursor sequences obtained from proteo-transcriptomics was uploaded to the ConoPrec tool ( http://conoserver.org/?page=conoprec ) to predict conotoxin mature peptides. Mature peptides were then loaded in the tool ConoMass 1 for mass computation ( http://conoserver.org/?page=ptmdiffmass ). This tool is useful for complex samples such as cone venom as it provides a prediction of the monoisotopic and average masses of each mature sequence, including the corresponding sequence and the number and type of post-translational modifications (PTMs) . Mass predictions were performed on non-reduced conotoxin sequences, including PTMs such as N-terminal amidation, pyroglutamylation, proline hydroxylation, and tryptophane bromination. Computed masses were then matched with an LC-MS experimental mass list for each sample.
Specimens of Cylinder canonicus were collected in February 2023 from the barrier reef surrounding the Island of Mayotte in the Mozambique channel under national and local permits/authorizations. Cone snails were then acclimated and maintained in marine aquariums in the laboratory. Specimens of C. canonicus were fed weekly or fortnightly locally collected gastropod mollusks of the family Nassariidae ( Tritia sp. ). Milked venom was obtained from three different C. canonicus specimens using previously described methods and venom gland extracts (sections) and tissues (RNA) were retrieved from two other specimens. Predatory-evoked milking venoms were obtained by luring cones with the live mollusk prey. Once the cone extends its yellow proboscis out, we intercept the sting with a collecting Eppendorf tube, lined with a fine slice of another prey’s foot over a parafilm. Milking venoms were translucid with a small white precipitate at the bottom of the tube. A whole venom gland was used for mRNA extraction, and a second venom gland was dissected into four sections (distal, distal-central, proximal-central, and proximal) before extracting the venom content for mass spectrometry and proteomic analyses. Finally, predatory-evoked venoms were milked from three specimens of C. canonicus (MV1, MV2 and MV4). Predatory behavior was initiated by facing a Nassariidae mollusk to each C. canonicus specimen and intercepting the injection with an Eppendorf tube covered with parafilm and a thin layer of prey mollusk foot. Milked samples were lyophilized and stored at −20 °C before use.
One Cylinder canonicus specimen was dissected on ice to extract the venom gland in the form of a long and thin duct. The venom duct was placed in 1 mL of TRIzol reagent (Invitrogen, Carlsbad, CA, USA) in an Eppendorf tube and the purified total ribonucleic acid (RNA) was extracted by following the manufacturer’s instructions . Next, mRNA was purified from the retrieved total RNA using the “Oligotex mRNA Mini Kit” (Qiagen, Valencia, CA, USA). Finally, mRNA extracts were submitted to Montpellier GenomiX (MGX, BioCampus Montpellier, CNRS) for sequencing. Complementary DNA (cDNA) libraries were constructed and sequenced using a high-throughput Illumina sequencer (Illumina Inc., San Diego, CA, USA) with a TruSeq Stranded mRNA Sample Prep Kit and a TruSeq Library indices following the manufacturer’s instructions. Paired-end sequencing (2 × 150 bp) yielded several millions of short-read sequences after filtering poor-quality reads. cDNA short raw reads were controlled for quality using FastQC v.0.11.9 ( http://www.bioinformatics.babraham.ac.uk/projects/fastqc ), trimmed, then assembled into longer contigs using Trinity v2.13.2. The assembled contigs were then translated into amino acids (six reading frames) in silico .
Conotoxin sequences were then retrieved using, in parallel, a manual and automatic approach. First, a locally-built transcriptomic platform (VenUM) was used to manually extract conopeptide precursors from the venom gland transcriptome of Cylinder canonicus . Registered conopeptide precursors found on ConoServer ( www.conoserver.org ) from species of the same clade, such as C. victoriae and C. textile , were used as queries to search against the transcriptome of C. canonicus to extract and annotate conopeptide precursors. Additionally, the ConoPrec tool from ConoServer was employed to identify cysteine frameworks, cleavage sites, and mature sequences . For further validation of conopeptide superfamilies and other protein annotations, protein BLAST searches were conducted using BLAST+ v2.16.0+ and the non-redundant protein sequences and Swiss-Prot r2024_03 databases . To classify gene superfamilies automatically, the bioinformatic tool ConoDictor v2.4.1 was utilized ( https://github.com/koualab/conodictor ) . This tool enables sequence classification and predicts superfamily assignments . Finally, multiple conopeptide sequence alignments were generated using the open-source software Jalview v.2.11.4.0 ( www.jalview.org ) and ClustalW v2.1 with default parameters ( www.clustal.org/clustal2/ ) .
LC-ESI-MS. Crude lyophilized venoms were resuspended in water and concentrations were evaluated by a NanoPhotometer ® (Implen GmbH, Munich, Germany). Electrospray ionization mass spectrometry coupled with liquid chromatography (LC-ESI-MS) experiments were performed following previous protocols . Separation by liquid chromatography was performed on an Acquity H-Class ultrahigh-performance liquid chromatography (UPLC) system. Around 20 µg of each venom sample were injected onto a Kinetex C 18 100 Å column (2.1 mm × 150 mm, 3 μm) (Phenomenex, Torrance, CA, USA) fitted with a precolumn for separation at 0.4 mL/min in a 0–80% gradient of a solvent B during 60 min (A: water (H 2 O) + 0.1% formic acid (FA); B: acetonitrile (ACN) + 0.1% formic acid). ESI-MS analyses were performed on a Waters Synapt G2-S (Waters Corporation, Milford, MA, USA) equipped with a time of flight (TOF) detection cellule and was used in positive mode across a masse range of 1000–3000 Dalton (Da) and scan time of 1 s. Each sample was infused in the ionization source at a flow rate of 5 µL/min. Then, ionization was performed with a capillary tension of 3 kV. Temperatures were set at 120 °C for the source and 350 °C for the desolvation gas. Gas flows were regulated at 850 L/h for the desolvation gas and at 6 L/h for the nebulizing gas. Total ion current (TIC) chromatograms and molecular masses were processed with Mass Lynx software (version 4.1, Waters, Corp., Milford, MA, USA). MALDI-TOF-MS. MALDI-TOF mass spectra were acquired using a RapifleX ® MALDI mass spectrometer (Bruker Daltonics, Billerica, MA, USA). All samples were diluted at 0.10 mg/mL. In parallel, a solution of α-cyano-4-hydroxycinnamic acid (HCCA, ACROS Organics, USA) matrix was saturated at 10 mg/mL by dissolving 10.0 mg of HCCA in 1 mL of 30% ACN in aqueous 0.1% TFA (70:30:0.1, ACN/H 2 O/TFA). Samples were spotted on an MTP 384 polished steel plate (Bruker Daltonics, Billerica, MA, USA) according to the dried-droplet spotting technique which consists of mixing 1 µL of the sample with 1 µL of HCCA matrix (α-Cyano-4-hydroxycinnamic acid) and spotting 1 µL of each sample in two different spots. External calibration was performed using a mixture of standard peptides (Bradykinin 757.399 Da, Angiotensin II 1046.542 Da, Angiotensin I 1296.685 Da, Substance P 1347.735 Da, Bombesin 1619.822 Da, ACTH clip 2093.086 Da, ACTH clip 2465.198 Da, Somatostatin 3147.471 Da). Mass spectrometry spectra were acquired in positive ion mode at a frequency of 5000 Hz, by applying 6000 shots/per sample spot (4 × 1500 shots) with a laser power of 30%. Ion detection was performed on a time of flight (TOF) detector in reflectron mode with a mass range of 500–5000 Da. FlexControl 3.0 software was used for data acquisition and FlexAnalysis 4.0 (Bruker Corporation, Billerica, MA, USA) for data treatment.
The milked predatory venom of three specimens of Cylinder canonicus , as well as dissected venoms of distal (D and DC) and proximal (PC and P) sections of the venom gland were prepared prior to proteomic analysis as described . A total of 50 µg of pooled milked (all 3 predatory venoms), distal (D + DC), and proximal (PC + P) venoms were first diluted in 89 µL of triethylammonium bicarbonate (TEAB) 100 mM under stirring during 30 min at room temperature. Disulfide bonds were reduced with dithiothreitol (DTT) at 10 mM for 30 min at 60 °C and alkylated with iodoacetamide (IAA) at 50 mM before 30 min incubation in the dark at room temperature. The venom samples were then enzymatically digested with 1.5 µg Trypsin (Gold, Promega, Madison, WI, USA) and incubated overnight at 30 °C. Samples were desalted on OMICS C 18 tips (OMIX Tips C 18 reverse-phase resin, Agilent Technologies Inc., Santa Clara, CA, USA) for purification and concentration of peptide/protein content, then dehydrated in a vacuum centrifuge. Venom samples were analyzed in nano-flow liquid chromatography coupled to tandem mass spectrometry (Nano-LC-MS/MS). Samples were resuspended in 20 μL of buffer A (0.1% formic acid), and 1 μL was loaded onto an analytical reversed-phase column (250 × 75 mm, Acclaim Pepmap 100 C 18 , Thermo Fisher Scientific). Samples then were separated with an Ultimate 3000 RSLC system (Thermo Fisher Scientific, Waltham, MA, USA) coupled to a Q Exactive HF-X instrument (Thermo Fisher Scientific, Waltham, MA, USA) via a nano-electrospray source (nanoESI), using a 123 min gradient of 6–40% of buffer B (80% ACN, 0.1% formic acid) and a flow rate of 300 nL/min . Raw data were loaded into PEAKS ® Studio 8 software (Bioinformatics Solutions Inc., Waterloo, Canada). Precursors and fragment ions were identified with a monoisotopic mass tolerance of 0.1 Da. Carbamidomethylation was set as the potential post-translation modification to be present. The full venom gland transcriptome of Cylinder canonicus was used as a database for peptide/protein sequencing. After spectra processing and protein identification, sequences were filtered by applying an FDR (False Discovery Rate) of 1%, at least 2 unique peptides were set to confirm a protein, and a de novo only ALC (average local confidence) score of 80%.
Proteomics and transcriptomics conotoxin results were further exploited to identify conotoxin detected in LC-MS of predatory venoms. ConoServer tools ConoPrec and ConoMass were used for mass prediction ( http://www.conoserver.org ) . First, the FASTA file of all conotoxin precursor sequences obtained from proteo-transcriptomics was uploaded to the ConoPrec tool ( http://conoserver.org/?page=conoprec ) to predict conotoxin mature peptides. Mature peptides were then loaded in the tool ConoMass 1 for mass computation ( http://conoserver.org/?page=ptmdiffmass ). This tool is useful for complex samples such as cone venom as it provides a prediction of the monoisotopic and average masses of each mature sequence, including the corresponding sequence and the number and type of post-translational modifications (PTMs) . Mass predictions were performed on non-reduced conotoxin sequences, including PTMs such as N-terminal amidation, pyroglutamylation, proline hydroxylation, and tryptophane bromination. Computed masses were then matched with an LC-MS experimental mass list for each sample.
|
Clinicopathological study of hepatic mesenchymal hamartoma and undifferentiated embryonal sarcoma of the liver: a single center study from Iran | 0e471322-72e5-492e-8681-47a6646019f2 | 8223305 | Anatomy[mh] | Undifferentiated embryonal sarcoma of the liver (UESL) and hepatic mesenchymal hamartoma (HMH) are two rare pathologic entities that are primarily seen in the pediatric population . UESL is a rare mesenchymal tumor accounting for 5–15% of liver malignancies in pediatrics [ – ]. On the other hand, HMH, accounting for 8% of liver tumors in children, comprises the majority of pediatric benign liver tumors after infantile hemangioma . UESL which is an aggressive tumor was first described in 1978 by Stocker and Ishak and is primarily diagnosed between 6 and 10 years of age without gender predominance . However, HMH is mainly diagnosed in children of less than 2 years of age and shows a slight male predominance in this age group . UESL usually arises from the right hepatic lobe with tumor size varying from 10 cm to 30 cm . Similarly, HMH is primarily seen in the right hepatic lobe in children and can have various sizes of up to 30 cm in diameter . Patients with UESL usually present with non-specific symptoms including anorexia, abdominal pain, fever, and nausea with subsequent findings of cystic and solid components in imaging studies [ , , , ]. Abdominal distention or mass is the most common clinical presentation of HMH which is usually seen as a multi-loculated cyst with a varying solid component on radiologic studies . The underlying pathologic mechanisms playing a role in the development of UESL and HMH are unclear. However, different studies have proposed a number of potential mechanisms. Comparative genome hybridization (CGH) studies of UESL have shown different patterns of chromosomal changes including losses of chromosome 9p, 11p, and 14 and gains of chromosome 1q, 5p, 6q, 8p, and 12q pointing to the potential role of chromosomal instability . Genetic alterations leading to the ectopic activation of chromosome 19q microRNA cluster (C19MC) are found in HMH . UESL is usually diagnosed based on the patient’s age, tumor location, and an immunohistochemistry panel of undifferentiated markers including vimentin, desmin, α 1 anti trypsin, CD10, and CD68 . However, HMH is usually diagnosed using clinical and histopathologic features alone . Overall, as a result of the low incidence of these pathologic entities, the clinicopathological features of UESL and MH are limited to case series and case reports. Considering the paucity of reports from the Middle Eastern region, this study was conducted to investigate the clinical, histological, and immunohistochemical features in a series of patients with UESL and HMH in a single referral center from Iran.
In this investigation, a retrospective evaluation of patients diagnosed with UESL and HMH at Shiraz University of Medical Sciences between 2012 and 2020 was conducted. The diagnosis was based on histopathological evaluation of tumor samples according to WHO classification of tumors of the digestive system. The specimens were obtained by surgical resection and were subsequently fixed in formalin and then embedded in paraffin. They were then stained using hematoxylin and eosin staining, periodic acid–Schiff (PAS) staining for UESL cases, and immunohistochemistical staining for all cases using the following antibodies: Vimentin (Máster Diagnóstica, rabbit monoclonal antibody, Clone SP20), HepPar 1 (Máster Diagnóstica, mouse monoclonal antibody, clone OCH1E5), Glypican 3 (Máster Diagnóstica, mouse monoclonal antibody, clone 1G12), Arginase-1 (Biocare Medical, rabbit monoclonal antibody, clone EP261), Ki67 (Máster Diagnóstica, rabbit monoclonal antibody, clone SP6), Desmin (Máster Diagnóstica, mouse monoclonal antibody, clone D33), SMA (Máster Diagnóstica, mouse monoclonal antibody, clone 1A4), CD56 (Máster Diagnóstica, rabbit monoclonal antibody, clone MRQ-42), CD10 (Máster Diagnóstica, mouse monoclonal antibody, clone 56C6), CD68 (Máster Diagnóstica, mouse monoclonal antibody, clone KP-1), BCL2 (Máster Diagnóstica, rabbit monoclonal antibody, clone EP36), PD-L1 (Máster Diagnóstica, rabbit monoclonal antibody, clone CAL10), C-Kit (Máster Diagnóstica, rabbit monoclonal antibody, clone EP10), CD34 (Máster Diagnóstica, mouse monoclonal antibody, clone QBEnd/10). Appropriate positive and negative controls were used throughout the experiments. The immunohistochemistry slides were subsequently evaluated by a pathologist. In the event that less than 1% of the cells in a slide showed immunoreactivity, the case was considered negative. The positive cases were subsequently graded based on staining intensity as weak, intermediate, and strong. If more than half of the cells of interest were stained, the staining was considered diffuse. Otherwise, in cases with a staining percentage between 1 and 50%, the staining was considered focal. The following information was collected for both groups of the patients: patient age, sex, presenting symptoms, tumor location, significant laboratory findings. Furthermore, for patients diagnosed with UESL extrahepatic metastasis, recurrence, disease stage, and radiologic findings were collected. The treatments received by the patients with UESL and the disease outcome were also collected by contacting the family members. In addition, the gross and microscopic pathologic findings were documented for all the cases. R ver 4.0.2 (2020-06-22) was used for statistical analysis. Considering the small sample size of the study, continuous variables were reported as the median and interquartile range (IQR). Mann-Whitney U test was used to compare the distribution of variables between two groups. The correlation between tumor size and age was assessed with Spearman’s ρ . Kaplan-Meier survival analysis for the patients with UESL. A p value < 0.05 was considered statistically significant.
The clinicopathologic characteristics and demographic data of the patients with UESL and HMH are presented in Tables and , respectively. Overall, there were 8 patients (3 males and 5 females) with UESL with a median age at diagnosis of 12.0 (IQR 4.6 to 13.3) years. In addition, 8 patients (5 males and 3 females) with HMH with a median age at diagnosis of 2.3 (IQR 1.4 to 2.5) years were identified. The age at diagnosis for those with UESL was significantly ( p = 0.002) higher than those with HMH. Abdominal pain was the most common (8/8) presenting symptom in patients with UESL, followed by nausea/vomiting (3/8) and fever (2/8). Abdominal distention (4/8) and accidental identification of an abdominal mass by the patient’s caregiver (3/8) were the most common complaints on initial presentation in patients with HMH. All the UESL and HMH patients identified had right liver lobe masses. Except for one patient with UESL, all of the UESL and HMH masses identified were unifocal. Initial laboratory abnormalities were more commonly observed in patients with UESL compared with those with MH. Anemia (4/8) and elevated hepatic transaminases levels (3/8) were seen in patients with UESL. Notably, elevated CA-125 levels were seen in one of the patients with UESL. Furthermore, one patient with HMH had elevated alpha-fetoprotein (AFP) levels. All patients diagnosed with HMH had undergone surgical resection of the liver mass and were alive after a median follow-up of 5.5 years without any complications. Most of the patients with UESL underwent gross tumor resection followed by adjuvant chemotherapy (7/8). Furthermore, 2/8 patients underwent radiation therapy as well. In patients with UESL, after a median follow-up time of 3 years, one patient had passed away and one of the patients was still under treatment due to recurrence. Overall, tumor recurrence was observed in two of the patients. No evidence of the disease after the treatment was detected among the other six patients with UESL. Patients with UESL had a mean survival time of 7.2 (95% CI 5.7 to 8.7) years estimated by Kaplan-Meier survival analysis. On gross examination, UESL was typically found as a well-defined mass with necrosis and hemorrhage. Gelatinous material in cases with myxoid change was also seen. The tumor size varied from 5.0 cm to 28.0 cm (median: 15.0 cm, IQR 10.0 cm to 19.0 cm) (Table ). There was no correlation between the tumor size and age at diagnosis (Spearman’s ρ = 0.584, p = 0.128). Histological findings in UESL included anaplastic stellate to spindle-shaped or epithelioid tumor cells with poorly defined, light eosinophilic cytoplasm (Fig. A-B). Nuclei were found to be irregular and hyperchromatic with numerous mitotic figures (Fig. C). The anaplastic cells were arranged loosely or compactly in a usually myxoid stroma (Fig. D-E). Most of the cases had bizarre multinucleated giant cells with abundant cytoplasm and atypical nuclear features (Fig. F). Multiple variably-sized, periodic acid-Schiff diastase resistant-positive eosinophilic hyaline globules were also frequently seen in the cytoplasm or extracellular stroma (Fig. G-H). Of note, collections of small round cells were also seen in two of the cases (Fig. I). Furthermore, hemangiopericytomatous pattern was identified in another one of the cases (Fig. J). Furthermore, concurrent HMH was also identified in two of UESL cases (Fig. K) (Table ). On gross evaluation, HMH cases were characterized as a well-defined solitary mass without any evidence of hemorrhage except in one of the cases. The tumor size ranged from 8.0 cm to 20.0 cm (median: 13.0 cm, IQR 11.4 cm to 15.5 cm). No correlation between the tumor size and age at diagnosis was noticed (Spearman’s ρ = − 0.331, p = 0.423). HMH was histologically characterized by the disordered arrangement of hepatic parenchyma, bile ducts, and mesenchyme consisting of spindled cells and myxoid stroma (Fig. A-C). Extramedullary hematopoiesis was detected in three of the cases. Furthermore, hemorrhage and severe hepatic steatosis were noted in one of the cases (Fig. D). Overall, 5 cases were predominantly cystic (Fig. A-B), while 3 cases were predominantly solid. (Table ). For immunohistochemical staining, we did not have access to the pathological specimens of two of the patients (HMH case 6 and UESL case 7) and therefore did not include them in the IHC studies. HMH associated with UESL in case 5 was included as a separate HMH case in the final analysis (Tables and ). A variable degree of Glypican 3 marker cytoplasmic staining with either strong or moderate immunoreactivity was found in UESL tumor cells. Diffuse strong cytoplasmic staining for this marker was also seen in four of HMH cases. Six UESL cases had strong cytoplasmic immunoreactivity for Desmin (3 diffuse, 3 focal) while five HMH cases showed strong cytoplasmic staining for this marker in the spindle cells (3 diffuse, 2 focal). Strong or moderate diffuse cytoplasmic staining for CD56 in all cases of UESL and also in the bile duct epithelium of HMH cases was observed. (Figs. and ).
UESL and HMH are rare hepatic lesions primarily seen in the pediatric and early adult populations. In this investigation, we, for the first time, provide a comprehensive clinicopathologic overview of these two entities in a case series from Iran. Overall, our findings show that patients with HMH are younger at presentation compared with those with UESL. Furthermore, both conditions had nonspecific initial clinical presentations with abdominal pain being present in all patients with UESL. These findings are in line with the previous investigations . Anemia and abnormal liver function tests were the most common findings in patients with UESL while patients with HMH had an overall normal baseline laboratory finding. Elevated tumor markers were observed in a number of patients; Elevated alpha fetoprotein levels in a HMH patient and an increased cancer antigen 125 (CA125) level was found in a patient with UESL. Elevated Erythropoietin levels have also been described in other patients with UESL which could be attributed to the mesenchymal origin of this tumor . All of the patients with UESL had undergone gross surgical resection of the tumor followed by chemotherapy except for one of the patients who had not received chemotherapy. In addition, two of the patients with UESL also received radiotherapy. In the past, the prognosis of liver sarcomas was poor overall. In the original report of 31 patients with UESL by Stocker and Ishak, only six patients were alive without any evidence of the disease . Over the past couple of years, mounting evidence suggests that radical surgical resection of the tumor supplemented by adjuvant chemotherapy leads to an overall improvement in the survival of these patients . Herein, we report encouraging results of following this protocol in patients with UESL. Six out of eight patients had no evidence of disease after a median follow-up time of 3 years. We were only able to retrieve the data regarding the chemotherapy regimen for three of the patients. However, based on our limited findings, all three of the patients who showed no evidence of disease on the latest follow up received vincristine and cyclophosphamide along with other chemotherapeutic agents. Treatment with vincristine, actinomycin-D, cyclophosphamide has been shown to be a successful therapeutic strategy in another report . Notably, one of the patients had no evidence of disease 6 years after the diagnosis only by gross tumor resection. This in turn highlights the paramount importance of surgical tumor resection in the treatment of this disorder. On gross evaluation, all of the masses were located in the right hepatic lobe with variable sizes of up to 28 cm. UESL masses usually showed hemorrhage and cystic changes with variable degrees of necrosis while HMH cases presented with unifocal solid or cystic structures. Small round cells were seen on histologic evaluation of two UESL cases which could be an important finding if present in liver biopsies from these patients and can mimic other pathologies. In addition, severe hepatic steatosis seen in one of HMH cases points to potential molecular defects leading to deregulation of cellular energetics as seen in other liver disorders . Furthermore, UESL was diagnosed in concurrence with HMH in two of the studied patients. This finding which has been also reported in a previous case series points to the potential malignant transformation of HMH in such patients . The immunohistochemical phenotypes of UESL and HMH have only been investigated in a few studies so far. Overall, in three case series, positive immunostaining for vimentin and Bcl-2 has been reported in most of the cases while positive staining for desmin, SMA, p53, pancytokeratin, Glypican-3, and calponin has been observed in some of the cases. Immunostaining for HepPar1, CD34, CD117 (C-kit), S100, HMB45, myogenin, ALK-1, and α-fetoprotein were found to be negative in the primary tumor cells [ – ]. In addition, positive immunostaining for desmin, vimentin, SMA, Glypican-3, Hep Par 1, and α-fetoprotein in different components of HMH has been reported [ – ]. Our investigation highlighting immune-reactivity for CD56 in UESL and different components of HMH expands the findings of previous case series studies and corroborates their findings for other markers as well. Furthermore, our findings demonstrating the expression of Bcl-2 and CD34 in various components of HMH not only broadens its immunophenotypic spectrum but also provides fresh impetus for further investigations regarding the malignant transformation of HMH since both these two markers showed moderate to strong immunoreactivity in different components of the HMH case found in association with UESL. The development of UESL after incomplete excision of HMH reported in the literature corroborates this hypothesis . Absence of PD-L1 expression in all of the UESL cases investigated in this study points to the potential lack of efficacy of immune checkpoint inhibitors targeting this pathway as a therapeutic target in these patients . However, positive staining for CD56 which is also reported in rhabdomyosarcoma and synovial sarcoma could have significant therapeutic implications in the management of UESL . CD56–chimeric antigen receptor T-cell therapy which has already shown promising results in the pre-clinical studies in other cancers could be used in the treatment of patients with a high CD56 expression in the tumor tissue who had shown disease recurrence with poor response to chemotherapy and radiotherapy (e.g. UESL case 6 in this study) . The main differential diagnoses of UESL in the pediatric and adolescent population include hepatoblastoma, HMH, embryonal rhabdomyosarcoma, hepatic angiosarcoma, and hepatocellular carcinoma . IHC staining could play a substantial role in accurate diagnosis particularly in situations where histopathologic clues are not helpful. Our study showing Glypican-3 expression in both UESL and HMH cases demonstrates that this marker cannot be used to differentiate the two from hepatocellular carcinoma and hepatoblastoma which have been shown to express this antigen . Nevertheless, our study showed absence of HepPar 1 and Arginase-1 in all of the UESL cases highlighting the importance of this maker in differentiating UESL from hepatocellular carcinoma for which these two markers have been shown to display a sensitivity of 84.4 and 96.0%, respectively . Positive staining for skeletal muscle differentiation markers such as myoD1 and myogenin along with cross-striation are distinctive findings in rhabdomyosarcoma . Furthermore, CD34 marker which was found to be negative in all of the cases in this investigation can be a very useful marker in the diagnosis of hepatic angiosarcoma which frequently stains positive for this marker . Features such as anaplasia and high grade mitotic activity (strongly positive staining for ki-67) as observed in this study are useful diagnostic features to distinguish UESL from HMH. In summary, this investigation reports the first case series of patients with UESL and HMH from Iran. Herein, we reported the clinicopathologic findings of sixteen patients from a single referral pediatric center. Although due to the rarity of these two pathologic entities we were not able to recruit a large number of patients in order to investigate the prognostic significance of different pathologic findings, we were able to identify significant histopathologic findings and novel IHC markers with diagnostic and therapeutic implications. Further investigations have to be conducted to shed light on the clinicopathologic and pathophysiologic basis of these two rare entities.
|
Establishing priorities for psychological interventions in pediatric settings: A decision-tree approach using the DISABKIDS-10 Index as a screening instrument | 7bcc1c8f-2702-4631-999c-8aae36301d47 | 5979027 | Pediatrics[mh] | Over the past few decades, scientific and technological advances in medicine have resulted in epidemiological changes marked by an increased prevalence of pediatric chronic health conditions (estimated prevalence rates between 15% and 18% ) and by the emergence of patients’ subjective perceptions of their physical, psychological and social functioning and well-being as the primary goals of clinical interventions . Evidence-based research has demonstrated that children and adolescents with chronic conditions are at a greater risk for psychosocial impairments; however, few patients present clinically significant psychological symptoms [ – ]. In addition, the practice of routine psychological assessments for all pediatric patients is not feasible due to the lack of time and human and financial resources in healthcare services . Self-report screening tools that can be easily and promptly administered, scored and interpreted by any healthcare provider (e.g., general practitioner or nurses), and not only by specialized mental health professionals, may represent a promising way to identify and refer at-risk children and adolescents for specialized psychological assessments and interventions, thus contributing to a more efficient allocation of health resources and to an improved patient satisfaction with the quality of received care . The most widely accepted models of child adjustment to pediatric chronic health conditions conceptualize adaptation as a multi-dimensional construct, including physical health, mental health (not just psychological maladjustment) and social functioning . Internalizing and externalizing problems have traditionally been assessed as specific psychological functioning outcomes because of their high prevalence in children with chronic conditions or disabilities . More recently, health-related quality of life (HrQoL), defined as “a multidimensional construct covering physical, emotional, mental, social, and behavioral components of well-being and function as perceived by patients and/or other observers” (, p. 344), has increasingly been acknowledged as a key health outcome in epidemiological and clinical studies, extending beyond the psychopathological conceptualization of child mental health and representing adaptation as a process that accounted for resiliency and variability on specific indicators . Pediatric chronic health conditions may impact both psychological problems and HrQoL and these two indicators of psychosocial adaptation are expected to influence each other reciprocally . In fact, empirical research has shown lower levels of perceived social support and poorer HrQoL in children/adolescents with chronic conditions compared to their healthy peers and has found that HrQoL impairments significantly account for more psychological problems . Additional risk factors for psychological problems include the diagnosis of a medical condition that affects brain function, male gender, younger age, low socioeconomic status (SES), single-parent household, and parents’ mental health problems [ , , ]. In primary care practice, physicians tend to underestimate psychosocial and functional impairments, and even severe psychological problems are often undetected and untreated, mostly because of lack of time and expertise in psychological assessment [ – ]. Two randomized controlled trials have demonstrated that providing patient-reported HrQoL information to primary care physicians improved patient-physician communication and increased detection and referral to mental health specialists . However, pediatric HrQoL measures are far from being used routinely in clinical practice, despite the recognition of their utility to facilitate patient-physician communication, to improve patient satisfaction with the quality of medical care, to estimate the healthcare needs of specific populations, to longitudinally monitor the disease status and effectiveness of treatment, to detect psychological problems secondary to medical conditions or treatments (so-called “hidden morbidities”), and to assist physicians in clinical decision-making and referral processes . In addition, HrQoL scores can be difficult to interpret in clinical practice because they are often presented as the mean values along the continuum from excellent to poor health , and population-based clinically meaningful cut-off points to identify at-risk patients are unavailable for most HrQoL instruments. In this context of routine monitoring and screening, short-form questionnaires that summarize scores into a single value (or index) have been found to be reliable and valid measures for global HrQoL assessment, while reducing response burden and saving administration costs . The DISABKIDS-10 Index is a commendable example of a short-form measure specifically designed to assess the perceived impact of chronic health conditions on children’s and adolescents’ physical, mental and social well-being and functioning. This chronic-generic instrument was built upon a non-categorical approach, which suggests that nosologically different health conditions may lead to similar impacts on patients’ HrQoL , thus allowing for comparisons across different diagnosis without sacrificing the sensitivity to specific psychosocial impairments resulting from the health condition. In addition, the DISABKIDS questionnaires are child-centered and developmentally appropriate, use subjective self-report whenever possible (proxy judgments can be used if the child is too young or too disabled to complete self-reports), are cross-culturally comparable, and emphasize health-enhancing aspects of HrQoL rather than merely listing symptoms, as recommended by the WHO Division of Mental Health . The current challenge is to maximize the use of HrQoL measures in pediatric healthcare services, by enhancing their interpretability as screening tools to detect psychosocial and functional disabilities secondary to the medical condition and to support the adequate referral to specialized mental healthcare. Thus, the current study aimed (1) to compare the levels of HrQoL and the prevalence of borderline/clinically significant psychological problems across pediatric patients with different chronic health conditions (i.e., asthma, epilepsy, cerebral palsy, type-1 diabetes, and overweight/obesity) and (2) to test a classification-tree model to identify risk profiles for the development of borderline/clinically significant psychological problems based on the DISABKIDS-10 Index as a HrQoL screening measure and on patients’ and families’ sociodemographic and clinical data.
Participants and procedures The participants were children and adolescents with chronic health conditions who were recruited at the outpatient pediatric services of three Portuguese public hospitals and 10 Cerebral Palsy Associations. For inclusion in the sample, the children/adolescents had to meet the following criteria: (1) age between 8 and 18 years at the time of recruitment; (2) diagnosis of asthma, cerebral palsy, type-1 diabetes, epilepsy or overweight/obesity according to the International Classification of Diseases-10 ; (3) absence of other comorbid chronic health conditions; (4) ability to understand and answer self-report questionnaires; and (5) accompanied by the parent currently assuming the primary caregiver role. In addition, pediatric patients with asthma or epilepsy were required to have the diagnosis for at least one year, children and adolescents with cerebral palsy were eligible if they had a minimum intelligence quotient (IQ) of 70, and children and adolescents with overweight/obesity were included when their body mass index (BMI) was above the 85 th percentile for same age and same sex peers according to the Centers for Disease Control and Prevention growth curves . These diagnoses were selected for inclusion in this study because they share non-nosological clinical features that may impact HrQoL and psychological functioning, such as central nervous system impairment (epilepsy and cerebral palsy), external visibility (obesity and cerebral palsy), unpredictability of crises/exacerbations (epilepsy and asthma), and need for treatments that require changes in daily routines (diabetes and obesity). Data collection occurred between March 2009 and December 2012, after the study had been approved by the institutions’ Ethics Committees/Direction Boards. Using the non-probabilistic convenience sampling method, children and adolescents who attended medical routine appointments in the period of sample collection were screened by health professionals based on their medical records and those who met the aforementioned inclusion criteria were invited to participate in the study. The study’s aims and procedures were explained in detail, and written consent forms were obtained from all parents and adolescents older than 13 years; younger children provided verbal informal assent. Children and adolescents who agreed to participate completed the self-report questionnaires in the institution they attended under the supervision of a trained research assistant, and their parents were asked to complete a sociodemographic and clinical datasheet. Measures Psychological problems To assess the children’s and adolescents’ psychological problems, we used the Difficulties scale of the Portuguese self-rated version of the Strengths and Difficulties Questionnaire (SDQ) , which is a brief behavioral questionnaire that was developed with reference to the main nosological categories recognized by the Diagnostic and Statistical Manual of Mental Disorders, 4 th edition (DSM-IV) . This scale comprises 20 items assessing emotional symptoms (e.g., “I am often unhappy, down-hearted or tearful”), conduct problems (e.g., “I fight a lot. I can make other people do what I want”), hyperactivity/inattention (e.g., “I am easily distracted, I find it difficult to concentrate”) and peer relationship problems (e.g., “Other children or young people pick on me or bully me”). The items were answered using a Likert-type response scale with three options (0 = not true , 1 = somewhat true , and 2 = certainly true ), providing a Total Difficulties score, with higher values indicating more psychological problems. Adequate reliability was obtained for the Difficulties scale (α = .76). The Total Difficulties sum score was dichotomized based on published cut-off values to differentiate between pediatric patients within the normal (0–15 points; n = 578) and the borderline/clinical range (16–40 points; n = 158). Health-related quality of life Children’s and adolescents’ HrQoL was assessed by the Portuguese self-report short version of the DISABKIDS-10 Index . This short-form questionnaire contains 10 items measuring the physical, mental and social impact of chronic health conditions (e.g., “Does your condition get you down?”) on 8- to 18-year-old patients’ lives and is answered on a 5-point Likert scale ranging from 1 ( never ) to 5 ( always ). According to its one-dimensional factor structure, a standardized composite score (0–100), representing the physical, mental and social domains of HrQoL, was calculated, with higher scores indicating better chronic-generic HrQoL. In the current sample, the questionnaire presented good reliability, with a Cronbach’s α value of .84. Sociodemographic and clinical data Sociodemographic data were reported by the parents and included their children’s age and gender, as well as parents’ education level, occupation, family structure and psychiatric history. Using a classification system specifically developed for the Portuguese context and based on the educational level and current job of the primary caregiver , the family SES was classified into low (e.g., unqualified employees in construction or manufacturing without completing the 9 th grade of school education), medium (e.g., employees in bureaus or banks, nurses, or teachers with intermediate or university courses) and high (e.g., senior officials of government, army, commerce or industry, physicians, or engineers with bachelor’s, master’s, doctorate or other post-graduate degrees). Due to the heterogeneous distribution of the SES levels observed in our sample, this variable was dichotomized into low ( n = 464) and medium/high ( n = 272). Clinical data were provided by the parents and/or physicians and included the use of medication and specific information for each clinical group. Specifically, the physicians classified asthma severity into 4 levels ( intermittent , mild persistent , moderate persistent and severe persistent ) according to the Global Initiative for Asthma guidelines , and epilepsy severity was classified into 7 levels (from not at all severe to extremely severe ) using the Global Assessment of Severity of Epilepsy Scale . Levels of function in patients with cerebral palsy were classified into 5 levels according to the Gross Motor Function Classification System . Values of glycated hemoglobin (HbA1c) at the time of assessment were obtained for patients with diabetes. The weight and height of patients with obesity were obtained from the parents and/or nutritionists. These markers of disease severity were dichotomized into mild (including patients with intermittent asthma, not at all to somewhat severe epilepsy, cerebral patients with level I of functioning with no limitations in walking, patients with diabetes with HbA1c ≤ 8%, and overweight children; n = 333 [45.2%]) and moderate/severe (including children/adolescents with mild, moderate and severe persistent asthma, moderately to extremely severe epilepsy, patients with cerebral palsy with levels II to V with movement restriction, patients with diabetes with HbA1c > 8%, and obese children; n = 315 [42.8%]). Data analyses The statistical analyses were conducted with the Statistical Package for the Social Sciences (SPSS v.20.0; IBM Corp., Armonk, NY). Except for sociodemographic and clinical variables, missing data that were random (Little’s Missing Completely at Random [MCAR] Tests: χ (12) 2 = 16.84, p = .16 for the SDQ Emotional Symptoms scale; χ (12) 2 = 12.29, p = .42 for the SDQ Conduct Problems scale; χ (8) 2 = 9.28, p = .32 for the SDQ Hyperactivity/Inattention scale; χ (12) 2 = 5.14, p = .95 for the SDQ Peer Relationship Problems scale; and χ (54) 2 = 39.68, p = .93 for the DISABKIDS-10 Index) and less than 5% of the values were replaced with the individual mean score for each variable. Descriptive statistics were obtained for sociodemographic and clinical variables and the sample characteristics were compared between diagnostic groups by using one-way analyses of variance (ANOVA) followed by post-hoc pairwise comparisons with Bonferroni correction for continuous variables and by χ 2 -tests for categorical variables. The classification-tree model was performed using the Chi-Squared Automatic Interaction Detection (CHAID) algorithm, which is a classification method that also accounts for interactions between the predictors . The CHAID method is based on chi-squared tests (χ 2 ) that compare the squared deviations between observed and expected frequencies, with Bonferroni adjusted p -values for multiple comparisons with an overall error rate of .05 . The algorithm starts by selecting the predictor that best discriminates the dependent variable (i.e., the predictor with the lowest p -value in the Chi-squared tests) and then splits the data set into two or more nodes (parent node). Subsequently, the method splits the new nodes into smaller nodes (child nodes) based on the variable that best discriminates each of them. The method ends when no more significant dependence relationships can be found between the dependent variable and the set of predictors. In the current study, nine eligible risk factors were included (age group, gender, diagnosis, HrQoL scores, medication, disease severity, SES, family structure and caregiver’s psychiatric history), and the target category was defined as having borderline/clinically significant psychological problems. The multicollinearity across the predictors was tested with preliminary correlation analysis and Variance Inflation Factor (VIF)/ Tolerance values. Because significant differences in most risk factors and the outcome variable were found between diagnostic groups, the health condition was forced into the model as the first split variable. Continuous independent variables were automatically recoded into discrete categories with minimal loss of information (interval = 10). The categories of each independent variable were merged if they were not significantly different with respect to the dependent variable; multiple branches were generated if the categories differed significantly ( p ≤ .05). The minimum sample size for parent and child nodes were specified as n = 40 and n = 20, respectively, and the tree depth was limited to 4 levels. The likelihood ratio χ 2 -statistic was used because of the sample size < 1000 cases. The cost of misclassifying a child/adolescent with a high risk of psychological problems as low risk was customized as twice the cost of misclassifying a low risk patient as high risk. To assess the predictive accuracy of the decision-tree model, the Receiver Operating Characteristic (ROC) curve was examined by plotting sensitivity (i.e., the proportion of patients with psychological problems who were correctly identified as positive) in the function of 1 –specificity (i.e., the proportion of patients without psychological problems who were correctly identified as negative) for different cut-off points of a parameter . The statistic used to summarize the ROC analysis was the area under the curve (AUC), which corresponds to the probability that a randomly selected patient with borderline/clinical psychological symptoms was identified by the classification-tree model as higher risk than a randomly selected patient without psychological problems. AUC values between 0.90–1.00 were considered to have excellent accuracy, 0.80–0.90 were considered good, 0.70–0.80 were considered fair, 0.60–0.70 were considered poor, and 0.50–0.60 were considered failed .
The participants were children and adolescents with chronic health conditions who were recruited at the outpatient pediatric services of three Portuguese public hospitals and 10 Cerebral Palsy Associations. For inclusion in the sample, the children/adolescents had to meet the following criteria: (1) age between 8 and 18 years at the time of recruitment; (2) diagnosis of asthma, cerebral palsy, type-1 diabetes, epilepsy or overweight/obesity according to the International Classification of Diseases-10 ; (3) absence of other comorbid chronic health conditions; (4) ability to understand and answer self-report questionnaires; and (5) accompanied by the parent currently assuming the primary caregiver role. In addition, pediatric patients with asthma or epilepsy were required to have the diagnosis for at least one year, children and adolescents with cerebral palsy were eligible if they had a minimum intelligence quotient (IQ) of 70, and children and adolescents with overweight/obesity were included when their body mass index (BMI) was above the 85 th percentile for same age and same sex peers according to the Centers for Disease Control and Prevention growth curves . These diagnoses were selected for inclusion in this study because they share non-nosological clinical features that may impact HrQoL and psychological functioning, such as central nervous system impairment (epilepsy and cerebral palsy), external visibility (obesity and cerebral palsy), unpredictability of crises/exacerbations (epilepsy and asthma), and need for treatments that require changes in daily routines (diabetes and obesity). Data collection occurred between March 2009 and December 2012, after the study had been approved by the institutions’ Ethics Committees/Direction Boards. Using the non-probabilistic convenience sampling method, children and adolescents who attended medical routine appointments in the period of sample collection were screened by health professionals based on their medical records and those who met the aforementioned inclusion criteria were invited to participate in the study. The study’s aims and procedures were explained in detail, and written consent forms were obtained from all parents and adolescents older than 13 years; younger children provided verbal informal assent. Children and adolescents who agreed to participate completed the self-report questionnaires in the institution they attended under the supervision of a trained research assistant, and their parents were asked to complete a sociodemographic and clinical datasheet.
Psychological problems To assess the children’s and adolescents’ psychological problems, we used the Difficulties scale of the Portuguese self-rated version of the Strengths and Difficulties Questionnaire (SDQ) , which is a brief behavioral questionnaire that was developed with reference to the main nosological categories recognized by the Diagnostic and Statistical Manual of Mental Disorders, 4 th edition (DSM-IV) . This scale comprises 20 items assessing emotional symptoms (e.g., “I am often unhappy, down-hearted or tearful”), conduct problems (e.g., “I fight a lot. I can make other people do what I want”), hyperactivity/inattention (e.g., “I am easily distracted, I find it difficult to concentrate”) and peer relationship problems (e.g., “Other children or young people pick on me or bully me”). The items were answered using a Likert-type response scale with three options (0 = not true , 1 = somewhat true , and 2 = certainly true ), providing a Total Difficulties score, with higher values indicating more psychological problems. Adequate reliability was obtained for the Difficulties scale (α = .76). The Total Difficulties sum score was dichotomized based on published cut-off values to differentiate between pediatric patients within the normal (0–15 points; n = 578) and the borderline/clinical range (16–40 points; n = 158). Health-related quality of life Children’s and adolescents’ HrQoL was assessed by the Portuguese self-report short version of the DISABKIDS-10 Index . This short-form questionnaire contains 10 items measuring the physical, mental and social impact of chronic health conditions (e.g., “Does your condition get you down?”) on 8- to 18-year-old patients’ lives and is answered on a 5-point Likert scale ranging from 1 ( never ) to 5 ( always ). According to its one-dimensional factor structure, a standardized composite score (0–100), representing the physical, mental and social domains of HrQoL, was calculated, with higher scores indicating better chronic-generic HrQoL. In the current sample, the questionnaire presented good reliability, with a Cronbach’s α value of .84. Sociodemographic and clinical data Sociodemographic data were reported by the parents and included their children’s age and gender, as well as parents’ education level, occupation, family structure and psychiatric history. Using a classification system specifically developed for the Portuguese context and based on the educational level and current job of the primary caregiver , the family SES was classified into low (e.g., unqualified employees in construction or manufacturing without completing the 9 th grade of school education), medium (e.g., employees in bureaus or banks, nurses, or teachers with intermediate or university courses) and high (e.g., senior officials of government, army, commerce or industry, physicians, or engineers with bachelor’s, master’s, doctorate or other post-graduate degrees). Due to the heterogeneous distribution of the SES levels observed in our sample, this variable was dichotomized into low ( n = 464) and medium/high ( n = 272). Clinical data were provided by the parents and/or physicians and included the use of medication and specific information for each clinical group. Specifically, the physicians classified asthma severity into 4 levels ( intermittent , mild persistent , moderate persistent and severe persistent ) according to the Global Initiative for Asthma guidelines , and epilepsy severity was classified into 7 levels (from not at all severe to extremely severe ) using the Global Assessment of Severity of Epilepsy Scale . Levels of function in patients with cerebral palsy were classified into 5 levels according to the Gross Motor Function Classification System . Values of glycated hemoglobin (HbA1c) at the time of assessment were obtained for patients with diabetes. The weight and height of patients with obesity were obtained from the parents and/or nutritionists. These markers of disease severity were dichotomized into mild (including patients with intermittent asthma, not at all to somewhat severe epilepsy, cerebral patients with level I of functioning with no limitations in walking, patients with diabetes with HbA1c ≤ 8%, and overweight children; n = 333 [45.2%]) and moderate/severe (including children/adolescents with mild, moderate and severe persistent asthma, moderately to extremely severe epilepsy, patients with cerebral palsy with levels II to V with movement restriction, patients with diabetes with HbA1c > 8%, and obese children; n = 315 [42.8%]).
To assess the children’s and adolescents’ psychological problems, we used the Difficulties scale of the Portuguese self-rated version of the Strengths and Difficulties Questionnaire (SDQ) , which is a brief behavioral questionnaire that was developed with reference to the main nosological categories recognized by the Diagnostic and Statistical Manual of Mental Disorders, 4 th edition (DSM-IV) . This scale comprises 20 items assessing emotional symptoms (e.g., “I am often unhappy, down-hearted or tearful”), conduct problems (e.g., “I fight a lot. I can make other people do what I want”), hyperactivity/inattention (e.g., “I am easily distracted, I find it difficult to concentrate”) and peer relationship problems (e.g., “Other children or young people pick on me or bully me”). The items were answered using a Likert-type response scale with three options (0 = not true , 1 = somewhat true , and 2 = certainly true ), providing a Total Difficulties score, with higher values indicating more psychological problems. Adequate reliability was obtained for the Difficulties scale (α = .76). The Total Difficulties sum score was dichotomized based on published cut-off values to differentiate between pediatric patients within the normal (0–15 points; n = 578) and the borderline/clinical range (16–40 points; n = 158).
Children’s and adolescents’ HrQoL was assessed by the Portuguese self-report short version of the DISABKIDS-10 Index . This short-form questionnaire contains 10 items measuring the physical, mental and social impact of chronic health conditions (e.g., “Does your condition get you down?”) on 8- to 18-year-old patients’ lives and is answered on a 5-point Likert scale ranging from 1 ( never ) to 5 ( always ). According to its one-dimensional factor structure, a standardized composite score (0–100), representing the physical, mental and social domains of HrQoL, was calculated, with higher scores indicating better chronic-generic HrQoL. In the current sample, the questionnaire presented good reliability, with a Cronbach’s α value of .84.
Sociodemographic data were reported by the parents and included their children’s age and gender, as well as parents’ education level, occupation, family structure and psychiatric history. Using a classification system specifically developed for the Portuguese context and based on the educational level and current job of the primary caregiver , the family SES was classified into low (e.g., unqualified employees in construction or manufacturing without completing the 9 th grade of school education), medium (e.g., employees in bureaus or banks, nurses, or teachers with intermediate or university courses) and high (e.g., senior officials of government, army, commerce or industry, physicians, or engineers with bachelor’s, master’s, doctorate or other post-graduate degrees). Due to the heterogeneous distribution of the SES levels observed in our sample, this variable was dichotomized into low ( n = 464) and medium/high ( n = 272). Clinical data were provided by the parents and/or physicians and included the use of medication and specific information for each clinical group. Specifically, the physicians classified asthma severity into 4 levels ( intermittent , mild persistent , moderate persistent and severe persistent ) according to the Global Initiative for Asthma guidelines , and epilepsy severity was classified into 7 levels (from not at all severe to extremely severe ) using the Global Assessment of Severity of Epilepsy Scale . Levels of function in patients with cerebral palsy were classified into 5 levels according to the Gross Motor Function Classification System . Values of glycated hemoglobin (HbA1c) at the time of assessment were obtained for patients with diabetes. The weight and height of patients with obesity were obtained from the parents and/or nutritionists. These markers of disease severity were dichotomized into mild (including patients with intermittent asthma, not at all to somewhat severe epilepsy, cerebral patients with level I of functioning with no limitations in walking, patients with diabetes with HbA1c ≤ 8%, and overweight children; n = 333 [45.2%]) and moderate/severe (including children/adolescents with mild, moderate and severe persistent asthma, moderately to extremely severe epilepsy, patients with cerebral palsy with levels II to V with movement restriction, patients with diabetes with HbA1c > 8%, and obese children; n = 315 [42.8%]).
The statistical analyses were conducted with the Statistical Package for the Social Sciences (SPSS v.20.0; IBM Corp., Armonk, NY). Except for sociodemographic and clinical variables, missing data that were random (Little’s Missing Completely at Random [MCAR] Tests: χ (12) 2 = 16.84, p = .16 for the SDQ Emotional Symptoms scale; χ (12) 2 = 12.29, p = .42 for the SDQ Conduct Problems scale; χ (8) 2 = 9.28, p = .32 for the SDQ Hyperactivity/Inattention scale; χ (12) 2 = 5.14, p = .95 for the SDQ Peer Relationship Problems scale; and χ (54) 2 = 39.68, p = .93 for the DISABKIDS-10 Index) and less than 5% of the values were replaced with the individual mean score for each variable. Descriptive statistics were obtained for sociodemographic and clinical variables and the sample characteristics were compared between diagnostic groups by using one-way analyses of variance (ANOVA) followed by post-hoc pairwise comparisons with Bonferroni correction for continuous variables and by χ 2 -tests for categorical variables. The classification-tree model was performed using the Chi-Squared Automatic Interaction Detection (CHAID) algorithm, which is a classification method that also accounts for interactions between the predictors . The CHAID method is based on chi-squared tests (χ 2 ) that compare the squared deviations between observed and expected frequencies, with Bonferroni adjusted p -values for multiple comparisons with an overall error rate of .05 . The algorithm starts by selecting the predictor that best discriminates the dependent variable (i.e., the predictor with the lowest p -value in the Chi-squared tests) and then splits the data set into two or more nodes (parent node). Subsequently, the method splits the new nodes into smaller nodes (child nodes) based on the variable that best discriminates each of them. The method ends when no more significant dependence relationships can be found between the dependent variable and the set of predictors. In the current study, nine eligible risk factors were included (age group, gender, diagnosis, HrQoL scores, medication, disease severity, SES, family structure and caregiver’s psychiatric history), and the target category was defined as having borderline/clinically significant psychological problems. The multicollinearity across the predictors was tested with preliminary correlation analysis and Variance Inflation Factor (VIF)/ Tolerance values. Because significant differences in most risk factors and the outcome variable were found between diagnostic groups, the health condition was forced into the model as the first split variable. Continuous independent variables were automatically recoded into discrete categories with minimal loss of information (interval = 10). The categories of each independent variable were merged if they were not significantly different with respect to the dependent variable; multiple branches were generated if the categories differed significantly ( p ≤ .05). The minimum sample size for parent and child nodes were specified as n = 40 and n = 20, respectively, and the tree depth was limited to 4 levels. The likelihood ratio χ 2 -statistic was used because of the sample size < 1000 cases. The cost of misclassifying a child/adolescent with a high risk of psychological problems as low risk was customized as twice the cost of misclassifying a low risk patient as high risk. To assess the predictive accuracy of the decision-tree model, the Receiver Operating Characteristic (ROC) curve was examined by plotting sensitivity (i.e., the proportion of patients with psychological problems who were correctly identified as positive) in the function of 1 –specificity (i.e., the proportion of patients without psychological problems who were correctly identified as negative) for different cut-off points of a parameter . The statistic used to summarize the ROC analysis was the area under the curve (AUC), which corresponds to the probability that a randomly selected patient with borderline/clinical psychological symptoms was identified by the classification-tree model as higher risk than a randomly selected patient without psychological problems. AUC values between 0.90–1.00 were considered to have excellent accuracy, 0.80–0.90 were considered good, 0.70–0.80 were considered fair, 0.60–0.70 were considered poor, and 0.50–0.60 were considered failed .
Descriptive statistics and analyses of variance After excluding 115 cases due to comorbidities with other chronic conditions and 8 cases due to missing values in a ratio > 5% of the data, the final sample comprised 736 children/adolescents with asthma ( n = 303), epilepsy ( n = 98), cerebral palsy ( n = 89), type-1 diabetes ( n = 84) or overweight/obesity ( n = 162). The sociodemographic and clinical characteristics for each diagnostic group are presented in . The diagnostic groups differed significantly in terms of the patients’ age, gender and family structure. Bonferroni post-hoc analyses showed that patients with obesity were significantly older than patients with epilepsy, with a mean difference ( MD ) of 1.06 ( p = .03). Pairwise comparisons with χ 2 -tests showed that the asthma group included significantly more boys compared to the diabetes [χ 2 (1) = 10.44, p < .01] and obesity groups [χ 2 (1) = 15.06, p < .01] and that children and adolescents with diabetes or obesity belonged to a single-parent household less often than did patients with asthma [χ 2 (1) = 7.45, p < .01; χ 2 (1) = 7.61, p < .01], epilepsy [χ 2 (1) = 11.73, p < .01; χ 2 (1) = 12.50, p < .01] or cerebral palsy [χ 2 (1) = 7.95, p < .01; χ 2 (1) = 7.42, p < .01]. No significant differences between diagnostic groups were found for family SES or the primary caregiver’s psychiatric history. Regarding the clinical characteristics of the sample, patients with epilepsy were medicated less often than those with asthma [χ 2 (1) = 18.20, p < .01] or diabetes were [χ 2 (1) = 11.01, p < .01], whereas children and adolescents with cerebral palsy or obesity used medication less often than did patients with asthma [χ 2 (1) = 248.82, p < .01; χ 2 (1) = 267.72, p < .01], epilepsy [χ 2 (1) = 78.44, p < .01; χ 2 (1) = 87.87, p < .01] or diabetes [χ 2 (1) = 105.74, p < .01; χ 2 (1) = 115.69, p < .01]. In addition, pediatric asthma patients had better HrQoL than patients with cerebral palsy ( MD = 8.29, p < .01) or obesity ( MD = 4.51, p = .03) and fewer borderline/clinically significant psychological symptoms compared to children/adolescents with epilepsy [χ 2 (1) = 14.80, p < .01] or obesity [χ 2 (1) = 4.54, p = .04], whereas patients with diabetes had fewer psychological symptoms than the epilepsy group [χ 2 (1) = 8.72, p < .01]. In terms of the specific clinical characteristics of the sample, more than half of pediatric asthma patients had intermittent asthma ( n = 166, 54.8%); the mean severity score for patients with epilepsy was 2.63, SD = 1.31; the great majority of children and adolescents with cerebral palsy had a mild form of the condition, including spastic subtypes ( n = 77, 91.7%) with no limitations in walking ( n = 56, 62.9% at functional level I) and a mean IQ of 93.11, SD = 18.30 (range from 70 to 147); the mean glycated hemoglobin level of children with diabetes was 7.91, SD = 2.30%; and the standardized body mass index (zBMI) of children and adolescents with overweight/obesity was M = 1.90, SD = 0.41. Decision-tree model predicting borderline/clinical psychological symptoms The preliminary analyses revealed weak to moderate correlations between the risk factors, with correlation coefficients ranging from .00 (between age group and disease severity) to -.55 (between diagnosis and medication). In addition the Variance Inflation Factor (VIF) values ranged from 1.03 to 1.78 (mean VIF = 1.08) and the Tolerance values were higher than .56, indicating that the model was not limited by multicollinearity problems. The decision-tree model predicting borderline/clinical psychological symptoms is shown in . The overall accuracy, i.e., the percentage of cases that were correctly predicted by the model was 78.1% (sensitivity = .72; specificity = .80; risk of misclassification = .28, SE = .02), with 11 terminal nodes derived from six predictors (diagnosis, HrQoL scores, family SES, family structure, age group, and use of medication). Children’s and adolescents’ gender, disease severity and caregiver’s psychiatric history were also included in the model as risk factors, but as they were non-significant predictors of psychological problems, they were automatically excluded from the model. Beyond diagnosis, which was forced into the model as the first split variable, HrQoL was the most important predictor of psychological problems among pediatric patients, generating two high-risk profiles that predicted more than 50.0% of borderline/clinical cases: 71.7% of children and adolescents with cerebral palsy, epilepsy or obesity and HrQoL standardized scores below the threshold of 57.5 points and 57.4% of patients with asthma or diabetes and HrQoL scores below 70 points had borderline/clinically significant psychological symptoms. In addition, two medium-risk profiles were derived from the interaction between HrQoL scores and use of medication in pediatric patients with cerebral palsy, epilepsy or obesity: 36.4% of children/adolescents with HrQoL scores between 57.5–80 points and 45.5% of patients with HrQoL scores between 80–87.5 and using medication presented borderline/clinical psychological symptoms. Within the group with asthma or diabetes and HrQoL scores between 70–87.5, children and adolescents with low SES presented a greater likelihood of psychological problems (22.0%) compared to those with medium/high SES (8.2%); for patients with medium/high SES, the last significant predictor was age group, with younger children (14.8%) being at higher risk for psychological problems than adolescents (0.0%). For pediatric asthma or diabetes patients with HrQoL scores above 87.5, the last predictor was family structure, with children and adolescents living in a single-parent household being more likely to present psychological problems (10.3%) than their peers living in a two-parent family (1.9%). For patients with cerebral palsy, epilepsy and obesity and HrQoL scores above 87.5, no additional significant dependence relationships were found between the set of predictors and the outcome variable. ROC curve for the decision-tree model The ROC curve for the decision-tree model is presented in . The decision-tree model presented good predictive accuracy, with AUC = .84 (95% CI = .80/.87).
After excluding 115 cases due to comorbidities with other chronic conditions and 8 cases due to missing values in a ratio > 5% of the data, the final sample comprised 736 children/adolescents with asthma ( n = 303), epilepsy ( n = 98), cerebral palsy ( n = 89), type-1 diabetes ( n = 84) or overweight/obesity ( n = 162). The sociodemographic and clinical characteristics for each diagnostic group are presented in . The diagnostic groups differed significantly in terms of the patients’ age, gender and family structure. Bonferroni post-hoc analyses showed that patients with obesity were significantly older than patients with epilepsy, with a mean difference ( MD ) of 1.06 ( p = .03). Pairwise comparisons with χ 2 -tests showed that the asthma group included significantly more boys compared to the diabetes [χ 2 (1) = 10.44, p < .01] and obesity groups [χ 2 (1) = 15.06, p < .01] and that children and adolescents with diabetes or obesity belonged to a single-parent household less often than did patients with asthma [χ 2 (1) = 7.45, p < .01; χ 2 (1) = 7.61, p < .01], epilepsy [χ 2 (1) = 11.73, p < .01; χ 2 (1) = 12.50, p < .01] or cerebral palsy [χ 2 (1) = 7.95, p < .01; χ 2 (1) = 7.42, p < .01]. No significant differences between diagnostic groups were found for family SES or the primary caregiver’s psychiatric history. Regarding the clinical characteristics of the sample, patients with epilepsy were medicated less often than those with asthma [χ 2 (1) = 18.20, p < .01] or diabetes were [χ 2 (1) = 11.01, p < .01], whereas children and adolescents with cerebral palsy or obesity used medication less often than did patients with asthma [χ 2 (1) = 248.82, p < .01; χ 2 (1) = 267.72, p < .01], epilepsy [χ 2 (1) = 78.44, p < .01; χ 2 (1) = 87.87, p < .01] or diabetes [χ 2 (1) = 105.74, p < .01; χ 2 (1) = 115.69, p < .01]. In addition, pediatric asthma patients had better HrQoL than patients with cerebral palsy ( MD = 8.29, p < .01) or obesity ( MD = 4.51, p = .03) and fewer borderline/clinically significant psychological symptoms compared to children/adolescents with epilepsy [χ 2 (1) = 14.80, p < .01] or obesity [χ 2 (1) = 4.54, p = .04], whereas patients with diabetes had fewer psychological symptoms than the epilepsy group [χ 2 (1) = 8.72, p < .01]. In terms of the specific clinical characteristics of the sample, more than half of pediatric asthma patients had intermittent asthma ( n = 166, 54.8%); the mean severity score for patients with epilepsy was 2.63, SD = 1.31; the great majority of children and adolescents with cerebral palsy had a mild form of the condition, including spastic subtypes ( n = 77, 91.7%) with no limitations in walking ( n = 56, 62.9% at functional level I) and a mean IQ of 93.11, SD = 18.30 (range from 70 to 147); the mean glycated hemoglobin level of children with diabetes was 7.91, SD = 2.30%; and the standardized body mass index (zBMI) of children and adolescents with overweight/obesity was M = 1.90, SD = 0.41.
The preliminary analyses revealed weak to moderate correlations between the risk factors, with correlation coefficients ranging from .00 (between age group and disease severity) to -.55 (between diagnosis and medication). In addition the Variance Inflation Factor (VIF) values ranged from 1.03 to 1.78 (mean VIF = 1.08) and the Tolerance values were higher than .56, indicating that the model was not limited by multicollinearity problems. The decision-tree model predicting borderline/clinical psychological symptoms is shown in . The overall accuracy, i.e., the percentage of cases that were correctly predicted by the model was 78.1% (sensitivity = .72; specificity = .80; risk of misclassification = .28, SE = .02), with 11 terminal nodes derived from six predictors (diagnosis, HrQoL scores, family SES, family structure, age group, and use of medication). Children’s and adolescents’ gender, disease severity and caregiver’s psychiatric history were also included in the model as risk factors, but as they were non-significant predictors of psychological problems, they were automatically excluded from the model. Beyond diagnosis, which was forced into the model as the first split variable, HrQoL was the most important predictor of psychological problems among pediatric patients, generating two high-risk profiles that predicted more than 50.0% of borderline/clinical cases: 71.7% of children and adolescents with cerebral palsy, epilepsy or obesity and HrQoL standardized scores below the threshold of 57.5 points and 57.4% of patients with asthma or diabetes and HrQoL scores below 70 points had borderline/clinically significant psychological symptoms. In addition, two medium-risk profiles were derived from the interaction between HrQoL scores and use of medication in pediatric patients with cerebral palsy, epilepsy or obesity: 36.4% of children/adolescents with HrQoL scores between 57.5–80 points and 45.5% of patients with HrQoL scores between 80–87.5 and using medication presented borderline/clinical psychological symptoms. Within the group with asthma or diabetes and HrQoL scores between 70–87.5, children and adolescents with low SES presented a greater likelihood of psychological problems (22.0%) compared to those with medium/high SES (8.2%); for patients with medium/high SES, the last significant predictor was age group, with younger children (14.8%) being at higher risk for psychological problems than adolescents (0.0%). For pediatric asthma or diabetes patients with HrQoL scores above 87.5, the last predictor was family structure, with children and adolescents living in a single-parent household being more likely to present psychological problems (10.3%) than their peers living in a two-parent family (1.9%). For patients with cerebral palsy, epilepsy and obesity and HrQoL scores above 87.5, no additional significant dependence relationships were found between the set of predictors and the outcome variable.
The ROC curve for the decision-tree model is presented in . The decision-tree model presented good predictive accuracy, with AUC = .84 (95% CI = .80/.87).
The decision-tree emerging from this study can be considered an innovative generic risk assessment tool to identify pediatric patients at greater risk of developing borderline/clinically significant psychological symptoms based on the DISABKIDS-10 Index as a screening HrQoL measure and a brief questionnaire assessing patients’ and families’ sociodemographic and clinical characteristics. The decision-tree model presented adequate validity and accuracy and allowed the identification of two high-risk (children/adolescents with asthma or diabetes and HrQoL scores below the threshold of 70.0; patients with epilepsy, cerebral palsy or obesity and HrQoL scores below 57.5) and two medium-risk profiles (children/adolescents with epilepsy, cerebral palsy or obesity and HrQoL between 57.5–80.0; children/adolescents with epilepsy, cerebral palsy or obesity, HrQoL between 80.0–87.5 and taking medication) that predicted 71.5% of borderline/clinical cases. Thus, these four groups of patients must be given priority regarding referral to specialized psychological assessment and interventions in healthcare settings. In the total sample, 21.5% of children and adolescents presented borderline/clinically significant psychological problems according to the SDQ cut-off values. This prevalence of patient-reported psychological problems is comparable to population norms from several European countries [ – ]. However, the prevalence of borderline/clinical psychological problems differed between diagnostic groups, with patients with epilepsy or obesity presenting significantly more symptoms than those with asthma or diabetes. In addition, patients with cerebral palsy or obesity had lower HrQoL scores than children and adolescents with asthma. Previous research has shown that children and adolescents with neurological disorders are more likely to have impaired HrQoL and more psychopathological problems compared with patients with other conditions with no impairment to the central nervous system . Recent studies have also documented poorer HrQoL and higher levels of depressive symptoms in overweight children and adolescents compared with normal-weight peers or patients with other chronic conditions . These findings may be explained by the weight self-stigma, victimization by peers, weight-related teasing and social marginalization that is often experienced by children and adolescents with obesity and that is reflected in their self-assessments of generic HrQoL as a multidimensional construct that includes social functioning and well-being. The CHAID algorithm clustered the group of children and adolescents with neurological disorders and obesity into the same node, in which 27.2% of patients had borderline/clinically significant psychological problems, in contrast with a prevalence of 16.3% in the asthma/diabetes node. The results indicated that even after controlling for diagnosis, the best predictor of borderline/clinical psychological problems was the child-reported HrQoL. However, the classification-tree model defined different cut-off points for the DISABKIDS-10 scores depending on the diagnostic group: children and adolescents with asthma or diabetes are at a greater risk of borderline/clinical psychological problems if they had HrQoL scores below 70.0, whereas patients with epilepsy, cerebral palsy or obesity were at high risk if they presented HrQoL scores below 57.5 and medium risk if they had HrQoL scores between 57.5–80.0. Patients with epilepsy, cerebral palsy or obesity and HrQoL scores between 80.0–87.5 were also at medium risk if they were taking medication. These cut-off points reflect the well-documented differences in HrQoL mean scores across chronic health conditions because neurological conditions and obesity are expected to have severe detrimental effects on HrQoL; thus, only very low HrQoL scores would be indicative of greater risk for comorbid psychological problems. Consistent with previous studies [ , , ], SES levels, family structure and patients’ age group were included as significant risk factors in the decision-tree model. However, although the CHAID algorithm subdivided the group of patients with asthma or diabetes and HrQoL scores above 70.0 based on significant differences between the sociodemographic groups, the proportion of borderline/clinical cases predicted by the generated profiles resembled those reported for normative samples. Male gender has been associated with higher levels of externalizing problems but not internalizing problems . Thus, it was automatically excluded as a risk factor in our decision-tree model because we considered psychological problems in general without differentiating between internalizing and externalizing symptoms. Our findings should be interpreted with caution due to some limitations in the study’s design and procedures. First, the study’s cross-sectional design precluded the establishment of causal links among variables. Thus, we cannot claim that lower HrQoL levels predict an increased likelihood of developing borderline/clinically significant psychological problems. Instead, we suggest that HrQoL and psychological problems should be considered two indicators of the broader construct of psychosocial adaptation that covary . Prospective studies are needed to test whether clinically significant changes in HrQoL would result in enhanced mental health. The second limitation was the non-probabilistic sample collection method and the consequent heterogeneous distribution of patients’ and parents’ sociodemographic and clinical characteristics. Specifically, the low percentage of participants with high SES and SDQ scores within the clinical range required the dichotomization of these variables; in addition, the disease severity markers were assessed with condition-specific scales and subsequently dichotomized, which increased the intragroup variability. A third limitation involves the use of the SDQ Total Difficulties score. Although there are recent recommendations for classifying SDQ scores into internalizing and externalizing problems in low-risk samples , and there is empirical evidence that specific diagnostics are differentially associated with emotional or behavioral problems , cut-off points to differentiate SDQ scores into normal, borderline and clinical internalizing and externalizing problems are unavailable. Fourth, the small sample size for each clinical group prevented the use of stratified analyses for specific diagnosis, because the CHAID algorithm uses multiway splits, and thus it needs rather large sample sizes to ensure accuracy. Fifth, other risk factors, such as parents’ education, number of children/siblings in the family, family routines and parents’ psychological symptoms and quality of life, were not included in the present study because the predictive accuracy of the CHAID algorithm is highly dependent of the sample size and number of predictors, and should be considered in further research with larger samples of children/adolescents with chronic health conditions. Sixth, this study relied solely on child-reported data and excluded parent-reports, although parents may be more prone to report worse HrQoL and more psychological problems than children and adolescents are . Finally, the DISABKIDS-10 Index is a generic cross-cultural HrQoL instrument, but only Portuguese children and adolescents with asthma, cerebral palsy, type-1 diabetes, epilepsy or overweight/obesity were included. Thus, the established cut-off points for interpreting HrQoL might be sample dependent. In addition, condition-specific instruments may be more sensitive to specific aspects of functioning that may be impacted directly by the condition and/or its treatments . Future research should replicate this classification-tree model in larger samples of patients from different countries and cultural backgrounds and with other specific chronic health conditions (e.g., pediatric cancer, juvenile rheumatoid arthritis, dermatitis, etc.) using both patient- and parent-reported data and combining generic and condition-specific instruments as screeners to establish the cross-cultural validity of HrQoL cut-off points and to maximize the sensitivity of the risk assessment tool. Further studies should also consider the examination of stratified classification-tree models according to specific diagnosis, including condition-specific HrQoL instruments as screeners as well as specific indicators of disease severity and/or control of symptoms. Despite these limitations, this is the first study to examine the interplay among clinical, sociodemographic and HrQoL data to predict the risk of developing borderline/clinically significant psychological problems, facilitating the development of a screening tool for healthcare providers. There is a growing need to identify at-risk populations in healthcare services, but the limitations of commonly used statistical methods and financial/human resource constraints have made this difficult. Specifically, logistic regression analyses have been traditionally used to determine the average effect of risk factors on the likelihood of developing borderline/clinical psychological problems and, thus, they were geared toward the average member of the population, with little sensitivity to potential differences between population subgroups . In contrast, the classification-tree analysis is a nonparametric technique that can be used without constraints regarding the functional form of the data and that accounts for interaction effects between predictors (i.e., it allows the identification of variables that may have a predictive value only for specific clusters of pediatric patients). The major advantage of using decision-tree modeling in our study was its ability to identify high-risk subgroups of pediatric patients through the segmentation of populations into meaningful and mutually exclusive subgroups whose members share similar clinical and sociodemographic characteristics that are associated with psychological problems. Decision-trees are visual, structured/hierarchized and easy to understand; these are requirements for routine monitoring and screening in healthcare settings. In addition, the variables that emerged as predictors of borderline/clinical cases are practical for use in a risk assessment tool because brief patient-report HrQoL measures and sociodemographic and clinical data can be easily and promptly obtained by any healthcare provider. In clinical practice, this decision-tree approach would consist of a series of questions that would follow various iterations as necessary, similar to many common diagnostic tools. According to our decision-tree model, the specific diagnostic would determine the influence of all other variables and must be considered first, followed by HrQoL assessment. Another major practical implication of this study is the establishment of condition-specific, clinically meaningful, cut-off points to interpret HrQoL data as assessed by the chronic-generic DISABKIDS-10 Index, which is an essential measurement property when using HrQoL instruments in clinical decision-making . This cost-effective generic risk assessment tool may contribute to overcoming barriers to psychosocial screening and monitoring and may become a useful tool to identify at-risk children and adolescents. For children and adolescents classified into low-risk profiles (less than 25% odds of presenting borderline/clinical psychological symptoms) clinical management should be based on other factors. However, the classification of a pediatric patient into a medium- or high-risk profile (25%-50% and more than 50% odds of presenting borderline/clinical psychological symptoms, respectively) should have a substantial impact on the clinical management and allocation of health resources, particularly with regard to referral for specialized psychological assessment and intervention. In addition to allowing healthcare providers to easily identify children and adolescents who are most likely to present borderline/clinically significant psychological problems, our findings suggest that psychosocial interventions aimed at improving generic physical, psychological and social functioning may be effective in promoting the mental health of pediatric patients.
S1 File Data set ( N = 736). (XLSX) Click here for additional data file.
|
Factors associated with non-pathogenic antibodies against desmoglein-3 in pemphigus foliaceus | 8aa06661-8f1f-4961-b99e-4658a193ec04 | 11342976 | Anatomy[mh] | Pemphigus are autoimmune diseases, in which intraepidermal acantholytic bullae are caused by the deposition of serum antibodies in desmogleins (Dsg), components of the epidermal desmosomes. , , Pemphigus vulgaris (PV) and pemphigus foliaceus (PF) are prevalent in southeastern Brazil. , In PV, patients are usually elderly, with initial mucosal involvement, and have autoantibodies against Dsg3. In the mucocutaneous form of PV, there is also the production of anti-Dsg1. PF (also known as Fogo Selvagem [FS] in Brazil) affects exclusively the skin of young adults, with anti-Dsg1 production. , , , , Dsg1 and 3 are transmembrane proteins of 160kDa and 130kDa, respectively, expressed by genes located on chromosome 18. Dsgs compensation theory explains the level of intraepidermal acantholysis in pemphigus. Anti-Dsg1 antibodies in PF cause subgranular acantholysis, as Dsg1 is expressed more intensily in the upper layers of the epidermis. Anti-Dsg3 antibodies cause suprabasal acantholysis, as Dsg3 is most often expressed in the lower layers of the epidermis, and in all epithelial layers in the mucous membranes. Moreover, as there is a lower expression of Dsg1 in the mucosa, it is not affected by anti-Dsg1 antibodies in PF, as Dsg3 compensates for epithelial cell cohesion in the mucosa. Subsequently, the hypothesis of Dsgs compensation, to explain the level of cleavage of epidermal acantholysis related to the production of anti-Dsg1 and 3, started to be discussed again, since exclusively suprabasal acantholysis in PV would not be explained in the mucocutaneous form, when anti-Dsg1 is also present. Other target molecules of epidermal adhesion are now identified in the pathogenesis of pemphigus, as well as mitochondrial proteins, cholinergic receptors, and other molecules, which act synergistically with anti-Dsg1 and Dsg3 antibodies in epidermal acantholysis. , , , , , , , , , As both PV and PF are prevalent in southeastern Brazil, there is an opportunity for outpatient follow-up of a large series of patients. , In the PF series, over a 25-year period, it was observed that 6,64 % of the patients, in addition to being positive for anti-Dsg1, also had reactive titers to Dsg3 in the ELISA test, without, however, showing a phenotype of mucosal lesions. Therefore, we aimed to analyze clinical-laboratory data in the group of patients with PF and reactivity to Dsg3 (PF-anti-Dsg3+). For this purpose, three groups were compared: PF-anti-Dsg3+, PF with anti-Dsg3 negativity and treatment-naïve [PF-anti-Dsg3(˗)], and treatment-naïve PV. For specific purposes, the following were evaluated: (i) Demographic and clinical data; (ii) Temporal evolution of anti-Dsg1 and 3 titers by ELISA and identification of Dsg1 and 3 by immunoblotting (IB) in the PF-anti-Dsg3+ group; and (iii) Expression of Dsg1 and 3 in paraffin-embedded skin biopsies by immunohistochemistry (IHC) and (iv) typification of HLA-DRB1 alleles associated with PF and PV in the PF-anti-Dsg3+ and PF-anti-Dsg3 groups. This is a comparative analytical study of three groups. The sample bank of the Dermatology Laboratory, Hospital das Clínicas, FMRP-USP, has the approval of the Research Ethics Committee (HCRP Process number 3605/2006). Patients and control individuals signed the free and informed consent form at the time of sample collection. Case series It consisted of 80 patients: 16 PF-anti-Dsg3+, 42 PF-anti-Dsg3 and 22 PV cases. The clinical findings were confirmed by histopathological examination and by direct immunofluorescence (DIF) and/or indirect immunofluorescence (IIF). Demographic, clinical and laboratory data (anti-Dsg1 and Dsg3 by ELISA and HLA alleles) were obtained from data recorded at the Laboratory of Dermatology. In the PF-anti-Dsg3+ group, three (20%) of the 15 patients (with information about treatment) were untreated at the time of blood collection, while 42 of the PF-anti-Dsg3 patients and 22 PV were all treatment-naïve. Immunoblotting (IB) assays were performed with 12 (75%) of the 16 serum samples from the PF-anti-Dsg3+ group, with some patients having more than one sample analyzed, collected at different times. IHC was performed on sections from the paraffin blocks of four patients from the PF-anti-Dsg3+ group and one from the PF-anti-Dsg3 group. ELISA for the detection of IgG antibodies against Dsg1 and 3 The manufacturers recommendations were followed. Values ≥ 20 U/mL, between 9‒20 U/mL and < 9 U/mL were considered positive, indeterminate and negative, respectively (MBL, Nagoya, Japan). IB with human epidermis extract and patient serum Briefly, skin fragments were incubated for 48 h in PBS with EDTA (Merck) and PMSF (Sigma-Aldrich) to separate the epidermis from the dermis. Protein extraction from the epidermis was carried out with a Tris-HCL, SDS, 2-mercaptoethanol solution, and then an EDTA, PMSF and proteinase inhibitor cocktail (Sigma-Aldrich) were added. After the electrophoretic run, the polyacrylamide gel was assembled in a sandwich with a 0.45 µm nitrocellulose membrane (BioRad) for protein transfer. After blocking with 3% skimmed milk in TBS, and consecutive washings, the serum samples (1:20) were incubated on nitrocellulose strips, followed by incubation with anti-human HRP IgG secondary antibody (BioRad). After washing, color reagent (Color-Plus HRP, BioRad) was added for colorimetric development. Expression of Dsgs1 and 3 in paraffin-embedded samples with IHC For IHC assays, the manufacturers recommendations were followed (HRP-Polymer MACH1, Biocare Medical, Concord, CA, USA). Briefly, 5μm sections were obtained and after antigen retrieval in citrate, inhibition of endogenous peroxidase and use of a blocking solution, were incubated with anti-Dsg1 and 3 monoclonal antibodies produced in mice (Abcam, MA, USA). After washing and incubation with Probe and HRP polymer, the reaction was terminated with diaminobenzidine (DAB). Statistical analysis The GraphPad Prism 9.2.0 software was used for statistical analysis and the generation of graphs. Frequencies between the groups were analyzed using Fisher or Chi-Square tests, and numerical data were analyzed using Kruskal-Wallis tests followed by Dunn’s multiple comparisons. The result was considered significant when p≤0.05. It consisted of 80 patients: 16 PF-anti-Dsg3+, 42 PF-anti-Dsg3 and 22 PV cases. The clinical findings were confirmed by histopathological examination and by direct immunofluorescence (DIF) and/or indirect immunofluorescence (IIF). Demographic, clinical and laboratory data (anti-Dsg1 and Dsg3 by ELISA and HLA alleles) were obtained from data recorded at the Laboratory of Dermatology. In the PF-anti-Dsg3+ group, three (20%) of the 15 patients (with information about treatment) were untreated at the time of blood collection, while 42 of the PF-anti-Dsg3 patients and 22 PV were all treatment-naïve. Immunoblotting (IB) assays were performed with 12 (75%) of the 16 serum samples from the PF-anti-Dsg3+ group, with some patients having more than one sample analyzed, collected at different times. IHC was performed on sections from the paraffin blocks of four patients from the PF-anti-Dsg3+ group and one from the PF-anti-Dsg3 group. The manufacturers recommendations were followed. Values ≥ 20 U/mL, between 9‒20 U/mL and < 9 U/mL were considered positive, indeterminate and negative, respectively (MBL, Nagoya, Japan). Briefly, skin fragments were incubated for 48 h in PBS with EDTA (Merck) and PMSF (Sigma-Aldrich) to separate the epidermis from the dermis. Protein extraction from the epidermis was carried out with a Tris-HCL, SDS, 2-mercaptoethanol solution, and then an EDTA, PMSF and proteinase inhibitor cocktail (Sigma-Aldrich) were added. After the electrophoretic run, the polyacrylamide gel was assembled in a sandwich with a 0.45 µm nitrocellulose membrane (BioRad) for protein transfer. After blocking with 3% skimmed milk in TBS, and consecutive washings, the serum samples (1:20) were incubated on nitrocellulose strips, followed by incubation with anti-human HRP IgG secondary antibody (BioRad). After washing, color reagent (Color-Plus HRP, BioRad) was added for colorimetric development. For IHC assays, the manufacturers recommendations were followed (HRP-Polymer MACH1, Biocare Medical, Concord, CA, USA). Briefly, 5μm sections were obtained and after antigen retrieval in citrate, inhibition of endogenous peroxidase and use of a blocking solution, were incubated with anti-Dsg1 and 3 monoclonal antibodies produced in mice (Abcam, MA, USA). After washing and incubation with Probe and HRP polymer, the reaction was terminated with diaminobenzidine (DAB). The GraphPad Prism 9.2.0 software was used for statistical analysis and the generation of graphs. Frequencies between the groups were analyzed using Fisher or Chi-Square tests, and numerical data were analyzed using Kruskal-Wallis tests followed by Dunn’s multiple comparisons. The result was considered significant when p≤0.05. Demographic and clinical data There was no statistical difference regarding gender in the three groups. As for age groups, the PF-anti-Dsg3 group was younger than the PV group (p = 0.0522). There was no statistical difference when comparing the PF-anti-Dsg3+ with the PF-anti-Dsg3 and the PV group (p = 0.5361 and p > 0.9999, respectively). The generalized form predominated in the PF-anti-Dsg3+ group, when compared to the PF-anti-Dsg3 group (p = 0.002). Serological data Regarding anti-Dsg3 antibodies, the PF-anti-Dsg3+ group had lower titers than the PV group (p < 0.0001; ). The distribution of anti-Dsg3 titers in the PF-anti-Dsg3+ group is depicted in A. B shows slight variation over time in anti-Dsg3 titers in six patients. In the IB performed with sera from 12 patients, three recognized Dsg1, and one recognized Dsg3 . PF10, which recognized Dsg3 in the IB, showed the 2nd highest anti-Dsg3 titer in the ELISA ( A). Expression of Dsgs 1 and 3 with IHC IHC showed internalization of Dsg1, with coarse clumps inside the cytoplasm, sometimes perinuclear, in skin samples from patients in the PF-anti-Dsg3 ( A) and PF-anti-Dsg3+ ( B). Dsg3 expression occurred in the keratinocyte envelope and intracytoplasmically, with emphasis on its expression throughout the epidermis. No internalization of Dsg3 was observed. HLA-DRB1 alleles associated with PF and PV in the PF-anti-Dsg3+ and PF-anti-Dsg3 groups; ) Five (31.3%) of 16 patients in the PF-anti-Dsg3+ group, and 16 (38.1%) of 42 in the PF-anti-Dsg3 group were typed. In the PF-anti-Dsg3+ group, five patients showed one or two susceptibility alleles for PF, and none of them had alleles associated with PV. In the PF-anti-Dsg3 group, 11 had alleles of susceptibility to PF, and two patients had alleles of susceptibility to PV in heterozygous form. ) There was no statistical difference regarding gender in the three groups. As for age groups, the PF-anti-Dsg3 group was younger than the PV group (p = 0.0522). There was no statistical difference when comparing the PF-anti-Dsg3+ with the PF-anti-Dsg3 and the PV group (p = 0.5361 and p > 0.9999, respectively). The generalized form predominated in the PF-anti-Dsg3+ group, when compared to the PF-anti-Dsg3 group (p = 0.002). Regarding anti-Dsg3 antibodies, the PF-anti-Dsg3+ group had lower titers than the PV group (p < 0.0001; ). The distribution of anti-Dsg3 titers in the PF-anti-Dsg3+ group is depicted in A. B shows slight variation over time in anti-Dsg3 titers in six patients. In the IB performed with sera from 12 patients, three recognized Dsg1, and one recognized Dsg3 . PF10, which recognized Dsg3 in the IB, showed the 2nd highest anti-Dsg3 titer in the ELISA ( A). ) IHC showed internalization of Dsg1, with coarse clumps inside the cytoplasm, sometimes perinuclear, in skin samples from patients in the PF-anti-Dsg3 ( A) and PF-anti-Dsg3+ ( B). Dsg3 expression occurred in the keratinocyte envelope and intracytoplasmically, with emphasis on its expression throughout the epidermis. No internalization of Dsg3 was observed. alleles associated with PF and PV in the PF-anti-Dsg3+ and PF-anti-Dsg3 groups; ) Five (31.3%) of 16 patients in the PF-anti-Dsg3+ group, and 16 (38.1%) of 42 in the PF-anti-Dsg3 group were typed. In the PF-anti-Dsg3+ group, five patients showed one or two susceptibility alleles for PF, and none of them had alleles associated with PV. In the PF-anti-Dsg3 group, 11 had alleles of susceptibility to PF, and two patients had alleles of susceptibility to PV in heterozygous form. The production of serum autoantibodies against Dsg3 is not expected in PF, since, in its pathogenesis, only anti-Dsg1 antibodies are implicated in subgranular acantholysis, with consequent formation of flaccid bullae on the skin. The PF phenotype does not include the mucosal lesions observed in PV due to antibodies against Dsg3. , , , , However, autoantibodies against Dsg3 have been rarely reported in PF cases. , , More recently, anti-Dsg3 antibodies have been reported in other forms of pemphigus. , Arteaga et al. (2002) described anti-Dsg3 in 7% of 276 patients with PF, demonstrating that anti-Dsg3 antibodies did not show a serological cross-reaction with Dsg1. Flores et al. (2012) reported anti-Dsg3 in 40% of 101 FS sera, and in 14% of controls from an endemic region for FS in Brazil. Oliveira et al. (2016) reported anti-Dsg3 in 4% of the patients in a PF series (including patients from the northeastern region of the state of São Paulo). The present series confirmed 6.64% of patients with PF with indeterminate or positive results for anti-Dsg3 in the ELISA test. When comparing the three groups, the age range of the PF-anti-Dsg3+ group tended to be older, comparable to that of the PV group (p>0.999), and similar to that of the PF-anti-Dsg3 group. The generalized clinical form predominated in the PF-anti-Dsg3+ group (p=0.0020). Regarding the anti-Dsg3 titers of the PF-anti-Dsg3+ group, they were lower when compared to those of the PV group (p<0.0001). The measurement of anti-Dsg3 in the PV group was carried out in treatment-naïve patients, while in the PF-anti-Dsg3+ group, three (20%) of the 15 patients (with information on treatment) were not receiving treatment (PF5, PF6 and PF14). Anti-Dsg3 titers in the PF-anti-Dsg3+ group showed no difference between treated and untreated patients ( A). Two patients stand out, in the sample of 16, with higher anti-Dsg3 titers, whose blood was collected while undergoing treatment. Patient PF2, a 63-year-old male, presented the generalized form, with 14 years evolution, and had keratoacanthoma and melanoma during the follow-up. Another patient PF10, 46 years-old male, had the generalized form, and a two-year history. There were coinciding factors in both cases, in addition to higher anti-Dsg3 titers, male gender, having had the onset of PF at an older age, and the generalized form of the disease. IB confirmed the production of anti-Dsg3 only in PF10 . The internalization of Dsgs, in the process of acantholysis, is observed on IHC by the formation of intracytoplasmic and perinuclear granules in keratinocytes, when anti-Dsg1 and 3 antibodies are used in PF and PV, respectively. The internalization of Dsg1 was observed in samples from the PF-anti-Dsg3 and PF-anti-Dsg3+ groups; however, there was no internalization of Dsg3. The low anti-Dsg3 titers in the PF-anti-Dsg3+ group could justify the lack of Dsg3 internalization, as well as the absence of mucosal lesions. In exclusively cutaneous PV, low anti-Dsg3 titers could explain the absence of mucosal lesions. Although HLA-DRB1 alleles were not typed in the whole group of PF-anti-Dsg3+, exclusive alleles of susceptibility to PF were determined in the five individuals that were typed, in homozygous or heterozygous forms, without the presence of alleles associated with PV. Recently, Sielski et al. (2022) demonstrated that cases of PV that contradicted the hypothesis of Dsgs compensation were related to the absence of DRB1 alleles of susceptibility to PV. Therefore, the absence of susceptibility alleles to PV in the PF-anti-Dsg3+ group could contribute to the non-pathogenicity of antibodies against Dsg3. The production of anti-Dsg3 in PF could be justified by the phenomenon of epitope spreading ‒ patients with a specific bullous disease have non-pathogenic autoantibodies against other molecules of the epidermis that do not cause the specific bullous disease. Its pathogenesis is explained by the exposure of other epidermal molecules during the inflammation process caused by acantholysis, in the case of pemphigus. , , , In this study, IHC showed the expression of Dsg3 in all layers of the epidermis. Thus, it could justify the greater exposure of Dsg3, with consequent production of anti-Dsg3. However, this expression in all layers of the epidermis was observed in samples from both groups – PF-anti-Dsg3 and PF-anti-Dsg3+. As the PF-anti-Dsg3 group consisted of treatment-naïve patients, it is not known whether anti-Dsg3 would be produced during pemphigus evolution. Moreover, there are rare reports of patients with clinically and laboratory-proven PF who migrate to the PV phenotype, or of patients with both PF and PV characteristics. , , , However, PV migrating to PF is more commonly observed. The present case series does not show similar cases. The presence of anti-Dsg3 antibodies in PF was related to an older age group (comparable to that of PV) and the generalized form of PF. The non-pathogenicity of anti-Dsg3 antibodies in PF can be attributed to low anti-Dsg3 titers, the lack of Dsg3 internalization as seen on IHC, and the absence of PV-associated HLA-DRB1 alleles. The project was funded by 10.13039/501100001807 FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo) (#2010/51729-2), SV received a PhD scholarship from 10.13039/501100002322 CAPES (Coordination for the Improvement of Higher Education Personnel) and TAJ received a PhD scholarship from FAPESP. Sebastian Vernal: Contributed to data survey, collection and interpretation of data; responsible for blotting assays. Tamiris Amanda Julio: Contributed with data survey, collection and interpretation of data; responsible for the ELISA assays. Fernando Henrique Alves: Contributed with data survey and collection of data, and participation in the propaedeutic and therapeutic conduct, and in IB assays. Aline Turatti: Contributed with data survey, collection and interpretation of data; responsible for DIF, IIF and IHC assays. Eduardo Antonio Donadi: Contributed with data survey, collection and interpretation of data; responsible for determining HLA-DR alleles. Ana Maria Roselino: Contributed with the design and planning of the study, collection and analysis of data; drafting and editing of the manuscript; participation in the propaedeutic and/or therapeutic conduct; and critical review of the literature. All authors approved the final version of the manuscript. None declared. |
Influences of aging and mating history in males on paternity success in the red flour beetle | 5dd97e9c-0ce3-43eb-9688-84689940ee1e | 11665990 | Forensic Medicine[mh] | Sexual selection is one of the mechanisms that has a major influence on evolution . Sexual selection refers to any selection that can arise due to the differences in fitness, associated with nonrandom success in the competition for access to gametes for fertilization . For example, to obtain mating opportunities with females, males often employ exaggerated weapon traits in male-male combats. Moreover, to persuade the female to accept mating, males often display gaudy ornamental traits and exhibit unique courtship behavior. These behaviors are displayed before copulation and are referred to as “pre-copulatory sexual selection” . Conversely, the sperm of every male is not necessarily used during fertilization in polyandrous species, resulting in male-male competition for egg fertilization inside the female, and this is referred to as “sperm competition” . Since this occurs after mating, it is referred to as “post-copulatory sexual selection” . Sperm competition leads to the evolution of traits that are beneficial in this competition. In post-copulatory sexual selection, "cryptic female choice" also plays an important role in addition to sperm competition as a male-male competition, and affects male paternity success . Cryptic female choice refers to nonrandom paternity success bias by females during or after mating. However, it is very difficult to separate the effects of cryptic female choice from those of sperm competition in paternity success . Males need to invest resources for reproduction during both pre- and post-copulatory sexual selection; however, the resource allocation is influenced by various factors and can vary. Therefore, understanding these factors is important for understanding sexual selection. Male aging is one factor that affects male investment in reproductive traits . Males are often injured or die in male-male combat. Considering that early adult males are expected to obtain several reproductive opportunities in the future, they might prioritize their survival and decrease investment into costly traits for reproduction. On the contrary, older adult are expected to obtain relatively reduced reproductive opportunities in the future, and as a result might increase their investment in reproduction, even at the risk of threatening their own survival . Thus, males need to make restrained investments in reproduction during early adulthood (young male) and increase this investment during late adulthood (old male), when their chances of reproduction dwindle . A previously conducted theoretical study has reported that, in several cases, increasing investment in sexual traits with age is a stable evolutionarily strategy . Older males in the broad-horned flour beetle species Gnatocerus cornutus , fight harder for females, in male-male combat, than younger males . The old male butterfly Hypolimnas bolina is also known to fight longer than young males . Moreover, the old male butterfly Bicyclus anynana is known to achieve more reproductive success than young males . These results demonstrate that male investment in reproductive traits in pre-copulatory sexual selection increases with aging. Therefore, male investment in post-copulatory sexual selection as well as pre-copulatory sexual selection is predicted to increase with age. However, male investment in pre-copulatory sexual selection could possibly differ from their investment in post-copulatory sexual selection. Male investment into post-copulatory sexual selection often negatively correlates with investment into pre-copulatory sexual selection in several animal species . Therefore, if this resource allocation tradeoff is unaffected by male aging, then investment in post-copulatory sexual selection may decrease with male age as a consequence of male investment in pre-copulatory sexual selection increasing with their age. Conversely, independent of male investment, it is possible that male fertilization success varies with male aging. For example, considering that the quality of gametes declines with age, indicating that it is important affected to evolutionary and ecological consequences on gamete performance and fitness . In particular, age-associated accumulation of new mutations in the germline decreases the genetic quality of gametes . Therefore, regardless of the increase or decrease in investment by males, altered sperm quality caused by aging can reduce male fertilization success. Nevertheless, these previous studies focused on humans, and their applicability to other species is not known. A meta-analysis of aging with regards to ejaculatory traits found no consistent trend across all organisms. In insects, the ejaculatory traits tend to improve with age in some species; however, some other species exhibit a decline in these traits, leading to varied results. These variations are thought to be influenced by the ecology and environment of the species. Furthermore, as mentioned above, male fertilization success is not only dependent on the changes in male factors (ejaculate volume, sperm mobility, and sperm size), but also on cryptic female choice. Therefore, the actual fertilization success of males with aging must be investigated in various insect species. Moreover, few empirical studies have reported that male fertilization success also depends on their mating experience. For example, among the older males belonging to the eastern mosquitofish ( Gambusia holbrooki ) species, males who experienced several copulations demonstrated increased paternity success (The proportion of offspring sired by a specific male among the offspring produced by a single female from multiple males), while males who had not experienced copulation showed reduced paternity success . Thus, the interaction between aging and mating history of males can affect post-copulatory sexual selection. However, research focusing on the interaction between aging and mating history with regards to post-copulatory sexual selection in male insects, particularly with regards to sperm competition is lacking. The objective of this study was to investigate the influence of male aging and mating history on male paternity success (The proportion of broods that a male parent can obtain through sperm competition in the red flour beetle Tribolium castaneum) . T. castaneum infests stored products . Several aspects of its reproduction are well understood; thereby making it a major model organism in studies of post-copulatory sexual selection . T . castaneum inhabits high-density populations with both sexes indulging in frequent and promiscuous copulation . During mating males transfer sperm into the female, and the females can store the sperms for several months within an elaborate spermatheca . Several previous studies have been conducted on post-copulatory sexual selection, such as sperm competition and cryptic female choice. Therefore, T . castaneum is an appropriate model for the investigation of post-copulatory sexual selection. In this study, we investigated whether paternity success is affected by male aging. If investment in post-copulatory as well as pre-copulatory sexual selections increase with male aging, paternity success should increase with male aging. Conversely, if sperm quality decreases with male aging and if the male becomes disadvantaged during sperm competition and cryptic female choice, their paternity success should decrease with male aging. If a virgin male invests more in reproduction as they age, the paternity success of older males should be higher in virgin males. In this study, experiments were conducted to test these hypotheses. Insects The beetle culture of the species T . castaneum used in this study was maintained in the laboratory, in a mixture of whole meal (Yoshikura Shokai, Tokyo) enriched with brewer’s yeast (Asahi Beer, Tokyo). The organisms were maintained at 28°C with a 16 h photoperiod (lights on at 07:00, lights off at 23:00). A modified method described in a previous study was used for rearing the organisms . The conditions mentioned above mimic the native environment of these beetles which are pests that infest stored grain. Wild-type strain males and females which are brown in color, were used in this experiment along with mutant strain males which were black in color. The wild-type strain had been maintained at the Universities of Tsukuba and Okayama in Japan for approximately 35 years . We focused on the sperm competition ability of males from this strain, which will hereafter be referred to as the “focal male.” The mutant strain, which is a phenotype that is frequently used as a marker in sperm competition studies in T . castaneum , is homozygous for an autosomal, semidominant black body color allele . The black mutant strain had been reared in conditions which were similar to those for the wild-type strain Okayama University for approximately 15 years. All the individuals used in the experiment were the youngest, at 10 days post-eclosion, considering the sexual maturation period of T . castaneum . Paternity success For investigating the relationship between male aging and investment in post-copulatory sexual selection in T . castaneum , paternity success was measured in terms of the results of sperm competition and cryptic female choice. Though offspring can originate from multiple males, in majority of the cases, paternity success is usually acquired by the last male to mate . Since the paternity success changes of the male that mated last could be easily tracked, this study focused on that male. In this experiment, a female belonging to the wild-type strain was sequentially mated with a mutant strain male and a focal male; following which the paternity success of the focal male that was the second to mate was investigated. The paternity success value of the second male has been referred to as P2, whereas the paternity success value of the first male is referred to as P1. The focal males were divided into two types to distinguish between the effects of aging on paternity success from those of mating experience. Focal male 1 was subjected to experiments following 10 days of age after eclosion, after which they were allowed to mate after every 100 days. Focal male 2, was aged and not allowed to mate until it reached 300 days after eclosion . Age vs. paternity success In the experiment using focal male 1, a virgin mutant black male (random age) and a virgin wild-type female (10-days-old and reared under monosexual condition until used for the experiment) were randomly chosen and placed into a circular plastic container (10 mm in diameter) along with food. The pair was allowed to mate freely for 24 h, following which, the black male was removed from the container, and one focal male 1 was placed into the container along with the female that had already mated with the black strain male. Similarly, the focal male 1and the same wild-type female was allowed to mate freely for 24 h. T . castaneum male beetles often fail to mate and therefore, the pair was placed together and allowed to copulate freely without controlling the number of copulations. Thus, two males were separately mated with one female. The exact number of copulations were not measured. Thus, the number of males that failed to mate at all times could be reduced by a considerable extent . After mating was complete, the female was isolated in a plastic container (45 mm in diameter, 10 mm in height) with sufficient food, and was allowed to lay eggs for ten days, for obtaining a sufficient number of offspring for the measurement of paternity success. The female lays approximately 10 eggs per day. The progeny (81 ± 2.2, mean number ± SE) were maintained at 28°C for 30 days. The adult body color of the progeny was scored to assign paternity success and generate P2 scores after they developed into adults. The focal male 1 that had measurable paternity success at 10 days of age was subjected to repeated paternity success measurements at 100, 200, and 300 days of age. The reproductive lifespan of T . castaneum is 0.5–1 year ; hence, the paternity success measurements were conducted up to 300 days of age. In cases where neither of the competing males sired offspring (i.e., female fertility was zero), the data were not considered for analysis as it could not be ascertained whether the copulations were successful and resulted in sperm transfer or storage, as females may also influence sperm retention. Around one fourth of the focal males 1 died before day 300, due to which, the number of focal males 1 who achieved paternity success at the age of 300 days was smaller compared with 10-day-old focal males 1. The sample sizes used for statistical analyses were 89 at 10 days, 81 at 100 days, 75 at 200 days, and 45 at 300 days, respectively. During the experiment period (from the eclosion of focal male 1 until its death), each focal male 1 was isolated in one well of a 48-well tissue culture plate (Cell Star, Greiner Bio-One, Kremsmünster, Austria) to prevent contact with other individuals, with a sufficient amount of feed for survival. The black males were reared under monosexual conditions with individuals of various ages. After the mating experiment, the black males used in the experiment were frozen and disposed. All experiments were performed between 12:00 hours and 18:00 hours in a room maintained at 28°C. Naïve male vs mated male Paternity success of focal males 2, who had not experienced copulation until the age of 300 days was measured to investigate the effects of the interactions between aging and mating history on their paternity success. The focal male 2 was housed individually under the same conditions as focal male 1, during the experiment period (from the eclosion of focal male 2 until it reaches 300 days old). In this experiment, paternity success was measured using the same method described above. The sample size used for the statistical analysis was 102. Statistical analysis Since our data were binary data with either wild-type or black beetles, a generalized linear mixed model with a binomial distribution was used to analyze paternity success. In this analysis, male age (10, 100, 200, and 300 days old) was an explanatory variable, and male ID was a random effect. The comparison of paternity success at 300 days of age was conducted between males whose paternity success had been measured multiple times (mated male) and males whose paternity success had never been measured before (naive male); comparisons with the control males were conducted for both, 10-day-old and 300-day-old males, respectively. Approximately 46% of all males had died before completion of 300 days in this study. Thus, paternity success could be influenced by the differences in longevity among males. To test this effect, “paternity success at older age” was considered as the paternity success determined at the last measurement before the death of the focal male (i.e., if males die at the age of 150 days, the paternity success at 100 days old was used, and if males lived passed 300 days, the paternity success at 300 days was used). The paternity success at older age was compared with the paternity success at young age (10-days-old). A GLM with binomial distribution was used to analyze the results for the paternity success; however, it was over dispersed. Hence, the estimated dispersion parameter was used for correcting the variance . Bonferroni correction was used for conducting multiple comparisons, when a significant effect was detected in each test. The software R ver.4.1.0 was used to conduct all analyses, using the statistical packages lme4 and car . The beetle culture of the species T . castaneum used in this study was maintained in the laboratory, in a mixture of whole meal (Yoshikura Shokai, Tokyo) enriched with brewer’s yeast (Asahi Beer, Tokyo). The organisms were maintained at 28°C with a 16 h photoperiod (lights on at 07:00, lights off at 23:00). A modified method described in a previous study was used for rearing the organisms . The conditions mentioned above mimic the native environment of these beetles which are pests that infest stored grain. Wild-type strain males and females which are brown in color, were used in this experiment along with mutant strain males which were black in color. The wild-type strain had been maintained at the Universities of Tsukuba and Okayama in Japan for approximately 35 years . We focused on the sperm competition ability of males from this strain, which will hereafter be referred to as the “focal male.” The mutant strain, which is a phenotype that is frequently used as a marker in sperm competition studies in T . castaneum , is homozygous for an autosomal, semidominant black body color allele . The black mutant strain had been reared in conditions which were similar to those for the wild-type strain Okayama University for approximately 15 years. All the individuals used in the experiment were the youngest, at 10 days post-eclosion, considering the sexual maturation period of T . castaneum . For investigating the relationship between male aging and investment in post-copulatory sexual selection in T . castaneum , paternity success was measured in terms of the results of sperm competition and cryptic female choice. Though offspring can originate from multiple males, in majority of the cases, paternity success is usually acquired by the last male to mate . Since the paternity success changes of the male that mated last could be easily tracked, this study focused on that male. In this experiment, a female belonging to the wild-type strain was sequentially mated with a mutant strain male and a focal male; following which the paternity success of the focal male that was the second to mate was investigated. The paternity success value of the second male has been referred to as P2, whereas the paternity success value of the first male is referred to as P1. The focal males were divided into two types to distinguish between the effects of aging on paternity success from those of mating experience. Focal male 1 was subjected to experiments following 10 days of age after eclosion, after which they were allowed to mate after every 100 days. Focal male 2, was aged and not allowed to mate until it reached 300 days after eclosion . In the experiment using focal male 1, a virgin mutant black male (random age) and a virgin wild-type female (10-days-old and reared under monosexual condition until used for the experiment) were randomly chosen and placed into a circular plastic container (10 mm in diameter) along with food. The pair was allowed to mate freely for 24 h, following which, the black male was removed from the container, and one focal male 1 was placed into the container along with the female that had already mated with the black strain male. Similarly, the focal male 1and the same wild-type female was allowed to mate freely for 24 h. T . castaneum male beetles often fail to mate and therefore, the pair was placed together and allowed to copulate freely without controlling the number of copulations. Thus, two males were separately mated with one female. The exact number of copulations were not measured. Thus, the number of males that failed to mate at all times could be reduced by a considerable extent . After mating was complete, the female was isolated in a plastic container (45 mm in diameter, 10 mm in height) with sufficient food, and was allowed to lay eggs for ten days, for obtaining a sufficient number of offspring for the measurement of paternity success. The female lays approximately 10 eggs per day. The progeny (81 ± 2.2, mean number ± SE) were maintained at 28°C for 30 days. The adult body color of the progeny was scored to assign paternity success and generate P2 scores after they developed into adults. The focal male 1 that had measurable paternity success at 10 days of age was subjected to repeated paternity success measurements at 100, 200, and 300 days of age. The reproductive lifespan of T . castaneum is 0.5–1 year ; hence, the paternity success measurements were conducted up to 300 days of age. In cases where neither of the competing males sired offspring (i.e., female fertility was zero), the data were not considered for analysis as it could not be ascertained whether the copulations were successful and resulted in sperm transfer or storage, as females may also influence sperm retention. Around one fourth of the focal males 1 died before day 300, due to which, the number of focal males 1 who achieved paternity success at the age of 300 days was smaller compared with 10-day-old focal males 1. The sample sizes used for statistical analyses were 89 at 10 days, 81 at 100 days, 75 at 200 days, and 45 at 300 days, respectively. During the experiment period (from the eclosion of focal male 1 until its death), each focal male 1 was isolated in one well of a 48-well tissue culture plate (Cell Star, Greiner Bio-One, Kremsmünster, Austria) to prevent contact with other individuals, with a sufficient amount of feed for survival. The black males were reared under monosexual conditions with individuals of various ages. After the mating experiment, the black males used in the experiment were frozen and disposed. All experiments were performed between 12:00 hours and 18:00 hours in a room maintained at 28°C. Paternity success of focal males 2, who had not experienced copulation until the age of 300 days was measured to investigate the effects of the interactions between aging and mating history on their paternity success. The focal male 2 was housed individually under the same conditions as focal male 1, during the experiment period (from the eclosion of focal male 2 until it reaches 300 days old). In this experiment, paternity success was measured using the same method described above. The sample size used for the statistical analysis was 102. Since our data were binary data with either wild-type or black beetles, a generalized linear mixed model with a binomial distribution was used to analyze paternity success. In this analysis, male age (10, 100, 200, and 300 days old) was an explanatory variable, and male ID was a random effect. The comparison of paternity success at 300 days of age was conducted between males whose paternity success had been measured multiple times (mated male) and males whose paternity success had never been measured before (naive male); comparisons with the control males were conducted for both, 10-day-old and 300-day-old males, respectively. Approximately 46% of all males had died before completion of 300 days in this study. Thus, paternity success could be influenced by the differences in longevity among males. To test this effect, “paternity success at older age” was considered as the paternity success determined at the last measurement before the death of the focal male (i.e., if males die at the age of 150 days, the paternity success at 100 days old was used, and if males lived passed 300 days, the paternity success at 300 days was used). The paternity success at older age was compared with the paternity success at young age (10-days-old). A GLM with binomial distribution was used to analyze the results for the paternity success; however, it was over dispersed. Hence, the estimated dispersion parameter was used for correcting the variance . Bonferroni correction was used for conducting multiple comparisons, when a significant effect was detected in each test. The software R ver.4.1.0 was used to conduct all analyses, using the statistical packages lme4 and car . Age vs. paternity success This experiment demonstrated paternity success data for males aged 10-days-old (0.79 ± 0.02, mean ± SE), 100-days-old (0.69 ± 0.03), 200-days-old (0.74 ± 0.04), and 300-days-old (0.65 ± 0.05). The results demonstrated a significant influence of age on paternity success ( F 3,286 = 3.84, p = 0.0102). Multiple comparisons with Bonferroni correction after this test revealed that there was no significant difference in paternity success at 10 days, 100 days, and 200 days, but at 300 days, the paternity success was significantly lower ( and ). To exclude the effect of males that died prior to the completion of 300 days, the paternity success immediately before death was used, and a significant difference was observed when it was compared with the paternity success at 10 days ( F 1,160 = 12.54, p = 0.0005). Similarly, the paternity success of older males (0.62 ± 0.04, mean ± SE) was significantly lower when compared with younger males ( and ). Naïve male vs mated males To exclude the influence of mating experience on paternity success, which was measured with male aging, "naive old males" that had never experienced mating until 300 days of age were included in the study and a comparison of the paternities of mated males and naïve old males revealed a significant difference ( F 2,233 = 4.73, p = 0.0097). The younger males demonstrated a significantly higher paternity success compared with the "naive older males" (0.77 ± 0.03, mean ± SE) ( and ). However, compared to the older males that had experienced multiple mating until 300 days, naive older males exhibited a significantly higher paternity success ( and ). Male longevity did not exert a significant effect on the lifetime paternity success of the male ( χ 2 1,79 = 0.20, p = 0.6521; ). However, male body size (precordial width) did exert a significant positive effect on their paternity success ( χ 2 1,76 = 10.41, p = 0.0013; ). There was no significant correlation between body size and longevity in males ( t 1,76 = −0.02, p = 0.9856; ). This experiment demonstrated paternity success data for males aged 10-days-old (0.79 ± 0.02, mean ± SE), 100-days-old (0.69 ± 0.03), 200-days-old (0.74 ± 0.04), and 300-days-old (0.65 ± 0.05). The results demonstrated a significant influence of age on paternity success ( F 3,286 = 3.84, p = 0.0102). Multiple comparisons with Bonferroni correction after this test revealed that there was no significant difference in paternity success at 10 days, 100 days, and 200 days, but at 300 days, the paternity success was significantly lower ( and ). To exclude the effect of males that died prior to the completion of 300 days, the paternity success immediately before death was used, and a significant difference was observed when it was compared with the paternity success at 10 days ( F 1,160 = 12.54, p = 0.0005). Similarly, the paternity success of older males (0.62 ± 0.04, mean ± SE) was significantly lower when compared with younger males ( and ). To exclude the influence of mating experience on paternity success, which was measured with male aging, "naive old males" that had never experienced mating until 300 days of age were included in the study and a comparison of the paternities of mated males and naïve old males revealed a significant difference ( F 2,233 = 4.73, p = 0.0097). The younger males demonstrated a significantly higher paternity success compared with the "naive older males" (0.77 ± 0.03, mean ± SE) ( and ). However, compared to the older males that had experienced multiple mating until 300 days, naive older males exhibited a significantly higher paternity success ( and ). Male longevity did not exert a significant effect on the lifetime paternity success of the male ( χ 2 1,79 = 0.20, p = 0.6521; ). However, male body size (precordial width) did exert a significant positive effect on their paternity success ( χ 2 1,76 = 10.41, p = 0.0013; ). There was no significant correlation between body size and longevity in males ( t 1,76 = −0.02, p = 0.9856; ). This study investigated whether aging and mating experiences in T . castaneum males affected their post-copulatory sexual selection. Repeated measurements of paternity success at four different ages revealed that paternity success decreased with increasing ages of males and the number of measurements. However, a comparison of the paternity success of older males who had mated four times with that of older males who had not mated until the conduction of this experiment demonstrated that the paternity success of naive males was higher compared with mated males. Moreover, the paternity success of young virgin males and old naive males were similar. These results indicated that, while male aging did not influence post-copulatory sexual selection, older males who had previously mated sometimes showed a reduced investment in post-copulatory sexual selection. Thus, male paternity success is greatly influenced by the interaction between male aging and mating experience. The results of this study are particularly interesting as they differed from those observed in a previous study in the fish Gambusia holbrooki . Although the fish species and insect species largely differ in terms of their in vitro fertilization in vivo fertilization respectively, this study is the first to detect increased paternity success in old males who did not experience copulation. A previously conducted theoretical study had indicated that males can increase their fitness by increasing their investment in reproductive traits with age , and this has been demonstrated in pre-copulatory sexual selection in several species . However, our study did not observe any significant difference in paternity success between virgin young and virgin old males in T . castaneum , and as such cannot support the theoretical prediction. A meta-analysis conducted on post-copulatory sexual traits, specifically ejaculation traits of insect species, has indicated that ejaculation traits tend to improve with age . However, several empirical studies have reported that in some animal species, males are disadvantaged with aging in post-copulatory sexual selection . To clear the confusion regarding this issue, further empirical studies on several other species are warranted. Several previous studies have indicated that decline in sperm quality with age leads to decreasing paternity success. Previous studies on humans have reported that the accumulation of harmful mutations in reproductive cells causes sperm quality to decline with age . A similar phenomenon might occur in insects as well. For example, it has been suggested that a decline in the seminal fluid quality due to aging leads to reduced paternity success in Drosophila melanogaster . Furthermore, paternity success might also be influenced by the quantity of sperm transferred. Previous research has demonstrated that starved males of the T . castaneum species transfer fewer sperm per spermatophore compared to fed males, and have lower paternity success . This indicates the importance of sperm quantity for paternity success in T . castaneum . Therefore, the age associated decline in paternity success might be a result of the reduction in the quantity of sperm transfer. Although the details of the sperm quality and the volume of the ejaculate in association with male aging in T . castaneum could not be revealed in the current study, it is assumed that the effect of male aging on these ejaculate traits is minor, since the paternity success of naive old males was comparable with that of young males. The results of this study, which demonstrated that paternity success did not differ between young and old males, was not in accordance with our hypothesis, according to which, investment of males in reproduction increases with their age. In several species, investment in pre-copulatory sexual traits has been reported to increase with male age ; Hence, it can be presumed that investment into pre-copulatory traits might increase with age compared with post-copulation sexual traits in T . castaneum . In a previously conducted study, T . castaneum males with high moving activity demonstrated increased mating success compared with males with low moving activity in . It would be interesting to investigate the effect of male aging on the moving activity in T . castaneum . Moreover, if sperm quality decreased with the age of males, it was possible that cryptic female choice influenced the reduction in paternity success; however, our results did not support this possibility. It is possible that the quality of sperm does not deteriorate with age or that the effects of aging on male ejaculation are not recognized by females, but this has not been ascertained. Furthermore, although it was possible that the negative effects of male aging on mating behavior adversely affected paternity success, it has not been proven. Since mating behavior was not observed in this study, the details are not clear, and shall be examined in future studies. Paternity success could be associated with male longevity. For example, males with shorter longevity might invest more in early reproduction, while those with longer longevity may invest lesser in early reproduction. However, no correlation between longevity and lifetime paternity success in T . castaneum males was observed in this study . Thus, the paternity success observed in our study might not have been influenced by the differences in male longevity. However, male paternity success did depend on their body size , indicating that larger males possess an advantage in achieving greater paternity success. This result indicates that either the larger males ejaculate more or that females prefer larger males; however, this cannot be ascertained from the results of this study. However, due to the lack of correlation between body size and longevity in males , it can be assumed that male body size played a minor role on the results of this study. A recent study reported that older males belonging to the G . holbrooki species, who had a greater experience of copulation exhibited higher paternity success compared with older males who were less experienced . This suggests that male mating history affects their paternity success. However, the results of this study demonstrated a significantly higher paternity success of naïve males compared with males who had previously experienced multiple matings ; thereby suggesting that paternity success is influenced by the interaction between their mating experience and male aging. Thus, our results contradicted the results obtained in the previous study . Males belonging to the G . holbrooki species, exhibited fierce male-male combat for mating opportunities . Thus, males who had extensive mating experience might have been preferred by females. In G . cornutus , the male possesses exaggerated weapon traits, and it has been reported that females do not always prefer males with large weapons that dominate in male-male combats . Conversely, males of the T . castaneum species do not engage in fierce male-male combat for females; instead, they actively engage in copulation . This indicates that the effects of the interaction between mating experience and aging in males on their paternity success might differ across species. Male mating history might not be an indicator of male quality in T . castaneum . The results of this study have indicated that male mating experience influences their paternity success to a greater extent compared with male aging in T . castaneum . Males with copulation experience tend to increase their longevity as they age, by investing less in reproduction and more in survival. On the contrary, older males that have no copulation experience (i.e., their fitness is zero) are expected to increase their investment in reproduction, thereby achieving greater reproductive success. This might explain the results of this study, according to which, males who did not have the opportunity of mating until 300 days of age achieved higher paternity success compared with those who had mated before. This study revealed the effects of male mating experience on paternity success in males of T . castaneum , a model organism for post-copulatory sexual selection. However, several factors could not be verified in this study. The exact cause of reduced paternity success of older males who had experienced multiple copulations remains unclear. For example, sperm quality and/or ejaculate volume may be compromised in older males who had experienced multiple copulations. It is advisable to conduct these investigations in the future to understand post-copulatory sexual selection. This study has demonstrated that, male paternity success (traits influenced by both sperm competition and cryptic female choice, either singly or in combination) in T . castaneum was not affected by aging. However, older males who had experienced multiple copulations experienced lesser paternity success, indicating that post-copulatory sexual selection is influenced by the interaction between aging and mating experience of males. Since paternity success differed in older males, depending on whether they had previously copulated or not, it can be assumed that reduced male investment in post-copulatory sexual selection, rather than aging, was the more influential factor in this result. Reduced volume of the ejaculate in older males who had experienced multiple matings, might have put them at a disadvantage in sperm competition with rival males along with cryptic female choice. The results do not provide detailed evidence for these possibilities, and future research is warranted. However, very few empirical studies have previously focused on the interaction between aging and mating experience in post-copulatory sexual selection, and hence our findings are crucial in understanding sexual selection. S1 Fig Relationship between longevity and their paternity success (mean). (DOCX) S2 Fig Relationship between body size (prothorax width) and their paternity success (mean). (DOCX) S3 Fig Relationship between body size (prothorax width) and their longevity. (DOCX) S1 File All relevant data of this study. (XLSX) |
Development of culturally sensitive pain neuroscience education materials for Hausa-speaking patients with chronic spinal pain: A modified Delphi study | 83484f58-cdc8-4fff-8b20-e92f09cd5287 | 8253446 | Patient Education as Topic[mh] | An educational program for teaching patients about pain has gained considerable attention in research [ – ] and refers to different terms such as Explain Pain , Therapeutic Neuroscience Education , and Pain Neuroscience Education (PNE) . PNE teaches people in pain about the biology and physiology of their pain experience, including processes such as normal biology of pain, pain modulation, pain matrix, peripheral and central sensitization, allodynia, and neuroplasticity . There is growing evidence for the value of PNE in decreasing pain, disability, fear-avoidance, pain catastrophization, movement restriction, and health care utilization in people struggling with pain [ , , ]. Consequently, PNE is increasingly used as part of physical-therapy for patients with chronic pain in clinical settings [ , , ]. Moreover, in 2015, the International Association for the Study of Pain endorsed a ‘‘call to action” which recognized an urgent need for all countries to improve access to pain management . The development and use of PNE is well-established among Caucasians [ , – ]. However, its access in many other world languages and cultures is lacking. In the last two decades, it has been reported that variations in culture, socioeconomic status, gender issues, and literacy levels have to be considered when developing educational tools for any population . Additionally, patients’ beliefs are a core part of pain perception and response, as such response to pain is influenced both by patients’ beliefs about it and the emotional significance attributed to it . A recent systematic review revealed evidence regarding the differences in pain beliefs, pain attitudes, and coping strategies among different cultures and ethnicities . Hence, a few attempts have been made to culturally adapt PNE materials. One of these attempts involved an internet-based method that required internet-access, literacy, and at least a smartphone (for access) among Brazilians . Another study among Turkish-immigrants in Belgium used home education (HE) materials that required literacy to read, and included Turkish-specific pictures and metaphors . There is an urgent need to develop culturally-sensitive PNE materials for different cultures and population groups worldwide in order to increase access to PNE interventions for patients with chronic pain. Furthermore, cultural sensitivity is defined by two dimensions: surface and deep structures. Surface structure involves matching intervention materials and messages to observable superficial characteristics of the target population which may involve using people, places, language, clothing, etc. that are familiar to and preferred by the target audience, whereas the deep structure encompasses the integration of cultural, social, historical, environmental, and psychological factors that influence the target health behavior in the proposed target population . Neither the translation nor the culturally-sensitive version of PNE is available in any of the indigenous African languages. Africa has several indigenous languages, of which Hausa is among the most popular . Hausa is one of the leading African languages in terms of number of speakers. In addition, it is the unofficial lingua franca in the West-African region and studies have reported a varying number of Hausa native speakers to be between 30 to over 50 million people . The existing PNE materials mainly developed in Australia [ , , ], Europe [ , , ], and America contain pictures, examples and metaphors that may not be appropriate for Hausa population. Some of the cases and scenarios do not relate to Hausa people due to differences in culture, religion, educational levels, or technological advancements. Furthermore, the conventional HE leaflets used by previous researchers [ – , , ] may not be feasible for use in Hausa population due to high levels of adult illiteracy rates, that is estimated to be around 43% . The PNE materials developed for first-generation Turkish-immigrants living in Belgium is the only one nearest to Hausa culture due to overlap in religion (Islam), and to some extent with regard to clothing and gender roles. Nevertheless, these materials used a lot of pictures, metaphors and stories that are not available in a Hausa context. Nijs et al. indicated that intellectual ability and health literacy of a patient should be taken into account before using a PNE program. Consequently, the lower literacy rate compared to that of Western populations is another reason to develop culturally-sensitive PNE materials for Hausa speaking patients suffering from chronic spinal pain (CSP). Finally, Hausa communities also differ from many of those where the existing PNE materials were developed in terms of religious and gender issues . Consequently, it has become important that health care programs should be culturally-sensitive and not just a mere translation of the available materials . The aim of the present study was to develop specific culturally-sensitive PNE Teaching Materials (TM) for Hausa women and men experiencing CSP, and general HE material by using a 2-phase sequential design of a focus group (FG) and modified Delphi-study.
Ethical approval The research protocol was approved by the Hospital ethics committee of Ghent University (B670201836558). Written and verbal informed consents were obtained from the literate (via mails) and non-literate (patients) experts respectively. Study design A focus group (FG) discussion of experts followed by an internet based 3-round modified Delphi-study . Procedure First, a FG discussion was carried out to generate the preliminary materials that were used during the Delphi rounds [ – ], following a similar procedure as described by Orhan et al. . Next, a 3-round modified Delphi-study was designed to gather inputs, corrections, and consensus of experts using a series of online questionnaires [ , , ]. Focus group The preliminary PNE teaching and HE materials were developed through FG discussions. Three to 4 experts are enough for a FG meetings when participants have specialized knowledge and/or experiences to discuss , even though using 4 to 6 experts is becoming increasingly popular because smaller groups are easier to recruit, host and are more comfortable for participants . Therefore, 4 experts were included in this FG discussion (these experts were different from those in the ‘Delphi expert panel’). Since there is no limit to the number of meetings in FG, as it can vary from 1 to several meetings based on the researchers’ need and saturation , this FG had 3 meetings and each meeting lasted for up to 1 hour . Experts in FG meetings are individuals that generally have superior knowledge about the topic, but there is no clear definition for such expertise . Therefore, all the 4 experts were physiotherapists who have experience in the PNE concept (with a minimum of 1 peer-reviewed publication in PNE). In addition, 1 of the experts was a native Hausa language speaker. The first author (NBM) moderated the meetings and recorded all discussions. The content of the existing PNE teaching and HE materials was discussed during the FG meetings. The main message of ‘Explain Pain’ was preserved. This content comprises the importance of pain, differences between acute and chronic pain, how pain originates in the nervous system, what makes pain to persist for long time, and the sensitivity of the nervous system . For the purpose of explaining this, contents, pictures, drawings, metaphors, and stories were also discussed in relation to Hausa culture. The FG resulted in the preliminary PNE teaching and HE materials to use in the first round of the Delphi-study. The following major adaptations to the existing materials were performed: (i) the development of HE materials in form of an oral interview between an expert and a journalist, so that Hausa speaking patients can listen to the interview, since majority of them cannot read ; (ii) modification of the pictures in the TM with drawings that depict African/Hausa people; (ii) due to cultural and religious peculiarities, separate PNE TM for male and female patients were developed ; (iv) metaphors, examples and stories that were thought to be inappropriate for Hausa speaking patients were changed or modified to fit Hausa contexts. The teaching and HE materials were produced in 2 languages; a Hausa version for Hausa speaking experts and an English version for experts that do not speak Hausa. Firstly, all the materials were developed in the English language and then translated into Hausa language by the first author who is fluent in English and a native Hausa speaker. The translated materials, together with the English materials, were sent to 2 Hausa language experts for corrections . This translation procedure was used after each round of the Delphi-rounds. All drawings in the materials were performed by a professional artist based on the feedback of the experts. Throughout the manuscript, TM refer to the PowerPoint slides developed for teaching patients about pain, whereas the HE material applies to the written script of the prospective oral interview that patients will listen to at home to supplement what they have learned from the TM. Delphi-study The 3-rounds Delphi ran from May 2018 to November 2018, and were conducted according to the recommendation of guidance on conducting and reporting of Delphi-studies (CREDES) . Delphi-experts There is no existing guideline on who an expert is, and how many experts should be recruited in a Delphi-study , but in this study, 4 key areas of expertise were defined; (1) PNE; (2) Hausa culture; (3) management of Hausa speaking patients with CSP; and (4) Hausa speaking patients with CSP. Therefore, the recruited experts were a combination of: (i) physiotherapists with some experience in PNE (1 published peer-reviewed paper on PNE) or Hausa culture (with one published peer-reviewed paper on Hausa cultural adaptation) or managing Hausa speaking patients with CSP (with at least 5-years of clinical experience); and (ii) Hausa speaking patients with CSP (with either neck or back pain that lasted for at least the previous 3 months). A total of 28 experts (with 7 experts from each of the 4 key expertise areas) were purposively sampled and invited to participate in the Delphi-study prior to round 1. Experts were identified and selected based on the network and personal contacts of the authors (NBM, MM, JM, BC). Round 1 Two weeks before the start of the first-round, the participants were sent e-mails containing the PNE TM (male and female) for them to study. After these 2-weeks, the questionnaires that consists of both open and closed-ended questions in English language regarding the submitted material were sent. The open-ended questions provided the experts with freedom to give any relevant feedback, while the closed-ended questions limited their responses to only PNE contexts, since many of the participants were not PNE experts. Experts who had limited computer literacy were guided on how to fill in the online questionnaires by one of the researchers (JM), who has experience in online surveys. The experts and patients were asked to complete and submit the questionnaires within 2 weeks. To increase the adherence rate of the experts, a reminder via e-mail with an additional 2-week grace period was sent to those that were unable to complete the questionnaires within the initial 2 weeks. The questionnaire for round 1 consisted of 7-items on demographics and a total of 42 (31 closed-ended and 11 open-ended) PNE materials-related questions that were adapted and modified from a previous study . The questions were divided into 9 sections (“ x” ) which consisted of ‘acute pain’, ‘pain biology’, ‘pain modulation’, ‘pain matrix’, ‘chronic pain’, ‘beliefs, thoughts and behaviors’, ‘central sensitization’, ‘implications’, and ‘general questions’. In each of the above mentioned sections (except the ‘general questions’), the following 4 multiple choice questions were asked using a 5-point Likert scale, (strongly agree–agree—don’t know–disagree—strongly disagree) : Do you think that these slides/pages provide relevant information about “x” (each section bears its name) ? Do you think that the stories used to describe the”x” are feasible for the Hausa population ? Do you think that visual information (pictures) for”x” in the TM is feasible for Hausa-patients ? Do you think the message is clear and patients will understand ? At the end of each section, an open-ended question was asked: If you have any suggestion(s) regarding the description of”x” , please write them in the box below . Three final and open-ended general questions were asked: What do you think about the order and the concept ? Is it understandable , logical ? General remarks ? Further suggestions ? The consensus level was predefined at ≥75% , which is the minimum consensus required for a decision to be made on a particular content. When 75% or more of the experts choose to ‘agree’ or ‘strongly agree’, then such content was retained. If 75% or more of the experts choose to ‘disagree’ or ‘strongly disagree’, such content was rejected. These contents were subsequently modified based on the experts’ suggestions and resubmitted in the next round. In the first round, only the 2 TM (male and female) were sent to experts in order to minimize participation fatigue and drop-outs due to bulkiness of the materials. Round 2 In this round, the modified TM together with the HE material were sent to all Delphi-experts that participated in round 1. In case of consensus regarding the inappropriateness of an item in the TM in round 1, such items were either modified or removed from both teaching and HE materials prior to sending them for round 2. Two different online questionnaires were sent to the experts during this round: one for the HE materials and another for the modified TM. Two weeks were given (plus two weeks grace) to the experts to complete and submit the questionnaires. The questions for the HE materials were similar to that of the TM in round 1. Except for ‘pain modulation’ and ‘central sensitization’ sections that were not included (since these 2 sections were not included in the HE material), instead, an ‘introduction’ section was included. The HE material questionnaire contained a total of 35 (25 closed-ended and 10 open-ended) questions. The questionnaire for the revised TM comprised a total of 31 (23 closed-ended and 8 open-ended) questions about changes/modifications done in the following sections ‘acute pain’, ‘pain biology’, ‘pain modulation’ and ‘beliefs, thoughts and behaviors’, because experts suggested for that. Additionally, closed-ended questions were asked concerning the male and female TM after the update: Do you think the updated material provides more relevant ‘a = information , b = pictures , c = stories and metaphors’ for educating Hausa-patients about pain than the round 1 material ? Do you think the message in the updated material is clear and patients will understand ? All the questions above were graded using the same 5-point Likert scale as in round 1 and the consensus level was maintained at ≥75%. Three general open-ended questions concerning the updated material were added: Is there anything; You would like us to add to this TM ? Specific that you would like us to modify or simplify again in this TM ? You will further suggest ? Round 3 After qualitatively analyzing the responses of the experts from Round 2, questions were asked to finalize the development of the materials. Six closed-ended and 2 open-ended questions were asked to finalize this round. The questions had the usual 5-point Likert scale as in the previous rounds. The questions and responses are presented in the Results section ( ). This round was completed within 2 weeks. Closure After round 3, the final teaching and HE materials were updated and developed (see – for the Hausa materials and – for the English versions). The finalized materials were sent to the experts along with an appreciation message for participation. The HE interview was then orally conducted and recorded ( ) between a professional Hausa journalist and the first author (NBM). Data analysis Content analysis was used to qualitatively analyze the data of each round. The first, second and last authors independently analyzed the comments of the experts. Based on the comments and suggestions of the experts, the authors identified the topics in relation to the comments of the experts, which enabled the authors to effect the necessary modifications and corrections on the materials
The research protocol was approved by the Hospital ethics committee of Ghent University (B670201836558). Written and verbal informed consents were obtained from the literate (via mails) and non-literate (patients) experts respectively.
A focus group (FG) discussion of experts followed by an internet based 3-round modified Delphi-study .
First, a FG discussion was carried out to generate the preliminary materials that were used during the Delphi rounds [ – ], following a similar procedure as described by Orhan et al. . Next, a 3-round modified Delphi-study was designed to gather inputs, corrections, and consensus of experts using a series of online questionnaires [ , , ].
The preliminary PNE teaching and HE materials were developed through FG discussions. Three to 4 experts are enough for a FG meetings when participants have specialized knowledge and/or experiences to discuss , even though using 4 to 6 experts is becoming increasingly popular because smaller groups are easier to recruit, host and are more comfortable for participants . Therefore, 4 experts were included in this FG discussion (these experts were different from those in the ‘Delphi expert panel’). Since there is no limit to the number of meetings in FG, as it can vary from 1 to several meetings based on the researchers’ need and saturation , this FG had 3 meetings and each meeting lasted for up to 1 hour . Experts in FG meetings are individuals that generally have superior knowledge about the topic, but there is no clear definition for such expertise . Therefore, all the 4 experts were physiotherapists who have experience in the PNE concept (with a minimum of 1 peer-reviewed publication in PNE). In addition, 1 of the experts was a native Hausa language speaker. The first author (NBM) moderated the meetings and recorded all discussions. The content of the existing PNE teaching and HE materials was discussed during the FG meetings. The main message of ‘Explain Pain’ was preserved. This content comprises the importance of pain, differences between acute and chronic pain, how pain originates in the nervous system, what makes pain to persist for long time, and the sensitivity of the nervous system . For the purpose of explaining this, contents, pictures, drawings, metaphors, and stories were also discussed in relation to Hausa culture. The FG resulted in the preliminary PNE teaching and HE materials to use in the first round of the Delphi-study. The following major adaptations to the existing materials were performed: (i) the development of HE materials in form of an oral interview between an expert and a journalist, so that Hausa speaking patients can listen to the interview, since majority of them cannot read ; (ii) modification of the pictures in the TM with drawings that depict African/Hausa people; (ii) due to cultural and religious peculiarities, separate PNE TM for male and female patients were developed ; (iv) metaphors, examples and stories that were thought to be inappropriate for Hausa speaking patients were changed or modified to fit Hausa contexts. The teaching and HE materials were produced in 2 languages; a Hausa version for Hausa speaking experts and an English version for experts that do not speak Hausa. Firstly, all the materials were developed in the English language and then translated into Hausa language by the first author who is fluent in English and a native Hausa speaker. The translated materials, together with the English materials, were sent to 2 Hausa language experts for corrections . This translation procedure was used after each round of the Delphi-rounds. All drawings in the materials were performed by a professional artist based on the feedback of the experts. Throughout the manuscript, TM refer to the PowerPoint slides developed for teaching patients about pain, whereas the HE material applies to the written script of the prospective oral interview that patients will listen to at home to supplement what they have learned from the TM. Delphi-study The 3-rounds Delphi ran from May 2018 to November 2018, and were conducted according to the recommendation of guidance on conducting and reporting of Delphi-studies (CREDES) . Delphi-experts There is no existing guideline on who an expert is, and how many experts should be recruited in a Delphi-study , but in this study, 4 key areas of expertise were defined; (1) PNE; (2) Hausa culture; (3) management of Hausa speaking patients with CSP; and (4) Hausa speaking patients with CSP. Therefore, the recruited experts were a combination of: (i) physiotherapists with some experience in PNE (1 published peer-reviewed paper on PNE) or Hausa culture (with one published peer-reviewed paper on Hausa cultural adaptation) or managing Hausa speaking patients with CSP (with at least 5-years of clinical experience); and (ii) Hausa speaking patients with CSP (with either neck or back pain that lasted for at least the previous 3 months). A total of 28 experts (with 7 experts from each of the 4 key expertise areas) were purposively sampled and invited to participate in the Delphi-study prior to round 1. Experts were identified and selected based on the network and personal contacts of the authors (NBM, MM, JM, BC).
The 3-rounds Delphi ran from May 2018 to November 2018, and were conducted according to the recommendation of guidance on conducting and reporting of Delphi-studies (CREDES) .
There is no existing guideline on who an expert is, and how many experts should be recruited in a Delphi-study , but in this study, 4 key areas of expertise were defined; (1) PNE; (2) Hausa culture; (3) management of Hausa speaking patients with CSP; and (4) Hausa speaking patients with CSP. Therefore, the recruited experts were a combination of: (i) physiotherapists with some experience in PNE (1 published peer-reviewed paper on PNE) or Hausa culture (with one published peer-reviewed paper on Hausa cultural adaptation) or managing Hausa speaking patients with CSP (with at least 5-years of clinical experience); and (ii) Hausa speaking patients with CSP (with either neck or back pain that lasted for at least the previous 3 months). A total of 28 experts (with 7 experts from each of the 4 key expertise areas) were purposively sampled and invited to participate in the Delphi-study prior to round 1. Experts were identified and selected based on the network and personal contacts of the authors (NBM, MM, JM, BC).
Two weeks before the start of the first-round, the participants were sent e-mails containing the PNE TM (male and female) for them to study. After these 2-weeks, the questionnaires that consists of both open and closed-ended questions in English language regarding the submitted material were sent. The open-ended questions provided the experts with freedom to give any relevant feedback, while the closed-ended questions limited their responses to only PNE contexts, since many of the participants were not PNE experts. Experts who had limited computer literacy were guided on how to fill in the online questionnaires by one of the researchers (JM), who has experience in online surveys. The experts and patients were asked to complete and submit the questionnaires within 2 weeks. To increase the adherence rate of the experts, a reminder via e-mail with an additional 2-week grace period was sent to those that were unable to complete the questionnaires within the initial 2 weeks. The questionnaire for round 1 consisted of 7-items on demographics and a total of 42 (31 closed-ended and 11 open-ended) PNE materials-related questions that were adapted and modified from a previous study . The questions were divided into 9 sections (“ x” ) which consisted of ‘acute pain’, ‘pain biology’, ‘pain modulation’, ‘pain matrix’, ‘chronic pain’, ‘beliefs, thoughts and behaviors’, ‘central sensitization’, ‘implications’, and ‘general questions’. In each of the above mentioned sections (except the ‘general questions’), the following 4 multiple choice questions were asked using a 5-point Likert scale, (strongly agree–agree—don’t know–disagree—strongly disagree) : Do you think that these slides/pages provide relevant information about “x” (each section bears its name) ? Do you think that the stories used to describe the”x” are feasible for the Hausa population ? Do you think that visual information (pictures) for”x” in the TM is feasible for Hausa-patients ? Do you think the message is clear and patients will understand ? At the end of each section, an open-ended question was asked: If you have any suggestion(s) regarding the description of”x” , please write them in the box below . Three final and open-ended general questions were asked: What do you think about the order and the concept ? Is it understandable , logical ? General remarks ? Further suggestions ? The consensus level was predefined at ≥75% , which is the minimum consensus required for a decision to be made on a particular content. When 75% or more of the experts choose to ‘agree’ or ‘strongly agree’, then such content was retained. If 75% or more of the experts choose to ‘disagree’ or ‘strongly disagree’, such content was rejected. These contents were subsequently modified based on the experts’ suggestions and resubmitted in the next round. In the first round, only the 2 TM (male and female) were sent to experts in order to minimize participation fatigue and drop-outs due to bulkiness of the materials.
In this round, the modified TM together with the HE material were sent to all Delphi-experts that participated in round 1. In case of consensus regarding the inappropriateness of an item in the TM in round 1, such items were either modified or removed from both teaching and HE materials prior to sending them for round 2. Two different online questionnaires were sent to the experts during this round: one for the HE materials and another for the modified TM. Two weeks were given (plus two weeks grace) to the experts to complete and submit the questionnaires. The questions for the HE materials were similar to that of the TM in round 1. Except for ‘pain modulation’ and ‘central sensitization’ sections that were not included (since these 2 sections were not included in the HE material), instead, an ‘introduction’ section was included. The HE material questionnaire contained a total of 35 (25 closed-ended and 10 open-ended) questions. The questionnaire for the revised TM comprised a total of 31 (23 closed-ended and 8 open-ended) questions about changes/modifications done in the following sections ‘acute pain’, ‘pain biology’, ‘pain modulation’ and ‘beliefs, thoughts and behaviors’, because experts suggested for that. Additionally, closed-ended questions were asked concerning the male and female TM after the update: Do you think the updated material provides more relevant ‘a = information , b = pictures , c = stories and metaphors’ for educating Hausa-patients about pain than the round 1 material ? Do you think the message in the updated material is clear and patients will understand ? All the questions above were graded using the same 5-point Likert scale as in round 1 and the consensus level was maintained at ≥75%. Three general open-ended questions concerning the updated material were added: Is there anything; You would like us to add to this TM ? Specific that you would like us to modify or simplify again in this TM ? You will further suggest ?
After qualitatively analyzing the responses of the experts from Round 2, questions were asked to finalize the development of the materials. Six closed-ended and 2 open-ended questions were asked to finalize this round. The questions had the usual 5-point Likert scale as in the previous rounds. The questions and responses are presented in the Results section ( ). This round was completed within 2 weeks. Closure After round 3, the final teaching and HE materials were updated and developed (see – for the Hausa materials and – for the English versions). The finalized materials were sent to the experts along with an appreciation message for participation. The HE interview was then orally conducted and recorded ( ) between a professional Hausa journalist and the first author (NBM). Data analysis Content analysis was used to qualitatively analyze the data of each round. The first, second and last authors independently analyzed the comments of the experts. Based on the comments and suggestions of the experts, the authors identified the topics in relation to the comments of the experts, which enabled the authors to effect the necessary modifications and corrections on the materials
After round 3, the final teaching and HE materials were updated and developed (see – for the Hausa materials and – for the English versions). The finalized materials were sent to the experts along with an appreciation message for participation. The HE interview was then orally conducted and recorded ( ) between a professional Hausa journalist and the first author (NBM).
Content analysis was used to qualitatively analyze the data of each round. The first, second and last authors independently analyzed the comments of the experts. Based on the comments and suggestions of the experts, the authors identified the topics in relation to the comments of the experts, which enabled the authors to effect the necessary modifications and corrections on the materials
Four experts who are all physiotherapists participated in the FG discussion (3 PhDs and 1 MSc holders). Although 28 experts were invited to participate in the Delphi-study, only 22 agreed. Nineteen out of 22, 18 out of 19, and all 18 experts participated in the rounds 1, 2 and 3 respectively, representing 86%, 94%, and 100% participation rate for each round. Five PNE experts, 5 CSP patients, 4 Hausa culture experts and 4 physiotherapists managing CSP patients completed round 3. The demographic characteristics of the experts in each round are presented in , while the participation flowchart can be seen in . Round 1 shows the responses of the experts to the closed-ended questions during round 1. For all the content of the materials, the experts reached a consensus of ≥75%, and as such, they were retained. However, the experts made suggestions for changes through the open-ended questions. The experts’ responses to the open-ended questions of round 1 are presented in . Since the materials were developed for Hausa speaking patients, suggestions from experts that were not familiar with the Hausa culture, especially those that contradicted suggestions of experts familiar with Hausa culture about the feasibility/cultural context of content were not used in the modification of the materials. Most of the suggestions and changes that were made were related to drawings (change or modification to fit the Hausa culture) and the simplification of information, e.g. giving an explanation of what a spinal cord is and giving some additional content like maladaptive beliefs. Round 2 presents the responses of the experts to closed-ended questions during round 2. During this round, the revised TM (male and female) and the HE material were reviewed by the experts. The experts reached a consensus of ≥75% for the entire content, with some parts reaching complete consensus (100%). Therefore, additional changes in content were only done in response to the open-ended questions. also presents the responses to the open-ended questions of experts in round 2, but suggestions requiring the use of video/animations were not considered because of resource limitations, and low technological advancements among the end-users. Round 3 The results of the closed and open-ended questions in the final round are presented in . Consensus was attained in all the closed-ended question items except one (see ). The aim of the study was to develop culturally-sensitive PNE materials for Hausa speakingpatients with CSP. Preliminary teaching and HE materials were developed through a FG discussion and subsequently the final materials were developed through a 3-round Delphi-study. The TM were adapted and modified from Orhan et al. , who conducted a similar study among Turkish-immigrants in Belgium, whose culture is considered to be the closest to a Hausa culture among all the culturally adapted PNE materials. In this study, an audio interview was developed as HE material, given the low literacy level in that region . This is supported by a previous study that recommended repetitions of PNE, in different forms (verbal or other) as it helps patients to understand the theory of neurophysiology. The PNE materials developed in this study is the first culturally-sensitive materials for an African language/culture. To our knowledge, this is also the first study that has tried to provide HE materials that can be used by the non-literates. This as necessary because the previous PNE materials were developed for populations that, as reports by UNESCO shows, have higher literacy levels than the general Hausa population . Therefore, this development is in-line with the recommendations for improving access to pain management for all . Although data on Hausa people being frequent listeners of audio talks is lacking, and a previous study has reported that about 89% of Nigerians to own an audio listening set . This may assist the listening of the HE interview among Hausa patients. Similarly, the development of the separate male and female TM has taken care of the importance attached to gender variations among Hausa people due to religion and culture. Varied gender treatment and differences exist in Hausa context, e.g. preventing girls from attending school; withdrawing girl-children from school; using girls for street hawking; and unequal treatment of children by the parent . Moreover, a previous review has reported variation of pain perception, emotion and understanding between males and females among different cultures , as such, a uniform material for both males and females may not be appropriate. Also both teaching and HE materials bear names, examples, metaphors, and drawings/pictures that are available and familiar for Hausa speaking patients, which is in conformity with the concept of cultural sensitivity . During round 1, the experts reached the minimum consensus level in almost all the closed-ended questions related to the TM. This may be a consequence of the adapted materials being modified in relation to the religious inclination of most Hausa people, and to some extent their culture. The experts’ responses to the open-ended questions during this round suggested the need for changes of some pictures and drawings to fit the culture and also simplification of some specific language terminologies used, which are very vital for the development of any culture sensitive tool . The experts’ consensus level increased in round 2, with 100% of experts agreeing on most of the content of the materials. This could indicate that the materials in round 2 were better accepted by the experts compared to the prior materials. Obviously, this trend is aimed for in tool development using experts’ opinions . During round 2, there were some suggestions regarding the use of videos and animations, but such additions were not included as they may not be appropriate for Hausa speaking patients due to low technological advancement, low literacy levels, and high poverty rates among the target population . During the final round, the experts reached consensus on all the contents of the teaching and the HE materials and there were no suggestions that warranted an additional round. The researchers considered the TM ready for application and the HE material was then recorded in form of an oral interview by the first author with a professional Hausa journalist.
shows the responses of the experts to the closed-ended questions during round 1. For all the content of the materials, the experts reached a consensus of ≥75%, and as such, they were retained. However, the experts made suggestions for changes through the open-ended questions. The experts’ responses to the open-ended questions of round 1 are presented in . Since the materials were developed for Hausa speaking patients, suggestions from experts that were not familiar with the Hausa culture, especially those that contradicted suggestions of experts familiar with Hausa culture about the feasibility/cultural context of content were not used in the modification of the materials. Most of the suggestions and changes that were made were related to drawings (change or modification to fit the Hausa culture) and the simplification of information, e.g. giving an explanation of what a spinal cord is and giving some additional content like maladaptive beliefs.
presents the responses of the experts to closed-ended questions during round 2. During this round, the revised TM (male and female) and the HE material were reviewed by the experts. The experts reached a consensus of ≥75% for the entire content, with some parts reaching complete consensus (100%). Therefore, additional changes in content were only done in response to the open-ended questions. also presents the responses to the open-ended questions of experts in round 2, but suggestions requiring the use of video/animations were not considered because of resource limitations, and low technological advancements among the end-users.
The results of the closed and open-ended questions in the final round are presented in . Consensus was attained in all the closed-ended question items except one (see ). The aim of the study was to develop culturally-sensitive PNE materials for Hausa speakingpatients with CSP. Preliminary teaching and HE materials were developed through a FG discussion and subsequently the final materials were developed through a 3-round Delphi-study. The TM were adapted and modified from Orhan et al. , who conducted a similar study among Turkish-immigrants in Belgium, whose culture is considered to be the closest to a Hausa culture among all the culturally adapted PNE materials. In this study, an audio interview was developed as HE material, given the low literacy level in that region . This is supported by a previous study that recommended repetitions of PNE, in different forms (verbal or other) as it helps patients to understand the theory of neurophysiology. The PNE materials developed in this study is the first culturally-sensitive materials for an African language/culture. To our knowledge, this is also the first study that has tried to provide HE materials that can be used by the non-literates. This as necessary because the previous PNE materials were developed for populations that, as reports by UNESCO shows, have higher literacy levels than the general Hausa population . Therefore, this development is in-line with the recommendations for improving access to pain management for all . Although data on Hausa people being frequent listeners of audio talks is lacking, and a previous study has reported that about 89% of Nigerians to own an audio listening set . This may assist the listening of the HE interview among Hausa patients. Similarly, the development of the separate male and female TM has taken care of the importance attached to gender variations among Hausa people due to religion and culture. Varied gender treatment and differences exist in Hausa context, e.g. preventing girls from attending school; withdrawing girl-children from school; using girls for street hawking; and unequal treatment of children by the parent . Moreover, a previous review has reported variation of pain perception, emotion and understanding between males and females among different cultures , as such, a uniform material for both males and females may not be appropriate. Also both teaching and HE materials bear names, examples, metaphors, and drawings/pictures that are available and familiar for Hausa speaking patients, which is in conformity with the concept of cultural sensitivity . During round 1, the experts reached the minimum consensus level in almost all the closed-ended questions related to the TM. This may be a consequence of the adapted materials being modified in relation to the religious inclination of most Hausa people, and to some extent their culture. The experts’ responses to the open-ended questions during this round suggested the need for changes of some pictures and drawings to fit the culture and also simplification of some specific language terminologies used, which are very vital for the development of any culture sensitive tool . The experts’ consensus level increased in round 2, with 100% of experts agreeing on most of the content of the materials. This could indicate that the materials in round 2 were better accepted by the experts compared to the prior materials. Obviously, this trend is aimed for in tool development using experts’ opinions . During round 2, there were some suggestions regarding the use of videos and animations, but such additions were not included as they may not be appropriate for Hausa speaking patients due to low technological advancement, low literacy levels, and high poverty rates among the target population . During the final round, the experts reached consensus on all the contents of the teaching and the HE materials and there were no suggestions that warranted an additional round. The researchers considered the TM ready for application and the HE material was then recorded in form of an oral interview by the first author with a professional Hausa journalist.
It was concluded that, PNE materials that could be used to teach Hausa speaking patients with CSP and an audio interview that Hausa speaking patients can listen to at home, were successfully developed, following a well-documented, consensus building procedure. Considering the composition of the expert panel that participated in the development (i.e. physiotherapists that are experts in PNE, Hausa culture, and management of Hausa speaking patients with CSP, supplemented with the Hausa speaking patients with CSP themselves), the materials hold the promise to have high face validity and also user-friendly. Practice implication The present Delphi-study may provide a direction for further research in which the effects of culturally-sensitive PNE materials can be piloted among Hausa speaking patients with CSP. Limitations During the focus group discussion, only physiotherapists with PNE knowledge were involved in the development of the initial PNE materials that were subsequently used during the Delphi rounds. The lack of other professionals involved in pain management might have affected the overall presentation of the PNE, however, we ensured preservation of the original PNE concept. Another potential limitation of this study is our inability to follow a standard translation procedure for the materials developed, this is because the content of the materials was changed after each of the Delphi rounds based on experts’ suggestions, and we lack resources and personnel to conduct standard translation procedures after each Delphi round. However, language experts with a minimum of PhD degrees in linguistics (Hausa language) were involved in the translation and they have been duly acknowledged. Additionally, there was variation in the language of the experts. Therefore, the experts that did not speak Hausa language had to study the English version of the document. Also, the experts recruited were predominantly Nigerians, and this is because Hausa people are predominantly found in Nigeria. Additionally, some of the patients recruited were not computer literate and not fluent in English language. Consequently, a research-assistant who was told not to influence their choices had to guide them on how to respond to the questionnaire.
The present Delphi-study may provide a direction for further research in which the effects of culturally-sensitive PNE materials can be piloted among Hausa speaking patients with CSP.
During the focus group discussion, only physiotherapists with PNE knowledge were involved in the development of the initial PNE materials that were subsequently used during the Delphi rounds. The lack of other professionals involved in pain management might have affected the overall presentation of the PNE, however, we ensured preservation of the original PNE concept. Another potential limitation of this study is our inability to follow a standard translation procedure for the materials developed, this is because the content of the materials was changed after each of the Delphi rounds based on experts’ suggestions, and we lack resources and personnel to conduct standard translation procedures after each Delphi round. However, language experts with a minimum of PhD degrees in linguistics (Hausa language) were involved in the translation and they have been duly acknowledged. Additionally, there was variation in the language of the experts. Therefore, the experts that did not speak Hausa language had to study the English version of the document. Also, the experts recruited were predominantly Nigerians, and this is because Hausa people are predominantly found in Nigeria. Additionally, some of the patients recruited were not computer literate and not fluent in English language. Consequently, a research-assistant who was told not to influence their choices had to guide them on how to respond to the questionnaire.
S1 File Pain education teaching slides (for males) in Hausa language. (PDF) Click here for additional data file. S2 File Pain education teaching slides (for females) in Hausa language. (PDF) Click here for additional data file. S3 File Home education audio interview in Hausa language. (MP3) Click here for additional data file. S4 File Pain education teaching slides (for males) in English language. (PDF) Click here for additional data file. S5 File Pain education teaching slides (for females) in English language. (PDF) Click here for additional data file. S6 File Home education audio interview (transcript) in English language. (PDF) Click here for additional data file. S1 Data (DOCX) Click here for additional data file. S2 Data (DOCX) Click here for additional data file. S3 Data (DOCX) Click here for additional data file.
|
Multimodal imaging for refractive surgery: | 6c18061e-0545-4e0c-8887-ffae14140cc9 | 7856969 | Ophthalmology[mh] | Prof. Renato Ambrósio Jr, MD, PhD Prof. Renato Ambrósio trained in Ophthalmology at the Instituto de Oftalmologia Tadeu Cvintal (São Paulo) in 1999, followed by a fellowship in Refractive Surgery and Cornea at the University of Washington (Seattle, WA) in October 2002, and PhD at the Faculdade de Medicina da Universidade de São Paulo in May 2004. In 2006, he was elected the President of the Brazilian Society of Administration in Ophthalmology, being in this position until July 2010. From 2012 until 2014, he was the last president of the Brazilian Society of Refractive Surgery before the incorporation with the Brazilian Society of Cataract and Implants for the creation of BRASCRS. He was also the vice-president of the Brazilian Council of Ophthalmology from 2013 until 2015 and is the current president of the International Society of Refractive Surgery (ISRS; 2020-22). In 2007, he founded The Rio de Janeiro Corneal Tomography and Biomechanics Study Group, from which over one hundred publications have originated. He is an affiliate professor of the Federal University of São Paulo (stricto sensu) and Adjunct Professor of the Federal University of the State of Rio de Janeiro (UNIRIO). Professor Ambrósio is a world-class refractive surgeon-scientist. His major areas of interest are corneal/refractive diagnostics, custom laser vision correction, refractive cataract surgery and therapeutic procedures for keratoconus. Besides his busy and proactive academic appointments, he is at private practices in Rio de Janeiro: Instituto de Olhos Renato Ambrósio and VisareRIO - Refracta Personal Laser.
|
Analysis of fatty acid-derived lipids in critically ill patients after cardiac surgery yields novel pathophysiologically relevant mediators with possible relevance for systemic inflammatory reactions | 35a1e5aa-56b6-4139-b2ed-0d8ea9d547b7 | 11826806 | Surgical Procedures, Operative[mh] | Background Patients in critical conditions may experience a wide range of pro-inflammatory clinical events, the majority of which result from major surgery or as result of infection-associated complications like sepsis. Following a lengthy and intricate surgery, individuals, particularly those undergoing heart surgery, are doomed to experience acute systemic inflammation. A number of these patients under inflammatory conditions develop acute liver failure, acute cardiac damage, or acute renal injury. Bioactive lipid mediators deriving from the arachidonic acid cascade and members of the sphingolipid family have shown to play a crucial role in inflammatory processes . Plasma levels of certain lipid mediators might therefore work as potential new biomarker candidates to early and reliably indicate changes in clinical stages as well as pathophysiological relevant mediators, triggering or counteracting specific inflammatory stages. As an example, S1P (d18:1) plays an important role in maintaining endothelial integrity, and Winkler et al. already concluded that low S1P (d18:1) levels might crucially contribute to capillary leakage . Taking the example of sepsis as specific cause for a pro-inflammatory state of patients’ immune system, different lipid mediators have been discussed as crucial or at least taking part in the course of the disease. In these cases, synthetization of inflammatory arachidonic acid pathway metabolites such as prostaglandin E2 (PGE 2 ) is reduced, pointing to a worse clinical outcome . Moreover, the generation of 5-lipoxygenase pathway product leukotriene C 4 (LTC 4 ) is significantly reduced during sepsis and elevated LTC 4 production was associated with higher patient survival. For reader’s overview, illustrates the simplified biosynthesis pathways of some relevant lipid mediators of the eicosanoid and sphingolipid families, which have been investigated in the present study. To this date, only a few studies have focused on fatty acid-derived lipids as potential biomarkers indicating inflammatory stages . Possibly due to methodological and analytical limitations, previous studies analyzed serum levels of lipids at definite time points. To the best of our knowledge, no study has analyzed a broad range of lipid profiles over extended time periods with lipid plasma level analysis following the patients stay at the ICU for several days. Therefore, the present study was aimed at simultaneously analyzing 53 lipid-signaling mediators from the plasma of nine patients hospitalized at Frankfurt University Hospital ICU for a period up to 28 days. We conclude that some lipid mediators may serve for early detection of new ongoing inflammation processes or might give insights to distinguish between different entities of pro-inflammatory states. To compare inflammatory reactions in patients with frequently used ex vivo models of inflammation, lipid mediator levels were compared to plasma lipid levels of healthy volunteers, stimulated ex vivo with the bacterial toxin lipopolysaccharide (LPS). Here, we present several lipid mediator-signaling pathways significantly regulated during periods of inflammation caused by different clinical events after prolonged cardiac surgery. Materials and methods 2.1 Study design and sample collection Cardiac surgery patients, who underwent either at least re-operation, double valve repair or replacement and large blood vessel surgery with complete cardio-pulmonary bypass (cbp) were included. Furthermore, patients were included where surgery was planned with extra-long cbp time. We aimed for patients with a suspected vigorous inflammatory response and a narrow age window. Nine ICU patients (see ) were finally included based on the following criteria: age > 18 years and < 80 years, post-operative stay at ICU for at least 8 days and more than 8 blood collections for lipid mediator analyses. Patients’ characteristics including further information on treatments can be found in . Samples including those for ex vivo experiments were collected in the period 2013–2020 at the Department of Anesthesiology, Intensive Care Medicine and Pain Therapy, Frankfurt University Hospital, Goethe University Frankfurt, Germany. In this pilot trial, the number of patients included was restricted by limited analytical capacity of LC/MS-MS, given the numerous lipid mediators to be analyzed daily over prolonged periods. Study-related blood drawing for determining IL-6 and leukocyte count was performed twice daily. Blood drawing for lipid mediator analysis was performed once daily for an average of 17.2 days (min. 8 days, max. 28 days.) Whole blood samples were collected daily from the included patients, starting on the first day after surgery. As the physiological and pathophysiological functions of most of the investigated 53 lipids have not yet been well characterized, their normal ranges in human plasma are largely unknown. The 95% and 2-fold upper and lower 95% normal range values of the lipids were therefore determined via LC-MS/MS analysis of the lipid mediator concentrations in plasma samples from ten healthy volunteers, who neither suffered from inflammatory diseases that could theoretically impact lipid plasma levels nor were under any drug therapy. Notably, due to unexpected technical limitations, not all normal lipid range values (38 out of 53 lipids) could be determined with the first blood sample collection. Therefore, a second blood sample collection and LC-MS/MS analysis, involving a further ten healthy volunteers, was conducted for the normal range determination of the 15 remaining plasma lipid levels, including eicosatetraenoic acids (HETEs), eicosatrienoic acids (DHET), phosphoglycerides (LPAs), epoxides of linoleic acid (EpOME), diols of linoleic acid (DiHOMEs) and hydroxyoctadecadienoic acids (HODEs). Prior to LC-MS/MS analysis, all person-related data in this study, except age and gender, were pseudo-anonymized. Lastly, the patients’ plasma lipid mediator levels were compared with those obtained from the stimulation of venous whole blood samples from the first healthy volunteer group using the bacterial component lipopolysaccharide (LPS) (see section 2.4). This experiment was aimed at addressing the question to what extent ex vivo models of inflammation reflect the in vivo situation of patients accurately and to what extent lipid mediator reactivity under pro-inflammatory conditions is similar among different patients and models of inflammation. 2.2 Assessment of ICU patients’ clinical parameters Clinical data were retrospectively collected from our clinical ICU database (MetaVision, iMDsoft, Israel), including routine diagnostic laboratory data (LAURIS, Nexus, Germany) and our imaging data system (PACS, Siemens Healthcare, Germany). A senior medical doctor screened every patient’s history mainly for signs of general inflammation, hemodynamic instability, progress in acute kidney injury and acute liver failure. Mainly, inflammation was measured using IL-6 as a significant diagnostic marker . In addition, to enable differentiation of septic phases from inflammatory reactions due to severe post-operative organ damage mainly resulting from left or right heart failure, special attention was given to microbiological test results (see ). We further screened especially for other disturbing factors which can potentially disrupt lipid levels such as different medical histories, co-medications with diverse pharmacotherapeutic treatment regimens and medical-technical interventions like the beginning of dialysis, artificial respiration and extracorporeal life support. 2.3 Sample recovery to determine ICU patients’ plasma lipid levels and healthy volunteers’ normal lipid ranges Citrated venous whole blood was collected daily at the same time point from nine ICU patients hospitalized at Frankfurt University Hospital and immediately centrifuged at 2000 g at 4°C for 15 minutes. Supernatants (plasma) of patients and samples of healthy volunteers for determination of the normal range were stored at -80°C before LC-MS/MS analysis. We used a sensitive LC-MS/MS method to determine 53 lipid mediators originating from different lipid signaling pathways. 2.4 Ex vivo whole blood LPS assay The venous whole blood samples of ten healthy male and female voluntary test subjects aged ≥18 served for plasma lipid profile assessment after ex vivo stimulation with the bacterial component lipopolysaccharide (LPS). For the ex vivo assay, 1 mL heparinized whole blood was incubated with LPS (100 ng/ml) for 8, 16, and 24 h at 37°C under gentle stirring (180 rpm). We have chosen 100 ng/ml LPS as it represents a commonly used low but already sufficient concentration to trigger the release of a broad spectrum of pro-inflammatory mediators, including both lipid mediators and cytokines (e.g . Notably, a broad range of lipids were analyzed after LPS treatment with a number of them playing well-documented roles in inflammation. However, LPS as stimulus cannot trigger stimulation/repression of biosynthesis of all lipid mediators. Some lipids analyzed in this study may not be involved in inflammation with a mode of regulation of biosynthesis not fully understood. Samples without LPS (0 h) were used as a control for the ex vivo LPS stimulation assay. Native control plasma of these volunteers also served to determine the normal ranges of the first series of 38 lipids investigated in this study. 2.5 LC-MS/MS analysis of lipid mediator levels in patients’ and healthy volunteers’ plasma to determine normal ranges and in plasma obtained from LPS-stimulated whole blood All lipids were analyzed via liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) at the Institute of Clinical Pharmacology, Frankfurt University Hospital, Frankfurt am Main, Germany. 2.5.1 Analysis of endocannabinoids Analysis of arachidonoylethanolamide (AEA), palmitoylethanolamide (PEA), 1- and 2-arachidonoylglycerol (1- and 2-AG), and oleoylethanolamide (OEA) was carried out as described elsewhere . Briefly, 50 µL plasma was liquid–liquid extracted. The residues were reconstituted with 50 µL of acetonitrile in glass vials, and 10 µL were injected into the LC-MS/MS system. This consisted of a hybrid triple quadrupole-ion trap QTrap 5500 mass spectrometer (Sciex, Darmstadt, Germany) equipped with a Turbo-V-source operating in negative ESI mode, an Agilent 1200 binary HPLC pump, column oven (40°C), and degasser (Agilent, Waldbronn, Germany), and an HTC Pal autosampler (Chromtech, Idstein, Germany). A cooling stack was used to store the samples at 4°C in the autosampler. HPLC analysis was carried out under gradient conditions using a Luna C18 column (150 mm L×2 mm ID, 5 µm particle size, Phenomenex, Aschaffenburg, Germany) and water and acetonitrile, both containing 0.01% ammonia as mobile phases. Analyst software was used to evaluate concentrations of the calibration standards, quality controls, and unknowns (version 1.6; Sciex, Darmstadt, Germany). Variations in accuracy, intra-day, and inter-day precision (n = 6 for each concentration, respectively) were <15% over the calibration range. The lower limits of quantification were 0.1 ng/mL for anandamide, 0.25 ng/mL for 2-AG, and 0.5 ng/mL for PEA and OEA. 2.5.2 Analysis of lysophosphatidic acids Sample extraction was performed with liquid–liquid extraction as already described . Therefore, 50 µL plasma was extracted twice with 500 µL of water-saturated n-butanol. The LC-MS/MS system was the same as described for endocannabinoids. For the chromatographic separation, a Luna C18 Mercury column was used (20 x 2 mm inner diameter, 5 µm particle size, and 100 Å pore size, Phenomenex, Aschaffenburg, Germany) with the same material precolumn. A linear gradient was run at a flow rate of 0.4 mL/min for the separation of the analytes with a total run time of 7 minutes. Mobile phase A was 50 mM ammonium acetate containing 0.2% formic acid, and mobile phase B was acetonitrile:isopropyl alcohol:formic acid (50:50:0.2, v/v/v). Quantification was performed using the internal standard method with Analyst software version 1.5 (Sciex, Darmstadt, Germany). Ratios of analyte peak area and internal standard area ( y -axis) were plotted against concentration ( x -axis), and calibration curves were calculated by linear regression with 1/x concentration weighting. The coefficient of correlation was at least 0.99. Variations in accuracy were less than 15% over the range of calibration. 2.5.3 Analysis of eicosanoids The following lipid mediators were analyzed using liquid chromatography-tandem-mass spectroscopy (LC-MS/MS): Leukotriene B4 (LTB 4 ); hydroxyeicosatetraenoic acids 5(S)-HETE, 12(S)-HETE, 15(S)-HETE and 20(S)-HETE); epoxyeicosatetraenoic; and dihydroxyeicosatrienoic acids (5,6-DHET; 8,9-DHET; 11,12-DHET;14,15-DHET). The LC-MS/MS system consisted of a 5500 QTrap mass spectrometer (Sciex, Darmstadt, Germany), operating in negative ESI mode, an Agilent 1200 HPLC system (Agilent, Waldbronn, Germany), and an HTC Pal autosampler (Chromtech, Idstein, Germany). Sample extraction of LTB 4 , HETEs, EETs, and DHETs was performed using liquid–liquid extraction: 200 µL of plasma were gently mixed with 20 µL of methanol and 20 µL of internal standard solution and extracted twice with 600 µL ethyl acetate. Samples for standard curve and quality control were prepared similarly: 200 µL PBS, 20 µL of standard solution, and 20µL internal standard solution were mixed and extracted with ethyl acetate. Working solutions of all analytes were prepared in methanol containing 0.1% BHT. The calibration standards were prepared by further dilution of the working standards. The organic phase was removed at 45°C under a gentle stream of nitrogen. The residues were reconstituted in 50 µL of methanol:water:BHT (50:50:10– 4 , v/v/v) prior to injection into the LC-MS/MS system. Chromatographic separation was achieved using a Gemini NX C18 column (150 mm × 2 mm ID, 5 µm, Phenomenex, Aschaffenburg, Germany) with a pre-column of the same material. A linear gradient was employed at a flow rate of 0.5 mL/min and a total run time of 17.5 minutes. Mobile phases were A water:ammonia (100:0.05, v/v) and B acetonitrile:ammonia (100:0.05, v/v). The gradient started at 85% A, changed to 10% A within 12 min, held for one min, and shifted back to 85% A in 0.5 min following 3.5 min equilibration. All data were acquired using Analyst software v1.6.2, and quantitation was performed by MultiQuant software v3.0 (both Sciex, Darmstadt, Germany) using the internal standard method (isotope-dilution mass spectrometry). Calibration curves were calculated by linear regression with 1/x or 1/x 2 weighting, and acceptance criteria were applied as described previously . Prostaglandins sample analysis was performed using liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) as described elsewhere . The LC-MS/MS system was the same as for endocannabinoid analysis. For the chromatographic separation, a Synergi Hydro-RP column and pre-column were used (150 x 2 mm ID, from Phenomenex, Aschaffenburg, Germany). A linear gradient was employed at a flow rate of 300 µL/min. Mobile phase A was water:formic acid (100:0.0025, v/v, pH 4.0) and mobile phase B was acetonitrile:formic acid (100:0.0025, v/v). The sample solvent was acetonitrile:water:formic acid (20:80:0.0025, v/v, pH 4.0). The total run time was 16 minutes, and the injection volume of samples was 20 µL. Quantitation was performed with Analyst software V1.5 (Sciex, Darmstadt, Germany) using the internal standard method (isotope-dilution mass spectrometry). Ratios of analyte peak area and internal standard peak area (y-axis) were plotted against concentration (x-axis), and calibration curves for each prostaglandin (PGE 2 , PGD 2 , PGF 2alpha ) were calculated by least square regression with 1/concentration 2 weighting. 2.5.4 Analysis of linoleic acid derivatives Lipid identification and quantification were performed as previously described . Briefly, stock solutions with analytes: 9-hydroxyoctadecadienoic acid (9-HODE), 13-HODE, 9,10-epoxy-12Z-octadecenoic acid (9,10-EpOME) 12,13-EpOME, 9,10-dihydroxy-12Z-octadecenoic acid (9,10-DiHOME), 12,13-DiHOME, and the corresponding internal standards were prepared in methanol. Sample pretreatment and LC-MS/MS analysis were performed as described in Section 2.5.3. 2.5.5 Analysis of sphingolipids For lipid extraction, 10 µL of plasma was mixed with 150 µL water, 150 µL extraction solution (citric acid 30 mM, disodium hydrogen phosphate 40 mM), and 20 µL internal standard solution as described elsewhere . Afterward, the amounts of sphingolipids were analyzed by liquid LC-MS/MS. An Agilent 1100 series binary pump (Agilent Technologies, Waldbronn, Germany) equipped with a Luna C8 column (150 mm x 2 mm ID, 3µm particle size, 100 Å pore size; Phenomenex, Aschaffenburg, Germany) was used for chromatographic separation. The column temperature was set at 35°C. The HPLC mobile phases consisted of water with 0.2% formic acid and 2 mM ammonium formate (mobile phase A) and acetonitrile:isopropanol:acetone (50:30:20, v/v/v) with 0.2% formic acid (mobile phase B). For separation, a gradient program was used at a flow rate of 0.3 mL/min. The MS/MS analysis was performed using a triple quadrupole mass spectrometer API4000 (Sciex, Darmstadt, Germany) equipped with a Turbo V Ion Source operating in positive electrospray ionization mode. Data acquisition was performed using Analyst software V 1.6, and quantification was performed with MultiQuant software V 3.0 (both Sciex, Darmstadt, Germany), employing the internal standard method (isotope-dilution mass spectrometry). Variations in the accuracy of the calibration standards were less than 15% over the whole calibration range, except for the lower limit of quantification, where a variation in accuracy of 20% was accepted. Using this method, we analyzed plasma levels of sphingolipids listed in . 2.6 Statistical analysis of all lipids in healthy volunteers and ICU patients All statistical analysis, including tests and graphs of the final data set (experimental values obtained from the laboratory analysis), was made using Graphpad Prism (version 9.5.0) and R software (version 3.5.1, open-source software 2018). The distribution of variables was pre-tested for normality using the Shapiro-Wilk-test to decide which kind of tests to apply later on (t-test or Wilcoxon-Mann-Whitney) and to test indirectly for symmetry to decide which kinds of location parameter (mean ± x-fold standard deviation or median ± upper and lower 95% range limits) should be taken for the descriptive analysis. For lipid plasma levels, the respective normal ranges were then defined as the median ± upper and lower normal range limit (NRL 95% ), indicating the range of lipid plasma levels shown by at least 95% of healthy humans. We used the median instead the mean, as the samples were asymmetric and not normally distributed. We calculated the 95% limit instead of quartiles because clinical laboratory plasma values are usually given as threshold ranges (normal or reference ranges) representing the plasma levels found in 95% of healthy subjects. The normal ranges for the classical lipids IL-6 and leukocytes were adopted from the reference ranges of Frankfurt University Hospital’s central clinical laboratory. For the lipids, we used as references the plasma sample of healthy volunteers (see sections 2.1, 2.3, and 2.4). We defined “significantly out of normal range” if plasma levels are outside the area defined by median NR ± 2-fold upper and lower limit NRL 95% . Lipids with at least one plasma peak level outside the range median NR ± 2-fold upper and lower limit NRL 95% , observed in at least five of the nine patients, were classified as “prevalently regulated”. Lipid plasma levels were illustrated with time-dependent graphs expressed as hours stayed at the ICU, starting at zero (day one after surgery. The NRL 95% normal range is indicated by dotted lines. The range median NR ± 2-fold upper and lower limit NRL 95% is shown as gray background. In few cases, where lipid plasma levels measurement errors are suspected, the data points are not included in the curves but are displayed as single data points for transparency reasons. For the ex vivo whole blood LPS assay, the pairwise Wilcoxon test (two-sample test for dependent, non-normally distributed samples) was used because several samples demonstrated an explicit deviation from normal distribution, while for others, there was a lack of sufficient data points for distribution analysis. Unstimulated controls were compared with samples that had been treated. “Significantly regulated” was used if at least one incubation period with LPS (6 h, 16 h, 24 h) versus the control was significantly different (p < 0.05). Due to the study design, no significance correction for multiple testing was applied. This was deliberate, as such a procedure would have changed the results negligibly. Study design and sample collection Cardiac surgery patients, who underwent either at least re-operation, double valve repair or replacement and large blood vessel surgery with complete cardio-pulmonary bypass (cbp) were included. Furthermore, patients were included where surgery was planned with extra-long cbp time. We aimed for patients with a suspected vigorous inflammatory response and a narrow age window. Nine ICU patients (see ) were finally included based on the following criteria: age > 18 years and < 80 years, post-operative stay at ICU for at least 8 days and more than 8 blood collections for lipid mediator analyses. Patients’ characteristics including further information on treatments can be found in . Samples including those for ex vivo experiments were collected in the period 2013–2020 at the Department of Anesthesiology, Intensive Care Medicine and Pain Therapy, Frankfurt University Hospital, Goethe University Frankfurt, Germany. In this pilot trial, the number of patients included was restricted by limited analytical capacity of LC/MS-MS, given the numerous lipid mediators to be analyzed daily over prolonged periods. Study-related blood drawing for determining IL-6 and leukocyte count was performed twice daily. Blood drawing for lipid mediator analysis was performed once daily for an average of 17.2 days (min. 8 days, max. 28 days.) Whole blood samples were collected daily from the included patients, starting on the first day after surgery. As the physiological and pathophysiological functions of most of the investigated 53 lipids have not yet been well characterized, their normal ranges in human plasma are largely unknown. The 95% and 2-fold upper and lower 95% normal range values of the lipids were therefore determined via LC-MS/MS analysis of the lipid mediator concentrations in plasma samples from ten healthy volunteers, who neither suffered from inflammatory diseases that could theoretically impact lipid plasma levels nor were under any drug therapy. Notably, due to unexpected technical limitations, not all normal lipid range values (38 out of 53 lipids) could be determined with the first blood sample collection. Therefore, a second blood sample collection and LC-MS/MS analysis, involving a further ten healthy volunteers, was conducted for the normal range determination of the 15 remaining plasma lipid levels, including eicosatetraenoic acids (HETEs), eicosatrienoic acids (DHET), phosphoglycerides (LPAs), epoxides of linoleic acid (EpOME), diols of linoleic acid (DiHOMEs) and hydroxyoctadecadienoic acids (HODEs). Prior to LC-MS/MS analysis, all person-related data in this study, except age and gender, were pseudo-anonymized. Lastly, the patients’ plasma lipid mediator levels were compared with those obtained from the stimulation of venous whole blood samples from the first healthy volunteer group using the bacterial component lipopolysaccharide (LPS) (see section 2.4). This experiment was aimed at addressing the question to what extent ex vivo models of inflammation reflect the in vivo situation of patients accurately and to what extent lipid mediator reactivity under pro-inflammatory conditions is similar among different patients and models of inflammation. Assessment of ICU patients’ clinical parameters Clinical data were retrospectively collected from our clinical ICU database (MetaVision, iMDsoft, Israel), including routine diagnostic laboratory data (LAURIS, Nexus, Germany) and our imaging data system (PACS, Siemens Healthcare, Germany). A senior medical doctor screened every patient’s history mainly for signs of general inflammation, hemodynamic instability, progress in acute kidney injury and acute liver failure. Mainly, inflammation was measured using IL-6 as a significant diagnostic marker . In addition, to enable differentiation of septic phases from inflammatory reactions due to severe post-operative organ damage mainly resulting from left or right heart failure, special attention was given to microbiological test results (see ). We further screened especially for other disturbing factors which can potentially disrupt lipid levels such as different medical histories, co-medications with diverse pharmacotherapeutic treatment regimens and medical-technical interventions like the beginning of dialysis, artificial respiration and extracorporeal life support. Sample recovery to determine ICU patients’ plasma lipid levels and healthy volunteers’ normal lipid ranges Citrated venous whole blood was collected daily at the same time point from nine ICU patients hospitalized at Frankfurt University Hospital and immediately centrifuged at 2000 g at 4°C for 15 minutes. Supernatants (plasma) of patients and samples of healthy volunteers for determination of the normal range were stored at -80°C before LC-MS/MS analysis. We used a sensitive LC-MS/MS method to determine 53 lipid mediators originating from different lipid signaling pathways. Ex vivo whole blood LPS assay The venous whole blood samples of ten healthy male and female voluntary test subjects aged ≥18 served for plasma lipid profile assessment after ex vivo stimulation with the bacterial component lipopolysaccharide (LPS). For the ex vivo assay, 1 mL heparinized whole blood was incubated with LPS (100 ng/ml) for 8, 16, and 24 h at 37°C under gentle stirring (180 rpm). We have chosen 100 ng/ml LPS as it represents a commonly used low but already sufficient concentration to trigger the release of a broad spectrum of pro-inflammatory mediators, including both lipid mediators and cytokines (e.g . Notably, a broad range of lipids were analyzed after LPS treatment with a number of them playing well-documented roles in inflammation. However, LPS as stimulus cannot trigger stimulation/repression of biosynthesis of all lipid mediators. Some lipids analyzed in this study may not be involved in inflammation with a mode of regulation of biosynthesis not fully understood. Samples without LPS (0 h) were used as a control for the ex vivo LPS stimulation assay. Native control plasma of these volunteers also served to determine the normal ranges of the first series of 38 lipids investigated in this study. LC-MS/MS analysis of lipid mediator levels in patients’ and healthy volunteers’ plasma to determine normal ranges and in plasma obtained from LPS-stimulated whole blood All lipids were analyzed via liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) at the Institute of Clinical Pharmacology, Frankfurt University Hospital, Frankfurt am Main, Germany. 2.5.1 Analysis of endocannabinoids Analysis of arachidonoylethanolamide (AEA), palmitoylethanolamide (PEA), 1- and 2-arachidonoylglycerol (1- and 2-AG), and oleoylethanolamide (OEA) was carried out as described elsewhere . Briefly, 50 µL plasma was liquid–liquid extracted. The residues were reconstituted with 50 µL of acetonitrile in glass vials, and 10 µL were injected into the LC-MS/MS system. This consisted of a hybrid triple quadrupole-ion trap QTrap 5500 mass spectrometer (Sciex, Darmstadt, Germany) equipped with a Turbo-V-source operating in negative ESI mode, an Agilent 1200 binary HPLC pump, column oven (40°C), and degasser (Agilent, Waldbronn, Germany), and an HTC Pal autosampler (Chromtech, Idstein, Germany). A cooling stack was used to store the samples at 4°C in the autosampler. HPLC analysis was carried out under gradient conditions using a Luna C18 column (150 mm L×2 mm ID, 5 µm particle size, Phenomenex, Aschaffenburg, Germany) and water and acetonitrile, both containing 0.01% ammonia as mobile phases. Analyst software was used to evaluate concentrations of the calibration standards, quality controls, and unknowns (version 1.6; Sciex, Darmstadt, Germany). Variations in accuracy, intra-day, and inter-day precision (n = 6 for each concentration, respectively) were <15% over the calibration range. The lower limits of quantification were 0.1 ng/mL for anandamide, 0.25 ng/mL for 2-AG, and 0.5 ng/mL for PEA and OEA. 2.5.2 Analysis of lysophosphatidic acids Sample extraction was performed with liquid–liquid extraction as already described . Therefore, 50 µL plasma was extracted twice with 500 µL of water-saturated n-butanol. The LC-MS/MS system was the same as described for endocannabinoids. For the chromatographic separation, a Luna C18 Mercury column was used (20 x 2 mm inner diameter, 5 µm particle size, and 100 Å pore size, Phenomenex, Aschaffenburg, Germany) with the same material precolumn. A linear gradient was run at a flow rate of 0.4 mL/min for the separation of the analytes with a total run time of 7 minutes. Mobile phase A was 50 mM ammonium acetate containing 0.2% formic acid, and mobile phase B was acetonitrile:isopropyl alcohol:formic acid (50:50:0.2, v/v/v). Quantification was performed using the internal standard method with Analyst software version 1.5 (Sciex, Darmstadt, Germany). Ratios of analyte peak area and internal standard area ( y -axis) were plotted against concentration ( x -axis), and calibration curves were calculated by linear regression with 1/x concentration weighting. The coefficient of correlation was at least 0.99. Variations in accuracy were less than 15% over the range of calibration. 2.5.3 Analysis of eicosanoids The following lipid mediators were analyzed using liquid chromatography-tandem-mass spectroscopy (LC-MS/MS): Leukotriene B4 (LTB 4 ); hydroxyeicosatetraenoic acids 5(S)-HETE, 12(S)-HETE, 15(S)-HETE and 20(S)-HETE); epoxyeicosatetraenoic; and dihydroxyeicosatrienoic acids (5,6-DHET; 8,9-DHET; 11,12-DHET;14,15-DHET). The LC-MS/MS system consisted of a 5500 QTrap mass spectrometer (Sciex, Darmstadt, Germany), operating in negative ESI mode, an Agilent 1200 HPLC system (Agilent, Waldbronn, Germany), and an HTC Pal autosampler (Chromtech, Idstein, Germany). Sample extraction of LTB 4 , HETEs, EETs, and DHETs was performed using liquid–liquid extraction: 200 µL of plasma were gently mixed with 20 µL of methanol and 20 µL of internal standard solution and extracted twice with 600 µL ethyl acetate. Samples for standard curve and quality control were prepared similarly: 200 µL PBS, 20 µL of standard solution, and 20µL internal standard solution were mixed and extracted with ethyl acetate. Working solutions of all analytes were prepared in methanol containing 0.1% BHT. The calibration standards were prepared by further dilution of the working standards. The organic phase was removed at 45°C under a gentle stream of nitrogen. The residues were reconstituted in 50 µL of methanol:water:BHT (50:50:10– 4 , v/v/v) prior to injection into the LC-MS/MS system. Chromatographic separation was achieved using a Gemini NX C18 column (150 mm × 2 mm ID, 5 µm, Phenomenex, Aschaffenburg, Germany) with a pre-column of the same material. A linear gradient was employed at a flow rate of 0.5 mL/min and a total run time of 17.5 minutes. Mobile phases were A water:ammonia (100:0.05, v/v) and B acetonitrile:ammonia (100:0.05, v/v). The gradient started at 85% A, changed to 10% A within 12 min, held for one min, and shifted back to 85% A in 0.5 min following 3.5 min equilibration. All data were acquired using Analyst software v1.6.2, and quantitation was performed by MultiQuant software v3.0 (both Sciex, Darmstadt, Germany) using the internal standard method (isotope-dilution mass spectrometry). Calibration curves were calculated by linear regression with 1/x or 1/x 2 weighting, and acceptance criteria were applied as described previously . Prostaglandins sample analysis was performed using liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) as described elsewhere . The LC-MS/MS system was the same as for endocannabinoid analysis. For the chromatographic separation, a Synergi Hydro-RP column and pre-column were used (150 x 2 mm ID, from Phenomenex, Aschaffenburg, Germany). A linear gradient was employed at a flow rate of 300 µL/min. Mobile phase A was water:formic acid (100:0.0025, v/v, pH 4.0) and mobile phase B was acetonitrile:formic acid (100:0.0025, v/v). The sample solvent was acetonitrile:water:formic acid (20:80:0.0025, v/v, pH 4.0). The total run time was 16 minutes, and the injection volume of samples was 20 µL. Quantitation was performed with Analyst software V1.5 (Sciex, Darmstadt, Germany) using the internal standard method (isotope-dilution mass spectrometry). Ratios of analyte peak area and internal standard peak area (y-axis) were plotted against concentration (x-axis), and calibration curves for each prostaglandin (PGE 2 , PGD 2 , PGF 2alpha ) were calculated by least square regression with 1/concentration 2 weighting. 2.5.4 Analysis of linoleic acid derivatives Lipid identification and quantification were performed as previously described . Briefly, stock solutions with analytes: 9-hydroxyoctadecadienoic acid (9-HODE), 13-HODE, 9,10-epoxy-12Z-octadecenoic acid (9,10-EpOME) 12,13-EpOME, 9,10-dihydroxy-12Z-octadecenoic acid (9,10-DiHOME), 12,13-DiHOME, and the corresponding internal standards were prepared in methanol. Sample pretreatment and LC-MS/MS analysis were performed as described in Section 2.5.3. 2.5.5 Analysis of sphingolipids For lipid extraction, 10 µL of plasma was mixed with 150 µL water, 150 µL extraction solution (citric acid 30 mM, disodium hydrogen phosphate 40 mM), and 20 µL internal standard solution as described elsewhere . Afterward, the amounts of sphingolipids were analyzed by liquid LC-MS/MS. An Agilent 1100 series binary pump (Agilent Technologies, Waldbronn, Germany) equipped with a Luna C8 column (150 mm x 2 mm ID, 3µm particle size, 100 Å pore size; Phenomenex, Aschaffenburg, Germany) was used for chromatographic separation. The column temperature was set at 35°C. The HPLC mobile phases consisted of water with 0.2% formic acid and 2 mM ammonium formate (mobile phase A) and acetonitrile:isopropanol:acetone (50:30:20, v/v/v) with 0.2% formic acid (mobile phase B). For separation, a gradient program was used at a flow rate of 0.3 mL/min. The MS/MS analysis was performed using a triple quadrupole mass spectrometer API4000 (Sciex, Darmstadt, Germany) equipped with a Turbo V Ion Source operating in positive electrospray ionization mode. Data acquisition was performed using Analyst software V 1.6, and quantification was performed with MultiQuant software V 3.0 (both Sciex, Darmstadt, Germany), employing the internal standard method (isotope-dilution mass spectrometry). Variations in the accuracy of the calibration standards were less than 15% over the whole calibration range, except for the lower limit of quantification, where a variation in accuracy of 20% was accepted. Using this method, we analyzed plasma levels of sphingolipids listed in . Analysis of endocannabinoids Analysis of arachidonoylethanolamide (AEA), palmitoylethanolamide (PEA), 1- and 2-arachidonoylglycerol (1- and 2-AG), and oleoylethanolamide (OEA) was carried out as described elsewhere . Briefly, 50 µL plasma was liquid–liquid extracted. The residues were reconstituted with 50 µL of acetonitrile in glass vials, and 10 µL were injected into the LC-MS/MS system. This consisted of a hybrid triple quadrupole-ion trap QTrap 5500 mass spectrometer (Sciex, Darmstadt, Germany) equipped with a Turbo-V-source operating in negative ESI mode, an Agilent 1200 binary HPLC pump, column oven (40°C), and degasser (Agilent, Waldbronn, Germany), and an HTC Pal autosampler (Chromtech, Idstein, Germany). A cooling stack was used to store the samples at 4°C in the autosampler. HPLC analysis was carried out under gradient conditions using a Luna C18 column (150 mm L×2 mm ID, 5 µm particle size, Phenomenex, Aschaffenburg, Germany) and water and acetonitrile, both containing 0.01% ammonia as mobile phases. Analyst software was used to evaluate concentrations of the calibration standards, quality controls, and unknowns (version 1.6; Sciex, Darmstadt, Germany). Variations in accuracy, intra-day, and inter-day precision (n = 6 for each concentration, respectively) were <15% over the calibration range. The lower limits of quantification were 0.1 ng/mL for anandamide, 0.25 ng/mL for 2-AG, and 0.5 ng/mL for PEA and OEA. Analysis of lysophosphatidic acids Sample extraction was performed with liquid–liquid extraction as already described . Therefore, 50 µL plasma was extracted twice with 500 µL of water-saturated n-butanol. The LC-MS/MS system was the same as described for endocannabinoids. For the chromatographic separation, a Luna C18 Mercury column was used (20 x 2 mm inner diameter, 5 µm particle size, and 100 Å pore size, Phenomenex, Aschaffenburg, Germany) with the same material precolumn. A linear gradient was run at a flow rate of 0.4 mL/min for the separation of the analytes with a total run time of 7 minutes. Mobile phase A was 50 mM ammonium acetate containing 0.2% formic acid, and mobile phase B was acetonitrile:isopropyl alcohol:formic acid (50:50:0.2, v/v/v). Quantification was performed using the internal standard method with Analyst software version 1.5 (Sciex, Darmstadt, Germany). Ratios of analyte peak area and internal standard area ( y -axis) were plotted against concentration ( x -axis), and calibration curves were calculated by linear regression with 1/x concentration weighting. The coefficient of correlation was at least 0.99. Variations in accuracy were less than 15% over the range of calibration. Analysis of eicosanoids The following lipid mediators were analyzed using liquid chromatography-tandem-mass spectroscopy (LC-MS/MS): Leukotriene B4 (LTB 4 ); hydroxyeicosatetraenoic acids 5(S)-HETE, 12(S)-HETE, 15(S)-HETE and 20(S)-HETE); epoxyeicosatetraenoic; and dihydroxyeicosatrienoic acids (5,6-DHET; 8,9-DHET; 11,12-DHET;14,15-DHET). The LC-MS/MS system consisted of a 5500 QTrap mass spectrometer (Sciex, Darmstadt, Germany), operating in negative ESI mode, an Agilent 1200 HPLC system (Agilent, Waldbronn, Germany), and an HTC Pal autosampler (Chromtech, Idstein, Germany). Sample extraction of LTB 4 , HETEs, EETs, and DHETs was performed using liquid–liquid extraction: 200 µL of plasma were gently mixed with 20 µL of methanol and 20 µL of internal standard solution and extracted twice with 600 µL ethyl acetate. Samples for standard curve and quality control were prepared similarly: 200 µL PBS, 20 µL of standard solution, and 20µL internal standard solution were mixed and extracted with ethyl acetate. Working solutions of all analytes were prepared in methanol containing 0.1% BHT. The calibration standards were prepared by further dilution of the working standards. The organic phase was removed at 45°C under a gentle stream of nitrogen. The residues were reconstituted in 50 µL of methanol:water:BHT (50:50:10– 4 , v/v/v) prior to injection into the LC-MS/MS system. Chromatographic separation was achieved using a Gemini NX C18 column (150 mm × 2 mm ID, 5 µm, Phenomenex, Aschaffenburg, Germany) with a pre-column of the same material. A linear gradient was employed at a flow rate of 0.5 mL/min and a total run time of 17.5 minutes. Mobile phases were A water:ammonia (100:0.05, v/v) and B acetonitrile:ammonia (100:0.05, v/v). The gradient started at 85% A, changed to 10% A within 12 min, held for one min, and shifted back to 85% A in 0.5 min following 3.5 min equilibration. All data were acquired using Analyst software v1.6.2, and quantitation was performed by MultiQuant software v3.0 (both Sciex, Darmstadt, Germany) using the internal standard method (isotope-dilution mass spectrometry). Calibration curves were calculated by linear regression with 1/x or 1/x 2 weighting, and acceptance criteria were applied as described previously . Prostaglandins sample analysis was performed using liquid chromatography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/MS) as described elsewhere . The LC-MS/MS system was the same as for endocannabinoid analysis. For the chromatographic separation, a Synergi Hydro-RP column and pre-column were used (150 x 2 mm ID, from Phenomenex, Aschaffenburg, Germany). A linear gradient was employed at a flow rate of 300 µL/min. Mobile phase A was water:formic acid (100:0.0025, v/v, pH 4.0) and mobile phase B was acetonitrile:formic acid (100:0.0025, v/v). The sample solvent was acetonitrile:water:formic acid (20:80:0.0025, v/v, pH 4.0). The total run time was 16 minutes, and the injection volume of samples was 20 µL. Quantitation was performed with Analyst software V1.5 (Sciex, Darmstadt, Germany) using the internal standard method (isotope-dilution mass spectrometry). Ratios of analyte peak area and internal standard peak area (y-axis) were plotted against concentration (x-axis), and calibration curves for each prostaglandin (PGE 2 , PGD 2 , PGF 2alpha ) were calculated by least square regression with 1/concentration 2 weighting. Analysis of linoleic acid derivatives Lipid identification and quantification were performed as previously described . Briefly, stock solutions with analytes: 9-hydroxyoctadecadienoic acid (9-HODE), 13-HODE, 9,10-epoxy-12Z-octadecenoic acid (9,10-EpOME) 12,13-EpOME, 9,10-dihydroxy-12Z-octadecenoic acid (9,10-DiHOME), 12,13-DiHOME, and the corresponding internal standards were prepared in methanol. Sample pretreatment and LC-MS/MS analysis were performed as described in Section 2.5.3. Analysis of sphingolipids For lipid extraction, 10 µL of plasma was mixed with 150 µL water, 150 µL extraction solution (citric acid 30 mM, disodium hydrogen phosphate 40 mM), and 20 µL internal standard solution as described elsewhere . Afterward, the amounts of sphingolipids were analyzed by liquid LC-MS/MS. An Agilent 1100 series binary pump (Agilent Technologies, Waldbronn, Germany) equipped with a Luna C8 column (150 mm x 2 mm ID, 3µm particle size, 100 Å pore size; Phenomenex, Aschaffenburg, Germany) was used for chromatographic separation. The column temperature was set at 35°C. The HPLC mobile phases consisted of water with 0.2% formic acid and 2 mM ammonium formate (mobile phase A) and acetonitrile:isopropanol:acetone (50:30:20, v/v/v) with 0.2% formic acid (mobile phase B). For separation, a gradient program was used at a flow rate of 0.3 mL/min. The MS/MS analysis was performed using a triple quadrupole mass spectrometer API4000 (Sciex, Darmstadt, Germany) equipped with a Turbo V Ion Source operating in positive electrospray ionization mode. Data acquisition was performed using Analyst software V 1.6, and quantification was performed with MultiQuant software V 3.0 (both Sciex, Darmstadt, Germany), employing the internal standard method (isotope-dilution mass spectrometry). Variations in the accuracy of the calibration standards were less than 15% over the whole calibration range, except for the lower limit of quantification, where a variation in accuracy of 20% was accepted. Using this method, we analyzed plasma levels of sphingolipids listed in . Statistical analysis of all lipids in healthy volunteers and ICU patients All statistical analysis, including tests and graphs of the final data set (experimental values obtained from the laboratory analysis), was made using Graphpad Prism (version 9.5.0) and R software (version 3.5.1, open-source software 2018). The distribution of variables was pre-tested for normality using the Shapiro-Wilk-test to decide which kind of tests to apply later on (t-test or Wilcoxon-Mann-Whitney) and to test indirectly for symmetry to decide which kinds of location parameter (mean ± x-fold standard deviation or median ± upper and lower 95% range limits) should be taken for the descriptive analysis. For lipid plasma levels, the respective normal ranges were then defined as the median ± upper and lower normal range limit (NRL 95% ), indicating the range of lipid plasma levels shown by at least 95% of healthy humans. We used the median instead the mean, as the samples were asymmetric and not normally distributed. We calculated the 95% limit instead of quartiles because clinical laboratory plasma values are usually given as threshold ranges (normal or reference ranges) representing the plasma levels found in 95% of healthy subjects. The normal ranges for the classical lipids IL-6 and leukocytes were adopted from the reference ranges of Frankfurt University Hospital’s central clinical laboratory. For the lipids, we used as references the plasma sample of healthy volunteers (see sections 2.1, 2.3, and 2.4). We defined “significantly out of normal range” if plasma levels are outside the area defined by median NR ± 2-fold upper and lower limit NRL 95% . Lipids with at least one plasma peak level outside the range median NR ± 2-fold upper and lower limit NRL 95% , observed in at least five of the nine patients, were classified as “prevalently regulated”. Lipid plasma levels were illustrated with time-dependent graphs expressed as hours stayed at the ICU, starting at zero (day one after surgery. The NRL 95% normal range is indicated by dotted lines. The range median NR ± 2-fold upper and lower limit NRL 95% is shown as gray background. In few cases, where lipid plasma levels measurement errors are suspected, the data points are not included in the curves but are displayed as single data points for transparency reasons. For the ex vivo whole blood LPS assay, the pairwise Wilcoxon test (two-sample test for dependent, non-normally distributed samples) was used because several samples demonstrated an explicit deviation from normal distribution, while for others, there was a lack of sufficient data points for distribution analysis. Unstimulated controls were compared with samples that had been treated. “Significantly regulated” was used if at least one incubation period with LPS (6 h, 16 h, 24 h) versus the control was significantly different (p < 0.05). Due to the study design, no significance correction for multiple testing was applied. This was deliberate, as such a procedure would have changed the results negligibly. Results 3.1 Clinical course of the patients included after cardiac surgery The average stay at the ICU was 43.2 days (min 8 days, max. 101 days). All patients received high dosages of vasopressors and inotropics at the beginning of intensive care therapy. The average SOFA score (modified SOFA using the Richmond Agitation-Sedation Scale in critically ill patients according to was 11.7 (min 8, max 15) at admission. There were four non-survivors . 3.2 In vivo lipid mediator profile in ICU patients after cardiac surgery In the present study we analyzed baseline values of all lipids investigated in this study since their normal ranges in human plasma are largely unknown. For ceramides, normal ranges based on plasma levels of healthy subjects were similar to those reported in literature. E.g. baseline ceramide level in our study are consistent with those published by others (i.e. Cer (d18:1/16:0): 0.220 ± 0.044 µM vs 0.24–0.30, Cer (d18:1/18:0): 0.067 ± 0.028 µM vs 0.001–0.11; Cer (d18:1/24:0): 4.490 ± 1.435 µM vs 1.15–3.34; Cer (d18:1/24:1): 0.746 ± 0.240 µM vs 0.71–1.14 . Comparisons of baseline values of other lipids investigated in this study could not be established, since their normal ranges in human plasma are largely unknown. shows examples of different lipid reactivity during the stages of systemic inflammation. Notably, the type of lipid is not mentioned at this stage, as the purpose of this figure is to introduce different types of lipid mediator reactivity with respect to normal ranges. shows an example of an unchanged lipid mediator. shows an example of a lipid mediator changed just above the normal range but still below the medianNR ± 2-fold upper and lower limit NRL95%. shows an example of a significantly regulated lipid mediator. Lipids that displayed plasma levels remaining consistently within the normal range during hospitalization, such as that shown in , were not evaluated further. We identified 22 of the 53 lipids which showed significant regulation during at least one time point on the ICU ward among at least 5 out of 9 patients, thus indicating potential clinical relevance. This group of prevalently regulated lipid mediators with potential clinical relevance could be further divided into lipids with either down-or upregulated levels during these general inflammatory conditions on ICU . We used IL-6 and leukocyte count as established markers of inflammation . Interestingly, the leukocyte count remained consistently elevated in most patients, pointing to the severity of illness the patients were suffering from, rather than specific clinical events. The relevant lipid plasma peaks were then compared with clinical phases. Regarding the type of plasma level regulation during the observation period and clinical events, we were able to define two major groups of lipids. The first group (dynamically upregulated lipids) features plasma levels rapidly changing from the normal range to the area above the twofold normal range upper NRL95%-limit. Based on visual interpretation for some of these lipids, changes in plasma levels seemed to have limited correlation with patients’ inflammatory status. and show representative graphs for 6-keto-PGF1 alpha and 2-AG. These two lipids partly correlated with inflammatory stages but rather seemed to correlate more to some specific clinical events. While 6-keto-PGF1 alpha had its peak shortly before initiation of continuous veno-venous hemodialysis (at least in 5 of the observed patients, see ), 2-AG peaked in all our nine patients during (right) heart failure in the context of either re-opening patient’s chest, implementation of veno-arterial ECMO or at least while significantly increasing the amount of catecholamines (see ). Besides these visual correlations, the remarkable increase of 6-keto-PGF1alpha in certain patients becomes more noteworthy considering that normal range levels of this lipid in healthy volunteers are below the detection limit of LC/MS-MS. Furthermore, Cer (d18:1/16:0), Cer (d18:1/18:1) and LacCer (d18:1/24:1) showed a tendency to accumulate in the plasma during the time of hospitalization and remained elevated for prolonged periods ( and , ). Notably, 6-keto-PGF1 alpha 1-AG, 2-AG and 20-HETE were the lipids that displayed the strongest maximal relative induction of biosynthesis compared with the plasma normal range concentrations in healthy persons. Among these, the strongest relative induction was seen with 6-keto-PGF1 alpha levels reaching up to 20 ng/ml and basal levels below the detection limit, followed by about 350-fold (2-AG), 200-fold (1-AG) and 40-fold maximal induction (20-HETE). A second group of lipids (significantly downregulated lipids) demonstrated plasma levels significantly below the normal range that, however remained impaired consistently and to similar degree during almost the entire time of hospitalization and were largely irrespective of clinical events. Examples of this type of lipid mediator, such as plasma levels of S1P (d18:1) and Lactosylceramide LacCer (d18:1/24:0), are shown in and . Several lipids displayed rather low normal-range concentrations, with lower NRL95% limits already close to the LC-MS/MS detection limit and levels that apparently were further reduced during hospitalization. This applies for example to LPA (18:2), LPA (18:3), LPA (20:4), 9,10-DiHOME, PEA, OEA and PGE 2 . Surprisingly, we did not observe any relevant elevation of the classical pro-inflammatory products of the cyclooxygenase-2 and 5-lipoxygenase pathway, including PGE 2 and 5-HETE or LTB 4 . All graphs of prevalently regulated lipids with potential clinical relevance (see ) not shown in – can be found in the supplemental section, - . 3.3 Ex vivo lipid mediator profile in LPS-stimulated human whole blood Next, we compared the changes of lipid mediators from patients with systemic inflammation with an ex vivo model of endotoxin-induced inflammation. Here, venous whole blood from healthy volunteers was stimulated with LPS for 6 h, 16 h, and 24 h. The lipid mediator profile included 38 lipids analyzed using LC-MS/MS. Notably, due to a technical failure during LC-MS/MS analysis, only 38 of the 53 lipid mediators could be analyzed, however including all important lipids in patients described above. Among the 38 included lipids we identified 27 lipid mediators with significantly changed plasma levels after incubation with LPS, as compared to the untreated control. and include graphs for selected lipids from this LPS-based assay that either had already shown significant regulation in the ICU patients or only showed reactivity in our LPS model but not in our patients. Graphs of lipid plasma levels after stimulation of whole blood with LPS not included in and can be found in - . Further details on the results of the LPS analysis at different time points can be found in . The earliest significant increase compared with the control after 6 hours was seen with prostaglandin PGE 2 , PGF 2alpha and Cer (d18:0/16:0) followed by PGD 2 , although these lipids were not regulated in our patients. The early and significant induction of classical pro-inflammatory mediators such as PGE 2 confirmed successful stimulation by LPS. Further significantly increased plasma level elevations were seen after 16 hours for several lipids that in most cases had shown strong equally directed regulation in our patients. Such similar effects were observed with the majority of ceramides. These lipids remained significantly elevated after 24 hours in the majority of cases. A late but significant increase after 24 hours was observed for Cer (d18:1/18:1) . As observed in our patients, three lipids were significantly downregulated in the whole blood assays after LPS treatment, including S1P (d18:1) (Sphingosin-1-phosphat), Cer (d18:1/24:0), and S1P (d18:0) (Sphinganine-1-Phosphat) . No effect was seen using the LPS assay for GlcCer (d18:1/24:1), 1-AG, 2-AG, and 6-keto-PGF1alpha, although these lipids were strongly regulated in our patients. Finally, all endocannabinoids and all lysophosphatidic acids displayed opposite changes in plasma levels in patients and the whole blood assay. Of the 38 lipids investigated, only 16 showed a similar reactivity both in our patients and in our ex vivo model of systemic inflammation. summarizes the effect of LPS on the biosynthesis of all lipid mediators in whole blood, including information on a possible regulation in patients and information on a possible similar reactivity under both conditions. Clinical course of the patients included after cardiac surgery The average stay at the ICU was 43.2 days (min 8 days, max. 101 days). All patients received high dosages of vasopressors and inotropics at the beginning of intensive care therapy. The average SOFA score (modified SOFA using the Richmond Agitation-Sedation Scale in critically ill patients according to was 11.7 (min 8, max 15) at admission. There were four non-survivors . In vivo lipid mediator profile in ICU patients after cardiac surgery In the present study we analyzed baseline values of all lipids investigated in this study since their normal ranges in human plasma are largely unknown. For ceramides, normal ranges based on plasma levels of healthy subjects were similar to those reported in literature. E.g. baseline ceramide level in our study are consistent with those published by others (i.e. Cer (d18:1/16:0): 0.220 ± 0.044 µM vs 0.24–0.30, Cer (d18:1/18:0): 0.067 ± 0.028 µM vs 0.001–0.11; Cer (d18:1/24:0): 4.490 ± 1.435 µM vs 1.15–3.34; Cer (d18:1/24:1): 0.746 ± 0.240 µM vs 0.71–1.14 . Comparisons of baseline values of other lipids investigated in this study could not be established, since their normal ranges in human plasma are largely unknown. shows examples of different lipid reactivity during the stages of systemic inflammation. Notably, the type of lipid is not mentioned at this stage, as the purpose of this figure is to introduce different types of lipid mediator reactivity with respect to normal ranges. shows an example of an unchanged lipid mediator. shows an example of a lipid mediator changed just above the normal range but still below the medianNR ± 2-fold upper and lower limit NRL95%. shows an example of a significantly regulated lipid mediator. Lipids that displayed plasma levels remaining consistently within the normal range during hospitalization, such as that shown in , were not evaluated further. We identified 22 of the 53 lipids which showed significant regulation during at least one time point on the ICU ward among at least 5 out of 9 patients, thus indicating potential clinical relevance. This group of prevalently regulated lipid mediators with potential clinical relevance could be further divided into lipids with either down-or upregulated levels during these general inflammatory conditions on ICU . We used IL-6 and leukocyte count as established markers of inflammation . Interestingly, the leukocyte count remained consistently elevated in most patients, pointing to the severity of illness the patients were suffering from, rather than specific clinical events. The relevant lipid plasma peaks were then compared with clinical phases. Regarding the type of plasma level regulation during the observation period and clinical events, we were able to define two major groups of lipids. The first group (dynamically upregulated lipids) features plasma levels rapidly changing from the normal range to the area above the twofold normal range upper NRL95%-limit. Based on visual interpretation for some of these lipids, changes in plasma levels seemed to have limited correlation with patients’ inflammatory status. and show representative graphs for 6-keto-PGF1 alpha and 2-AG. These two lipids partly correlated with inflammatory stages but rather seemed to correlate more to some specific clinical events. While 6-keto-PGF1 alpha had its peak shortly before initiation of continuous veno-venous hemodialysis (at least in 5 of the observed patients, see ), 2-AG peaked in all our nine patients during (right) heart failure in the context of either re-opening patient’s chest, implementation of veno-arterial ECMO or at least while significantly increasing the amount of catecholamines (see ). Besides these visual correlations, the remarkable increase of 6-keto-PGF1alpha in certain patients becomes more noteworthy considering that normal range levels of this lipid in healthy volunteers are below the detection limit of LC/MS-MS. Furthermore, Cer (d18:1/16:0), Cer (d18:1/18:1) and LacCer (d18:1/24:1) showed a tendency to accumulate in the plasma during the time of hospitalization and remained elevated for prolonged periods ( and , ). Notably, 6-keto-PGF1 alpha 1-AG, 2-AG and 20-HETE were the lipids that displayed the strongest maximal relative induction of biosynthesis compared with the plasma normal range concentrations in healthy persons. Among these, the strongest relative induction was seen with 6-keto-PGF1 alpha levels reaching up to 20 ng/ml and basal levels below the detection limit, followed by about 350-fold (2-AG), 200-fold (1-AG) and 40-fold maximal induction (20-HETE). A second group of lipids (significantly downregulated lipids) demonstrated plasma levels significantly below the normal range that, however remained impaired consistently and to similar degree during almost the entire time of hospitalization and were largely irrespective of clinical events. Examples of this type of lipid mediator, such as plasma levels of S1P (d18:1) and Lactosylceramide LacCer (d18:1/24:0), are shown in and . Several lipids displayed rather low normal-range concentrations, with lower NRL95% limits already close to the LC-MS/MS detection limit and levels that apparently were further reduced during hospitalization. This applies for example to LPA (18:2), LPA (18:3), LPA (20:4), 9,10-DiHOME, PEA, OEA and PGE 2 . Surprisingly, we did not observe any relevant elevation of the classical pro-inflammatory products of the cyclooxygenase-2 and 5-lipoxygenase pathway, including PGE 2 and 5-HETE or LTB 4 . All graphs of prevalently regulated lipids with potential clinical relevance (see ) not shown in – can be found in the supplemental section, - . Ex vivo lipid mediator profile in LPS-stimulated human whole blood Next, we compared the changes of lipid mediators from patients with systemic inflammation with an ex vivo model of endotoxin-induced inflammation. Here, venous whole blood from healthy volunteers was stimulated with LPS for 6 h, 16 h, and 24 h. The lipid mediator profile included 38 lipids analyzed using LC-MS/MS. Notably, due to a technical failure during LC-MS/MS analysis, only 38 of the 53 lipid mediators could be analyzed, however including all important lipids in patients described above. Among the 38 included lipids we identified 27 lipid mediators with significantly changed plasma levels after incubation with LPS, as compared to the untreated control. and include graphs for selected lipids from this LPS-based assay that either had already shown significant regulation in the ICU patients or only showed reactivity in our LPS model but not in our patients. Graphs of lipid plasma levels after stimulation of whole blood with LPS not included in and can be found in - . Further details on the results of the LPS analysis at different time points can be found in . The earliest significant increase compared with the control after 6 hours was seen with prostaglandin PGE 2 , PGF 2alpha and Cer (d18:0/16:0) followed by PGD 2 , although these lipids were not regulated in our patients. The early and significant induction of classical pro-inflammatory mediators such as PGE 2 confirmed successful stimulation by LPS. Further significantly increased plasma level elevations were seen after 16 hours for several lipids that in most cases had shown strong equally directed regulation in our patients. Such similar effects were observed with the majority of ceramides. These lipids remained significantly elevated after 24 hours in the majority of cases. A late but significant increase after 24 hours was observed for Cer (d18:1/18:1) . As observed in our patients, three lipids were significantly downregulated in the whole blood assays after LPS treatment, including S1P (d18:1) (Sphingosin-1-phosphat), Cer (d18:1/24:0), and S1P (d18:0) (Sphinganine-1-Phosphat) . No effect was seen using the LPS assay for GlcCer (d18:1/24:1), 1-AG, 2-AG, and 6-keto-PGF1alpha, although these lipids were strongly regulated in our patients. Finally, all endocannabinoids and all lysophosphatidic acids displayed opposite changes in plasma levels in patients and the whole blood assay. Of the 38 lipids investigated, only 16 showed a similar reactivity both in our patients and in our ex vivo model of systemic inflammation. summarizes the effect of LPS on the biosynthesis of all lipid mediators in whole blood, including information on a possible regulation in patients and information on a possible similar reactivity under both conditions. Discussion Eicosanoids, sphingolipids, and endocannabinoids belong to the group of bioactive lipids crucially involved in several inflammatory diseases such as bronchial asthma, local inflammation, allergy, and arthritis . However, their role in systemic inflammatory responses in critically ill patients is only partly understood. We initially hypothesized that plasma levels of at least some lipids might be changed during e.g. sepsis periods, theoretically providing a possible rationale for their use as new biomarkers or for identifying crucial signaling pathways and new therapeutic strategies. Therefore, plasma levels of several lipids from nine ICU patients following cardiac surgery were analyzed daily during hospitalization and compared with clinically established biomarkers of inflammation such as leukocytes and IL-6. We found that 22 of the 53 lipids analyzed by LC-MS/MS were significantly changed during clinical events. Among these, 2 lipid mediators showed a possible increase during acute kidney injury or (right) heart failure respectively, while the established inflammation parameters such as IL-6 only showed a congruency to a low degree with 11 other lipid mediators. Another group of 9 lipids displayed plasma levels that were consistently lower than the normal range. Several studies investigating the role of lipids in sepsis have confirmed the crucial role of certain lipids such as ceramides, PGE 2 , LTB 4 and S1P (d18:1) in the pathogenesis of this severe immune dysregulation . However, to our knowledge, no previous study has analyzed such a large number of different lipid levels using a highly sensitive LC/MS-MS methodology at that many time points throughout the patients stay at the ICU. Moreover, our study reflects lipid levels in hospitalized patients under clinically realistic conditions rather than using standardized and simplified animal models of disease or even ex vivo or in vitro models of sterile inflammation. Some studies investigated the role of eicosanoids and sphingolipids in sepsis. The role of elevated levels of certain cyclooxygenase-2-derived prostanoids with effects on the hemodynamic and inflammatory processes during sepsis has been published . Accordingly, we could confirm these findings by showing that the COX-2-derived metabolite 6-keto-PGF 1alpha is elevated in almost all patients. 6-keto-PGF 1alpha is a stable hydrolysis product of prostacyclin (PGI 2 ). Given the short half-life of PGI2 in plasma (a few minutes), its synthesis is more accurately reflected by measurement of the 6-keto-PGF 1alpha metabolite. PGI 2 has several functions that include the promotion of platelet aggregation and thrombosis, triggering endothelial adhesion and extravasation of neutrophils at sites of inflammation, and suppressing vasoconstrictor responses to pressor hormones, thereby crucially contributing to the clinical appearance of sepsis and as observed here, acute kidney injury. The increased use of vasopressors and the subsequent vasopressor desensitization might play a significant role in our observed patients. Anyhow, our study attaches importance to the pathophysiological role of PGI 2 and strongly encourages further research into the role of this lipid and 6-keto-PGF 1alpha . In accordance with a study by Bruegel et al., we did not observe any relevant elevation of prostaglandin E 2 or 5-lipoxygenase pathway products (e.g. 5-HETE or LTB 4 ) in the patients, particularly as these synthesis pathways are suppressed under septic conditions . Increased levels of Cer (d18:1/16:0), Cer (d18:1/18:1) and Cer (d18:0/18:0) were tracked in the present study, supporting a number of studies that described increased levels of ceramide in septic patients, which were associated with lower survival rate and poor prognosis . This aligns well with recent studies suggesting that ceramides are pathophysiologic relevant players in cardiovascular diseases involved in regulation of blood pressure, vascular tone and heart function . However, Cer (d18:1/16:0), and to a lower degree, Cer (d18:1/18:1), displayed a tendency to accumulate in certain patients, which excludes these metabolites as biomarkers for indicating transitions between clinical stages. On the other hand, such findings raise the important question of their largely unknown potential (patho)physiological role in sepsis. Furthermore, Winkler et al. found inversely decreased S1P (d18:1) concentrations in septic patients. In our study, S1P(d18:1) levels were impaired irrespectively of any clinical stage, excluding S1P(d18:1) as a potential marker to indicate septic stages. Similar to a previous report by colleagues investigating 2-AG levels in patients during endotoxic shock , we found strongly increased concentrations of the endocannabinoids 1- and 2-AG in our patients. The role of endocannabinoids in our setting is largely unclear, as anti-inflammatory and pro-inflammatory effects have been reported . We could show that an increase in endocannabinoids is not restricted to the presence of endotoxin and bacterial infections. We can only speculate, how 2-AG levels are linked to (right) heart failure, but it could be an important signaling mediator, causing the typical biological effects of endocannabinoids at cannabinoid receptor 1, such as reduction in anxiety, analgesia, euphoria and impairment of short-term memory and also enhancement of hemorrhagic and endotoxin-induced hypotension . The increase in 14,15-DHET, which is a direct metabolite and therefore index of 14,15-EET levels, might act as a mechanistic player in the occurrence of any hypotension. We further demonstrated that LacCer (d18:1/24:0) levels were significantly and consistently reduced in the plasma of the patients included in this study. The limited data suggests that lactosylceramides are more likely to be considered as pro-inflammatory mediators triggering nitric oxide synthesis, oxidative stress and neutrophil activation . However, more studies are needed to investigate whether the impaired levels of lactosylceramides might represent an endogenous anti-inflammatory reaction countering an inflammatory process. For biomarker candidates the degradation/elimination half-life in plasma is of relevance as too long half-life values may prevent the lipid to indicate rapid changes in clinical stages. However, less is known about the half-life of the lipid mediators analyzed in this study. In a study investigating the biosynthetic pathway of very-long and long-chain sphingolipids in 293T cells the turnover rate of different ceramide species could be determined. The authors found that due to their more rapid metabolism into sphingolipids, ceramides containing very-long chain fatty acids such as Cer (d18:1/26:0) and Cer (d18:1/24:1) have a much shorter half-life than long-chain ceramides containing C16:0 and C18:0 fatty acids [Cer (d18:1/16:0), Cer (d18:1/18:1)]. The reported turnover rates for the respective ceramides range from 4,5 h to approximately 24 h for long-chain [i.e. Cer (d18:1/16:0), Cer (d18:1/18:0)] and very-long chain ceramides [i.e. Cer (d18:1/24:0), Cer(d26:1)], respectively. This relatively long half-life may explain persistent plasma level of Cer (d18:1/16:0) and Cer (d18:1/18:1) in certain patients. In contrast to these lipids, the plasma half-life of 1- and 2-AGs and 6-keto-PGF 1alpha are in the range of minutes, with the longest reported half-life being 30 min for 6-keto-PGF 1alpha . Their short half-life may represent an important prerequisite to function as potential biomarker candidates. Notably, the lipid levels analyzed in the plasma of LPS treated whole blood only partly showed similar alterations compared with changes of lipid plasma levels of the patients in this study. This supports the theory, that ex vivo inflammation models such as the LPS assay used in our study only provide rather limited opportunities for predicting changes of lipid mediators at least in critically ill patients. Retrospectively, some study design settings are, however, limiting the amount and level of information that could be finally obtained from our study. Collecting only one blood sample daily for lipid mediator analysis made it impossible to evaluate the exact rapidity of the plasma lipid mediator level alterations upon transition between clinical stages. Moreover it prevented a comparison with classical biomarkers such as IL-6. Furthermore, the change in lipid plasma levels was partly heterogeneous among the included patients, possibly due to their different medical histories, co-medications, diverse pharmacotherapeutic treatment regimens, and other medical-technical interventions at the ICU such as dialysis, artificial respiration, and extracorporeal life support and other medical-technical interventions at the ICU. Remarkably, the dynamics of lipid mediators may be influenced by norepinephrine infusions affecting the release of certain lipid metabolites . In some instances, we observed depleted lipid mediator plasma levels, which might have been due to the well-recognized globally impaired capability of the mediator-synthesizing machinery, such as had been already reported for certain lipids and leukocytes . It remains unclear whether there is in fact sufficient “regulatory space” for these lipids to achieve significantly downregulated levels as, for example, non-enzymatic production may also cause a low minimum basal lipid level which cannot be exceeded. Moreover, the same blood collection time interval was applied for all mediators investigated suggesting that possible individual circadian fluctuations in lipid biosynthesis/regulation could not be recorded. Unfortunately, it was technically impossible to collect subject-related baseline lipid levels prior to surgery and the lipid values of patients therefore had to be compared to normal range of healthy volunteers. However, although a number of parameters potentially disrupted lipid plasma levels of patients, certain lipid mediators, including 6-keto-PGF 1alpha, Cer (d18:1/16:0), Cer (d18:1/18:1), 1-AG, 2-AG, and Cer (d18:0/18:0) showed clear and largely consistent changes during the clinical stages in the majority of patients monitored and potentially affect the course of the disease. To partially compensate for these disruptive effects, the level to reach “significant” regulation was therefore elevated in our study (± 2-fold upper and lower limit NRL 95% ). Nevertheless, a number of mediators showed very strong regulation during inflammatory stages highly exceeding these already expanded normal ranges. Regarding the ex vivo LPS assay we had to use heparinized blood as the ex vivo biosynthesis of a number of lipids is dependent on calcium influx into immune cells making the use of citrated blood impossible because of calcium depletion. Thus, lipid biosynthesis after the ex vivo stimulation of whole blood using LPS may not fully reflect the situation in patients. To archive time-dependent effects after pro-inflammatory stimulation in accordance with our patient study design, simulated samples we compared to untreated controls at one time point (0h). Thus we cannot fully exclude potentially small deviations due to time-dependent induction/suppression of biosynthesis of lipid mediators purely caused by incubation at 37°C. However, as strong induction/suppression of the relevant pro-inflammatory lipids after stimulation with LPS are well documented and because of the heterogeneity of the samples originating from different donors broadening the scatter of measured control values such effects should be negligible. Furthermore, a large number of prior publications already confirm the findings from the LPS-assay. However, to achieve optimal comparability to the results from ICU patients, we decided repeating this LPS assay using the same analytic LC/MS-MS methodology for both conditions identified in this study. A systematically evaluation of selected lipids with markedly higher patient numbers during sepsis according to actual definitions, acute kidney injury and extracorporeal procedures seems to be the next big step, while monitoring patients’ pharmacotherapy. Regarding the question of impact and novelty of our study in this field of research, we conclude that some of our findings confirm other published studies investigating lipid profiles during sepsis and inflammation, predominantly in animal models and partly in patients. Thus, subsequent deeper studies on these mediators in specific diseases such as e.g. pneumogenic sepsis or postoperative heart failure are encouraged. However, our study is the first to provide insights into the dynamics of plasma lipid level changes during hospitalization and clinical events. Despite all limitations and technical problems, we were able to identify several new lipid-based mediators with potential pathophysiological relevance and possible value as biomarkers. Conclusions Based on these findings, the following main conclusions can be drawn. First, we confirmed increased or decreased levels of certain lipids during inflammatory periods. Moreover, in several cases, these lipid levels remained consistently outside the normal range for extended periods, irrespective of any clinical event. Thus, our study strongly questions the proposed use of these lipids as biomarkers for indicating a transition between clinical stages. However, these lipids may have general yet undefined pathophysiological relevance for critically ill patients. We confirm the down- and upregulation of other types of lipids such as 1-AG, 2-AG and 6-keto-PGF 1alpha . These lipids showed high dynamics during the observation period, displayed remarkably elevated plasma levels and might therefore function as highly volatile candidates for markers indicating transitions of clinical stage, such as induction of systemic inflammation, aggravated acute kidney injury or (right) heart failure. Taken in its entirety, our comprehensive pilot study provides extended and deeper insights into the role of lipid signaling in hospitalized ICU patients after cardiac surgery. Our study strongly encourages more targeted studies including a higher number of patients allowing sub-stratifications analyses and evaluation of the function and use of selected lipids as players and potential biomarkers in the critically ill. The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. The study design was approved by the local ethics committee at Frankfurt University Hospital (AZ: 265/09 and 476/13). The ethics committee at Frankfurt University Hospital also granted written approval of an extension of the studies mentioned herein including the determination of hospitalized patients’ lipids in plasma. The ethics committee also agreed in writing to the experiments using LPS-stimulated whole blood and the subsequent analysis of lipid mediators. HN, JR, AU, DS, GG, KZ, PP, ES, MP, PM, EU and TJM contributed to the conception or design of the work and sample acquisition; JR, UH, CA, NF, RG, ES, PM, TJM, VR and TB contributed to the acquisition, analysis, or interpretation of data; and JR, AU, UH, DS, MP, NF, PP, GG, ES, TB, VR and TJM have drafted the work or substantively revised it. All authors contributed to the article and approved the submitted version. |
Effect of different boron contents within boron-doped hydroxyapatite-chitosan nano-composite on the microhardness of demineralized enamel | f9f04dd3-1b6e-4eb9-b2e7-3617bceb219d | 11580543 | Dentistry[mh] | Dental enamel, consisting of structured hydroxyapatite crystals, acts as a protective barrier for the underlying dentin and dental pulp . Tooth decay involves the demineralization and erosion of hard tooth tissue due to acidic conditions and low pH levels . Factors such as diet, oral hygiene practices, and medical history contribute to the development and progression of tooth decay . Various non-invasive treatments are available to prevent tooth decay, which can help reduce the incidence of cavities. Fluoride-based therapies, such as fluoride varnish and fluoride-containing mouthwashes, are widely recognized as the leading approach in preventive dentistry. These treatments work by promoting the remineralization of hard tissues, converting hydroxyapatite into fluorapatite . Despite the availability of numerous fluoride interventions, tooth decay remains a prevalent chronic condition globally . However, there are some concerns regarding the widespread use of fluoride-containing compounds, due to the high risk of dental fluorosis development, especially in children . Recent studies are now focusing on new remineralization materials and techniques to enhance the effectiveness of decay control. For instance, the application of calcium and phosphate on demineralized enamel surfaces is being explored due to their similarities to tooth tissue properties . Hydroxyapatite contains calcium and phosphate with a structure akin to tooth enamel and possesses antibacterial properties. Nano-hydroxyapatite at a concentration of 10% has demonstrated the ability to remineralize tooth enamel by penetrating interprismatic spaces and interacting with interprismatic proteins . Chitosan, a polymer derived from chitin, exhibits antimicrobial properties and has been found to prevent decay effectively . Chitosan is a biocompatible and non-toxic material with diverse medical properties, including enhancing immune response, promoting wound healing, and exhibiting antimicrobial activity . Another significant characteristic of chitosan is its capacity to absorb hydrogen ions in acidic environments through the amino group in its structure. This absorption creates a positive charge that enables chitosan to bind to negatively charged surfaces like tooth enamel, soft tissues, and cell membranes . Boron, a semimetal element present in the human body and tooth enamel, plays a vital role in bone and tooth development . Research has shown conflicting results regarding the impact of boron on enamel remineralization, emphasizing the importance of determining the ideal concentration for leveraging its remineralization properties . Given the significance of remineralization in maintaining healthy teeth and improving oral health outcomes, this study is proposed to investigate the effects of different percentage of boron in boron-doped hydroxyapatite-chitosan nanocomposites on the microhardness of demineralized enamel. This research aims to bridge existing gaps in understanding and contribute to advancing dental care practices for better oral health and quality of life. Null hypothesis of this study is that the microhardness of B@HApC group is not significantly different from the control group.
This in vitro study received ethical approval from the committee of Hamadan University of Medical Sciences (IR.UMSHA.REC.1401.1009). The study utilized healthy human premolar teeth extracted during orthodontic treatment and adhesive discs to examine the remineralization capability of the synthesized adhesives. The remineralization agent consists of 15 mM Ca 2+ (CaCl 2 ), 9.0 mM phosphate (KH 2 PO 4 ), and 1 ppm of fluoride (NaF). On the other hand, the demineralization agent contains 2.2 mM NaH 2 PO 4 , 5.0 mM CaCl 2 , and 50 mM acetic acid. Additionally, artificial saliva was formulated with 4.0 g/L of NaCl, 4.0 g/L of KCl, 795.0 g/L of CaCl 2 .2H2O, 78.0 g/L of NaHPO 4 .2H 2 O, 0.007 g/L of Na 2 S.2H 2 O, and water and urea as needed . Inclusion criteria: Pre-molar teeth free from cracks and decay lesions. Exclusion criteria: Teeth showing signs of decay, cracks, or lesions, as well as samples that fractured during cutting. Sample size calculation Based on the variance in means between the initial two groups, a test power of 95%, and a 99% confidence level, a sample size of 13 per group was calculated using the following formula. To accommodate for potential sample loss, a total of 15 samples per group were included in the study . [12pt]{minimal}
$$\:n=}_{1-}+{Z}_{1-})}^{2}({{}_{1}}^{2}+{{}_{2}}^{2})}{{({}_{1}-{}_{2})}^{2}}$$ Synthesis of boron-doped hydroxyapatite chitosan nano-composite The sol-gel method is employed in an alkaline setting to prepare the composite . Initially, 20 mL of tetraethyl orthosilicate (TEOS) is combined with 2.8 mL of nitric acid and 9.9 mL of distilled water, undergoing acid hydrolysis for one hour. At this stage, precise quantities of boric acid are introduced to obtain boron nano-composites with concentrations of 5%, 10%, and 15% for each group, followed by a 30-min stirring period. Subsequently, 2.15 mL of triethyl phosphate (TEP) is incorporated into the solution, along with a specified amount of Ca(NO 3 ) 2 . Further, 2 mL of 0.5–1% acetic acid is added while stirring to 0.5% chitosan. Upon the addition of ammonia and chitosan, the mixture is stirred until a gel-like substance forms. Additionally, a composite lacking boron is also formulated. The resulting gels are maintained at 60 ℃ for 24 h. Samples preparation Thirty dental samples, collected within the last three months, were immersed in 0.5% chloramine and stored in distilled water at 37 °C until the commencement of the study. The tissues surrounding the teeth were meticulously cleaned with a sterile blade, and an Olympus stereomicroscope (Shinjuku, Tokyo, Japan) was utilized to inspect for any cracks or specific lesions. Subsequently, the dental crown was partitioned into facial and palatal sections using a diamond saw-disc (Tizkavan, Tehran, Iran), and embedded in acrylic resin (Acropars, Tehran, Iran), before being returned to distilled water. The samples were then exposed to a demineralization solution with a pH of 4.6 for 8 h, followed by immersion in an artificial saliva solution for one hour. This was succeeded by placement in a remineralization solution with a pH of 7 for 15 h. This sequential process was repeated over a period of 14 days to establish a demineralized enamel surface, with the solutions being refreshed every two days. All samples are randomly selected and divided into four groups ( N = 15): Group 1: Hydroxyapatite chitosan nanocomposite (without boron). Group 2: Boron-doped hydroxyapatite chitosan nanocomposite with a concentration of 5%. Group 3: Boron-doped hydroxyapatite chitosan nanocomposite with a concentration of 10%. Group 4: Boron-doped hydroxyapatite chitosan nanocomposite with a concentration of 15%. The nanocomposites are blended with distilled water to achieve a slurry mixture (2 mg powder in 3 mL distilled water). Subsequently, using a microbrush, they are applied twice daily at 9 AM and 5 PM for a duration of 4 min on the demineralized enamel surface after drying the teeth directly. Following application, the teeth are rinsed and immersed in artificial saliva for 48 h. This procedure is repeated over a span of 28 days, and upon completion, the samples are preserved in distilled water for an additional 48 h . Microhardness test The microhardness of the samples is assessed through the Vickers test method. To prepare a smooth surface on the teeth for indent placement during the test, silicon carbide paper with a grit size of 600, utilized under running water, is employed. Microhardness measurement is conducted using a microhardness tester (Future Tech, FM-810, Kawasaki, Japan), by applying a pyramid-shaped diamond indenter with a force of 300 g for 10 s at three distinct points. The microhardness of the samples is recorded both before and after the application of the nanocomposite intervention . SEM evaluation The surface morphology of the enamel was analyzed using a scanning electron microscope (SEM) equipped with a sputtering module (TESCAN MIRA, operating at 15KV in Czech Republic). A single sample from each group was chosen, dehydrated with ethanol, coated with a thin layer of gold, and scrutinized at magnifications of 10,000×, 50,000×, and 100,000×. Nanoparticle characterization X-Ray Diffraction (XRD) analysis was conducted to study the crystal structure of the nanocomposites. The XRD instrument used was the Philips PW1730 (Netherlands), operating at a wavelength of A1.5405 and a power of 40KW/30 mA. The examination of the structure and crystalline phase of the nanoparticles involved scanning in the range of 10 to 80 degrees. The synthesis of the nanocomposite and the identification of functional groups in the sample were evaluated using the FT-IR (Fourier Transform Infrared Spectroscopy) technique. FTIR spectra of the synthesized nanocomposites were captured within the frequency range of 4000 –400 cm − 1 employing FT-IR equipment (Perkin Elmer, Spectrum One, USA). The resulting spectra provided insights into the functional groups and chemical bonding structure present in the sample. The morphology, shape, particle size, and distribution at the nanoscale were analyzed using a Transmission Electron Microscope (TEM) model (Philips, specifically the EM208S, Netherlands). Statistical analysis Quantitative variables were described using mean and standard deviation, while the Analysis of Variance (ANOVA) test was utilized to compare quantitative variables among the study groups. Within-group comparisons were conducted using the Paired t test. Post hoc analyses, such as Tukey’s test, were applied for pairwise mean microhardness comparisons between groups. The significance level for all tests was set at p < 0.05. Statistical analyses were carried out using SPSS software version 27.
Based on the variance in means between the initial two groups, a test power of 95%, and a 99% confidence level, a sample size of 13 per group was calculated using the following formula. To accommodate for potential sample loss, a total of 15 samples per group were included in the study . [12pt]{minimal}
$$\:n=}_{1-}+{Z}_{1-})}^{2}({{}_{1}}^{2}+{{}_{2}}^{2})}{{({}_{1}-{}_{2})}^{2}}$$
The sol-gel method is employed in an alkaline setting to prepare the composite . Initially, 20 mL of tetraethyl orthosilicate (TEOS) is combined with 2.8 mL of nitric acid and 9.9 mL of distilled water, undergoing acid hydrolysis for one hour. At this stage, precise quantities of boric acid are introduced to obtain boron nano-composites with concentrations of 5%, 10%, and 15% for each group, followed by a 30-min stirring period. Subsequently, 2.15 mL of triethyl phosphate (TEP) is incorporated into the solution, along with a specified amount of Ca(NO 3 ) 2 . Further, 2 mL of 0.5–1% acetic acid is added while stirring to 0.5% chitosan. Upon the addition of ammonia and chitosan, the mixture is stirred until a gel-like substance forms. Additionally, a composite lacking boron is also formulated. The resulting gels are maintained at 60 ℃ for 24 h.
Thirty dental samples, collected within the last three months, were immersed in 0.5% chloramine and stored in distilled water at 37 °C until the commencement of the study. The tissues surrounding the teeth were meticulously cleaned with a sterile blade, and an Olympus stereomicroscope (Shinjuku, Tokyo, Japan) was utilized to inspect for any cracks or specific lesions. Subsequently, the dental crown was partitioned into facial and palatal sections using a diamond saw-disc (Tizkavan, Tehran, Iran), and embedded in acrylic resin (Acropars, Tehran, Iran), before being returned to distilled water. The samples were then exposed to a demineralization solution with a pH of 4.6 for 8 h, followed by immersion in an artificial saliva solution for one hour. This was succeeded by placement in a remineralization solution with a pH of 7 for 15 h. This sequential process was repeated over a period of 14 days to establish a demineralized enamel surface, with the solutions being refreshed every two days. All samples are randomly selected and divided into four groups ( N = 15): Group 1: Hydroxyapatite chitosan nanocomposite (without boron). Group 2: Boron-doped hydroxyapatite chitosan nanocomposite with a concentration of 5%. Group 3: Boron-doped hydroxyapatite chitosan nanocomposite with a concentration of 10%. Group 4: Boron-doped hydroxyapatite chitosan nanocomposite with a concentration of 15%. The nanocomposites are blended with distilled water to achieve a slurry mixture (2 mg powder in 3 mL distilled water). Subsequently, using a microbrush, they are applied twice daily at 9 AM and 5 PM for a duration of 4 min on the demineralized enamel surface after drying the teeth directly. Following application, the teeth are rinsed and immersed in artificial saliva for 48 h. This procedure is repeated over a span of 28 days, and upon completion, the samples are preserved in distilled water for an additional 48 h .
The microhardness of the samples is assessed through the Vickers test method. To prepare a smooth surface on the teeth for indent placement during the test, silicon carbide paper with a grit size of 600, utilized under running water, is employed. Microhardness measurement is conducted using a microhardness tester (Future Tech, FM-810, Kawasaki, Japan), by applying a pyramid-shaped diamond indenter with a force of 300 g for 10 s at three distinct points. The microhardness of the samples is recorded both before and after the application of the nanocomposite intervention .
The surface morphology of the enamel was analyzed using a scanning electron microscope (SEM) equipped with a sputtering module (TESCAN MIRA, operating at 15KV in Czech Republic). A single sample from each group was chosen, dehydrated with ethanol, coated with a thin layer of gold, and scrutinized at magnifications of 10,000×, 50,000×, and 100,000×.
X-Ray Diffraction (XRD) analysis was conducted to study the crystal structure of the nanocomposites. The XRD instrument used was the Philips PW1730 (Netherlands), operating at a wavelength of A1.5405 and a power of 40KW/30 mA. The examination of the structure and crystalline phase of the nanoparticles involved scanning in the range of 10 to 80 degrees. The synthesis of the nanocomposite and the identification of functional groups in the sample were evaluated using the FT-IR (Fourier Transform Infrared Spectroscopy) technique. FTIR spectra of the synthesized nanocomposites were captured within the frequency range of 4000 –400 cm − 1 employing FT-IR equipment (Perkin Elmer, Spectrum One, USA). The resulting spectra provided insights into the functional groups and chemical bonding structure present in the sample. The morphology, shape, particle size, and distribution at the nanoscale were analyzed using a Transmission Electron Microscope (TEM) model (Philips, specifically the EM208S, Netherlands).
Quantitative variables were described using mean and standard deviation, while the Analysis of Variance (ANOVA) test was utilized to compare quantitative variables among the study groups. Within-group comparisons were conducted using the Paired t test. Post hoc analyses, such as Tukey’s test, were applied for pairwise mean microhardness comparisons between groups. The significance level for all tests was set at p < 0.05. Statistical analyses were carried out using SPSS software version 27.
Figure displays the XRD pattern of the synthesized hydroxyapatite (HAp) sample. The significant peaks in the pattern are observed at 2θ angles of 26.82°, 30.82°, 49.78°, and 64.27°, corresponding to the 002, 211, 123, and 304 reflection planes, respectively, of the HAp crystal structure. A distinct peak around 30.82°, representing the (211) plane of the HAp crystal lattice, suggests a well-defined and predominant hexagonal crystal structure in the HAp sample, characteristic of the standard HAp phase (JCPDS 84-1998). The TEM micrograph of the synthesized nanomaterial, as depicted in Fig. , illustrates nanoparticles with a semi-needle shape . The examined sample exhibited nearly identical Fourier-transform infrared (FT-IR) spectra, displaying (See Fig. ) the characteristic bands of boron doped hydroxyapatite chitosan (B@HApC). The major peaks observed at 1090 –1035 cm − 1 correspond to PO 4 (V 3 ), while the peak at 960 corresponds to PO 4 (V 1 ). Minor peaks at 472 cm − 1 are associated with PO4 (V 2 ). Furthermore, the peaks at 1400 indicate carbon interaction with nanoparticles, corresponding to CO 3 . The stretching band at 3400 cm − 1 corresponds to the OH group of hydroxyapatite . The increase in B content in the nanocomposite was indicated by the intensity of the OH − related peak, suggesting the replacement of OH − with an ion provided by H 3 BO 3 . Additionally, peak at 1241 cm − 1 , and 1371 cm − 1 indicated the presence of BO 3 3− . The asymmetric stretching bonds within the chitosan structure (C-O-C) is registered at 1149 cm − 1 . Microhardness test The microhardness results of the samples prior to the nanocomposite application are detailed in Table . Notably, the control group exhibits the highest mean microhardness at 293.19 ± 21.80, while the 5% group demonstrates the lowest value at 264.47 ± 52.23. Following the nanocomposite application, the microhardness values are outlined in Table . The 15% group displays the highest microhardness value with a mean and standard deviation of 328.8 ± 22.00, whereas the 5% group indicates the lowest mean value at 304.60 ± 25.48. The results of the paired t-test for the variance in mean microhardness of demineralized enamel before and after the nanocomposite intervention are summarized in Table . The analysis reveals a significant variance in mean values across the control, 5%, 10%, and 15% groups (p-value < 0.05). The 15% group exhibits the largest mean difference at 49.42 ± 23.21, while the control group shows the smallest mean difference at 15.20 ± 14.32. Based on the findings presented in Table and the Tukey HSD test results, pairwise comparisons of mean microhardness of demineralized enamel pre- and post-intervention were conducted. The study indicates a significant variance in mean microhardness between the 15% boron-doped nanocomposite group and the control group. Conversely, no significant differences were observed between the other groups and the control group. Moreover, pairwise comparisons between the 5%, 10%, and 15% groups did not reveal significant variances in mean microhardness, suggesting that the presence of boron in the hydroxyapatite-chitosan nanocomposite at concentrations exceeding 15% notably enhanced the microhardness of demineralized enamel samples. However, statistically, increasing the boron concentration from 5 to 10% or from 10 to 15% did not result in a significant impact on enhancing the microhardness of demineralized enamel. Figure illustrates the average variance in microhardness between demineralized enamel before and after the intervention, highlighting that the most significant difference in microhardness is observed in the 15% group. SEM evaluation Scanning Electron Microscope (SEM) was employed to analyze the surface morphology of the enamel, capturing SEM micrographs of the enamel surface in various groups. The results are depicted in Fig. . The SEM analysis of the sample surfaces unveiled distinct features. The control group exhibited a porous enamel surface. In contrast, the 5%, 10%, and 15% groups displayed a mineral layer comprising hydroxyapatite crystals that enveloped the porous enamel surface. Notably, elevating the boron concentration to 15% resulted in the formation of a thicker and more uniform mineral layer.
The microhardness results of the samples prior to the nanocomposite application are detailed in Table . Notably, the control group exhibits the highest mean microhardness at 293.19 ± 21.80, while the 5% group demonstrates the lowest value at 264.47 ± 52.23. Following the nanocomposite application, the microhardness values are outlined in Table . The 15% group displays the highest microhardness value with a mean and standard deviation of 328.8 ± 22.00, whereas the 5% group indicates the lowest mean value at 304.60 ± 25.48. The results of the paired t-test for the variance in mean microhardness of demineralized enamel before and after the nanocomposite intervention are summarized in Table . The analysis reveals a significant variance in mean values across the control, 5%, 10%, and 15% groups (p-value < 0.05). The 15% group exhibits the largest mean difference at 49.42 ± 23.21, while the control group shows the smallest mean difference at 15.20 ± 14.32. Based on the findings presented in Table and the Tukey HSD test results, pairwise comparisons of mean microhardness of demineralized enamel pre- and post-intervention were conducted. The study indicates a significant variance in mean microhardness between the 15% boron-doped nanocomposite group and the control group. Conversely, no significant differences were observed between the other groups and the control group. Moreover, pairwise comparisons between the 5%, 10%, and 15% groups did not reveal significant variances in mean microhardness, suggesting that the presence of boron in the hydroxyapatite-chitosan nanocomposite at concentrations exceeding 15% notably enhanced the microhardness of demineralized enamel samples. However, statistically, increasing the boron concentration from 5 to 10% or from 10 to 15% did not result in a significant impact on enhancing the microhardness of demineralized enamel. Figure illustrates the average variance in microhardness between demineralized enamel before and after the intervention, highlighting that the most significant difference in microhardness is observed in the 15% group.
Scanning Electron Microscope (SEM) was employed to analyze the surface morphology of the enamel, capturing SEM micrographs of the enamel surface in various groups. The results are depicted in Fig. . The SEM analysis of the sample surfaces unveiled distinct features. The control group exhibited a porous enamel surface. In contrast, the 5%, 10%, and 15% groups displayed a mineral layer comprising hydroxyapatite crystals that enveloped the porous enamel surface. Notably, elevating the boron concentration to 15% resulted in the formation of a thicker and more uniform mineral layer.
Addressing early-stage caries with restorative materials can potentially result in long-term complications like marginal leakage and secondary caries. Consequently, contemporary dentistry emphasizes non-invasive strategies for remineralizing dental enamel and averting demineralization when caries is at its nascent phase. The current focus lies in identifying approaches or materials that can decelerate the demineralization process or expedite the remineralization process. Boron, classified as a semimetal element, plays a role in bone metabolism and is recognized for its potential antimicrobial properties, making it a promising candidate for dental applications . Presently, the utilization of boron in the form of boron-doped hydroxyapatite chitosan (B-HApsC) has garnered attention due to its osteoinductive effects, given the significance of boron as an essential element in the human body . While the main mechanism of B-HApsC involves deposition, akin to nano hydroxyapatite, boron in mineralized tissues appears to exist in a form capable of generating reversible boroesters that regulate mineralized tissue formation and upkeep, consequently bolstering bone and teeth strength. Our study indicates that higher concentrations of boron synergistically enhance deposition . Epidemiological studies and animal experiments have highlighted the presence of boron in soil, drinking water, and food. Its anti-caries effect is further accentuated when combined with fluoride. Moreover, studies have underscored the significant role of boron in bone metabolism and regenerative medicine . Taking into account these properties of boron, our study delved into investigating, for the first time, the impact of varying concentrations of boron integrated into boron-doped hydroxyapatite chitosan nanocomposite on demineralized enamel surfaces. The mineralization process in this study was executed using conventional remineralization and demineralization solutions, adhering to the pH-cycling protocol . Research indicates that pH-cycling is a dependable approach for inducing lesion advancement and replicating artificial enamel demineralization, effectively simulating conditions akin to caries development . Surface microhardness (SMH) serves as a dependable and swift approach for assessing enamel remineralization and demineralization status. In our study, we opted for the Vickers indenter over the Knoop indenter. The Vickers indenter creates a square-shaped impression on the enamel surface, facilitating easy detection and measurement . The Vickers test was utilized for intergroup comparisons . To mitigate any potential bias, the test was conducted at three distinct points on the enamel surface, and the average of these measurements was employed for subsequent statistical analysis . In order to examine the surface morphology, SEM images of the dental samples were obtained. SEM is a common and accepted method for observing surface changes in dental enamel and conducting topographical analyses . SEM imaging was utilized to examine the surface morphology of the dental samples, a widely acknowledged method for observing surface alterations in dental enamel and conducting topographical analyses . The study results indicated that the enhancement in microhardness of remineralized enamel was statistically significant solely in the group incorporating 15% boron nanocomposite compared to the control group ( p < 0.05). Nonetheless, clinical assessment of the samples pre- and post-intervention also unveiled an augmentation in microhardness in the 5% and 10% groups. The n-HApC nanocomposite has exhibited superior fluoride capacity compared to n-HAp alone. Research studies have highlighted the effectiveness of the n-HApC nanocomposite as a fluoride-releasing agent, attributed to its efficacy, cost-efficiency, and biocompatibility . Memarpour et al. have proposed that using nanohydroxyapatite before applying a fluoride varnish can serve as an efficient approach to fill porous spaces in demineralized enamel and facilitate enamel remineralization . Moreover, Masaeli et al. have demonstrated that the incorporation of nanohydroxyapatite and chitosan enhances the wear resistance and hardness of glass ionomer cement . Scribante et al. explored the biomimetic impact of nano-hydroxyapatite on demineralized tooth enamel prior to bonding orthodontic brackets. Their findings revealed that the application of nano-hydroxyapatite enhanced the bond strength of brackets in comparison to demineralized enamel . The research conducted by Zhang et al. demonstrated that the use of chitosan and bioactive glass can enhance the remineralization of white spot lesions, particularly with a brief exposure to polyacrylic acid . These findings align with the outcomes of our study. Additionally, Hyraishi et al. observed that incorporating enamel and dentin in an S-PRG solution with boron led to elevated pH levels in samples immersed in both the S-PRG and high boron solutions. SEM analysis indicated demineralization in the control group, whereas the enamel surface remained unaffected in the S-PRG and boron solution groups . Durand et al. posited that a reduction in boron content can result in enamel hypoplasia, manifesting as irregularities in the enamel matrix. The primary change stems from a decrease in matrix thickness, leading to alterations in tooth contour. Consequently, the tooth experiences morphological shifts and becomes more prone to decay . Tasli et al. illustrated that human dental stem cells cultured in the presence of sodium pentaborate pentahydrate displayed heightened alkaline phosphatase (ALP) activity, along with increased expression of genes and proteins related to bone and tooth formation. This underscores an enhanced odontogenic and osteogenic potential of the cells . Moreover, another study indicated that boron can activate the mitogen-activated protein kinase (MAPK) signaling pathway, which plays a significant role in stimulating cell proliferation at low concentrations and inhibiting it at high concentrations . Given the importance of these signaling pathways in the differentiation of mesenchymal cells into osteoblasts and their proliferation, these findings hold relevance in the realm of bone biology . Hakki et al. provided evidence showcasing the beneficial impact of boron supplements on mineral tissues in an animal model . Additionally, Nielsen’s research revealed that boron supplementation enhanced the fracture strength of rat femurs . Gümüsderelioğlu’s study delved into the effects of encapsulated boron on the proliferation and differentiation of pre-osteoblastic cells (MC3T3-E3) in laboratory settings, demonstrating a stimulating effect of boron on osteoblasts . Multiple studies have highlighted that incorporating boric acid in antimicrobial formulations enhances their antimicrobial effectiveness. This can be linked to boron’s essential role in tooth enamel remineralization. The augmentation of tooth enamel, a critical mineral component, can be credited to the presence of boron, with its effects being further amplified in synergy with hydroxyapatite-chitosan . The aforementioned studies align with the outcomes of our research, validating the antibacterial properties of boric acid and its role in enhancing the strength of mineral tissues, such as tooth enamel. In the B@HApC groups (5%, 10%, 15%), boron-doped crystals effectively coat the porous surface of hydroxyapatite-chitosan enamel prisms, augmenting the microhardness of tooth enamel. SEM analysis corroborated the microhardness results. The porous appearance resulting from enamel surface demineralization, inadequately addressed by hydroxyapatite-chitosan in the control group nanocomposite, was notably covered by a mineral layer in the intervention groups. This signifies the efficacy of boron in enhancing the microhardness of demineralized tooth enamel.
In summary, the findings of this study indicate that elevating the boron concentration in the B@HApC composite results in enhanced microhardness of demineralized enamel. In vitro assessments of dental samples and SEM images in the treated groups revealed the most substantial increase in enamel microhardness in the 15% group, followed by the 10% and 5% groups. Statistical analysis has confirmed a rise in the microhardness of demineralized enamel at concentrations of 15% and above. Research limitations Limitations of the research include the inability of surface microhardness evaluation to accurately replicate the clinical and histological depth of lesions. Additionally, challenges were encountered in obtaining sufficient and representative samples within a constrained timeframe. Another limitation of study was the absence of positive control group.
Limitations of the research include the inability of surface microhardness evaluation to accurately replicate the clinical and histological depth of lesions. Additionally, challenges were encountered in obtaining sufficient and representative samples within a constrained timeframe. Another limitation of study was the absence of positive control group.
|
Guideline-based care for psychiatric electroceuticals: Results from a National Survey of Board-Certified Psychiatrists | 33ce9c58-268f-4c29-b159-6bb4133139f8 | 11096671 | Psychiatry[mh] | CLINICAL PRACTICE AND RESEARCH GUIDELINES FOR PSYCHIATRIC ELECTROCEUTICAL INTERVENTIONS (PEIs) Clinical practice guidelines—’statements that include recommendations intended to optimize patient care that are informed by a systematic review of evidence and an assessment of the benefits and harms of alternative care options’ —may assist both practitioners’ and patients’ decisions in selecting the most effective and safest clinical care. Due to administrative cost-cutting, the influence of evidence-based medicine attempts to standardize interventions in the face of the growing complexity of healthcare, and rapid advancements in medical technologies, guidelines have become increasingly common and sought out by practitioners, governments and healthcare associations. Clinical practice guidelines may reduce morbidity and mortality, improve patient quality of life, promote distributive justice and improve efficiency and quality of healthcare. However, flawed guidelines can result in the delivery of ineffective, harmful or wasteful interventions. Therefore, efforts have been made to assess quality of healthcare guidelines. Even though guidelines may not evolve at the same pace as the scientific knowledge that justifies them, correctly applied guidelines can improve patient outcomes. – Neuromodulation is a promising and rapidly developing field within psychiatry. PEIs are neuromodulation techniques that utilize magnetic or electric stimuli for treating a range of psychiatric conditions. Electroconvulsive therapy (ECT) and repetitive transcranial magnetic stimulation (rTMS) are approved by the US Food and Drug Administration (FDA) for the treatment of major depressive disorder, while deep brain stimulation (DBS) and adaptive brain implants (ABIs) have not been FDA-approved for treating depression. In this paper, we review the current state of guidelines for PEIs and then present results from a national survey of psychiatrists, seeking to identify key considerations for future guideline refinement. The 2010 American Psychiatric Association’s (APA) guidelines on treatment for depression briefly discuss ECT as a treatment option, but it has not published comprehensive ECT guidelines since 2001. The Clinical TMS Society in 2016 and McClintock et al. in 2018 published consensus statements on rTMS. These ECT and rTMS guidelines summarize data on safety and efficacy, including indications for their use, risks and treatment procedures, discusses the importance of standardized training and professional certification for their safe use, and cite areas where data are not sufficient for recommendations to be made. , , There also is an effort to establish research guidelines regarding the ethical and safe research use of interventions that are not FDA-approved or that are used for novel indications. A 2014 paper presented international consensus on guidelines for stereotactic neurosurgery for psychiatric disorders which discussed the need for establishing evidence for neurosurgical interventions including DBS and recommended that independent institutional review boards should be established and work alongside regulatory agencies to review and oversee research protocols.
PSYCHIATRISTS’ MAIN CONSIDERATIONS FOR GUIDELINES Seeking to identify critical considerations for developing new and updated guidelines, we administered a standardized online survey to a large national sample of board-certified psychiatrists ( N = 495) using a full factorial design. Near the end of the survey, we gave psychiatrists several options and asked them to select the one that should be the main consideration when developing practical guidelines for the PEI modality (ECT, rTMS, DBS or ABIs) they were assigned to. Overall, most psychiatrists selected either providing evidence of the safety and efficacy of these interventions (46.8%) or selecting patients who would be good candidates for these interventions (19.9%) as their main considerations when developing guidelines for these interventions ( ). Less than five percent of psychiatrists selected consultation with a second psychiatrist before initiation of treatment (4.4%) and establishing treatment approaches (i.e., electrode placement, administration and safeguards) for subpopulations (2.0%). Differentiating responses by intervention reveals several interesting variations ( ). Most notably, greater percentages of psychiatrists assigned to either non-FDA-approved modality (DBS, ABIs) than of those assigned to an FDA-approved modality (ECT, rTMS) reported providing evidence of the safety and efficacy of PEIs as their main consideration when developing practical guidelines for PEIs, while greater percentages of the latter reported improving systems of care delivery as their main consideration. We also examined the extent to which familiarity with PEIs influenced psychiatrists’ main consideration for practical guidelines. We operationalized familiarity as the number of four different PEIs—ECT, rTMS, DBS and vagus nerve stimulation—that psychiatrists have referred patients for or have administered to patients directly for either treatment or clinical trials in the last 5 years. Approximately 19.4% of our sampled psychiatrists reported no familiarity with any of these four PEIs, 27.7% reported familiarity with one, 38.8% with two, 10.5% with three and 3.6% reported familiarity with all four PEIs. Psychiatrists’ familiarity with PEIs does shape their selection of main consideration when developing practical guidelines for PEIs. Greater experience shifts psychiatrists’ main consideration from intervention safety and efficacy to improved patient selection and treatment optimization. Specifically, for each additional PEI referred for or administered in the prior 5 years, our sampled psychiatrists were 33% more likely to choose selecting good patient candidates than providing evidence of safety and effectiveness as their main consideration ( ). Also, for each additional PEI referred for or administered in the prior 5 years, our sampled psychiatrists are 253% (2.53 times) more likely to choose establishing treatment approaches for subgroups than providing evidence of safety and effectiveness as their main consideration. These results represent the relative effect of PEI familiarity while accounting for all other experimental, covariate and control variables.
GAPS IN CURRENT GUIDELINES We are not surprised that nearly half of all psychiatrists overall (46.8%)—and even higher percentages of those assigned to non-FDA-approved PEIs (61.6% and 72.4% for DBS and ABI, respectively)—prioritize the collection of evidence on the efficacy and safety of these interventions. These attitudes reflect the primary goals of existing clinical and research guidelines that provide evidence-based guidance. – , Yet, with rapid advancement in PEI technology, such as different coil designs and treatment protocols in rTMS, it is important to acknowledge that current guidelines do not incorporate the most current evidence-based information on these topics. Selecting patients who would be good candidates for these interventions was the second most common selection for psychiatrists overall and was selected more often by those assigned to ECT compared to those assigned to DBS and ABIs. While current PEI guidelines list specific indications for use – they do not reflect recent findings that might have changed considerations for selecting patients. Notably, while psychiatrists assigned to FDA-approved PEIs were more likely to consider improving systems of care delivery than were those assigned to experimental interventions, existing clinical guidelines do not explicitly address this. – , One reason for this difference is that proof of safety and efficacy is needed before a treatment can be approved by the FDA. Once approved, securing insurance coverage to allow sufficient access to the intervention predominates concerns and becomes the key consideration (e.g., improving systems of care delivery). Furthermore, only the APA’s ECT guidelines mention consultation, stating that consultation with another physician should be considered when patients have particularly high-risk conditions and when the physician is unsure whether the patient has the capacity to give informed consent. They do not explicitly mention consultation between psychiatrists . Regarding treatment approaches for subpopulations , some guidelines focus on considerations of procedure modification for specific at-risk populations, , while others identify subpopulations in which evidence of treatment is insufficient for recommendations. For example, the APA’s ECT guidelines note that procedure should be modified for populations such as the elderly, pregnant and those with specific medical conditions. McClintock et al. listed specific populations (such as adolescents with mood disorders and women with perinatal depression, among others) in which there is insufficient evidence supporting rTMS use. Nuttin et al. discussed the importance of increased care for research involving vulnerable populations. Finally, existing guidelines’ discussion of establishing the timing of these interventions within the overall course of treatment is sparse and limited to indicating that approved PEIs (ECT and rTMS) are typically used following the failure of first-line treatments. , Guidance regarding PEIs may be unique in psychiatry because these interventions may be less well-known and understood by psychiatrists.
LOOKING TO THE FUTURE Insights from research on neuromodulation and the development of new treatment modalities present psychiatrists with a particular challenge: keeping up to date on the most recent science to select the most effective and safest therapeutic approaches. A better understanding of psychiatrists’ main considerations for PEI guidelines is important for highlighting areas where current guidelines have not provided needed insight for clinicians, indicating gaps in evidence or signalling that updates to current guidelines are needed. This data also can highlight important areas to cover when establishing future PEI guidelines. Furthermore, approaches for expanding psychiatrists’ awareness of existing guidelines should be considered. Guidelines, however, are inherently imperfect. The dynamic nature of science may render them rapidly out-of-date. Additionally, they only provide a ‘jumping off point’ for clinical decision-making, and must be critically interpreted adaptable to specific patients’ needs and preferences.
|
Adipose-Tissue-Derived Mesenchymal Stem Cells Mediate PD-L1 Overexpression in the White Adipose Tissue of Obese Individuals, Resulting in T Cell Dysfunction | f8729bc6-6895-45b3-82d2-1852d64f30e2 | 8534339 | Anatomy[mh] | Effective responses to intracellular pathogens or tumors are usually dependent on Natural killer (NK) cells and/or T cells, especially CD8+ cytotoxic T cells. NK cell killing can be regulated by killer cell inhibitory receptors (KIR) after being bound to HLA class I molecules . Exhaustion is another way of inhibiting cytotoxic functions in either NK or CD8+ cytolytic T cells (CTL), due to increased protein expression of inhibitory receptors such as PD-1, CTLA-4, TIM-3, LAG-3, or TIGIT on T cells, or NKG2A on NK cells, and their respective ligands PD-L1 or PD-L2, CD 155, CD112, CD113, and HLAG/E on antigen-presenting cells . These molecules are called immune checkpoints and the PD-1/PD-L1 axis is known as the strongest exhaustion inducer . Moreover, tumors have been shown to use this pathway in order to escape antitumor immunity with the help of pro-inflammatory cytokines, particularly IFNγ . Τhe expression of PD-L1 in tumors is now considered a good prognostic factor, due to the great improvement in patient outcome with immune checkpoint inhibitors . A Nobel Price of Medicine was awarded in 2018 to James Allison and Tasuku Honjo for discovering immune checkpoints, which have been shown to have strong beneficial effects in cancer therapy after PD-L1 or PD-1 blockade. Several studies have demonstrated a negative correlation between CD4+ or CD8+ T cell frequencies and the severity of SARS-CoV-2 infection . Indeed, total CD3+, CD8+, or CD4+ T cell frequencies below 800, 300, or 400/μL, respectively, were negatively associated with survival of COVID-19 . In addition to reduced NK or T cell frequencies, functional exhaustion of these cells was demonstrated to contribute to SARS-CoV-2-mediated defective immune cell responses . Exhaustion was first described upon long-term infection with viruses , but the hallmark of SARS-CoV-2 is its ability to use this inhibitory pathway in very early stages of infection. Moreover, NKG2A and PD-1 have been shown to be upregulated by SARS-CoV-2 in NK and T cells, respectively, and to be downregulated upon recovery from COVID-19, demonstrating that SARS-CoV-2 is an inducer of exhaustion . Recent reports have demonstrated the resistance of mesenchymal stem cell (MSC) or ASC to SARS-CoV-2 infection and the beneficial effects of healthy MSC transplantation in severe COVID-19 patients . However, as opposed to ASC, adipocytes express the ACE2 receptor at high density, and even more when they belong to obese individuals , suggesting a mechanism that could account for the higher susceptibility of obese individuals to SARS-CoV-2 infection. Obesity with a BMI of >30 Kg/m 2 is an independent risk factor for severe forms of COVID-19, as demonstrated in several clinical departments including ours, where 25% of hospitalized patients and 35% of ICU patients below the age of 70 were found to be obese . Indeed, obesity is associated with comorbidities such as hypertension, hypercholesterolemia, cardiovascular diseases, atherosclerosis, type 2 diabetes, chronic inflammatory diseases, and various cancers . Besides the elevation of chronic inflammation markers inside adipose tissues, plasma, and/or metabolically altered organs, immune dysfunction has also been described during obesity . Indeed, higher rates of infection, delayed wound healing, and vaccination failure are often described in obese individuals. The impact of obesity on immunity has been related to various mechanisms, such as metabolic alterations in T cells, acceleration of thymic aging, alteration of lymphoid organ architecture, reduction in the maintenance of memory T cells, impairment of cell-mediated immune responses, abnormal lymphoproliferative responses, and decreased cytotoxic activities related to reduced NK and CD8+ T cell frequencies, especially among elderly individuals . Moreover, increased T cell exhaustion due to enhanced PD-1 expression has recently been reported to occur in blood T cells among obese individuals suffering from cancer . As white adipose tissue (WAT) is known to initiate inflammation in obese individuals, we investigated whether increased adiposity and inflammation in obesity could result in exhaustion of WAT-infiltrating T cells. Here, we showed that obesity is associated with PD-L1 overexpression inside WAT and implicated adipose-tissue-derived mesenchymal stem cells (ASC) and IFNγ secretion in this upregulation. We also demonstrated that overexpression of PD-L1 has spread to cells present in the ASC environment, particularly to adipocytes.
2.1. Animals, Diet, and Experimental Design All procedures were performed with the approval of the Regional Committee of Ethics for Animal Experiments (CECCAP), registered under number C2EA15 by the French Ministry of Higher Education and Research, in accordance with European directive 2010/63. The experimental protocol received the number 2018061411298503. Four-week-old C57Bl/6J male mice (Envigo; Gannat, France) were randomly housed, two per cage, at 21 °C with a normal light/dark cycle, free access to water, and a standard chow diet (Genestil, Royaucourt, France) for a one-week acclimatization. Next, mice were randomly assigned to four different groups ( n = 5–6/group). One group was fed a standard diet (ST0;10% fat and 35.6% maltodextrin) and another was fed a high-fat/high-sucrose diet (HF0; 39.4% fat and 16.6% maltodextrin +16.6% sucrose, Envigo). Food was given three times a week. Body weight was recorded weekly. 2.2. Tissue Collection and Blood and Plasma Analyses Before sampling, the mice were fasted for 6 h and then weighed. Blood was collected by retro-orbital sampling and mice were euthanized by cervical dislocation. Adipose tissues were quickly removed, weighed, frozen in liquid nitrogen, and stored at −80 °C. Blood glucose concentrations (Accucheck Performa glucometer; Roche Diabetes Care, Meylan, France,), plasma levels of insulin (Mouse Ultrasensitive ELISA, Eurobio, Courta-boeuf, 92973, France), leptin, and adiponectin (ELISA kits, CrystalChem Europe, Zaandam, Netherlands) were measured. 2.3. Isolation and Expansion of Adipose-Tissue-Derived Mesenchymal Stem Cells Subcutaneous or visceral AT samples were isolated from residues of bariatric surgery from obese subjects (BMI > 30 kg/m 2 ), or visceral surgery from lean controls with the approval of the Committee for the Protection of Human Subjects of the “Hospices Civils de Lyon” number 12/111 and with the patient’s consent, according to ethical principles for medical research involving human subjects, as described in the World Medical Association’s Helsinki Declaration. AT samples (50–100 mg) were mechanically dissected and incubated in 2 g/L of collagenase type Ia solution (Sigma Aldrich, Saint-Quentin Fallavier, France) in DMEM:F12 medium (50/50 vol/vol) (Invitrogen, Thermofisher Scientific, Waltham, MA, USA) for 40 min at 37 °C by mixing. Isolation of ASC was performed as previously reported . Adipose-tissue-derived mesenchymal stem cells (ASC) were expanded in a culture medium composed of DMEM:F-12 supplemented with 10% FCS, 2 mmol/L L-glutamine, and 100 U/mL penicillin–streptomycin. Culture medium was changed by half, three times a week, during 3–4 passages before being used or stored in liquid nitrogen. The function of ASC was validated as recommended by the French Cell Therapy Society and checked for expression of CD90, CD105, and CD73 molecules. 2.4. Isolation of Blood Mononuclear Cells (MNC) Blood samples from the Blood Bank Center of Lyon, Lyon, 69007, France) were obtained from healthy human blood donors. MNC were separated from blood through Ficoll-Histopaque density gradient centrifugation (Sigma-Aldrich) and stored in liquid nitrogen. 2.5. Coculture Assays ASC were harvested and seeded in 24-well plates (100,000 cells/well) for 24 h in RPMI supplemented with 10% FCS. Then, 24 h later, MNC were added in the presence or absence of phytohemagglutinin (PHA, 5 µg/mL, Sigma-Aldrich). ASC/MNC ratios of 1:5 were used. For mRNA analyses, the culture period was 24 h, whereas it was 48 h when conditioning media were collected. In that latter case, IL-17A neutralizing antibodies (Secukinumab, Novartis Pharma S.A.S., 92506, Rueil-Malmaison, France) at 50 μg/mL or IFNγ-neutralizing antibodies (Invitrogen, 16-7318-81) at 50 μg/mL were eventually added from the beginning of cultures. For blocking experiments, ASC/MNC cocultures were treated with PD-L1 neutralizing antibodies (R&D Systems, Minneapolis, MN, USA; AF156) at 1 or 10 μg/mL, or with irrelevant polyclonal goat IgG (sc-8828; Santa-Cruz Biotechnology/INC, Heidelberg, Germany) at 10 μg/mL, or Secukinumab at 50 μg/mL. 2.6. mRNA Measurements Total RNA was extracted from cocultures using the Tri Isolation ReagentTM (Roche Diagnostics, Meylan, France) and frozen at –80 °C, and RT-qPCR was performed from reversely transcribed RNA into cDNA, as previously described . mRNA was quantified relative to the housekeeping gene TBP (TATA BOX binding protein), using the mathematical method depending on ΔCT and the amplification efficiency of the transcripts, as described by Plaffl et al. . The individual primer sequences used for RT-qPCR are provided in . 2.7. Flow Cytometry Fluorescein isothiocyanate (FITC)–, phycoerythrin (PE)- Allophycocyanin (APC)-, and Alexa-fluor- directly conjugated anti-human antibodies were used to stain the various cells tested: CD14 (21270146; ImmunoTools, Friesoythe, Germany), CD73 (21270734; ImmunoTools), CD8 (21270083; ImmunoTools), CD274 (12-5982-82; Invitrogen, Illkirch, France), CD279 (FAB7115G; R&D Systems, Biotechne, Abingdon, UK), and Granzyme B (IC2906G; R&D Systems). Analyses were performed using a LSR II 3 lasers cytofluorometer and Diva software (BD Biosciences, Le-Pont-de-Claix, France). For Granyme B intracellular staining, cells were permeabilized using the BD Cytofix/Cytoperm permeabilizing kit (AB_2869008; Beckton Dickinson, Le-Pont-de-Claix, France), and cells were analyzed 6 h after interaction with adipocytes. 2.8. Differentiation of Obese Subcutaneous ASC into AD Ob-ASC were differentiated into AD as previously described . 2.9. Immunohistochemistry ATs were fixed in formaldehyde 3.7% after isolation from bariatric surgery. They were then embedded in paraffin before being cut with the microtome MICROT002 and mounted onto microscope slides for analysis, after staining with anti-human PD-L1 (R&D Systems; MAB1561) at 1:50 dilution. 2.10. Statistical Analysis Statistical analyses were performed using one-way ANOVA with or without repeated measures, or mixed-effects analyses (if some values were missing), followed by post hoc Fisher’s LSD tests. In some experiments, two-tailed unpaired t-tests were used. Prism software was employed for analysis (Prism 8, GraphPad software, San Diego, CA, USA). Differences were considered as statistically significant when p was <0.05.
All procedures were performed with the approval of the Regional Committee of Ethics for Animal Experiments (CECCAP), registered under number C2EA15 by the French Ministry of Higher Education and Research, in accordance with European directive 2010/63. The experimental protocol received the number 2018061411298503. Four-week-old C57Bl/6J male mice (Envigo; Gannat, France) were randomly housed, two per cage, at 21 °C with a normal light/dark cycle, free access to water, and a standard chow diet (Genestil, Royaucourt, France) for a one-week acclimatization. Next, mice were randomly assigned to four different groups ( n = 5–6/group). One group was fed a standard diet (ST0;10% fat and 35.6% maltodextrin) and another was fed a high-fat/high-sucrose diet (HF0; 39.4% fat and 16.6% maltodextrin +16.6% sucrose, Envigo). Food was given three times a week. Body weight was recorded weekly.
Before sampling, the mice were fasted for 6 h and then weighed. Blood was collected by retro-orbital sampling and mice were euthanized by cervical dislocation. Adipose tissues were quickly removed, weighed, frozen in liquid nitrogen, and stored at −80 °C. Blood glucose concentrations (Accucheck Performa glucometer; Roche Diabetes Care, Meylan, France,), plasma levels of insulin (Mouse Ultrasensitive ELISA, Eurobio, Courta-boeuf, 92973, France), leptin, and adiponectin (ELISA kits, CrystalChem Europe, Zaandam, Netherlands) were measured.
Subcutaneous or visceral AT samples were isolated from residues of bariatric surgery from obese subjects (BMI > 30 kg/m 2 ), or visceral surgery from lean controls with the approval of the Committee for the Protection of Human Subjects of the “Hospices Civils de Lyon” number 12/111 and with the patient’s consent, according to ethical principles for medical research involving human subjects, as described in the World Medical Association’s Helsinki Declaration. AT samples (50–100 mg) were mechanically dissected and incubated in 2 g/L of collagenase type Ia solution (Sigma Aldrich, Saint-Quentin Fallavier, France) in DMEM:F12 medium (50/50 vol/vol) (Invitrogen, Thermofisher Scientific, Waltham, MA, USA) for 40 min at 37 °C by mixing. Isolation of ASC was performed as previously reported . Adipose-tissue-derived mesenchymal stem cells (ASC) were expanded in a culture medium composed of DMEM:F-12 supplemented with 10% FCS, 2 mmol/L L-glutamine, and 100 U/mL penicillin–streptomycin. Culture medium was changed by half, three times a week, during 3–4 passages before being used or stored in liquid nitrogen. The function of ASC was validated as recommended by the French Cell Therapy Society and checked for expression of CD90, CD105, and CD73 molecules.
Blood samples from the Blood Bank Center of Lyon, Lyon, 69007, France) were obtained from healthy human blood donors. MNC were separated from blood through Ficoll-Histopaque density gradient centrifugation (Sigma-Aldrich) and stored in liquid nitrogen.
ASC were harvested and seeded in 24-well plates (100,000 cells/well) for 24 h in RPMI supplemented with 10% FCS. Then, 24 h later, MNC were added in the presence or absence of phytohemagglutinin (PHA, 5 µg/mL, Sigma-Aldrich). ASC/MNC ratios of 1:5 were used. For mRNA analyses, the culture period was 24 h, whereas it was 48 h when conditioning media were collected. In that latter case, IL-17A neutralizing antibodies (Secukinumab, Novartis Pharma S.A.S., 92506, Rueil-Malmaison, France) at 50 μg/mL or IFNγ-neutralizing antibodies (Invitrogen, 16-7318-81) at 50 μg/mL were eventually added from the beginning of cultures. For blocking experiments, ASC/MNC cocultures were treated with PD-L1 neutralizing antibodies (R&D Systems, Minneapolis, MN, USA; AF156) at 1 or 10 μg/mL, or with irrelevant polyclonal goat IgG (sc-8828; Santa-Cruz Biotechnology/INC, Heidelberg, Germany) at 10 μg/mL, or Secukinumab at 50 μg/mL.
Total RNA was extracted from cocultures using the Tri Isolation ReagentTM (Roche Diagnostics, Meylan, France) and frozen at –80 °C, and RT-qPCR was performed from reversely transcribed RNA into cDNA, as previously described . mRNA was quantified relative to the housekeeping gene TBP (TATA BOX binding protein), using the mathematical method depending on ΔCT and the amplification efficiency of the transcripts, as described by Plaffl et al. . The individual primer sequences used for RT-qPCR are provided in .
Fluorescein isothiocyanate (FITC)–, phycoerythrin (PE)- Allophycocyanin (APC)-, and Alexa-fluor- directly conjugated anti-human antibodies were used to stain the various cells tested: CD14 (21270146; ImmunoTools, Friesoythe, Germany), CD73 (21270734; ImmunoTools), CD8 (21270083; ImmunoTools), CD274 (12-5982-82; Invitrogen, Illkirch, France), CD279 (FAB7115G; R&D Systems, Biotechne, Abingdon, UK), and Granzyme B (IC2906G; R&D Systems). Analyses were performed using a LSR II 3 lasers cytofluorometer and Diva software (BD Biosciences, Le-Pont-de-Claix, France). For Granyme B intracellular staining, cells were permeabilized using the BD Cytofix/Cytoperm permeabilizing kit (AB_2869008; Beckton Dickinson, Le-Pont-de-Claix, France), and cells were analyzed 6 h after interaction with adipocytes.
Ob-ASC were differentiated into AD as previously described .
ATs were fixed in formaldehyde 3.7% after isolation from bariatric surgery. They were then embedded in paraffin before being cut with the microtome MICROT002 and mounted onto microscope slides for analysis, after staining with anti-human PD-L1 (R&D Systems; MAB1561) at 1:50 dilution.
Statistical analyses were performed using one-way ANOVA with or without repeated measures, or mixed-effects analyses (if some values were missing), followed by post hoc Fisher’s LSD tests. In some experiments, two-tailed unpaired t-tests were used. Prism software was employed for analysis (Prism 8, GraphPad software, San Diego, CA, USA). Differences were considered as statistically significant when p was <0.05.
3.1. PD-L1 and/or PD-1 Are Overexpressed in the WAT of Obese Mice Whereas blood T cells from obese individuals or mice have been shown to overexpress the PD-1 checkpoint , here we investigated the possibility that WAT-infiltrating T cells could also express markers of exhaustion among the obese. With this goal in mind, we used a high-fat-diet-induced obesity (DIO) mouse model, wherein postweaning C57BL/6 male mice were fed either a control (chow) or hypercaloric diet for 16 weeks, and we measured the mRNA expression levels of PD-L1/PD-1exhaustion markers. As shown in A, the efficiency of the hypercaloric diet was assessed through (i) increased body weight and white adipose tissue mass; (ii) metabolic consequences, such as increased fasting plasma insulin and blood glucose; and (iii) adipokine disturbances, such as decreased adiponectin over leptin protein levels ( A). In addition, PD-L1 was overexpressed at the mRNA level in both the visceral (vis) and subcutaneous (sc) WAT of obese, in comparison with lean mice ( B). However, increased PD-1 expression was observed in the vis but not sc WAT of mice fed with the hypercaloric diet ( B), which suggested more pronounced inhibition of T cell functions in vis. To support these results, immunochemistry was used to stain obese or lean visceral WAT for PD-L1. As shown in C, a significant increase in PD-L1 expression was observed in the vis WAT of obese mice. 3.2. Inflammation-Mediated by Adipose-Tissue-Derived Mesenchymal Stem Cells from Obese Individuals Contributes to PD-L1 Upregulation in Cocultures with Mononuclear Cells Having previously demonstrated that adipose-tissue-derived mesenchymal stem cells (ASC) from obese individuals (ob-ASC), rather than from lean individuals, can promote inflammation through Th17 cell and monocyte cell activation , we then investigated whether inflammation-mediated by ob-ASC could play a role in upregulating PD-L1 and/or PD-1 expression using a human in vitro coculture model. As shown in A, phytohemagglutinin A (PHA) activation of mononuclear cells (MNC) enhanced PDL1 mRNA expression levels, but the co-presence of ASC resulted in potentiating the expression of PDL1 mRNA, whether ASC were issued from lean or obese donors. However, ob-ASC upregulated PDL1 mRNA expression at a statistically significant higher level than lean ASC. In addition, a two-way ANOVA test demonstrated (i) that both the body mass index (BMI) and the activation status were statistically significant independent factors, (ii) as well as the interaction between these two factors ( A). Then, to validate those results at the protein level and investigate whether increased PD-L1 levels occurred in MNC or ASC themselves, we co-stained monocytes or ASC with the CD14 or CD73 marker, respectively, and the PD-L1 molecules. As shown in B, flow cytofluorometry supported the mRNA data. Indeed, an increased surface expression of PD-L1 occurred, following PHA activation of MNC (i.e., 29.4–35.1% of positive PD-L1 cells), which was potentiated by the presence of ob-ASC (29.4–59.4%), but with a weaker potentiating effect in the presence of lean ASC (29.4–43.2%). CD14 labelling showed that PD-L1 was expressed at the basal level in monocytes with approximately 2/3 of the CD14 positive cells expressing both CD14 and PD-L1 (i.e., 7.7% of cells were doubly positive for CD14 and PD-L1, versus 3.6% singly positive for CD14). However, following 48 h of activation with PHA, almost the whole population of CD14-positive cells became doubly positive for PD-L1, even though their ratio diminished from 7.7% to 4.05% of total cells, probably due to the proliferation of PHA-activated T cells. The presence of obese or lean ASC did not induce any upregulation of the surface expression level of PD-L1 in monocytes. However, it resulted in the upregulation of the intensity of CD14 fluorescence, suggesting the transition of monocytes towards a population of “classical monocytes”, as previously described . Moreover, PD-L1 expression in CD73+ ob-ASC was higher than in lean ASC (i.e., from 3.0% to 6.4%). However, the prominent population of cells in which PD-L1 expression was upregulated was CD73-negative (i.e., from 33.8% to 48.7%). This latter population was likely to correspond to activated T lymphocytes, as T cells were the most preponderant cells in those cocultures, on the one hand, and are known to express PD-L1 expression following antigenic stimulation, on the other hand . Altogether, those results demonstrated that PD-L1 expression is upregulated following activation with PHA at both the transcriptional and protein levels, and was considerably amplified by the presence of ob-ASC. 3.3. Activation of T Cells Upregulates PD-1 Expression but Is Not Influenced by the Presence of Adipose-Derived Stem Cells (ASC) Because PD-1 needs to be linked by PD-L1 to exert its inhibitory effect on T cells, we then investigated whether the presence of obese ASC could also upregulate its expression. As shown in A, PHA activation of MNC increased the levels of PD1 mRNA expression, but the presence of obese or lean ASC did not further influence its expression. Supporting these results, and as shown in B, PD-1 surface expression was enhanced in PHA-activated MNC (i.e., from 27% to 33.79%). Those levels barely increased from 33.79% to 35.29% or 37.56% in the presence of lean or ob-ASC, respectively. The levels of CD8+ doubly positive cells for PD-1 increased, i.e., from 3.5% in the resting state to 4.7% with PHA, and were barely increased in the presence of ob-ASC from 4.7% to 5.45%. Therefore, these results suggested that PD-1 is likely to be a marker of activation, with obese ASC influencing PD-L1, but not—or only slightly—PD-1 expression. 3.4. PD-L1 Blockade Partially Restores Th1 Cell Secretion and Improves Ob-ASC Mediated Th17 Cell Activation, without Affecting Pro-Inflammatory Cytokine Secretion by Accessory Cells To then analyze the consequences of PD-L1 overexpression, we measured the effects of PD-L1 blockade on the pro-inflammatory cytokine expression. As previously reported and as shown in , the presence of ob-ASC resulted in increasing IL-17A, IL-1Β, and IL-6, together with decreasing TNFα and IL-2 mRNA expression levels, as compared with PHA-activated MNC. The modification of the cytokine pattern has been previously reported by us to be related to (i) the immunomodulatory effect of ASC on Th1 responses, (ii) the promotion by ob-ASC of Th17 double positive cells, secreting both IL-17 and IFNγ, and (iii) the activation of pro-inflammatory secretion by accessory cells such as monocytes and ob-ASC . The addition of neutralizing anti-PD-L1 mAbs in those cocultures resulted in the partial restoration of Th1 cell responses, as assessed by increased TNFα and IL-2 mRNA expression levels. An increase in IL-17A and IFNγ mRNA expression by Th17 cells was also observed. However, IL1 Β and IL6 mRNA expression levels were not modified, thus demonstrating a specific PD-L1 inhibitory effect on T cell functions. The specificity of the anti-PD-L1 effects was supported by further experiments comparing IL-17 and PD-L1 blockade ( ). 3.5. Ob-ASC-Mediated Inflammation Results in Increasing PD-L1 Expression in Freshly Differentiated Adipocytes Because the micro-environment plays an important role in the propagation of inflammation in WAT, we then asked whether conditioned media (CM) collected from ob-ASC and MNC cocultures could propagate exhaustion. Therefore, we cultured freshly differentiated adipocytes from ob-ASC with CM harvested from PHA-activated cocultures. We observed a strong upregulation of the expression of PDL1 mRNA, up to almost 4-fold. Having previously demonstrated that Th17 cells play an important role in the propagation of inflammation in this coculture model , we then asked whether cytokines secreted by those T cells, i.e., IL-17A and/or IFNγ, could play a role in PDL1 mRNA overexpression in adipocytes. Therefore, neutralizing antibodies directed against IFNγ or IL-17A were added or not to these cocultures. The results clearly showed that CM-mediated PDL1 mRNA overexpression was dependent on IFNγ secretion only, since blockade of IFNγ but not IL-17A prevented CM effects on PDL1 mRNA overexpression ( A). To validate these results, we reciprocally added IFNγ, or IL-17A recombinant protein to adipocytes and observed that IFNγ, but not IL-17A, significantly upregulated PDL1 mRNA expression ( B).
Whereas blood T cells from obese individuals or mice have been shown to overexpress the PD-1 checkpoint , here we investigated the possibility that WAT-infiltrating T cells could also express markers of exhaustion among the obese. With this goal in mind, we used a high-fat-diet-induced obesity (DIO) mouse model, wherein postweaning C57BL/6 male mice were fed either a control (chow) or hypercaloric diet for 16 weeks, and we measured the mRNA expression levels of PD-L1/PD-1exhaustion markers. As shown in A, the efficiency of the hypercaloric diet was assessed through (i) increased body weight and white adipose tissue mass; (ii) metabolic consequences, such as increased fasting plasma insulin and blood glucose; and (iii) adipokine disturbances, such as decreased adiponectin over leptin protein levels ( A). In addition, PD-L1 was overexpressed at the mRNA level in both the visceral (vis) and subcutaneous (sc) WAT of obese, in comparison with lean mice ( B). However, increased PD-1 expression was observed in the vis but not sc WAT of mice fed with the hypercaloric diet ( B), which suggested more pronounced inhibition of T cell functions in vis. To support these results, immunochemistry was used to stain obese or lean visceral WAT for PD-L1. As shown in C, a significant increase in PD-L1 expression was observed in the vis WAT of obese mice.
Having previously demonstrated that adipose-tissue-derived mesenchymal stem cells (ASC) from obese individuals (ob-ASC), rather than from lean individuals, can promote inflammation through Th17 cell and monocyte cell activation , we then investigated whether inflammation-mediated by ob-ASC could play a role in upregulating PD-L1 and/or PD-1 expression using a human in vitro coculture model. As shown in A, phytohemagglutinin A (PHA) activation of mononuclear cells (MNC) enhanced PDL1 mRNA expression levels, but the co-presence of ASC resulted in potentiating the expression of PDL1 mRNA, whether ASC were issued from lean or obese donors. However, ob-ASC upregulated PDL1 mRNA expression at a statistically significant higher level than lean ASC. In addition, a two-way ANOVA test demonstrated (i) that both the body mass index (BMI) and the activation status were statistically significant independent factors, (ii) as well as the interaction between these two factors ( A). Then, to validate those results at the protein level and investigate whether increased PD-L1 levels occurred in MNC or ASC themselves, we co-stained monocytes or ASC with the CD14 or CD73 marker, respectively, and the PD-L1 molecules. As shown in B, flow cytofluorometry supported the mRNA data. Indeed, an increased surface expression of PD-L1 occurred, following PHA activation of MNC (i.e., 29.4–35.1% of positive PD-L1 cells), which was potentiated by the presence of ob-ASC (29.4–59.4%), but with a weaker potentiating effect in the presence of lean ASC (29.4–43.2%). CD14 labelling showed that PD-L1 was expressed at the basal level in monocytes with approximately 2/3 of the CD14 positive cells expressing both CD14 and PD-L1 (i.e., 7.7% of cells were doubly positive for CD14 and PD-L1, versus 3.6% singly positive for CD14). However, following 48 h of activation with PHA, almost the whole population of CD14-positive cells became doubly positive for PD-L1, even though their ratio diminished from 7.7% to 4.05% of total cells, probably due to the proliferation of PHA-activated T cells. The presence of obese or lean ASC did not induce any upregulation of the surface expression level of PD-L1 in monocytes. However, it resulted in the upregulation of the intensity of CD14 fluorescence, suggesting the transition of monocytes towards a population of “classical monocytes”, as previously described . Moreover, PD-L1 expression in CD73+ ob-ASC was higher than in lean ASC (i.e., from 3.0% to 6.4%). However, the prominent population of cells in which PD-L1 expression was upregulated was CD73-negative (i.e., from 33.8% to 48.7%). This latter population was likely to correspond to activated T lymphocytes, as T cells were the most preponderant cells in those cocultures, on the one hand, and are known to express PD-L1 expression following antigenic stimulation, on the other hand . Altogether, those results demonstrated that PD-L1 expression is upregulated following activation with PHA at both the transcriptional and protein levels, and was considerably amplified by the presence of ob-ASC.
Because PD-1 needs to be linked by PD-L1 to exert its inhibitory effect on T cells, we then investigated whether the presence of obese ASC could also upregulate its expression. As shown in A, PHA activation of MNC increased the levels of PD1 mRNA expression, but the presence of obese or lean ASC did not further influence its expression. Supporting these results, and as shown in B, PD-1 surface expression was enhanced in PHA-activated MNC (i.e., from 27% to 33.79%). Those levels barely increased from 33.79% to 35.29% or 37.56% in the presence of lean or ob-ASC, respectively. The levels of CD8+ doubly positive cells for PD-1 increased, i.e., from 3.5% in the resting state to 4.7% with PHA, and were barely increased in the presence of ob-ASC from 4.7% to 5.45%. Therefore, these results suggested that PD-1 is likely to be a marker of activation, with obese ASC influencing PD-L1, but not—or only slightly—PD-1 expression.
To then analyze the consequences of PD-L1 overexpression, we measured the effects of PD-L1 blockade on the pro-inflammatory cytokine expression. As previously reported and as shown in , the presence of ob-ASC resulted in increasing IL-17A, IL-1Β, and IL-6, together with decreasing TNFα and IL-2 mRNA expression levels, as compared with PHA-activated MNC. The modification of the cytokine pattern has been previously reported by us to be related to (i) the immunomodulatory effect of ASC on Th1 responses, (ii) the promotion by ob-ASC of Th17 double positive cells, secreting both IL-17 and IFNγ, and (iii) the activation of pro-inflammatory secretion by accessory cells such as monocytes and ob-ASC . The addition of neutralizing anti-PD-L1 mAbs in those cocultures resulted in the partial restoration of Th1 cell responses, as assessed by increased TNFα and IL-2 mRNA expression levels. An increase in IL-17A and IFNγ mRNA expression by Th17 cells was also observed. However, IL1 Β and IL6 mRNA expression levels were not modified, thus demonstrating a specific PD-L1 inhibitory effect on T cell functions. The specificity of the anti-PD-L1 effects was supported by further experiments comparing IL-17 and PD-L1 blockade ( ).
Because the micro-environment plays an important role in the propagation of inflammation in WAT, we then asked whether conditioned media (CM) collected from ob-ASC and MNC cocultures could propagate exhaustion. Therefore, we cultured freshly differentiated adipocytes from ob-ASC with CM harvested from PHA-activated cocultures. We observed a strong upregulation of the expression of PDL1 mRNA, up to almost 4-fold. Having previously demonstrated that Th17 cells play an important role in the propagation of inflammation in this coculture model , we then asked whether cytokines secreted by those T cells, i.e., IL-17A and/or IFNγ, could play a role in PDL1 mRNA overexpression in adipocytes. Therefore, neutralizing antibodies directed against IFNγ or IL-17A were added or not to these cocultures. The results clearly showed that CM-mediated PDL1 mRNA overexpression was dependent on IFNγ secretion only, since blockade of IFNγ but not IL-17A prevented CM effects on PDL1 mRNA overexpression ( A). To validate these results, we reciprocally added IFNγ, or IL-17A recombinant protein to adipocytes and observed that IFNγ, but not IL-17A, significantly upregulated PDL1 mRNA expression ( B).
Brown adipocytes and differentiated 3T3 cell lines have been shown to express PD-L1 . Additionally, we demonstrated that white adipocyte precursors can also overexpress PD-L1 following interaction with MNC, especially when they were collected from obese individuals ( ). Interestingly, we demonstrated that inflammation mediated by ob-ASC and MNC cocultures spread PD-L1 overexpression to cells present in their environment, especially adipocytes, under the influence of IFNγ secretion ( ). A role for IFNγ in the induction of exhaustion has already been reported by others, particularly in a tumoral model , thus demonstrating a regulatory role for IFNγ, in addition to its role in activating effectors such as macrophages or other accessory cells. Moreover, our data suggested that the negative modulation of Th1 cells that occurred in the presence of ob-ASC was partly related to IFNγ-mediated PD-L1 overexpression, as the PD-L1 blockade partially restored Th1 cell cytokine secretion ( ). IL-2 is a cytokine that plays a major role in the maturation of precytotoxic T cells into cytotoxic T cells (CTL) in order to allow them to subsequently exert their cytotoxic function . We have previously reported that the presence of mesenchymal stem cells inhibit pre-CTL differentiation into CTL effectors, which is restored by the addition of IL-2 . Therefore, our results suggest that ob-ASC may inhibit the maturation of pre-CTL into CTL effectors and subsequent CTL function through IL-2 negative modulation. Moreover, PD-L1 overexpression may partly account for this inhibition, as demonstrated by the positive effects of the PD-L1 blockade on IL2 mRNA expression ( ). Of interest, CTL activity was reduced in the presence of obese adipocytes in comparison with lean adipocytes, as demonstrated by granzyme B intracellular staining ( ), thus suggesting that ASC-mediated spreading of PD-L1 overexpression towards obese adipocytes may inhibit obese adipocytes’ ability to activate CTL effectors. PD-1 was also found to be upregulated following activation of MNC with PHA ( ). However, this upregulation was not potentiated by the presence of ASC, suggesting that PD-1 is a marker of activation, in agreement with several studies reporting a regulatory function of PD-1 or other immune checkpoints, notably CTLA-4 [ , , , ]. Indeed, PD-1 was demonstrated to be transiently expressed during acute viral infections in order to regulate T cell functions. However, in chronic infection with LCMV or HIV, high levels of PD-1 were shown to be sustained over time and to contribute to T cell exhaustion . In our coculture model, we were not able to compare acute and chronic stimulation. However, when mice were fed a hypercaloric regimen for 16 weeks, which corresponded to a chronic situation, we observed an upregulation of PD-1 along with PD-L1 in the vis WAT of obese mice in comparison to the vis WAT of lean mice ( ). However, in the sc WAT of obese mice, PD-L1 but not PD-1 was upregulated, probably due to the higher levels of pro-inflammatory cytokine secretion known to be present in vis WAT . Altogether, these data lead us to suggest that increased expression of PD-L1 on adipocytes and of PD-1 on T cells may trigger exhaustion and inhibit T cell functions, in particular IL-2 secretion and subsequent CTL function. This could be of high importance in situations where viruses infiltrate WAT. Infection with SARS-CoV-2 could represent one such situation, as AT is known to express ACE2, a receptor for the spike protein of SARS-CoV-2 . Interestingly enough, the expression of ACE2 has been shown to be upregulated in the epicardial fat of obese versus lean mice . Therefore, increased exhaustion of T cells in the inflammatory environment present in obese visceral WAT could help SARS-CoV-2 to reside inside it, which may lead us to consider fat as a reservoir for this virus. It would then become easy for the virus to invade and spread in proximal tissues or organs, as suggested by Nishimura et al. for the H5N1 virus . In that case, anti-PD-1 or anti-PD-L1 biologics could become a promising alternative approach to help obese individuals achieve a better immune response against the severe form of COVID-19, as shown in the graphical abstract. Accordingly, when PD-L1 was neutralized with specific antibodies, we observed a partial restoration of cytokines expressed by Th1 cells, notably IL-2, which plays a critical role in precytolytic (CTL) differentiation into CTL effectors ( ). TNFα, IL-17A, and IFNγ also increased, demonstrating the inhibitory effect of PD-L1 on T cell function whatever the T cell subset. Moreover, IL-1β and IL-6 pro-inflammatory cytokines, which were shown to be secreted by monocytes and ob-ASC in our model , did not increase after PD-L1 blockade, thus demonstrating the specificity of the PD-L1 blockade.
This study was performed first in vivo in animals and then in in vitro coculture assays with human ASC and MNC. These assays have been used to investigate the mechanisms leading to increased expression of immune checkpoints, as observed in vivo. Although these coculture assays have been previously used to demonstrate the role of ASC, Th17 cells, and monocytes in initiating inflammation among AT from obese individuals , they have some limitations relative to the cell isolation and 2D culture processes and thus do not exactly correspond to the in vivo situation. Then, we demonstrated that inflammation may lead to exhaustion of T cells following interaction of MNC with ASC relative t increased expression of immune checkpoints, and suggested, in , that T cell exhaustion in WAT could contribute to the higher susceptibility of obese individuals to viruses infecting WAT, as seen during the COVID-19 pandemic. However, we did not directly test this hypothesis in the present study, such as by investigating the cytolytic activity against adipocytes previously infected by viruses, and notably SARS-CoV-2, in the presence or absence of anti-PD-L1. Indeed, such investigation would require additional experiments with specific SARS-CoV-2-infected adipose tissue, which is not available in our laboratory due to regulatory aspects.
In this study, we analyzed the impact of obesity on immune checkpoint expression in WAT, using a mouse model of diet-induced obesity. We observed that PD-L1 is overexpressed in WAT of obese mice, and is associated with increased expression of PD-1 in visceral WAT as compared with lean mice. Using human in vitro cocultures with ASC from obese individuals and MNC, we also observed an overexpression of PD-L1 with respect to cocultures with ASC from nonobese individuals, Moreover, conditioned media harvested from these cocultures enhanced PD-L1 expression in freshly differentiated adipocytes, depending on the presence of IFNγ. As adipocytes differentiated from obese ASC impaired cytotoxic activity, our results suggest that PD-L1 overexpression may occur in the visceral WAT of obese individuals under IFNγ secretion and may lead to T cell dysfunction, notably decreased cytolytic activity. Such a mechanism could shed some light on why adipose tissue-infiltrating viruses such as SARS-CoV-2 can worsen disease in obese individuals.
|
Measuring professional stigma towards patients with a forensic mental health status: protocol for a Delphi consensus study on the design of a questionnaire | 19ac237e-7ace-47d3-a72c-e4d95ad727b2 | 9438202 | Forensic Medicine[mh] | Forensic mental healthcare (FMHC) is aimed at improving patients’ mental health, reducing their risk of recidivism and ultimately a secure reintegration into society. In general terms, FMHC offers treatment to individuals who are both mentally disordered and whose behaviour has led or could lead again to offending. FMHC focuses on rehabilitative activities, as well as individualised care pathways, in order to increase the possibilities of a successful reintegration and return to their social environment. Treatment is typically provided on a continuum from highly specialised FMHC wards (within penitentiary settings) to (supported) community mental healthcare services (CMHC). CMHC services, however, seem reluctant to admit patients stigmatised by the label ‘forensic’. As a consequence, patients in FMHC may become subject to prolonged inpatient admissions, fostered institutionalisation and eventually a frustrated rehabilitation. To improve the rehabilitation options for patients in FMHC, a better liaison and understanding between FMHC and CMHC is needed. A first step in this direction could be to understand the attitudes CMHC professionals have towards patients with a forensic status. Research has shown, for instance, that CMHC professionals mention stereotypical pictures of ‘criminals’ and ‘dangerous criminals’ when asked about patients with a forensic status. Others found that patients with a history of offending were particularly associated with stereotypes of dangerousness and aggression, and they could count on less public sympathy than non-offending patients. Further believed most of the public (including police officers and psychiatrists) that these patients would not voluntarily undergo treatment and they were opposed to the idea to let them receive community-based treatment. Stigma and stigmatising attitudes are widespread. It involves stereotyping and devaluing individuals based on their belonging to a certain social group. In this regard, individuals with a mental illness are often associated with dangerousness, rarity, responsibility, incompetence, weakness of character, dependence, unpredictability, inferiority and vulnerability. Patients with a forensic status may be subject to simultaneous or multiple stigmas , as they also have a history of criminal offending. Hence, they further may be considered evil, mean, unintelligent, psychologically maladjusted, immature, inconsiderate and dishonest. Stereotypes refer to the beliefs or ‘knowledge’ structures about the characteristics and behaviours of a group of people. They are the cognitive component underlying stigma and stigmatising attitudes. Prejudice , understood as ‘the emotional reaction or feelings that people have toward a group or member of a group’, is the affective component. For instance, the stereotype of dangerousness may lead to feelings of fear or may be experienced as anxiety. Prejudice towards individuals with a mental illness includes fear, pity and anger, but this may vary per mental illness. For instance, the majority of the public feel sorry for individuals with mental illness, particularly for those with depression; however, they report uneasiness, uncertainty and fear towards individuals with schizophrenia and rejection towards individuals with drug abuse and alcoholism. Importantly, prejudice involves an active (cognitively and affectively) evaluative response, resulting in a negative emotional reaction. This means that people can be aware of stereotypes but not endorse them. This is especially important when fighting discrimination, the behavioural component of stigma. Discrimination is the unfair or unjust behaviours towards a social group or its member(s) (out-group) or exclusively favourable behaviour towards the members of one’s own group (in-group). Discriminatory behaviours exist along a continuum from subtle to overt and when it concerns individuals with a mental illness, withholding help, avoidance, segregation and coercion are most often described. Others also mentioned rejection, social distance, and exclusion. Although mental healthcare (MHC) professionals might be expected to have more positive attitudes towards individuals with a mental illness, research has shown that they too are susceptible to the negative attitudes endorsed in the general public. Despite their training, professional knowledge and experience with people with mental illness, they report, for instance, a desire for social distance comparable to the public. Psychiatrists seem to have more negative attitudes than general practitioners and clinical psychologists; however, when comparing the attitudes of students, doctors and nurses, the nurses held the least favourable attitudes towards patients with a mental illness. Regarding long-term treatment outcomes, psychiatrists seem more pessimistic than the general public, and also other medical professionals express low expectations of recovery. Lammie and colleagues assessed practitioner attitudes towards patients in medium and low secure forensic mental health settings. Even though the overall responses were positive, a significant minority of professionals reported to hold negative attitudes like recovery pessimism, pity, fear, anger, a desire for social distance, avoidance and blame. Notably, the negative attitudes were expressed more subtle. Meaning that professionals with mental health training seem to show positive explicit attitudes, but negative implicit attitudes, which may reflect unconscious emotions related to mental illness. Stigmatising attitudes towards individuals with mental illness have been associated with negative outcomes such as reduced self-esteem, social isolation, chronic stress, delayed help-seeking and loss of personal relationships. Also a history of criminal offending may have negative consequences including hindered access to services like housing and education, fewer employment opportunities and reduced social networks and supports. Of note, reverse outcomes have been shown to decrease the likelihood of recidivism and increase the likelihood of successful community re-entry. Here it is important to distinguish public stigma —which refers to the reaction of the general population or large social groups towards another or smaller social group, thereby endorsing stereotypes about and acting against them from self-stigma or internalised stigma —which refers to the extent to which an individual turns negative stereotypes and prejudice against oneself. Stigmatisation of a group of people can thus result in the internalisation of the stigmatising beliefs. This on its turn can affect recovery and negatively impact mental illness coping mechanisms and treatment engagement. Self-stigma has, furthermore, been associated with more severe psychiatric symptoms and a history of incarceration and homelessness, reduced coping strategies and feelings of shame, guilt, anger and distrust of others, as well as a risk factor for reoffending. Stigmatisation among professionals or professional stigma can be even more detrimental than by the public. It can have a significant impact on treatment outcomes and the patient’s quality of life. Among long-term patients with impoverished relationships, 76% named their healthcare professional as the most important person in their lives. Professionals’ negative attitudes may reduce treatment-seeking behaviours because patients anticipate their discrimination towards them, and the negative affective reactions and desire of social distance can lead to augmented disempowerment. The distinction between public/professional and self-stigma is important for understanding, explaining and building strategies to change stigmatising attitudes. Increased awareness of stereotypes or knowledge about FMHC, for instance, might be instrumental in combating prejudice or discrimination. A better understanding of CMHC professionals’ attitudes towards patients with a forensic status may therefore give indications on how to improve the liaison between FMHC and CMHC. Measures such as education programmes and awareness-raising events can be suggested to reduce stigmatising attitudes, and eventually increase the rehabilitation options for patients with a forensic status. To the author’s knowledge, there is no instrument specifically designed for the assessment of professional stigma towards patients with a forensic mental health status. Stigma assessment is complex as it involves an individual’s attitude towards a target population, and this attitude might be influenced by experiences, prejudices, stereotypes and knowledge. A Delphi study, as means for consensus building, allows to consider this interplay of factors through the involvement of experts that understand (1) the perspective of the perceiver (ie, professionals working in CMHC), (2) the target population (ie, patients, professionals and academics experienced in FMHC) and (3) stigma as an empirical construct (ie, academics investigating stigma). Departing from the many instruments that assess the attitudes towards individuals with mental illnesses, and in a lesser extent towards individuals with a history of criminal offending, this method enables to utilise the knowledge from international experts to select the most relevant items for the assessment of CMHC professionals’ attitudes towards patients with a forensic status.
The aim of this study is to reach expert consensus on items to assess stigmatising attitudes among community MHC professionals towards patients with a forensic mental health status. By means of a modified Delphi approach, consensus is sought on items that were selected and adapted from instruments that assess stigma towards individuals with either a mental illness or a history of criminal offending.
This study will be conducted using a modified version of the Delphi technique. The Delphi technique is an iterative multistage approach to seek consensus among ‘experts’ on a certain subject. Rather than having experts to meet physically, the Delphi technique can be conducted online, which allows the involvement of international experts. Within the context of mental health research, the Delphi technique has been applied for a great variety of purposes, among which the development of questionnaires. Contrary to a classical Delphi study, the first stage will not consist of a complete open round to obtain all qualitative input. Instead, we will apply a modified Delphi study, meaning that we will depart from a preselected longlist of items drawn from various stigma assessment instruments, and ask the experts to complete the list in case important items are missing. The anticipated rounds for achieving consensus are presented in . Development of the Delphi questionnaire Literature review—search strategy and study selection To identify the instruments that measure stigma among the public, health professionals and students, a targeted literature review was conducted in PubMed using the following terms ‘stigma*’ OR ‘stereotyp*’ OR ‘prejud*’ OR ‘attitude’ OR ‘discrim*’. The search strategy was further constructed by combining these with terms related to mental illness (ie, ‘mental* OR psychiatr* OR psychol*AND (disorder* OR illness*)’) or criminal background (ie, ‘offend*’ OR ‘forensic’ OR ‘prison*’ OR ‘secure unit’ OR ‘crim*’ OR ‘justice’) and assessment (ie, ‘assess* OR measure* OR question* OR instrument’). Finally, a third search included all terms. To obtain the most recent scientific evidence, the search was limited to studies published in 2011 or later. Additionally, we reviewed related papers referenced in selected studies, especially development articles, and consulted websites (ie, Indigo Network, www.indigo-group.org ) related to stigma assessment. The study eligibility criteria were as follows: Type of studies : quantitative studies with statistical analysis and with a validated measurement instrument, including papers on the development and psychometric evaluation of instruments relevant to our study. Construct of interest : only studies measuring public stigma or professional stigma were eligible. Stigma could be measured in a broad sense, so measures of beliefs, attitudes and behaviours were included. Target population : samples composed of Mental Health Practitioners (psychiatrists and psychologists), General Practitioners, Primary care and/or medical students. The population stigmatised had to be adults with mental illness a/o a history of criminal offending. Language : only English and Spanish papers were selected. Excluded were studies with non-validated or non-specified measurement instruments, studies focussing on the assessment of perceived stigma, associative stigma and stigma towards specific disorders, or studies assessing the impact of an intervention aimed at reducing stigma. Finally, also studies whose sample were children or adolescents, or whose stigma was directed towards this type of population were discarded of the eligibility process. Literature review—results The three searches together yielded 6939 articles, after removing duplicates. Inspection of abstracts and titles found that 6769 did not fulfil the inclusion criteria. A total of 170 articles were identified as potentially relevant, but 13 articles could not be retrieved and 79 were later excluded on closer examination of the full text as they did not match the inclusion criteria. Thus, a total of 78 articles were finally included. A preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow chart reflecting the study selection is presented in . Among the selected studies, 47 measured professional stigma , 15 measured public stigma and 4 measured both; 6 articles were psychometric evaluations and the rest (6) were instrument development or validation papers. The target populations were mainly patients with mental illness, and only one paper studied stigma towards forensic psychiatric patients; highlighting the gap of literature in this field. The most used scales were Community Attitudes towards Mental Illness Scale, followed by The Mental Illness: Clinicians’ Attitude and its different versions, Opinions About Mental Illness Scale and Opening Minds Stigma Scale for Healthcare Providers. The Attribution Questionnaire-27 and modified versions of Bogardus Social Distance Scale were also commonly used, but these scales were discarded because of the use of vignettes (AQ-27) and because the factor ‘Social Distance’ was already included in other questionnaires considered more appropriate for the purpose of our study (ie, Community Attitudes towards Mental Illness Scale). An overview of the instruments that were considered for the development of our Delphi questionnaire is presented in , indicating also the respective items that were selected and/or adapted. 10.1136/bmjopen-2022-061160.supp1 Supplementary data Structure of the Delphi questionnaire For the structure of the questionnaire, we followed the conceptualisation as proposed by Fox et al , taking into account items related to stereotypes, prejudices and discrimination. All items of the identified instruments were listed and categorised accordingly. Subsequently, all items were put in random order. To shorten the initial list of 468 items, each of the authors scored on a 7-point scale how relevant each item was for the purpose of the Delphi study. Overall, 79 items were selected (mean score of 5.33 or higher). To have a list with consistent wording (eg, type of care or patients), 70 items were reworded. Six items were rephrased; basically, these entailed comparisons between patients with a mental illness and ‘normal people’, we changed them to compare patients with a mental illness and patients with a forensic status. For one item (ie, ATP 36), we included two rephrased items. Finally, five items were added by the authors; these items were based on experiences in daily practice and considered missing in the existing instruments. Participants Our general approach is to invite five categories of experts: academics with knowledge about stigma assessment, academics with knowledge about patients with a forensic mental health status, healthcare professionals (eg, psychiatrists, nurses, psychologists, social workers, general practitioners) working in CMHC, healthcare professionals working in FMHC, and patients who are in the position of being or have been transferred from FMHC to CMHC. With regard to the groups of academics and professionals, an initial list of potential participants has been created following the purposive sampling approach. The authors (ie, GE-R and EV) approached their contacts in the field of FMHC in Europe and the CMHC in Catalonia, Spain. All contacts were asked to present five more potential candidates that met one or more of the following inclusion criteria: either a listed author in at least one publication related to (1) stigma towards patients with a forensic status; (2) stigma towards patients with a mental illness; (3) stigma towards (ex-)offenders; (4) stigma assessment; (5) conceptualisation of stigma; (6) care pathways or treatment in FMHC. and/or with clinical experience in patient care in (1) CMHC or (2) FMHC. For the identification of the stigma academics, (recurrent) authors of publications about stigma towards individuals with mental illness, (ex-)offenders, or patients with forensic mental health status were listed. With respect to the group of patients, an initial list of potential candidates has been created based on their transfer (history) of FMHC to CMHC. Although there is no widespread consensus about the appropriate sample size per participant category, a sample of 10–18 participants has been suggested. On the other hand, the more participants, the higher the reliability of the composite consensus. We will therefore aim for a minimum overall participation of 50 experts. Recruitment Except for the patients, potential participants will be contacted via their work email address, which is either publicly available or provided by the authors’ contacts. They will receive an email explaining the purpose of the Delphi study and an invitation to participate. Experts who confirm their willingness to participate, receive a second email with a link to the internet-based questionnaire and an explanatory letter with instructions on how to complete the questionnaire. The patient candidates will be approached by their (former) treating psychologist (author GE), who will explain the purpose of the study and invite the patients to participate, stressing the completely voluntary nature of participation. Patients who confirm to participate will receive the questionnaire and the instructions printed on paper. The introductory page of the questionnaire includes a consent clause, explaining that by clicking/marking the ‘I agree’ button, they consent to participate in the Delphi study. In all communications, we will explain the voluntary nature of the study, state that withdrawal is allowed at any time without any consequence for the participant and how personal data protection rights can be exercised. Confidentiality will be protected and individual data will not be shared with other participants or third parties. Each participant will be allocated an automatic random identification number, which will enable us to include the participant’s individual results in the feedback rounds. All other feedback will contain aggregate data to protect the participants’ identities and opinions. Structure of the Delphi procedure The Delphi method will consist of several iterative rounds in order to reach consensus, with different activities taking place in each of the consecutive rounds (see ). Round 1 In the first round, participants will receive a web-based or printed questionnaire with a list of potential items (i=85) randomly ordered to avoid biases. They will be asked to indicate the relevance of each item for the assessment of stigma by CMHC professionals towards patients with a forensic status, by giving a score on a 7-point Likert scale (1=not important at all to 7=extremely important). They will further be asked if they agree with the wording of the items (yes/no/do not know); thereby providing the opportunity to make suggestions for alternative wording. Finally, we will ask the participants to add important items that they consider missing and to include any additional comments in an open text box. Round 1 is foreseen to start in March 2022. Participants will be given 4 weeks to complete round 1. Reminders will be sent to non-responders every week following distribution. Round 2 The responses from round 1 will be aggregated and analysed (cf. data analysis). The aggregated anonymous results (ie, group median and IQR), the participant’s own responses and a narrative summary of the suggestions for rephrasing and additional comments will be sent as feedback together with an explanatory introduction for the second round. Items with consensus on inclusion or exclusion will be identified. Newly suggested items (ie, considered missing), newly reworded items and the remaining items will be presented using the same method as in round 1 (ie, 7-point Likert scale). Participants will again be asked if the rewording is adequate (yes/no/do not know) and to make suggestions for improvement. Participants will have the opportunity to leave additional comments. Of note, we will no longer ask for missing items. Round 3 After analysis of the responses of round 2, participants will receive feedback from rounds 1 and 2 (ie, aggregated anonymous results, narrative summary and own responses), indicating the items that reached consensus on inclusion or exclusion. The items will again be presented on a 7-point Likert scale for reconsideration. Additional comments will be allowed but improvement of phrasing will no longer be sought. Using the a priori established consensus thresholds (cf. data analysis), we will decide if a fourth round will be needed to reach consensus. If indicated, round 3 will be repeated; otherwise, the Delphi study will end with the consolidated list based on the outcomes of round 3. The Delphi study is foreseen to be finished by December 2022; notwithstanding, this will depend on the number of rounds needed to reach consensus. Data analysis To determine consensus, we will use the quantitative data obtained from the 7-point Likert scale. We will calculate descriptive statistics, including central tendency (median) and distribution (IQR) for all participants and per expert category. Following a multigroup consensus approach, the consensus thresholds will be defined as ≥60% of the participants of at least four of the five expert groups ranked the item in the top three (5–7; ie, inclusion) or bottom three (1–3; ie, exclusion) Likert categories. As a secondary measure, we will use the total number of items on which consensus on inclusion has been reached. For the stigma assessment questionnaire to be manageable, we will use a threshold of 30 items. For the reworded items, a ‘yes minus no’ score will be calculated (ie, the number of participants who answered a ‘yes’ on a specific item minus the number of participants who answered a ‘no’). For the modified items with low scores on ‘yes minus no’, new formulations will be proposed based on the suggestions from the participants. These will be included in the questionnaire of the following round (until round 3). We will conduct thematic content analyses for the qualitative data (ie, the missing items and additional comments). Similar newly suggested items will be combined or reformulated to avoid duplicates. Data collection and management All rounds will be conducted using Qualtrics software. Qualtrics is a secure web application for developing surveys with more complex response formats, methods of distribution or data management. The software complies with the General Data Protection Regulation and with the regulations necessary to process and store protected health information. Qualtrics is ISO 27001 certified and FredRAMP licensed. Qualtrics is a SaaS (software as a service), the software and data are hosted on Information and Communication Technology servers that are accessed via the Internet. Databases extracted from Qualtrics software will be securely stored on the server of Parc Sanitari Sant Joan de Déu. Only pseudonymised data will be exported to SPSS and Excel for further quantitative and qualitative analyses. Patient and public involvement Patients will participate as an expert panel in the Delphi study.
Literature review—search strategy and study selection To identify the instruments that measure stigma among the public, health professionals and students, a targeted literature review was conducted in PubMed using the following terms ‘stigma*’ OR ‘stereotyp*’ OR ‘prejud*’ OR ‘attitude’ OR ‘discrim*’. The search strategy was further constructed by combining these with terms related to mental illness (ie, ‘mental* OR psychiatr* OR psychol*AND (disorder* OR illness*)’) or criminal background (ie, ‘offend*’ OR ‘forensic’ OR ‘prison*’ OR ‘secure unit’ OR ‘crim*’ OR ‘justice’) and assessment (ie, ‘assess* OR measure* OR question* OR instrument’). Finally, a third search included all terms. To obtain the most recent scientific evidence, the search was limited to studies published in 2011 or later. Additionally, we reviewed related papers referenced in selected studies, especially development articles, and consulted websites (ie, Indigo Network, www.indigo-group.org ) related to stigma assessment. The study eligibility criteria were as follows: Type of studies : quantitative studies with statistical analysis and with a validated measurement instrument, including papers on the development and psychometric evaluation of instruments relevant to our study. Construct of interest : only studies measuring public stigma or professional stigma were eligible. Stigma could be measured in a broad sense, so measures of beliefs, attitudes and behaviours were included. Target population : samples composed of Mental Health Practitioners (psychiatrists and psychologists), General Practitioners, Primary care and/or medical students. The population stigmatised had to be adults with mental illness a/o a history of criminal offending. Language : only English and Spanish papers were selected. Excluded were studies with non-validated or non-specified measurement instruments, studies focussing on the assessment of perceived stigma, associative stigma and stigma towards specific disorders, or studies assessing the impact of an intervention aimed at reducing stigma. Finally, also studies whose sample were children or adolescents, or whose stigma was directed towards this type of population were discarded of the eligibility process. Literature review—results The three searches together yielded 6939 articles, after removing duplicates. Inspection of abstracts and titles found that 6769 did not fulfil the inclusion criteria. A total of 170 articles were identified as potentially relevant, but 13 articles could not be retrieved and 79 were later excluded on closer examination of the full text as they did not match the inclusion criteria. Thus, a total of 78 articles were finally included. A preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow chart reflecting the study selection is presented in . Among the selected studies, 47 measured professional stigma , 15 measured public stigma and 4 measured both; 6 articles were psychometric evaluations and the rest (6) were instrument development or validation papers. The target populations were mainly patients with mental illness, and only one paper studied stigma towards forensic psychiatric patients; highlighting the gap of literature in this field. The most used scales were Community Attitudes towards Mental Illness Scale, followed by The Mental Illness: Clinicians’ Attitude and its different versions, Opinions About Mental Illness Scale and Opening Minds Stigma Scale for Healthcare Providers. The Attribution Questionnaire-27 and modified versions of Bogardus Social Distance Scale were also commonly used, but these scales were discarded because of the use of vignettes (AQ-27) and because the factor ‘Social Distance’ was already included in other questionnaires considered more appropriate for the purpose of our study (ie, Community Attitudes towards Mental Illness Scale). An overview of the instruments that were considered for the development of our Delphi questionnaire is presented in , indicating also the respective items that were selected and/or adapted. 10.1136/bmjopen-2022-061160.supp1 Supplementary data Structure of the Delphi questionnaire For the structure of the questionnaire, we followed the conceptualisation as proposed by Fox et al , taking into account items related to stereotypes, prejudices and discrimination. All items of the identified instruments were listed and categorised accordingly. Subsequently, all items were put in random order. To shorten the initial list of 468 items, each of the authors scored on a 7-point scale how relevant each item was for the purpose of the Delphi study. Overall, 79 items were selected (mean score of 5.33 or higher). To have a list with consistent wording (eg, type of care or patients), 70 items were reworded. Six items were rephrased; basically, these entailed comparisons between patients with a mental illness and ‘normal people’, we changed them to compare patients with a mental illness and patients with a forensic status. For one item (ie, ATP 36), we included two rephrased items. Finally, five items were added by the authors; these items were based on experiences in daily practice and considered missing in the existing instruments.
To identify the instruments that measure stigma among the public, health professionals and students, a targeted literature review was conducted in PubMed using the following terms ‘stigma*’ OR ‘stereotyp*’ OR ‘prejud*’ OR ‘attitude’ OR ‘discrim*’. The search strategy was further constructed by combining these with terms related to mental illness (ie, ‘mental* OR psychiatr* OR psychol*AND (disorder* OR illness*)’) or criminal background (ie, ‘offend*’ OR ‘forensic’ OR ‘prison*’ OR ‘secure unit’ OR ‘crim*’ OR ‘justice’) and assessment (ie, ‘assess* OR measure* OR question* OR instrument’). Finally, a third search included all terms. To obtain the most recent scientific evidence, the search was limited to studies published in 2011 or later. Additionally, we reviewed related papers referenced in selected studies, especially development articles, and consulted websites (ie, Indigo Network, www.indigo-group.org ) related to stigma assessment. The study eligibility criteria were as follows: Type of studies : quantitative studies with statistical analysis and with a validated measurement instrument, including papers on the development and psychometric evaluation of instruments relevant to our study. Construct of interest : only studies measuring public stigma or professional stigma were eligible. Stigma could be measured in a broad sense, so measures of beliefs, attitudes and behaviours were included. Target population : samples composed of Mental Health Practitioners (psychiatrists and psychologists), General Practitioners, Primary care and/or medical students. The population stigmatised had to be adults with mental illness a/o a history of criminal offending. Language : only English and Spanish papers were selected. Excluded were studies with non-validated or non-specified measurement instruments, studies focussing on the assessment of perceived stigma, associative stigma and stigma towards specific disorders, or studies assessing the impact of an intervention aimed at reducing stigma. Finally, also studies whose sample were children or adolescents, or whose stigma was directed towards this type of population were discarded of the eligibility process.
The three searches together yielded 6939 articles, after removing duplicates. Inspection of abstracts and titles found that 6769 did not fulfil the inclusion criteria. A total of 170 articles were identified as potentially relevant, but 13 articles could not be retrieved and 79 were later excluded on closer examination of the full text as they did not match the inclusion criteria. Thus, a total of 78 articles were finally included. A preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow chart reflecting the study selection is presented in . Among the selected studies, 47 measured professional stigma , 15 measured public stigma and 4 measured both; 6 articles were psychometric evaluations and the rest (6) were instrument development or validation papers. The target populations were mainly patients with mental illness, and only one paper studied stigma towards forensic psychiatric patients; highlighting the gap of literature in this field. The most used scales were Community Attitudes towards Mental Illness Scale, followed by The Mental Illness: Clinicians’ Attitude and its different versions, Opinions About Mental Illness Scale and Opening Minds Stigma Scale for Healthcare Providers. The Attribution Questionnaire-27 and modified versions of Bogardus Social Distance Scale were also commonly used, but these scales were discarded because of the use of vignettes (AQ-27) and because the factor ‘Social Distance’ was already included in other questionnaires considered more appropriate for the purpose of our study (ie, Community Attitudes towards Mental Illness Scale). An overview of the instruments that were considered for the development of our Delphi questionnaire is presented in , indicating also the respective items that were selected and/or adapted. 10.1136/bmjopen-2022-061160.supp1 Supplementary data
For the structure of the questionnaire, we followed the conceptualisation as proposed by Fox et al , taking into account items related to stereotypes, prejudices and discrimination. All items of the identified instruments were listed and categorised accordingly. Subsequently, all items were put in random order. To shorten the initial list of 468 items, each of the authors scored on a 7-point scale how relevant each item was for the purpose of the Delphi study. Overall, 79 items were selected (mean score of 5.33 or higher). To have a list with consistent wording (eg, type of care or patients), 70 items were reworded. Six items were rephrased; basically, these entailed comparisons between patients with a mental illness and ‘normal people’, we changed them to compare patients with a mental illness and patients with a forensic status. For one item (ie, ATP 36), we included two rephrased items. Finally, five items were added by the authors; these items were based on experiences in daily practice and considered missing in the existing instruments.
Our general approach is to invite five categories of experts: academics with knowledge about stigma assessment, academics with knowledge about patients with a forensic mental health status, healthcare professionals (eg, psychiatrists, nurses, psychologists, social workers, general practitioners) working in CMHC, healthcare professionals working in FMHC, and patients who are in the position of being or have been transferred from FMHC to CMHC. With regard to the groups of academics and professionals, an initial list of potential participants has been created following the purposive sampling approach. The authors (ie, GE-R and EV) approached their contacts in the field of FMHC in Europe and the CMHC in Catalonia, Spain. All contacts were asked to present five more potential candidates that met one or more of the following inclusion criteria: either a listed author in at least one publication related to (1) stigma towards patients with a forensic status; (2) stigma towards patients with a mental illness; (3) stigma towards (ex-)offenders; (4) stigma assessment; (5) conceptualisation of stigma; (6) care pathways or treatment in FMHC. and/or with clinical experience in patient care in (1) CMHC or (2) FMHC. For the identification of the stigma academics, (recurrent) authors of publications about stigma towards individuals with mental illness, (ex-)offenders, or patients with forensic mental health status were listed. With respect to the group of patients, an initial list of potential candidates has been created based on their transfer (history) of FMHC to CMHC. Although there is no widespread consensus about the appropriate sample size per participant category, a sample of 10–18 participants has been suggested. On the other hand, the more participants, the higher the reliability of the composite consensus. We will therefore aim for a minimum overall participation of 50 experts.
Except for the patients, potential participants will be contacted via their work email address, which is either publicly available or provided by the authors’ contacts. They will receive an email explaining the purpose of the Delphi study and an invitation to participate. Experts who confirm their willingness to participate, receive a second email with a link to the internet-based questionnaire and an explanatory letter with instructions on how to complete the questionnaire. The patient candidates will be approached by their (former) treating psychologist (author GE), who will explain the purpose of the study and invite the patients to participate, stressing the completely voluntary nature of participation. Patients who confirm to participate will receive the questionnaire and the instructions printed on paper. The introductory page of the questionnaire includes a consent clause, explaining that by clicking/marking the ‘I agree’ button, they consent to participate in the Delphi study. In all communications, we will explain the voluntary nature of the study, state that withdrawal is allowed at any time without any consequence for the participant and how personal data protection rights can be exercised. Confidentiality will be protected and individual data will not be shared with other participants or third parties. Each participant will be allocated an automatic random identification number, which will enable us to include the participant’s individual results in the feedback rounds. All other feedback will contain aggregate data to protect the participants’ identities and opinions.
The Delphi method will consist of several iterative rounds in order to reach consensus, with different activities taking place in each of the consecutive rounds (see ). Round 1 In the first round, participants will receive a web-based or printed questionnaire with a list of potential items (i=85) randomly ordered to avoid biases. They will be asked to indicate the relevance of each item for the assessment of stigma by CMHC professionals towards patients with a forensic status, by giving a score on a 7-point Likert scale (1=not important at all to 7=extremely important). They will further be asked if they agree with the wording of the items (yes/no/do not know); thereby providing the opportunity to make suggestions for alternative wording. Finally, we will ask the participants to add important items that they consider missing and to include any additional comments in an open text box. Round 1 is foreseen to start in March 2022. Participants will be given 4 weeks to complete round 1. Reminders will be sent to non-responders every week following distribution. Round 2 The responses from round 1 will be aggregated and analysed (cf. data analysis). The aggregated anonymous results (ie, group median and IQR), the participant’s own responses and a narrative summary of the suggestions for rephrasing and additional comments will be sent as feedback together with an explanatory introduction for the second round. Items with consensus on inclusion or exclusion will be identified. Newly suggested items (ie, considered missing), newly reworded items and the remaining items will be presented using the same method as in round 1 (ie, 7-point Likert scale). Participants will again be asked if the rewording is adequate (yes/no/do not know) and to make suggestions for improvement. Participants will have the opportunity to leave additional comments. Of note, we will no longer ask for missing items. Round 3 After analysis of the responses of round 2, participants will receive feedback from rounds 1 and 2 (ie, aggregated anonymous results, narrative summary and own responses), indicating the items that reached consensus on inclusion or exclusion. The items will again be presented on a 7-point Likert scale for reconsideration. Additional comments will be allowed but improvement of phrasing will no longer be sought. Using the a priori established consensus thresholds (cf. data analysis), we will decide if a fourth round will be needed to reach consensus. If indicated, round 3 will be repeated; otherwise, the Delphi study will end with the consolidated list based on the outcomes of round 3. The Delphi study is foreseen to be finished by December 2022; notwithstanding, this will depend on the number of rounds needed to reach consensus.
In the first round, participants will receive a web-based or printed questionnaire with a list of potential items (i=85) randomly ordered to avoid biases. They will be asked to indicate the relevance of each item for the assessment of stigma by CMHC professionals towards patients with a forensic status, by giving a score on a 7-point Likert scale (1=not important at all to 7=extremely important). They will further be asked if they agree with the wording of the items (yes/no/do not know); thereby providing the opportunity to make suggestions for alternative wording. Finally, we will ask the participants to add important items that they consider missing and to include any additional comments in an open text box. Round 1 is foreseen to start in March 2022. Participants will be given 4 weeks to complete round 1. Reminders will be sent to non-responders every week following distribution.
The responses from round 1 will be aggregated and analysed (cf. data analysis). The aggregated anonymous results (ie, group median and IQR), the participant’s own responses and a narrative summary of the suggestions for rephrasing and additional comments will be sent as feedback together with an explanatory introduction for the second round. Items with consensus on inclusion or exclusion will be identified. Newly suggested items (ie, considered missing), newly reworded items and the remaining items will be presented using the same method as in round 1 (ie, 7-point Likert scale). Participants will again be asked if the rewording is adequate (yes/no/do not know) and to make suggestions for improvement. Participants will have the opportunity to leave additional comments. Of note, we will no longer ask for missing items.
After analysis of the responses of round 2, participants will receive feedback from rounds 1 and 2 (ie, aggregated anonymous results, narrative summary and own responses), indicating the items that reached consensus on inclusion or exclusion. The items will again be presented on a 7-point Likert scale for reconsideration. Additional comments will be allowed but improvement of phrasing will no longer be sought. Using the a priori established consensus thresholds (cf. data analysis), we will decide if a fourth round will be needed to reach consensus. If indicated, round 3 will be repeated; otherwise, the Delphi study will end with the consolidated list based on the outcomes of round 3. The Delphi study is foreseen to be finished by December 2022; notwithstanding, this will depend on the number of rounds needed to reach consensus.
To determine consensus, we will use the quantitative data obtained from the 7-point Likert scale. We will calculate descriptive statistics, including central tendency (median) and distribution (IQR) for all participants and per expert category. Following a multigroup consensus approach, the consensus thresholds will be defined as ≥60% of the participants of at least four of the five expert groups ranked the item in the top three (5–7; ie, inclusion) or bottom three (1–3; ie, exclusion) Likert categories. As a secondary measure, we will use the total number of items on which consensus on inclusion has been reached. For the stigma assessment questionnaire to be manageable, we will use a threshold of 30 items. For the reworded items, a ‘yes minus no’ score will be calculated (ie, the number of participants who answered a ‘yes’ on a specific item minus the number of participants who answered a ‘no’). For the modified items with low scores on ‘yes minus no’, new formulations will be proposed based on the suggestions from the participants. These will be included in the questionnaire of the following round (until round 3). We will conduct thematic content analyses for the qualitative data (ie, the missing items and additional comments). Similar newly suggested items will be combined or reformulated to avoid duplicates.
All rounds will be conducted using Qualtrics software. Qualtrics is a secure web application for developing surveys with more complex response formats, methods of distribution or data management. The software complies with the General Data Protection Regulation and with the regulations necessary to process and store protected health information. Qualtrics is ISO 27001 certified and FredRAMP licensed. Qualtrics is a SaaS (software as a service), the software and data are hosted on Information and Communication Technology servers that are accessed via the Internet. Databases extracted from Qualtrics software will be securely stored on the server of Parc Sanitari Sant Joan de Déu. Only pseudonymised data will be exported to SPSS and Excel for further quantitative and qualitative analyses.
Patients will participate as an expert panel in the Delphi study.
The Delphi consensus study has received ethical approval from the ethics committee of Fundación Sant Joan de Déu (reference number C.I. PIC-186-21) and the institutional research board of Parc Sanitari Sant Joan de Déu (reference number C.R. 66-2021-09). Dissemination of the results will be through peer-reviewed publications, presentations, symposiums and workshops at (inter-)national academic conferences and a summary of the results will be shared with the participants, and key persons in community as well as forensic MHC.
Reviewer comments Author's
manuscript
|
Three-year evaluation of the nosocomial infections in pediatrics: bacterial and fungal profile and antimicrobial resistance pattern | 9c706d55-4a9f-4e09-a4ee-35bc75b791d1 | 8851736 | Pediatrics[mh] | A nosocomial infection (NI) (also known as hospital-acquired infection) is a localized or a systemic infection occurring with an adverse reaction to infectious agents that develops in 48 h or more after admission . NIs could lead to considerably higher mortality rates, length of the hospital stay and costs, and represent a serious public health concern worldwide . The leading bacteria related to NIs are Staphylococcus aureus , coagulase-negative staphylococci (CoNS), Streptococcus pneumoniae , Escherichia coli , Pseudomonas aeruginosa, Haemophilus influenzae, Klebsiella pneumoniae, Acinetobacter , and Enterococci . Nowadays, antibiotics remain the leading therapy for treating bacterial infections. However, by the unreasonable use of antibiotics, certain strains of multidrug-resistant (MDR) bacteria have emerged by selection pressure; consequently, bacteria that have been once sensitive, re-emerged as resistant to different antibiotics and create limited therapeutic options, increased risks of treatment failure and poor patient management . Knowledge of proper antimicrobial prescription policy of a particular setting in addition to the investigation of causative agents and their antimicrobial susceptibility profile, is essential to improve the management and reduction of the rate of NIs . The aim of the current study was the evaluation of the frequency and antimicrobial susceptibility of NIs in an Iranian children medical center during three years.
This cross-sectional study was carried out in the referral hospital of Children’s Medical Center, Tehran, Iran between March 2017 and February 2020. Ethical approval (IR.TUMS.CHMC.REC.1399.037) was obtained from the Ethical Committee of Tehran University of Medical Sciences, Tehran, Iran. All patients who admitted to the medical wards of Children’s Medical Center, Tehran, Iran for more than 48 h and had the evidence of NIs with positive blood, wounds and sterile fluids culture of gram-positive/gram-negative bacteria and fungi were included in this study. Duplicate isolates from one patient were excluded from the study. In vitro phenotypic characterization of bacteria or fungi was carried out using standard culture and biochemical tests as described previously . The disk diffusion method or minimal inhibitory concentration (MIC) was used to test each isolate for in vitro antimicrobial susceptibility based on the Clinical and Laboratory Standards Institute criteria . The following antibiotics disks from MAST Categories Ltd., Merseyside, UK, were used: imipenem (10 µg), ampicillin (10 µg), cefotaxime (30 µg), clindamycin (2 µg), Trimethoprim−sulfamethoxazole (1.25/23.75 mg), ceftazidime (30 µg), nitrofurantoin (200 µg), ceftriaxone (30 µg), erythromycin (15 µg), gentamycin (10 µg), cefepime (30 µg), penicillin (10 µg), linezolid (30 µg), cefoxitin (30 µg). Staphylococcus aureus ATCC 25,923 was used for quality control of the test. The MICs of vancomycin and colistin were determined by E-test methods. Statistical analysis of the results was performed using SPSS 13.0 (SPSS Inc. Chicago, IL, USA). The results were presented as mean, frequency and standard deviation for quantitative and percentage and frequency for qualitative data.
In the current study, a total of 718 patients included, among which 61.3% were male (N = 440). The median age of the patients was 2.5 years (IQR: 1 month–3 years). Among the patients, 27.2% had underlying heart disease (N = 195) and 16.3% had seizures (N = 117). Intrinsic and acquired immunodeficiency was also reported in a number of patients (N = 35, 4.9%, N = 59, 8.2%, respectively). Three hundred and eighty-four patients (53.5%) utilized catheters, and 101 of them (14.1%) had endotracheal tube during their hospitalization. The frequency of isolated microorganisms among the studied patients based on the sources of their isolation was mentioned in Table . Klebsiella pneumonia and Candida spp . were the most prevalent isolates (N = 125, 17.4%, N = 121, 16.9%, respectively), followed by P. aeruginosa (N = 72, 10%) and CoNS (N = 69, 9.6%). Also, most of the samples were isolated from blood (N = 495, 69%), followed by sterile fluids (N = 165, 23%) and finally wounds (N = 58, 8%). Klebsiella pneumonia was the most frequent organism isolated from blood and wounds, and Candida spp . was the most frequent organism isolated from sterile fluids. There was a slight decrease in the total number of isolates each year compared to the previous year (the first year: 272 patients, 37.9%; the second year: 234 patients, 32.6%; the third year: 212 patients, 29.5%). Morganella morganii and Haemophilus spp. specimens were isolated only in the first year of the study (2017). During these years, Serratia marcescens (n = 13, n = 12, n = 6, respectively) and S. aureus (n = 19, n = 10, n = 6, respectively) showed a decreasing trend. While the Enterococcus spp. (n = 21, n = 13, n = 13, respectively) and P. aeruginosa (n = 30, n = 21, n = 21, respectively) after a 2-year downward trend, in 2019, remained stable. The frequency of Pseudomonas spp. (n = 6, n = 17, n = 25, respectively) and Enterobacter spp. (n = 5, n = 16, n = 17, respectively) represented an increasing trend. Most of the isolates were collected from hospitalized patients at neonatal intensive care unit (NICU) and pediatric intensive care unit (PICU) (N = 109, 15.2%, N = 100, 13.9%, respectively) and the most isolated microorganisms from them were K. pneumonia (N = 29, 26.6%) and Candida spp. (N = 25, 25%), respectively. Antibiotic susceptibility frequencies of evaluated microorganisms were depicted in Table . Escherichia coli , Acinetobacter baumannii , S. marcescens , K. pneumonia and Pseudomonas spp. strains showed 100% sensitivity to colistin. Pseudomonas aeroginusa strains as a whole showed significant sensitivity to the studied and the most sensitive antibiotics were imipenem (80.4%) and ceftazidime (80.8%). Subsequently, the highest sensitivity to ceftazidime was observed in Pseudomonas spp. (79.2%), while A. baumannii strains showed 94.8% resistance to this antibiotic. Vancomycin resistance was not reported among S. aureus isolates in this study. Clindamycin had the least effect on CoNS strains (18.6%). Staphylococcus aureus strains were highly resistant to gentamycin (100%), ciprofloxacin (100%) and penicillin (85.7%). Methicillin-resistant S . aureus (MRSA) was found in 45.5% of the isolates. However, next to vancomycin, nitrofourantoin and imipenem (each n = 1/1, 100%), and Trimethoprim−sulfamethoxazole (n = 18/23, 78.3) were the most effective antimicrobial agents on it. High levels of resistance to gentamycin were also showed among S. marcescens (n = 19/22, 86.4%), Enterococcus spp. (n = 13/17, 76.5%), A. baumannii (n = 29/38), 76.3%), and Pseudomonas spp. (n = 16/21, 76.2%) strains. All of the tested isolates of Streptococcus spp. were 100% sensitive to ampicillin and penicillin (each n = 3/3), and vancomycin (n = 4/4), but fully resistant to erythromycin (n = 2/2) and Trimethoprim−sulfamethoxazole (n = 1/1). Escherichia coli showed a high level of resistance to cefotaxime (n = 28/33, 87.5%), Trimethoprim−sulfamethoxazole (n = 25/30, 83.3%), cefepime (n = 23/28, 82.2%), and imipenem (n = 7/9, 77.8%), but 100% sensitivity to nitrofourantoin (n = 14/14). Acinetobacter baumannii strains also displayed more than 90% resistance to the almost all antibiotics studied including imipenem, cefepime, Trimethoprim−sulfamethoxazole, meropenem, piperacillin/ tazobactam, amikacin, ciprofloxacin, and cefotaxime. Likewise, K. pneumonia (n = 64/84, 76.2%) and S. marcescens (n = 18/20, 90%) strains were resistant to piperacillin/ tazobactam. However, this antibiotic was mostly effective on Pseudomonas spp . (n = 18/20, 90%). All of the tested isolates of S. maltophilia were susceptible to Trimethoprim−sulfamethoxazole (n = 29/29, 100%) and showed high susceptibility rate to ciprofloxacin (n = 27/28, 96.4%). The isolates of Enterobacter spp . showed 73.7% sensitivity to amikacin (n = 14/19).
In this study, we evaluated the microorganisms isolated from NIs over three consecutive years which generally had a slow decreasing trend. The present study showed K. pneumoniae (N = 125, 17.4%), Candida spp. (N = 121, 16.9%), and P. aeruginosa (N = 72, 10%) as the most frequent microorganisms which cause NIs among the studied children. Of course other frequent NI-causing bacteria were reported in our study including CoNS (9.6%), A. baumannii (7.1%), Psuedomonas spp. (6.7%), and Enterococcus spp. (6.5%). 61% of isolated organisms were gram-negative bacteria, which was about three times more than the number of gram-positive bacteria isolated in our study (22.1%). Likewise, high rate of gram-negative bacteria was reported in Feleke et al. (53.2%) study . Also unlike the study of Feleke et al. and the one reported from Jimma which mentioned S. aureus and E. coli as their the most common isolates , in the present study, K. pneumoniae was the most common bacteria isolated. Similarly, in the study accomplished by Mahmoudi et al., K. pneumoniae (n = 263, 27.5%) was reported as the most frequent bacteria. In a study by Bouza et al. , E. coli (35.3%) was the most commonly isolated microorganism, and Klebsiella spp. were reported as 9.8% of the pathogens. Gupta et al. reported that S. aureus and CoNS as the most common isolated gram-positive bacteria which is in line with our results. Nouri et al. reported the high prevalence of gram-negative bacteria (77.9%) in NIs and low prevalence of gram-positive bacteria (22.1%), exactly as ours, and the most common bacterium causing NIs among the latter was S. aureus . 67% of isolated strains was from ICUs (mostly NICU and PICU) (N = 482), which was compatible with the results of our previous study . Also, Alvares et al. reported nosocomial pneumonia as the third most common NI in their pediatric intensive care unit . Candida spp. strains were isolated frequently from PICU (25%) and emergency ICU (24.4%). Surgical and ICU patients are at higher risk of rising nosocomial fungal infections . In critically ill patients, the disseminated candida infections are the principal causes of morbidity and mortality both in immunocompetent and immunocompromised patients . A. baumannii strains were considerably resistant to almost all tested antibiotics except for colistin (100% sensitivity), which is similar to previous studies . Sohail et al. also showed that only 0.1% of the isolated strains were resistant to colistin. The results of study reported by Vahdani et al. showed antibiotic-resistant A. Baumannii infections with high resistant rate to ceftazidime (96%), followed by ceftizoxime (95%), ceftriaxone (93%), ciprofloxacin (85%), and trimethoprim/sulfamethoxazole (85%). Along with the significance of MDR A. baumannii in NIs, the increasing reports of outbreaks caused by carbapenem-resistant A. baumannii in recent years have become another frightening reality . In the present study, K. pneumonia strains were highly resistant to cefotaxime (95.6%), while showed 100% susceptibility to colistin, vancomycin, ampicillin, ceftazidime and clindamycin. Sensitivity to gentamycin reported as low as 37.5% among K. pneumonia strains in our study. Compared with the results of the study by Ares et al. , the resistance rates of isolates in the current study against studied antibiotics, especially carbapenems, were considerably high. This difference in the resistance patterns of K. pneumoniae could be due to the different prevalent clones in Iran and other countries in addition to differences in antibiotic treatment regimens in different areas . All E. coli isolates tested in this study were sensitive to nitrofurantoin and colistin, while showing significant resistance to the other antibiotics compared to our previous study . However, the resistance of this microorganism to imipenem (77.8% in comparison with 8%) has increased significantly compared to the mentioned study. High resistance to ampicillin has been reported in other studies, as well . The frequency of MRSA (43%) was more than the amount reported by our previous study (26%) , Nigussie et al. and Latif et al. (38.5% and 31.25%, respectively) . In this study, P. aeruginosa strains were highly sensitive to amikacin (81.8%), imipenem (80.4%), piperacillin/tazobactam (78.4%), Trimethoprim−sulfamethoxazole and ciprofloxacin (75%). However, resistance rates of P. aeruginosa to gentamicin (27.3%), amikacin (18.2%) and ceftazidime (19.2%) were higher than our recent study . In addition, lower resistance rate for cefepime was reported by Larru et al. (4.3%) and Ares et al. (8.5%) , compared to the percentage of 38% in the current study. There are only a limited number of studies describing the S. maltophilia infection in children . Treatment of nosocomial S. maltophilia infections is complicated due to high rates of antibiotic resistance . We reported 100% resistant S. maltophilia isolate to gentamycin, imipenem, and penicillin (n = 1/1). However, treatment of S. maltophilia infection is difficult due to antimicrobial resistance to a variety of agents; trimethoprim-sulfamethoxazol can continue to be the first choice for the treatment of S. maltophilia. In the study performed by Alsuhaibani et al., the most effective antibiotic against S. maltophilia isolates was Trimethoprim−sulfamethoxazole (94.1%), which is consistent with our data (100%). Also in the study by Sun et al. , the resistance rate of S. maltophilia strains to cefpime, cefotaxime, ceftazidime and gentamicin was 45.1%, 94.1%, 60.8% and 82.4%, respectively. Regarding the frequency of resistance to vancomycin, no cases were reported among S. aureus , while 64% of Enterococcus spp. were resistant to vancomycin that is similar to our recent previous study and is higher than our previous studies in our hospital during 2009–2010 . Since NIs are an important determinant in hospital, improving of the prevention and treatment of NIs is still highly needed .
High frequency of antimicrobial resistance to the commonly tested antibiotics is a concerning alarm. Therefore, effective infection control programs and rational antibiotic use policies should be established promptly.
|
The clinicopathological and prognostic significances of IGF-1R and Livin expression in patients with colorectal cancer | a60380bf-471d-40aa-a006-a0279e113291 | 9354317 | Anatomy[mh] | Colorectal cancer (CRC) is the third most common malignant tumor and the second most common cause of death worldwide. According to the GLOBOCAN project of the World Health Organization (WHO) Cancer Research Center, the number of new CRC cases in 2020 is about 1.93 million worldwide, and the number of deaths is about 940,000 . By 2030, the global burden of CRC will increase by 60%: the number of new cases will exceed 2.2 million, and the number of deaths will exceed 1.1 million . The incidence of CRC varies globally, with colon cancer being the highest rates in European regions, Australia/New Zealand, and Northern America and with rectal cancer being the highest rates in Eastern Asia, while CRC incidence tends to be low in most regions of Africa and in South Central Asia . Age-standardized incidence and mortality rates for CRC in the United States of America has decreased significantly recently, but increased in China . Several risk factors, including age, hereditary components, chronic intestinal inflammation, obesity, excessive alcohol and red meat consumption, smoking, and lack of physical exercise, have been identified for CRC . Despite the advancement in comprehensive therapy, the long-term survival of CRC patients remains unsatisfactory . The poor therapeutic outcome of patients with CRC is mainly due to local recurrence and distant metastasis . Therefore, understanding the pathogenesis of CRC and finding effective therapeutic targets is imperative . Insulin-like growth factor 1 receptor (IGF-1R) belongs to the tyrosine kinase receptor family. The elevated expression and activity of IGF-1R are observed in many cancer types and are associated with tumor cell proliferation, survival, anti-apoptosis, and drug resistance . IGF-1R triggers various intracellular signaling cascades in colonic mucosal cells, enhancing cell cycle progression and inhibiting apoptosis . IGF-1R inhibition results in the G1 cell cycle arrest and a significant decrease in CRC cell proliferation, survival, and radioresistance . A previous study demonstrated that IGF-1R is overexpressed in human colorectal cancer HT-29 cell lines. Cyclolignan picropodophyllin (PPP, an IGF-1R inhibitor) inhibits the proliferation and migration of HT-29 cell lines in a dose-dependent manner and decreases the expression/phosphorylation of IGF-1R, extracellular signal-regulated kinase 1/2 (ERK1/2), and Akt . The overexpression of the inhibitor of apoptosis protein (IAP) family, including baculoviral IAP repeat-containing protein-7 (BIRC7), facilitates apoptosis evasion in CRC . BIRC7 is also referred to as Livin or melanoma inhibitor of apoptosis. BIRC7/Livin also interacts with the second mitochondria-derived activator of caspases through the baculoviral IAP repeat domain to promote the caspase activation in the cytochrome c pathway . BIRC7/Livin potentiates migration and invasion of CRC cells partially via nuclear factor-kappa B (NF-κB)-mediated epithelial-mesenchymal transition . Although both IGF-1R and Livin are potential biomarkers for CRC, only a few studies have investigated their role in normal mucosa-colorectal adenoma-colorectal cancer progression. The analysis of differential gene expression between diverse cancer types and adjacent normal tissues across The Cancer Genome Atlas (TCGA) cohort revealed higher IGF-1R and Livin gene expression in patients with colon adenocarcinoma (COAD) compared to the adjacent normal tissues from the Sangerbox, GEPIA2, and UALCAN portal datasets. The current study also used TCGA-COAD to explore the clinical significance, patient survival, and correlation between IGF-1R and Livin expression and CRC. In agreement with the TCGA study, CRC ( n = 60), colorectal adenoma ( n = 30), and adjacent normal tissues ( n = 10) and their clinicopathological data were selected, and the expression and distribution of IGF-1R and Livin proteins in various intestinal mucosal tissues was detected, and the correlation between IGF-1R and Livin expression and clinicopathological parameters, occurrence, development, and clinical prognosis of CRC patients was analyzed. These data may reveal a novel oncogenic function and the clinical value of IGF-1R and Livin in CRC.
Patients and surgical specimens A total of 60 patients with primary CRC who underwent surgical resection at the General Surgery Department of the Second Affiliated Hospital of Dalian Medical University (Dalian, China) comprised the CRC group. The cohort was in Dukes’ stages A, B, C, and D ( n = 15 cases). None of the patients had received chemotherapy, radiotherapy, or other anti-cancer treatment before the operation. Biopsies from 10 cases of adjacent non-tumor tissues comprised the control group. Thirty patients who underwent colonoscopy and endoscopic polypectomy, endoscopic mucosal resection (EMR), or endoscopic submucosal dissection (ESD) at the Department of Gastroenterology of the Second Affiliated Hospital of Dalian Medical University were regarded as the colorectal adenoma group, including 8 cases of tubular adenoma and 22 cases of villous tubular adenoma. All tissue specimens were confirmed pathologically, and all the clinicopathological data were complete. Samples from patients with diabetes, acromegaly, dwarfism, cachexia, severe organ failure, or other malignant tumors were excluded. This research protocol has been reviewed by the Ethics Committee of the Second Affiliated Hospital of Dalian Medical University and agreed that the project would be carried out on the premise of protecting the rights and interests of subjects (approval No. 132, ethical review 2021, the Second Affiliated Hospital of Dalian Medical University). Data processing and analysis For gene expression, clinical significance, and survival analysis, the TCGA-COAD datasets from Sangerbox ( http://sangerbox.com/ ), GEPIA2 ( http://gepia2.cancer-pku.cn/ ), and UALCAN portal ( http://ualcan.path.uab.edu/ ) were used. The significant difference was estimated by Student’s t-test considering unequal variance for gene expression and clinical significance analysis. The normalized average counts from the adjacent normal tissue samples served as the threshold to filter the samples. Then, only samples with counts above the threshold were used for survival analysis. Survival plots were drawn by comparing the top 30% (high-expression group) and the bottom 30% filtered samples (low-expression group). Survival curves of overall survival (OS) and disease-free survival (DFS) were estimated by Kaplan–Meier analysis. The P -values were calculated by log-rank test. The correlation analysis was performed with the GEPIA2 dataset in the “colorectal adenocarcinoma” and “adjacent normal tissues” from the TCGA cohort by Spearman’s chi-square test analysis, using the non-log scale for calculation and the log-scale axis for visualization. Immunohistochemistry (IHC) The specimens were fixed in formalin and embedded with paraffin before slicing into 3-μm-thick sections. After deparaffinization using xylene and dehydration using graded ethanol, the sections were immersed into citrate buffer at pH 7.2–7.4 for antigen retrieval. Then, the sections were probed with IGF-1R (1:200; Abcam Cat# ab39522, RRID:AB_2122268) and Livin (1:200; Proteintech Cat# 27543–1-AP, RRID:AB_2880901) antibodies at 4 °C overnight and 37 °C for 1 h, followed by phosphate-buffered saline (PBS) washes. Subsequently, the sections were incubated with a biotinylated secondary antibody (Zhongshan Golden Bridge Biotech Co. Ltd., Beijing, China) at 37 °C for 30 min. The immunoreaction was developed by incubation with diaminobenzidine (DAB) for 10 min. Hematoxylin was used for counterstaining and alcohol gradient for dehydration. Finally, the sections were sealed with neutral gum and observed under a light microscope (Olympus Bx-51, Japan). All the slides were separately examined and scored by two experienced pathologists blinded to the study design. Staining results: IGF-1R and Livin were observed as yellow or brownish-yellow staining in the cell membrane. Five random fields were examined at high magnification for each sample. The staining intensity (no staining is 0, slight staining is 1 point, moderate staining is 2 points, and deep staining is 3 points) and ratios of positive cells (0–25% refers to 0 points, 26–50% is 1 point, 51–75% is 2 points, > 76% is 3 points) were determined using the couple score and semiquantitative method. The final results were multiplied by two scores mentioned above: 0–1 point, negative (−); 2–4 points, weak positive ; 5–7 points, positive ; 8–9 points, strong positive . Statistical analysis The data were analyzed using SPSS 20.0 statistical software (IBM Co., Armonk, NY, USA). Mann–Whitney U test or Kruskal–Wallis test was used for enumeration data, and Spearman’s chi-squared test was used for correlation analysis. The test standard α = 0.05 and P < 0.05 indicated a statistically significant difference.
A total of 60 patients with primary CRC who underwent surgical resection at the General Surgery Department of the Second Affiliated Hospital of Dalian Medical University (Dalian, China) comprised the CRC group. The cohort was in Dukes’ stages A, B, C, and D ( n = 15 cases). None of the patients had received chemotherapy, radiotherapy, or other anti-cancer treatment before the operation. Biopsies from 10 cases of adjacent non-tumor tissues comprised the control group. Thirty patients who underwent colonoscopy and endoscopic polypectomy, endoscopic mucosal resection (EMR), or endoscopic submucosal dissection (ESD) at the Department of Gastroenterology of the Second Affiliated Hospital of Dalian Medical University were regarded as the colorectal adenoma group, including 8 cases of tubular adenoma and 22 cases of villous tubular adenoma. All tissue specimens were confirmed pathologically, and all the clinicopathological data were complete. Samples from patients with diabetes, acromegaly, dwarfism, cachexia, severe organ failure, or other malignant tumors were excluded. This research protocol has been reviewed by the Ethics Committee of the Second Affiliated Hospital of Dalian Medical University and agreed that the project would be carried out on the premise of protecting the rights and interests of subjects (approval No. 132, ethical review 2021, the Second Affiliated Hospital of Dalian Medical University).
For gene expression, clinical significance, and survival analysis, the TCGA-COAD datasets from Sangerbox ( http://sangerbox.com/ ), GEPIA2 ( http://gepia2.cancer-pku.cn/ ), and UALCAN portal ( http://ualcan.path.uab.edu/ ) were used. The significant difference was estimated by Student’s t-test considering unequal variance for gene expression and clinical significance analysis. The normalized average counts from the adjacent normal tissue samples served as the threshold to filter the samples. Then, only samples with counts above the threshold were used for survival analysis. Survival plots were drawn by comparing the top 30% (high-expression group) and the bottom 30% filtered samples (low-expression group). Survival curves of overall survival (OS) and disease-free survival (DFS) were estimated by Kaplan–Meier analysis. The P -values were calculated by log-rank test. The correlation analysis was performed with the GEPIA2 dataset in the “colorectal adenocarcinoma” and “adjacent normal tissues” from the TCGA cohort by Spearman’s chi-square test analysis, using the non-log scale for calculation and the log-scale axis for visualization.
The specimens were fixed in formalin and embedded with paraffin before slicing into 3-μm-thick sections. After deparaffinization using xylene and dehydration using graded ethanol, the sections were immersed into citrate buffer at pH 7.2–7.4 for antigen retrieval. Then, the sections were probed with IGF-1R (1:200; Abcam Cat# ab39522, RRID:AB_2122268) and Livin (1:200; Proteintech Cat# 27543–1-AP, RRID:AB_2880901) antibodies at 4 °C overnight and 37 °C for 1 h, followed by phosphate-buffered saline (PBS) washes. Subsequently, the sections were incubated with a biotinylated secondary antibody (Zhongshan Golden Bridge Biotech Co. Ltd., Beijing, China) at 37 °C for 30 min. The immunoreaction was developed by incubation with diaminobenzidine (DAB) for 10 min. Hematoxylin was used for counterstaining and alcohol gradient for dehydration. Finally, the sections were sealed with neutral gum and observed under a light microscope (Olympus Bx-51, Japan). All the slides were separately examined and scored by two experienced pathologists blinded to the study design. Staining results: IGF-1R and Livin were observed as yellow or brownish-yellow staining in the cell membrane. Five random fields were examined at high magnification for each sample. The staining intensity (no staining is 0, slight staining is 1 point, moderate staining is 2 points, and deep staining is 3 points) and ratios of positive cells (0–25% refers to 0 points, 26–50% is 1 point, 51–75% is 2 points, > 76% is 3 points) were determined using the couple score and semiquantitative method. The final results were multiplied by two scores mentioned above: 0–1 point, negative (−); 2–4 points, weak positive ; 5–7 points, positive ; 8–9 points, strong positive .
The data were analyzed using SPSS 20.0 statistical software (IBM Co., Armonk, NY, USA). Mann–Whitney U test or Kruskal–Wallis test was used for enumeration data, and Spearman’s chi-squared test was used for correlation analysis. The test standard α = 0.05 and P < 0.05 indicated a statistically significant difference.
IGF-1R is frequently overexpressed in primary CRC tumors The analysis of the IGF-1R gene expression between diverse cancer types and adjacent normal tissues across The Cancer Genome Atlas (TCGA) cohort revealed high gene expression in COAD ( n = 458) compared to the adjacent normal tissues ( n = 41) ( P < 0.05, Fig. A) from Sangerbox dataset. Similarly, upregulated IGF-1R gene expression was also observed in primary COAD in GEPIA2 dataset ( n = 275, Fig. B) and UALCAN portal dataset ( n = 286, P < 0.05, Fig. C). Furthermore, the IHC results showed significantly higher levels of IGF-1R in tubular adenoma, villous adenoma, and CRC tissues compared to the adjacent normal tissues, as assessed by the Mann–Whitney U test, and the difference between adenoma and CRCs was statistically significant (Fig. D, Table ). These results demonstrated that IGF-1R is commonly overexpressed in CRC and may play a critical role in precancerous polyps growth and malignant transformation. Livin expression is upregulated in CRC Next, the Livin (also known as BIRC7) expression in primary CRC tissues was analyzed from TCGA cohort. The gene expression was significantly upregulated in primary COAD compared to the adjacent normal tissues in the Sangerbox dataset ( P < 0.001, Fig. A). Specifically, the upregulation of Livin gene expression in COAD was validated in two datasets (GEPIA2 and UALCAN) of paired tumor and adjacent normal samples from the TCGA study (Fig. B and C). Livin IHC was conducted in a cohort of 100 patients with localized CRCs, adenoma, and adjacent normal tissues to further analyze Livin expression and validate these findings. The staining showed that the level of Livin protein in the colorectal adenoma and cancer groups was significantly higher than that of the adjacent normal tissues; however, no significant difference was detected between the colorectal adenoma and cancer groups (Fig. D, Table ). These results demonstrated that the high level of Livin protein is related to the occurrence of CRC, and it may not be a crucial factor in the progression of adenoma to CRC. Overexpression of IGF-1R is an independent predictor of oncogenic function in CRC patients Next, the clinicopathological significance of the IGF-1R gene and protein expression was evaluated in CRC patients. A significant correlation was established between IGF-1R gene expression and clinicopathologic features, including age (61–80 years), gender (male), histological subtype (adenocarcinoma), individual cancer stages (stage 3), nodal metastasis status (N1 and N2), and TP53-mutant (Fig. A-E and G). In addition, the promoter methylation level of IGF-1R was closely related to COAD (Fig. F). Also, the correlation between IGF-1R protein expression and clinicopathological features in CRC was evaluated. Consistent with the TCGA study, IGF-1R protein overexpression was associated with the depth of invasion, Dukes’ stage, lymph node metastasis, and distant metastasis in patients with CRC (all P < 0.05, Table ), and with the increase in Dukes’ stage, the expression level of IGF-1R protein increased significantly (Fig. H, Table ). However, no correlation was established between IGF-1R protein expression and age, gender, diameter, location, histological subtype, and differentiation (Table ). These results demonstrated that IGF-1R might be used as a biomarker to evaluate the condition of CRC patients. Clinical application value of Livin expression in CRC The association between Livin gene and protein expression and the clinicopathological features of CRC was analyzed. As shown in Fig. A–G, a significant correlation was established between Livin gene expression and clinicopathological features, including age (41–60, 61–80, and 81–100 years), gender (male and female), histological subtype (adenocarcinoma), individual cancer stages (1, 2, and 3), nodal metastasis status (N0, N1, and N2), and TP53 mutation status (TP53-mutant and TP53-non-mutant), while no correlation was established in the promoter methylation level of Livin. Moreover, Spearson’s chi-square test analysis showed that Livin protein overexpression was an independent risk factor for stage and metastasis but was not associated with age, gender, tumor size, primary tumor location, pathological type, histological type, or depth of invasion (Fig. H, Table ). IGF-1R and Livin are not prognostic markers of CRC The correlation between IGF-1R and Livin expression and patients’ OS and DFS was estimated by Kaplan–Meier analysis. The survival map showed that both IGF-1R and Livin are not significantly correlated with COAD from the TCGA dataset (Fig. A and B). the patient cases from TCGA-COAD with survival data were sorted into the high- and low-expression of IGF-1R and Livin groups. As shown in the Kaplan–Meier survival curves, CRC patients with high IGF-1R or Livin expression showed a longer OS and DFS than those with low expression based on the log-rank test ( P > 0.05 for both OS and DFS, Fig. C-F). Moreover, Spearman’s chi-squared test analysis established a marked correlation between IGF-1R and Livin expression in COAD ( R = 0.2, P = 0.00048, Fig. ).
The analysis of the IGF-1R gene expression between diverse cancer types and adjacent normal tissues across The Cancer Genome Atlas (TCGA) cohort revealed high gene expression in COAD ( n = 458) compared to the adjacent normal tissues ( n = 41) ( P < 0.05, Fig. A) from Sangerbox dataset. Similarly, upregulated IGF-1R gene expression was also observed in primary COAD in GEPIA2 dataset ( n = 275, Fig. B) and UALCAN portal dataset ( n = 286, P < 0.05, Fig. C). Furthermore, the IHC results showed significantly higher levels of IGF-1R in tubular adenoma, villous adenoma, and CRC tissues compared to the adjacent normal tissues, as assessed by the Mann–Whitney U test, and the difference between adenoma and CRCs was statistically significant (Fig. D, Table ). These results demonstrated that IGF-1R is commonly overexpressed in CRC and may play a critical role in precancerous polyps growth and malignant transformation.
Next, the Livin (also known as BIRC7) expression in primary CRC tissues was analyzed from TCGA cohort. The gene expression was significantly upregulated in primary COAD compared to the adjacent normal tissues in the Sangerbox dataset ( P < 0.001, Fig. A). Specifically, the upregulation of Livin gene expression in COAD was validated in two datasets (GEPIA2 and UALCAN) of paired tumor and adjacent normal samples from the TCGA study (Fig. B and C). Livin IHC was conducted in a cohort of 100 patients with localized CRCs, adenoma, and adjacent normal tissues to further analyze Livin expression and validate these findings. The staining showed that the level of Livin protein in the colorectal adenoma and cancer groups was significantly higher than that of the adjacent normal tissues; however, no significant difference was detected between the colorectal adenoma and cancer groups (Fig. D, Table ). These results demonstrated that the high level of Livin protein is related to the occurrence of CRC, and it may not be a crucial factor in the progression of adenoma to CRC.
Next, the clinicopathological significance of the IGF-1R gene and protein expression was evaluated in CRC patients. A significant correlation was established between IGF-1R gene expression and clinicopathologic features, including age (61–80 years), gender (male), histological subtype (adenocarcinoma), individual cancer stages (stage 3), nodal metastasis status (N1 and N2), and TP53-mutant (Fig. A-E and G). In addition, the promoter methylation level of IGF-1R was closely related to COAD (Fig. F). Also, the correlation between IGF-1R protein expression and clinicopathological features in CRC was evaluated. Consistent with the TCGA study, IGF-1R protein overexpression was associated with the depth of invasion, Dukes’ stage, lymph node metastasis, and distant metastasis in patients with CRC (all P < 0.05, Table ), and with the increase in Dukes’ stage, the expression level of IGF-1R protein increased significantly (Fig. H, Table ). However, no correlation was established between IGF-1R protein expression and age, gender, diameter, location, histological subtype, and differentiation (Table ). These results demonstrated that IGF-1R might be used as a biomarker to evaluate the condition of CRC patients.
The association between Livin gene and protein expression and the clinicopathological features of CRC was analyzed. As shown in Fig. A–G, a significant correlation was established between Livin gene expression and clinicopathological features, including age (41–60, 61–80, and 81–100 years), gender (male and female), histological subtype (adenocarcinoma), individual cancer stages (1, 2, and 3), nodal metastasis status (N0, N1, and N2), and TP53 mutation status (TP53-mutant and TP53-non-mutant), while no correlation was established in the promoter methylation level of Livin. Moreover, Spearson’s chi-square test analysis showed that Livin protein overexpression was an independent risk factor for stage and metastasis but was not associated with age, gender, tumor size, primary tumor location, pathological type, histological type, or depth of invasion (Fig. H, Table ).
The correlation between IGF-1R and Livin expression and patients’ OS and DFS was estimated by Kaplan–Meier analysis. The survival map showed that both IGF-1R and Livin are not significantly correlated with COAD from the TCGA dataset (Fig. A and B). the patient cases from TCGA-COAD with survival data were sorted into the high- and low-expression of IGF-1R and Livin groups. As shown in the Kaplan–Meier survival curves, CRC patients with high IGF-1R or Livin expression showed a longer OS and DFS than those with low expression based on the log-rank test ( P > 0.05 for both OS and DFS, Fig. C-F). Moreover, Spearman’s chi-squared test analysis established a marked correlation between IGF-1R and Livin expression in COAD ( R = 0.2, P = 0.00048, Fig. ).
The present study demonstrated that IGF-1R and Livin genes are highly expressed in CRC and the other cancer types, such as cholangiocarcinoma and kidney, liver, and lung cancers, through the TCGA database . Also, the expression of IGF-1R and Livin genes was increased in the 275 CRC cases from the GEPIA2 dataset and 286 colorectal cancer cases from the UALCAN portal dataset. IGF-1R may play a role in the neoplastic progression of CRC because it is related to both cellular proliferation and differentiation . However, low-level IGF-1R expression is found in some invasive cancers, which promotes cancer cell dedifferentiation to a mesenchymal type morphology with loss of cell adhesion . The expression of Livin is associated with tumor development and progression in CRC by increasing tumor cell motility and inhibiting apoptosis . The occurrence and development of CRC is a multi-stage complex process, with the dynamic evolution of normal epithelial, proliferative polyp, adenoma, cancer, and metastasis. The results of IHC staining showed a significant correlation between the expression of IGF-1R protein and neoplastic progression from normal mucosa to adenomatous polyps and finally to CRC. The high expression of Livin protein is related to the occurrence of CRC, but it may not be a critical factor in the progression of adenoma to CRC. Furthermore, the clinical importance of IGF-1R and Livin expression in CRC patients was investigated at the mRNA and protein levels in two independent cohorts and found that both molecules were overexpressed. Expression of IGF-1R increases with tumor size in CRC . Livin mRNA expression is strongly related to CRC invasive depth but not to clinical tumor stage, differentiation, lymph node metastasis, tumor morphological category and pathological type, and patient’s age and gender . According to the TCGA study, the levels of IGF-1R and Livin were associated significantly with age, gender, histological subtype, individual cancer stages, nodal metastasis status, and TP53 mutation status in CRC relative to the adjacent normal tissues. TP53 is a well-known tumor suppressor gene that encodes p53 and is frequently inactivated by mutation or deletion in most human CRCs . In addition, the promoter methylation level of IGF-1R, but not Livin, was closely related to CRC. Moreover, our results showed that high levels of IGF-1R and Livin were significantly correlated with stage and metastasis. However, no correlation was established between IGF-1R and Livin protein expression and other clinicopathological features, such as age, gender, diameter, location, differentiation, and histological subtype from sixty patients with CRC. Our results are inconsistent with those reported in the TCGA database as well as other studies, possibly due to regional and population differences, as well as insufficient sample size and methodology in this study. Several studies have shown that CRC patients with negative expressions of IGF-1R and Livin had significantly higher accumulative survival rate and longer mean survival duration than those with positive expression of IGF-1R and Livin . Multivariate analysis of potential prognostic factors showed that IGF-1R expression and worsened performance status are independent predictors of poor outcomes . Additionally, high expression of Livin may influence the prognosis of CRC . Both survival map and Kaplan–Meier survival curves showed that high expression levels of IGF-1R and Livin were not correlated significantly with the short OS of patients with CRC. The clinical outcome of CRC varies greatly depending on the aggressiveness of individual tumors. Many patients experience disease recurrence following radical surgery. Nonetheless, the TCGA dataset demonstrated that neither IGF-1R nor Livin serves as an independent prognostic marker for CRC patients, which was opposite to the previous findings. Thus, additional prognostic biomarkers may accurately assess the risk to guide personalized chemotherapy. These results showed a marked correlation between IGF-1R and Livin expression in COAD by Spearman’s chi-square test. In future studies, confocal fluorescence staining would be employed to study the co-localization of IGF-1R and Livin proteins in CRC tissues to determine whether double-positive expression promotes the occurrence and development of CRC more than single-positive expression.
In summary, IGF-1R and Livin oncogenes are highly expressed in CRC patients. Both can be considered biomarkers for the stage, metastasis, and risk of carcinogenesis. IGF-1R may drive neoplastic initiation and progression along the colorectal normal mucosa-polyp-cancer sequence. In addition, these findings indicated that IGF-1R and Livin are not independent prognostic markers for patients with CRC. Nonetheless, further investigations are needed to substantiate the current findings.
Additional file 1: Fig. S1. The negative and positive control expression of IGF-1R and Livin proteins. (40×, scale = 20 μm).
|
Unique anatomy of a rare type II dens invaginatus in a maxillary lateral incisor: a case report | b44c455f-816d-43c2-99e1-341939900dbb | 11954236 | Surgery[mh] | Dens invaginatus (DI) arises from excessive proliferation or distortion of the enamel organ into the dental papilla during the morpho-differentiation phase of tooth development . This condition can result in the formation of a pocket or anatomic dead space on the tooth surface following eruption . The thin lining of the invagination in DI tends to trap bacteria, leading to early caries and consequently pulp necrosis . The anomaly originates in the early stages of tooth development, specifically during the cap stage, where abnormal pressure or growth disturbances causes the enamel organ to invaginate into the dental papilla. As the tooth continues to develop, the invaginated enamel organ proliferates and differentiates alongside the surrounding dental tissues, potentially extending into the root and forming a complex internal structure that may encompass enamel, dentin, and pulp tissue [ , – ]. The etiology of DI remains controversial and unclear, with genetic or mechanical theories being the most widely accepted explanations . A genetic predisposition is suggested by familial clustering and the high prevalence of DI in specific ethnic groups . It is hypothesized that external forces from adjacent tooth germs, along with trauma and infections, may contribute to the invagination of enamel into the dental papilla . Furthermore, factors such as focal growth retardation or acceleration of the tooth bud, as well as restrictions imposed by the dental arch on the enamel organ, have also been considered as potential contributors . Reported prevalence rates of DI vary significantly, ranging from 0.04% to12% across different studies . This variation is likely attributed to racial differences, as well as differing diagnostic criteria and investigation methods. DI predominantly occurs in permanent teeth, although it has occasionally been observed in deciduous teeth; among tooth types, maxillary lateral incisors are most commonly affected, followed by maxillary central incisors and canines . The morphological features of this invagination exhibit considerable variability in terms of depth and structural complexity. Several classification systems have been proposed to characterize this anomaly. Among these, the most widely adopted classification system was established by Oehlers , who categorized DI into three types based on radiographic characteristics: Type I, where the invagination is confined to the crown and does not extends beyond the cemento-enamel junction (CEJ); Type II, where the invagination extends beyond the CEJ but does not reach the entire root length; Type IIIa, where the invagination penetrates through the root and opens in the periodontal ligament space via a pseudo-foramen; and type IIIb, where invagination traverses the root and contacts the periapical region at the apical foramen (Fig. ). Oehlers also identified various crown shapes associated with the three types of DI, including normal crowns with a pronounced lingual or palatal pit, as well as conical, barrel-shaped, or peg-shaped crowns featuring an incisal pit. Chen et al. reported that Type I (87.5%) was the most prevalent form in a Chinese subpopulation, significantly higher than the incidence of Type II (12.23%) and Type III (0.72%). DI increases the risk of developing pulpal pathosis, which may be attributed to the absence of enamel in the invagination and the presence of remnants of the dental papilla or periodontal connective tissue . Previous scholars have reported that the enamel within the invagination could be hypoplastic . This condition heightens the risk of microbial penetration and subsequent infection of the pulp tissue, potentially resulting in pulp necrosis. In this report, we present a case of Type II DI affecting a maxillary lateral incisor, which ultimately required extraction due to refractory apical periodontitis. Pre-treatment evaluation was performed using cone-beam computed tomographic (CBCT), and post-extraction analysis was conducted using micro-computed tomography (micro-CT). The unique anatomy of this tooth may offer valuable insights for appropriate dental management strategies.
A 20-year-old male presented to the Department of Dentistry at the Affiliated Hospital of Nantong University on March 18, 2024, with a 12-month history of recurrent pain and swelling in the left anterior maxillary region. Prior to this visit, the patient had sought treatment at a local rural clinic, where he was prescribed ornidazole and amoxicillin (500 mg each, twice daily) for 10 days. However, the symptoms persisted despite adherence to the antibiotic regimen. The patient denied any history of trauma or previous dental interventions in the affected area. No periapical radiographs were documented during the prior treatment. Intraoral examination revealed that tooth #22 exhibited a barrel-shaped morphology, characterized by a depression at the incisal edge and the absence of the palatal fossa and developing groove. A small dark pit was observed at the foramen coecum. A palpable swelling, tender upon palpation, was noted on the anterior maxilla, adjacent to the apical area of tooth #22. The tooth demonstrated sensitivity to percussion and exhibited Grade I mobility. Thermal testing indicated that tooth #22 was non-vital, although the color of the crown remained normal. Examination of the contralateral lateral incisors revealed normal findings. Given the complexity of the symptoms and the need for detailed anatomical evaluation to establish an accurate diagnosis, a CBCT scan (3D eXam I, KaVo Dental GmbH, Germany) was performed with adherence to the “as low as reasonably achievable” (ALARA) principle for radiation dose optimization. The CBCT images (Fig. A-C) revealed a well-defined radiolucent area measuring approximately 10.3 × 9.9 × 7.0 mm at the apex of tooth #22, with the radiolucent margin extending from the apex to the middle third of the root. The apical lesion exhibited an oval shape with a distinct cortical border. The enamel invagination extended beyond the CEJ and reached the mid-root level, indicating a type II DI. Due to the voxel size limitation of 250 μm, it was unclear whether the invaginated canal communicated with the main root canal. Based on the clinical and CBCT findings, a diagnosis of type II DI with associated pulpal necrosis and chronic apical periodontitis was established. Given the complexity of the root/canal structure and the extent of bone defect, the prognosis of conservative treatment was considered uncertain. The patient was informed of the available treatment options, which included extraction or endodontic therapy with surgical intervention. Considering the complexity of treatment procedure, the high dental cost, and the uncertain success rate, the patient opted for extraction. Consequently, tooth #22 was extracted (Fig. D-F), and the periapical lesion was thoroughly removed. The excised periapical lesion was submitted for histopathological examination, which confirmed the diagnosis of a periapical granuloma (Fig. ). The granulomatous tissue was predominantly infiltrated with inflammatory cells, including lymphocytes, plasma cells, neutrophil, and macrophages, without evidence of an epithelium-lined lumen. To gain a more detailed understanding of the anatomy of DI, the extracted tooth was scanned using a micro-CT device (SkyScan1174; Bruker-microCT, Kontich, Belgium) with a voxel size of 32 μm. Both the tooth structures and the root canal system were reconstructed in three dimensions (3D). The 2D (Fig. G-I) and 3D (Fig. J-L) micro-CT images revealed that the invagination communicated with the external oral cavity via an incisal pit at the foramen coecum and with the main root canal at the mid-root level, serving as a conduit for pulp infection. The enamel surrounding the invagination exhibited uneven thickness, with an enamel and dentin defect at the apical end forming a connection to the pulp cavity. The enamel and the underlying dentine were completely folded into the root, creating a concentric structure and a “pseudo root canal” at the center, while compressing the original pulp cavity into a concentric, sheet-like structure, thereby complicating access to the main root canal during root canal therapy. Additionally, apical ramifications with five accessory canals were also identified at the root apex (Fig. L).
Early detection of DI is crucial to prevent complications such as pulpitis and/or periapical periodontitis. The invagination acts as a potential pathway for microorganisms and irritants to reach the pulp, potentially leading to inflammation and necrosis . DI is a dental anomaly with potential genetic and ethnic predispositions . During tooth development, ectomesenchymal signaling pathways are crucial in controlling the growth and folding of the enamel organ. A deficiency in specific molecules can lead to irregularly shaped teeth and defects in the developing tooth germ. Therefore, the hypothesis that genetic factors could be responsible for DI is plausible . DI has been association with various syndromes, including those with genetic origins such as Williams, Nance-Horan, and Ekman-Westborg-Julin syndromes . Pokala and Acs documented a case where an individual with a deletion at chromosome 7q32 exhibited DI alongside other dental anomalies, such as hypodontia. These findings suggest a potential genetic basis for DI, highlighting the importance of considering genetic and possibly ethnic factors in its etiology. Additionally, the prevalence of DI varies among ethnic groups, suggesting an ethnic predisposition. For instance, certain populations, including the Chinese populations , exhibit a higher incidence of DI and barrel/shovel-shaped maxillary lateral incisors, which could be attributed to genetic variations or environmental factors specific to those groups. Understanding these predispositions is essential for early diagnosis and management, enabling clinicians to identify high-risk individuals and implement preventive measures effectively . Although teeth with DI often exhibit atypical crown shapes, such as conical, peg-shaped, barrel-shaped, or dilated forms , these morphological abnormalities do not consistently indicate the presence of a DI lesion. DI is frequently discovered incidentally during radiographic examinations, where the invagination appears as a deep, enamel-lined fissure, often associated with the formation of a radiolucent pocket . This condition results in complex root canal morphology, posing significant challenges for endodontic treatment. Oehlers’ classification, based on 2D radiographic images, may underestimate the true extent and complexity of the invagination, as it is a 3D structure with varying extensions in different planes . Therefore, CBCT is particularly valuable for the accurate diagnosis and management of DI . Treatment strategies for DI, such as root canal therapy, removal of the invagination, intentional replantation, and surgical intervention, are determined based on the classification of DI and the condition of the pulp and periapical tissues . Sealing the invaginations is recommended as a prophylactic measure when the pulp is not infected. For lesions that are not extensive, the application of flowable composite is generally adequate for sealing. In more severe cases, particularly those near the pulp, the use of mineral trioxide aggregates or bioceramic materials may be more appropriate . In instances involving deep periodontal pockets, periodontal flap surgery may also be indicated. However, a study has demonstrated a failure rate of 13.4% for prophylactic invagination treatments after a follow-up period of six months or longer, with all reported failures occurring in cases of Type II DI . Consequently, periodic follow-up examinations are essential to prevent the development of periapical disease. In this case, given the absence of a radicular groove and the patient’s age of 25, with periapical periodontitis symptoms emerging only in the past year, we speculate that early prophylactic sealing of the invagination post-eruption could have prevented pulp necrosis and the development of a periapical granuloma, thereby preserving the tooth. However, the pit at the occlusal foramen coecum (Fig. F), which is the entrance to the invagination, along with the mildly barreled tooth shape, may be easily overlooked during dental examination. Even if identified, these features may not necessarily lead to the consideration of DI and the need for sealing. Once the invagination becomes infected, more complex treatment modalities are required. Advances in pulp biology and recent clinical research on vital pulp therapy have provided new insights into biologically driven treatment options . In cases where uninflamed pulp tissue remains and the pulp has not undergone complete necrosis, pulpotomy may be considered, particularly in immature teeth, to promote continued root development . Under magnification, hemostasis is achieved, the degree of inflammation is assessed, and inflamed pulp tissue is excised to allow the remaining healthy tissue to recover and heal . However, if the inflammation progresses, root canal treatment may ultimately be necessary. In recent decades, pulp revascularization has emerged as a promising treatment for immature teeth affected by DI, with several successful cases reported [ – ]. This technique specifically designed for immature permanent teeth with necrotic pulp, aiming to promote root development by harnessing the differentiation potential of remaining stem cells . There is currently no consensus on whether the invagination should be removed, particularly in Type II and III DI. While retaining the invagination may enhance root strength, residual debris within the invagination can impede effective cleaning and obturation of the root canal system, potentially compromising treatment outcomes . In instances where the invagination does not communicate with the root canal system via accessory canal(s) or dentin defects, the pulp may remain vital and healthy, allowing for the isolated treatment of invagination . Regardless of the presence of an apical lesion, nonsurgical endodontic treatment should be prioritized . Although, the complex anatomy of DI may reduce the success rate of such treatments, surgical intervention should not be considered as the initial approach. To improve success rates, the use of an operating microscope and CBCT is essential for thorough cleaning, shaping, and filling of the root canal system . Recent advances in contemporary endodontics have introduced splints based on internal anatomical information derived from CBCT, facilitating guided and minimally invasive access to the pulp cavity . Following root canal treatment, periodic follow-up is necessary to assess therapeutic efficacy. If treatment failure is detected, endodontic surgery or intentional replantation may be considered. Should these interventions fail and the lesion remains uncontrollable, extraction may become the final option . In this case, it is regrettable that the patient chose extraction over attempting therapeutic intervention. The preoperative CBCT examination and post-extraction micro-CT analysis reveal that the primary challenge in treating this case lies in effective removal of the pulp and pathological tissues from the root canal system, particularly in the coronal and middle third levels. Figure B and J show that the invagination is entirely embedded in the center of the tooth, making its removal extremely challenging. The central pseudo root canal was expected to contain periodontal tissues rather than the dental pulp, and it does not communicate with the external periodontal space. The blind canal terminates apically with enamel and dentin defects that connect to the main root canal through gaps or tubules. A more prudent strategy would be to preserve the invagination and establish a pathway through it to access the main root canal and apex. However, the pulp chamber was severely compressed by the invagination, complicating the ability of instruments to reach the working length and thoroughly clean the entire root canal space. Large amounts of debris and uninstrumented areas may remain in the irregular area of the pulp chamber after canal instrumentation. Theoretically, sufficient canal irrigation or deeper penetration with small files using ultrasonics is recommended to remove tissues from the irregular space of the root canal system. Additionally, the presence of apical ramification and a periapical granuloma suggests that conventional root canal treatment is unlikely to completely eliminate infectious debris within the root canal system. Consequently, the prognosis is poor without the intervention of apical surgery. As illustrated in Fig. L, micro-CT imaging revealed the presence of multiple (five) apical accessory canals that were not detectable using CBCT. These accessory canals may contribute to persistent apical inflection and the development of a periapical granuloma. Apicoectomy with a 3 mm apical resection can effectively remove the majority of the apical delta and infected tissue, which are challenging to clean and shape using conventional endodontic techniques. However, root-end resection may expose irregular canal spaces and complex anatomical structures associated with the invagination. Meticulous retropreparation and retrofilling of these areas are critical to achieving predictable periapical healing. Although no endodontic treatment or apical surgery was performed in this case, it highlights the importance of a detailed analysis of the root canal system in cases of DI, particularly in differentiating between an invaginated “pseudo root canal” and the true main root canal system. This study has several limitations that warrant acknowledged. First, the analysis is based on a single case of DI with atypical tooth anatomy, which limits the generalizability of the findings to a broader population. Second, while the patient’s decision to undergo extraction facilitated detailed micro-CT evaluation, the clinical relevance of the observed anatomical variations remains theoretical, as no conservative therapeutic interventions were pursued. A clearer understanding of endodontic treatment options could have significantly influenced the patient’s decision regarding tooth extraction. Informing the patient about the potential for tooth preservation through advanced endodontic procedures, such as root canal therapy, and/or apical surgery, might have led to a preference for less invasive options. Educating the patient on the success rates, benefits, and potential complications of these treatments could have provided a more comprehensive perspective, potentially favoring tooth preservation over extraction. Additionally, the costs and time involved in endodontic treatment versus extraction might have influenced the decision. Finally, demographic factors such as ethnicity, age, and gender significantly influence the presentation of this dental anomaly [ , , ]. To mitigate potential selection bias and enhance the generalizability of the findings, future research should prioritize multi-centered, in vivo CBCT studies involving larger, ethnically diverse cohorts across a wider age range. In conclusion, this case report highlights the intricate and unique morphological characteristics of the root canal system in a maxillary lateral incisor with Type II DI. The invaginated radicular region is particularly susceptible to establishing pathways of communication with both the external oral environment and the internal root canal system, thereby increasing the risk of endodontic and periodontal infections.
|
Effect of Nitrogen Fertilizer on the Rhizosphere and Endosphere Bacterial Communities of Rice at Different Growth Stages | c92feb34-d975-4553-af4b-f33da2e241f5 | 11678815 | Microbiology[mh] | Plant root microorganisms encompass a diverse array of microbes associated with different regions of the root system, including the rhizoplane, rhizosphere, and endosphere et al. . These microorganisms form complex interactions with their host plants , significantly influencing plant growth, development, nutrient uptake, and disease resistance . Rhizosphere microorganisms inhabit the soil surrounding the roots, where they play a crucial role in nutrient cycling. Their metabolic activities, such as respiration and organic acid production, contribute to the dissolution of insoluble minerals, thereby enhancing the availability of phosphorus and other essential nutrients for plant uptake. On the other hand, endosphere microorganisms, residing within the root tissues, influence plant growth and development by secreting bioactive substances, including phytohormones . Current research emphasizes the roles of beneficial root microorganisms, such as nitrogen-fixing bacteria, mycorrhizal fungi, plant growth-promoting rhizobacteria (PGPR), and biocontrol bacteria. These microbes contribute to plant growth directly by supplying nutrients and modulating hormonal pathways or indirectly by enhancing plant immunity and suppressing pathogens . Notably, specific bacterial genera, such as Variovorax , have been shown to promote root elongation and even mitigate growth inhibition caused by other microbial strains . Therefore, it is of great practical significance to deeply explore the response of plant root microbial communities to agronomic measures. Rice, a staple food for nearly half of the world’s population, ranks among the three most important global food crops . Nitrogen (N) is a crucial nutrient for plant growth, significantly influencing the soil microenvironment and crop productivity. N fertilizer application has become a fundamental agronomic practice to boost rice yields, but optimizing its use is critical for achieving a balance between food security and environmental sustainability . Research indicates that N fertilizer profoundly affects the microbial diversity and composition of plant root systems. For instance, Wang et al. reported that rhizosphere bacterial diversity in Bothriochloa ischaemum increased with low to moderate N fertilizer application but decreased or plateaued at high N levels. Similarly, Fan et al. found that long-term N addition reduced the diversity of wheat root-associated bacteria. Chen et al. observed that optimized N application in rice rhizospheres decreased specific denitrifying bacteria while increasing nitrifying bacteria, accompanied by higher Shannon, Pielou, and Simpson indices. Additionally, there was an increase in the relative abundance of phyla such as Proteobacteria and Actinobacteria, while Firmicutes and Bacteroidetes declined under efficient N application. At present, most previous studies have only studied a certain growth stage of a crop or only investigated the response of rhizosphere microorganisms to nitrogen application, ignoring the dynamic changes in microorganisms in the whole growth cycle of the crop and in different regions of the root system. It has been demonstrated that the microbial community composition of a plant root system dynamically adjusts with the stage of plant growth and development, and plants will recruit different rhizosphere microorganisms at different growth stages to meet their moderate functional requirements. Monteiro et al. found that the community structure of nitrogen-fixing bacteria in the rhizosphere changed significantly during the early stage of growth of Chrysopogon zizanioides (L.) Roberty, using denaturing gradient gel electrophoresis and stabilizing after 3 months of growth. Chaparro et al. found that the root microbiota of rice at the seedling stage were significantly different from those of other growth stages. Zhang et al. studied the pattern of change in the root microbiome during the entire rice reproductive period and found that the root microbiome changed gradually with the developmental period of rice and began to stabilize after entering the reproductive growth stage. The relative abundance of δ-Proteobacteria in the root increased significantly, whereas the relative abundance of β-Proteobacteria, Firmicutes, and γ-Proteobacteria declined. However, relatively few studies have been conducted on whether changes in nitrogen demand in rice at different growth stages affect the structure of a root microbial community. Therefore, it is important to study the spatial dynamics of a root microbial community and its mechanisms throughout the growth stages of a crop under different N application levels. The application of nitrogen fertilizer significantly alters the composition, diversity, and functional potential of bacterial communities in the rhizosphere and endosphere of rice plants, with these effects varying across different growth stages due to dynamic shifts in plant nutrient demands, root exudate profiles, and microenvironmental conditions. Specifically, we hypothesize that nitrogen fertilization enhances the relative abundance of nitrogen-cycling bacteria (e.g., nitrifiers and denitrifiers) in the rhizosphere, particularly during vegetative growth stages when nitrogen uptake is highest; induces distinct successional patterns in bacterial communities between the rhizosphere and endosphere across growth stages due to differences in root physiology and nutrient partitioning; and decreases overall microbial diversity in the rhizosphere at later stages of growth due to the potential nutrient imbalances and dominance of specific nitrogen-responsive taxa. In this study, we employed high-throughput sequencing technology to investigate the bacterial community structure in the rhizosphere and endosphere of rice at different growth stages under varying nitrogen (N) fertilizer levels. Using Huaidao No. 5, a conventional japonica rice variety, as the model plant, we aimed to (1) evaluate the response of rhizosphere and endosphere bacterial community structure and diversity to different N fertilizer treatments and (2) explore the dynamics of these bacterial communities across distinct growth stages of rice. The findings from this research provide a theoretical foundation for enhancing nitrogen utilization efficiency in rice through tailored fertilization strategies. By advancing our understanding of the interplay between rice growth stages, nitrogen management, and root-associated microbiomes, this work contributes valuable insights toward the sustainable development of agroecosystems. 2.1. α-Diversity of Bacterial Communities in the Rhizosphere and Endosphere of Rice The Chao1 index in the rhizosphere did not conform to a normal distribution, and the Kruskal–Wallis test of nonparametric tests was used, resulting in no significant differences between any of the treatments ( C). The rhizosphere’s Shannon index was significantly higher in the N1 treatment than in the N0 treatment at both JS and TS, and there was no significant difference at MS ( p < 0.05). Regarding growth stages, there was a significant difference between the JS and TS Shannon indices, which were not significantly different from MS ( A). As for the endosphere, both the Shannon and Chao1 indices showed a tendency of decreasing and then increasing as the growth stage progressed, and the N1 treatment was significantly higher than the N0 treatment ( B,D). 2.2. β-Diversity of Bacterial Communities in the Rhizosphere and Endosphere of Rice PCoA was used to visualize the differences in the bacterial communities . Based on the location of the root zone, the rhizosphere and endosphere bacterial communities were divided into two distinct clusters, in which the rhizosphere bacterial communities were clustered closer together and the endosphere bacterial communities were clustered farther apart. This suggests that the location of the root zone is the most important factor influencing a bacterial community. B,C revealed significant differences in bacterial communities at different growth stages of the rhizosphere and endosphere. The PERMANOVA results showed that the growth stage was the main factor in bacterial community variation, explaining 51.4% and 65.9% of the variation in rhizosphere and endosphere bacterial communities . 2.3. Compositions of Rhizosphere and Endosphere Bacterial Communities Actinobacteria, Proteobacteria, Firmicutes, Bacteroidetes, Nitrospirae, Gemmatimonadetes, Acidobacteria, and Chloroflexi were the dominant phyla in the individual treatments of the rhizosphere. Actinobacteria, Proteobacteria, Firmicutes, Bacteroidetes, Acidobacteria, Candidatus_Saccharibacteria, and Spirochaetes were the dominant phyla in the individual treatments of the endosphere ( A). The LDA results for the rhizosphere and endosphere bacterial community compositions under the N1 and N0 treatments across three growth stages were as follows. In the rhizosphere, at the jointing stage, the N1 treatment significantly enriched Actinobacteria, Proteobacteria, Bacteroidetes, Gemmatimonadetes, Acidobacteria, Verrucomicrobia, Steroidobacter, Nocardioides, Blastococcus, Chlorobi, and Anaeromyxobacter . Conversely, the N0 treatment was enriched in Bradyrhizobium, Sporacetigenium, Rhizobium, Pseudomonas, Bacillus, Clostridium, Chryseobacterium , and Firmicutes ( A). At the tasseling stage, N1 was enriched in Proteobacteria, Nitrospirae, Bacteroidetes, Verrucomicrobia, Acidobacteria, Gemmatimonadetes, and Chloroflexi, while N0 showed enrichment in Mycobacterium, Agromyces, Chryseobacterium, Firmicutes, Nocardioides, and Actinobacteria ( B). At the maturity stage, N1 was enriched in Bacteroidetes, Adhaeribacter , Gemmatimonadetes, Acidobacteria, and Massilia . N0, on the other hand, was significantly enriched in Actinobacteria ( C). In the endosphere, at the jointing stage, N1 treatment led to enrichment in Proteobacteria, Bacteroidetes, Devosia, Stigmatella, Spirochaetes, Acidobacteria, and Anaeromyxobacter , while N0 was enriched in Methylosinus, Bradyrhizobium, Microbacterium, Bosea, Hyphomicrobium, Pleomorphomonas, Rhizobium , and Actinobacteria ( D). At the tasseling stage, N1 was enriched with Chryseobacterium , Bacteroidetes, Chitinophaga, Devosia , Candidatus_Saccharibacteria, Sphingomonas, and Asticcacaulis . N0 treatment favored Firmicutes, Microbacterium, Sterolibacterium, Pleomorphomonas, Mycobacterium, Bradyrhizobium, Methylosinus, Hyphomicrobium , and Rhizobium ( E). At the maturity stage, N1 treatment enriched Actinobacteria, Mycobacterium , Candidatus_Saccharibacteria, Niastella, Devosia, Streptomyces , Sphingomonas , Caulobacter , Clostridium , Chloroflexi, and Sphingobium , whereas N0 treatment enriched Brevundimonas, Sterolibacterium, Rhizomicrobium, Methylosinus, Pleomorphomonas , Hyphomicrobium , Rhizobium, Bacteroidetes, Chryseobacterium , and Proteobacteria ( F). These results illustrate the distinct bacterial taxa recruited by rice in the rhizosphere and endosphere at different growth stages, with compositions varying significantly between the N1 and N0 treatments, highlighting the impact of nitrogen fertilization on microbial community assembly. 2.4. Effects on the Abundance of Functional Genes for Nitrogen Metabolism N1 treatment significantly enhanced the total abundance of JS and TS functional genes in the rhizosphere, but significantly reduced the total abundance of TS and MSfunctional genes in the endosphere . LDA analysis showed that the N1 treatment significantly altered the abundance of individual nitrogen metabolism genes in both the rhizosphere and endosphere. The endosphere exhibited a higher abundance of these functional genes compared to the rhizosphere as the growth stage progressed. In the rhizosphere, at the jointing stage, N1 significantly enriched glnB and nifZ , while N0 enriched fixK, nac, ntrB, nifU , and glnK. At the tasseling stage, N1 was enriched in ntrX, ntrY , and ptsN , while N0 was enriched in nac, ntrB, and nifU . At the maturity stage, N1 was enriched in glnK , and N0 in nifN . In the endosphere, at the jointing stage, N1 treatment significantly enriched glnB, ptsN , and ntrC , while N0 enriched fixK, glnK, nac, ntrY, and ntrX . At the tasseling stage, N1 showed significant enrichment in glnB, nifU, glnK, ptsN, ntrB, and ntrC , whereas N0 was enriched in a broad range of nitrogen fixation genes, including nac, nifK, nifD, nifB, nifE, nifW, nifQ, nifH, nifN, nifT, nifX, nifZ, and nixK . At the maturity stage, N1 was enriched in nifU, glnB, ntrC, ptsN, ntrB, and glnK , while N0 was enriched in ntrY, nifD, nifQ, nifB, ntrX, nifK, nifN, nifX, nifE, nifT, nifW, and nifZ . These results indicate that the N1 and N0 treatments differentially affected specific nitrogen metabolism genes across growth stages, particularly in the endosphere, where the abundance of functional genes was higher compared to the rhizosphere (LDA > 3, p < 0.05) . The Chao1 index in the rhizosphere did not conform to a normal distribution, and the Kruskal–Wallis test of nonparametric tests was used, resulting in no significant differences between any of the treatments ( C). The rhizosphere’s Shannon index was significantly higher in the N1 treatment than in the N0 treatment at both JS and TS, and there was no significant difference at MS ( p < 0.05). Regarding growth stages, there was a significant difference between the JS and TS Shannon indices, which were not significantly different from MS ( A). As for the endosphere, both the Shannon and Chao1 indices showed a tendency of decreasing and then increasing as the growth stage progressed, and the N1 treatment was significantly higher than the N0 treatment ( B,D). PCoA was used to visualize the differences in the bacterial communities . Based on the location of the root zone, the rhizosphere and endosphere bacterial communities were divided into two distinct clusters, in which the rhizosphere bacterial communities were clustered closer together and the endosphere bacterial communities were clustered farther apart. This suggests that the location of the root zone is the most important factor influencing a bacterial community. B,C revealed significant differences in bacterial communities at different growth stages of the rhizosphere and endosphere. The PERMANOVA results showed that the growth stage was the main factor in bacterial community variation, explaining 51.4% and 65.9% of the variation in rhizosphere and endosphere bacterial communities . Actinobacteria, Proteobacteria, Firmicutes, Bacteroidetes, Nitrospirae, Gemmatimonadetes, Acidobacteria, and Chloroflexi were the dominant phyla in the individual treatments of the rhizosphere. Actinobacteria, Proteobacteria, Firmicutes, Bacteroidetes, Acidobacteria, Candidatus_Saccharibacteria, and Spirochaetes were the dominant phyla in the individual treatments of the endosphere ( A). The LDA results for the rhizosphere and endosphere bacterial community compositions under the N1 and N0 treatments across three growth stages were as follows. In the rhizosphere, at the jointing stage, the N1 treatment significantly enriched Actinobacteria, Proteobacteria, Bacteroidetes, Gemmatimonadetes, Acidobacteria, Verrucomicrobia, Steroidobacter, Nocardioides, Blastococcus, Chlorobi, and Anaeromyxobacter . Conversely, the N0 treatment was enriched in Bradyrhizobium, Sporacetigenium, Rhizobium, Pseudomonas, Bacillus, Clostridium, Chryseobacterium , and Firmicutes ( A). At the tasseling stage, N1 was enriched in Proteobacteria, Nitrospirae, Bacteroidetes, Verrucomicrobia, Acidobacteria, Gemmatimonadetes, and Chloroflexi, while N0 showed enrichment in Mycobacterium, Agromyces, Chryseobacterium, Firmicutes, Nocardioides, and Actinobacteria ( B). At the maturity stage, N1 was enriched in Bacteroidetes, Adhaeribacter , Gemmatimonadetes, Acidobacteria, and Massilia . N0, on the other hand, was significantly enriched in Actinobacteria ( C). In the endosphere, at the jointing stage, N1 treatment led to enrichment in Proteobacteria, Bacteroidetes, Devosia, Stigmatella, Spirochaetes, Acidobacteria, and Anaeromyxobacter , while N0 was enriched in Methylosinus, Bradyrhizobium, Microbacterium, Bosea, Hyphomicrobium, Pleomorphomonas, Rhizobium , and Actinobacteria ( D). At the tasseling stage, N1 was enriched with Chryseobacterium , Bacteroidetes, Chitinophaga, Devosia , Candidatus_Saccharibacteria, Sphingomonas, and Asticcacaulis . N0 treatment favored Firmicutes, Microbacterium, Sterolibacterium, Pleomorphomonas, Mycobacterium, Bradyrhizobium, Methylosinus, Hyphomicrobium , and Rhizobium ( E). At the maturity stage, N1 treatment enriched Actinobacteria, Mycobacterium , Candidatus_Saccharibacteria, Niastella, Devosia, Streptomyces , Sphingomonas , Caulobacter , Clostridium , Chloroflexi, and Sphingobium , whereas N0 treatment enriched Brevundimonas, Sterolibacterium, Rhizomicrobium, Methylosinus, Pleomorphomonas , Hyphomicrobium , Rhizobium, Bacteroidetes, Chryseobacterium , and Proteobacteria ( F). These results illustrate the distinct bacterial taxa recruited by rice in the rhizosphere and endosphere at different growth stages, with compositions varying significantly between the N1 and N0 treatments, highlighting the impact of nitrogen fertilization on microbial community assembly. N1 treatment significantly enhanced the total abundance of JS and TS functional genes in the rhizosphere, but significantly reduced the total abundance of TS and MSfunctional genes in the endosphere . LDA analysis showed that the N1 treatment significantly altered the abundance of individual nitrogen metabolism genes in both the rhizosphere and endosphere. The endosphere exhibited a higher abundance of these functional genes compared to the rhizosphere as the growth stage progressed. In the rhizosphere, at the jointing stage, N1 significantly enriched glnB and nifZ , while N0 enriched fixK, nac, ntrB, nifU , and glnK. At the tasseling stage, N1 was enriched in ntrX, ntrY , and ptsN , while N0 was enriched in nac, ntrB, and nifU . At the maturity stage, N1 was enriched in glnK , and N0 in nifN . In the endosphere, at the jointing stage, N1 treatment significantly enriched glnB, ptsN , and ntrC , while N0 enriched fixK, glnK, nac, ntrY, and ntrX . At the tasseling stage, N1 showed significant enrichment in glnB, nifU, glnK, ptsN, ntrB, and ntrC , whereas N0 was enriched in a broad range of nitrogen fixation genes, including nac, nifK, nifD, nifB, nifE, nifW, nifQ, nifH, nifN, nifT, nifX, nifZ, and nixK . At the maturity stage, N1 was enriched in nifU, glnB, ntrC, ptsN, ntrB, and glnK , while N0 was enriched in ntrY, nifD, nifQ, nifB, ntrX, nifK, nifN, nifX, nifE, nifT, nifW, and nifZ . These results indicate that the N1 and N0 treatments differentially affected specific nitrogen metabolism genes across growth stages, particularly in the endosphere, where the abundance of functional genes was higher compared to the rhizosphere (LDA > 3, p < 0.05) . 3.1. The Growth Stage Is the Primary Factor Affecting Rhizosphere and Endosphere Bacterial Communities The composition and diversity of microbial communities in rice roots differ significantly between spatial structures, with the root zone location being a key factor in shaping the bacterial diversity of Huaidao No. 5. The rhizosphere’s bacterial diversity is consistently higher than that of the endosphere. This disparity is largely due to the “gate valve” role played by a root surface, which selectively enriches bacterial taxa from the rhizosphere into the endosphere, resulting in a distinct microbial community structure between these two regions . These findings align with those reported by Jiang et al. , Monteiro et al. , and Xu et al. . Diversity metrics, such as the Shannon and Chao1 indices, followed a trend of decreasing and then increasing as the growth stage progressed ( B,D). The rhizosphere, being more dynamic and directly exposed to a soil environment, shows rapid responses to nitrogen (N) application due to shifts in its chemistry, including nitrogen, phosphate, and organic matter levels. This exposure leads to a quicker response of sensitive species in the rhizosphere community to N inputs . In contrast, the endosphere provides a relatively isolated and stabilized environment, making it less sensitive to external environmental changes. Rhizosphere bacteria, being directly connected to external soil, have greater access to nutrients and are generally more adaptable. While nitrogen application increases available nitrogen sources for rhizosphere bacteria, endosphere bacteria primarily rely on nitrogen assimilated by the host plant, resulting in a comparatively weaker response to nitrogen inputs . The PCoA analysis results highlighted root zone location as the most significant factor influencing the rice root bacterial community structure ( A). When examining the effects of nitrogen application and growth stage on rhizosphere and endosphere communities, PERMANOVA analysis further identified growth stage as the predominant factor in structuring these bacterial communities , with notable shifts occurring as the growth stage advanced. These findings align with studies by Edwards et al. and Zhang et al. , who observed similar patterns of microbiome dynamics across the reproductive period in rice. Early in plant growth, root secretions attract a large influx of soil microorganisms to the rhizosphere, with some colonizing the plant interior. This leads to high variability in the microbial community during the plant’s nutrient growth stage. However, as rice enters the reproductive growth stage, the rhizosphere microbial community stabilizes, likely due to fewer root exudate-induced fluctuations . In Huaidao No. 5 rice, Actinobacteria, Proteobacteria, Firmicutes, Bacteroidetes, and Acidobacteria were identified as dominant phyla in both the rhizosphere and endosphere. This finding is consistent with previous studies on the rice rhizosphere microbiome . Proteobacteria, Bacteroidetes, and Firmicutes are generally regarded as eutrophic taxa, capable of rapid growth in environments enriched with labile carbon . The LDA results indicated that rice selectively recruits various beneficial bacterial genera at different growth stages . Key genera such as Devosia, Anaeromyxobacter , and Clostridium possess nitrogen-fixing capabilities and play vital roles in nitrogen cycling within soil and root systems . Other genera, including Devosia, Massilia, Chryseobacterium, Sphingomonas, Streptomyces, Caulobacter, Asticcacaulis, and Sphingobium , are known as plant growth-promoting (PGPR) rhizobacteria . These PGPR rhizobacteria enhance plant growth and nutrient uptake and boost stress tolerance through mechanisms like nitrogen fixation, phosphorus solubilization, antibiotic production, and phytohormone secretion. As the rice growth stage progresses, the biomarker type and relative abundance of microbes in the rhizosphere and endosphere also shift significantly. Notably, Devosia emerged as a persistent biomarker across growth stages in the endosphere under N1 treatment. This genus, classified as α-Proteobacteria, effectively colonizes plant roots and exhibits multifunctional plant growth-promoting attributes. Devosia harbors nitrogen-fixing genes, like nifH, and secretes phytohormones (e.g., indoleacetic acid, IAA), solubilizes minerals like phosphorus, and supports overall plant growth through diverse mechanisms . In summary, as rice enters different stages of its reproductive period, its root-associated microbiota dynamically shift to recruit beneficial bacteria, thereby maintaining a supportive root environment and promoting plant health and growth. 3.2. Functional Gene Abundance of Nitrogen Metabolism at Different Growth Stages in Response to Nitrogen Fertilization Functional genes involved in nitrogen metabolism in agricultural soils provide critical insights into the species and abundance of functional microorganisms that drive specific nitrogen transformation pathways . The results of LDA analysis showed that the types or abundances of functional genes for nitrogen metabolism were significantly higher in the endosphere than the rhizosphere as the growth stage progressed, and this difference was particularly significant at maturity. This may be because the heightened demand for nutrients in mature rice plants spurs active nitrogen metabolism within the roots, while the enclosed endosphere environment intensifies the expression of nitrogen metabolism genes. Conversely, rhizosphere microorganisms may compete with or regulate nitrogen utilization within the roots, leading to a differential effect on gene abundance . Among these functional genes, glnB and ptsN emerged as biomarkers throughout the entire growth stage in the endosphere. The gene glnB , which encodes the PII protein, displayed the highest relative abundance across treatments. PII proteins play an essential role in sensing intracellular nitrogen levels (by monitoring metabolites like α-ketoglutarate, ATP, and ADP) and respond by initiating nitrogen fixation under low-nitrogen conditions, thereby enabling bacteria to utilize atmospheric nitrogen . The gene ptsN encodes the EIIA ( Ntr ) protein, which is also critical for nitrogen metabolism and interacts with PII proteins from glnB to regulate nitrogen-related gene expression . Notably, the relative abundance of ptsN remained fairly stable across growth stages but accounted for a large proportion of all functional genes . The results of this study align with previous research, such as that by Xu et al. , who reported that nitrogen application did not significantly impact the relative abundance of the nitrogen-fixing gene nifH . In this study, nitrogen application at 270 kg N ha −1 similarly showed no substantial effect on the abundance of nitrogen metabolism genes, though the growth stage did influence the types and abundances of these genes. By contrast, Li et al. and Li et al. reported an increase in nifH expression with nitrogen addition. These differing findings may be attributed to background nitrogen levels in the soil; when nitrogen-fixing bacteria can readily absorb soil nitrogen, their need to fix atmospheric nitrogen decreases, leading to the downregulated expression of nitrogen fixation genes . Moreover, nitrogen application alters the composition of rhizosphere microbial communities, and external factors such as temperature, moisture, and soil conditions further influence nitrogen utilization efficiency . These environmental variables impact the structure of rhizosphere microbial communities and, consequently, the expression of nitrogen metabolism-related genes, highlighting the complex interplay between the growth stage, soil conditions, and microbial nitrogen cycling in rice. The composition and diversity of microbial communities in rice roots differ significantly between spatial structures, with the root zone location being a key factor in shaping the bacterial diversity of Huaidao No. 5. The rhizosphere’s bacterial diversity is consistently higher than that of the endosphere. This disparity is largely due to the “gate valve” role played by a root surface, which selectively enriches bacterial taxa from the rhizosphere into the endosphere, resulting in a distinct microbial community structure between these two regions . These findings align with those reported by Jiang et al. , Monteiro et al. , and Xu et al. . Diversity metrics, such as the Shannon and Chao1 indices, followed a trend of decreasing and then increasing as the growth stage progressed ( B,D). The rhizosphere, being more dynamic and directly exposed to a soil environment, shows rapid responses to nitrogen (N) application due to shifts in its chemistry, including nitrogen, phosphate, and organic matter levels. This exposure leads to a quicker response of sensitive species in the rhizosphere community to N inputs . In contrast, the endosphere provides a relatively isolated and stabilized environment, making it less sensitive to external environmental changes. Rhizosphere bacteria, being directly connected to external soil, have greater access to nutrients and are generally more adaptable. While nitrogen application increases available nitrogen sources for rhizosphere bacteria, endosphere bacteria primarily rely on nitrogen assimilated by the host plant, resulting in a comparatively weaker response to nitrogen inputs . The PCoA analysis results highlighted root zone location as the most significant factor influencing the rice root bacterial community structure ( A). When examining the effects of nitrogen application and growth stage on rhizosphere and endosphere communities, PERMANOVA analysis further identified growth stage as the predominant factor in structuring these bacterial communities , with notable shifts occurring as the growth stage advanced. These findings align with studies by Edwards et al. and Zhang et al. , who observed similar patterns of microbiome dynamics across the reproductive period in rice. Early in plant growth, root secretions attract a large influx of soil microorganisms to the rhizosphere, with some colonizing the plant interior. This leads to high variability in the microbial community during the plant’s nutrient growth stage. However, as rice enters the reproductive growth stage, the rhizosphere microbial community stabilizes, likely due to fewer root exudate-induced fluctuations . In Huaidao No. 5 rice, Actinobacteria, Proteobacteria, Firmicutes, Bacteroidetes, and Acidobacteria were identified as dominant phyla in both the rhizosphere and endosphere. This finding is consistent with previous studies on the rice rhizosphere microbiome . Proteobacteria, Bacteroidetes, and Firmicutes are generally regarded as eutrophic taxa, capable of rapid growth in environments enriched with labile carbon . The LDA results indicated that rice selectively recruits various beneficial bacterial genera at different growth stages . Key genera such as Devosia, Anaeromyxobacter , and Clostridium possess nitrogen-fixing capabilities and play vital roles in nitrogen cycling within soil and root systems . Other genera, including Devosia, Massilia, Chryseobacterium, Sphingomonas, Streptomyces, Caulobacter, Asticcacaulis, and Sphingobium , are known as plant growth-promoting (PGPR) rhizobacteria . These PGPR rhizobacteria enhance plant growth and nutrient uptake and boost stress tolerance through mechanisms like nitrogen fixation, phosphorus solubilization, antibiotic production, and phytohormone secretion. As the rice growth stage progresses, the biomarker type and relative abundance of microbes in the rhizosphere and endosphere also shift significantly. Notably, Devosia emerged as a persistent biomarker across growth stages in the endosphere under N1 treatment. This genus, classified as α-Proteobacteria, effectively colonizes plant roots and exhibits multifunctional plant growth-promoting attributes. Devosia harbors nitrogen-fixing genes, like nifH, and secretes phytohormones (e.g., indoleacetic acid, IAA), solubilizes minerals like phosphorus, and supports overall plant growth through diverse mechanisms . In summary, as rice enters different stages of its reproductive period, its root-associated microbiota dynamically shift to recruit beneficial bacteria, thereby maintaining a supportive root environment and promoting plant health and growth. Functional genes involved in nitrogen metabolism in agricultural soils provide critical insights into the species and abundance of functional microorganisms that drive specific nitrogen transformation pathways . The results of LDA analysis showed that the types or abundances of functional genes for nitrogen metabolism were significantly higher in the endosphere than the rhizosphere as the growth stage progressed, and this difference was particularly significant at maturity. This may be because the heightened demand for nutrients in mature rice plants spurs active nitrogen metabolism within the roots, while the enclosed endosphere environment intensifies the expression of nitrogen metabolism genes. Conversely, rhizosphere microorganisms may compete with or regulate nitrogen utilization within the roots, leading to a differential effect on gene abundance . Among these functional genes, glnB and ptsN emerged as biomarkers throughout the entire growth stage in the endosphere. The gene glnB , which encodes the PII protein, displayed the highest relative abundance across treatments. PII proteins play an essential role in sensing intracellular nitrogen levels (by monitoring metabolites like α-ketoglutarate, ATP, and ADP) and respond by initiating nitrogen fixation under low-nitrogen conditions, thereby enabling bacteria to utilize atmospheric nitrogen . The gene ptsN encodes the EIIA ( Ntr ) protein, which is also critical for nitrogen metabolism and interacts with PII proteins from glnB to regulate nitrogen-related gene expression . Notably, the relative abundance of ptsN remained fairly stable across growth stages but accounted for a large proportion of all functional genes . The results of this study align with previous research, such as that by Xu et al. , who reported that nitrogen application did not significantly impact the relative abundance of the nitrogen-fixing gene nifH . In this study, nitrogen application at 270 kg N ha −1 similarly showed no substantial effect on the abundance of nitrogen metabolism genes, though the growth stage did influence the types and abundances of these genes. By contrast, Li et al. and Li et al. reported an increase in nifH expression with nitrogen addition. These differing findings may be attributed to background nitrogen levels in the soil; when nitrogen-fixing bacteria can readily absorb soil nitrogen, their need to fix atmospheric nitrogen decreases, leading to the downregulated expression of nitrogen fixation genes . Moreover, nitrogen application alters the composition of rhizosphere microbial communities, and external factors such as temperature, moisture, and soil conditions further influence nitrogen utilization efficiency . These environmental variables impact the structure of rhizosphere microbial communities and, consequently, the expression of nitrogen metabolism-related genes, highlighting the complex interplay between the growth stage, soil conditions, and microbial nitrogen cycling in rice. 4.1. Experimental Design The field experiment was conducted in ShantouTown, Yangzhou City, Jiangsu Province (119°52′N, 32°31′E), using the rice variety Huaidao No. 5. Two nitrogen (N) application treatments were established as the following: no N application (N0) and conventional N application (N1, 270 kg N ha −1 ). The experiment utilized a randomized block design with three replicates per treatment. Each plot measured 9 m 2 (3 m × 3 m), with a transplanting density of 30 cm × 16 cm, and two seedlings per hill. In addition to nitrogen, phosphate fertilizer (P 2 O 5 , 12%) at 1000 kg ha −1 and potash fertilizer (K 2 O, 60%) at 200 kg ha −1 were uniformly applied across all plots. The experimental field was characterized by conventional soil with the following baseline physicochemical properties: pH 7.42, ammonium nitrogen (NH 4 + ) 10.7 mg/kg, nitrate nitrogen (NO 3 − ) 9.53 mg/kg, total nitrogen (TN) 1.57 g/kg, organic matter (OM) 51.02 g/kg, and electrical conductivity (EC) 0.4 mS/cm. Other field management practices, such as weeding and pest control, were performed according to standard high-yield cultivation techniques. 4.2. Rice Root Sample and Rhizosphere Soil Sample Collection Rice rhizosphere soil was collected according to the method described by Edwards et al. . Roots with 1 mm soil attached were placed in a triangular vial containing 50 mL of sterile phosphate buffer and vortexed for 10 min to clean all soil from the root surface. The soil cleaned from the roots was freeze-dried and stored as rhizosphere soil in a −80 °C freezer. Root surface decontamination of the above washed rice roots was performed with reference to the method described by Wang et al. . First, washed rice roots were placed in sterile 50 mL centrifuge tubes, and 30 mL of sterile phosphate buffer (containing 0.02% Tween 20) was added to submerge the samples. The samples were processed in an ultrasonic cleaner (SB-5200) at 40 kHz for 1 min, and the procedure was repeated five times, followed by a 5 min immersion in 2% sodium hypochlorite solution, then washing again with sterile water. The final washing solution was used as a template for PCR amplification to check whether surface sterilization was complete. After surface sterilization, the plant samples were lyophilized, pulverized, sieved, and stored at −80° C until DNA extraction. 4.3. DNA Extraction and Illumina NovaSeq Sequencing Root and soil samples were extracted using the FastDNA ® SPIN Kit for soil (MP Biomedicals, Santa Ana, CA, USA), which was described in the kit manual. The DNA of the samples was detected by 2% agarose gel electrophoresis and stored at −20 °C for subsequent analysis. Primers 799F (5′-AACMGGATTAGATACCCKG-3′) and 1193R (5′-ACGTCATCCCCACCTTCC-3′), which contained index sequences, were used to amplify the V5–V7 region of the 16S rRNA gene of soil and root bacteria . PCR amplification products were purified and sent to Shanghai Tianhao Biotechnology Co., Ltd. (Shanghai, China), for sequencing using the NovaSeq 6000 platform (Illumina, San Diego, CA, USA) with the bipartite sequencing strategy of SP-Xp (PE250). Raw sequencing data were processed using QIIME2 software (2023.5) . Firstly, the primer fragments of the sequences were excised and the unmatched primers were removed. Then, according to the default settings, the DADA2 plug-in was invoked to perform quality control, denoising, splicing , and chimera removal on the sequences to form the characteristic sequences of Amplicon Sequence Variants (ASVs) and the ASV abundance table. Species annotation was performed using a Naive Bayes classifier pre-trained on the Silva132 database, according to default parameters (confidence level 0.8). To correct for differences in bacterial community diversity due to sequencing depth, a random leveling process was performed for each sample. The leveling depth was based on the lowest sequence count in the sequenced samples. 4.4. Statistical Analysis The data were processed using Microsoft Excel 2016, and the bacterial Chao1 index and Shannon indices were calculated using QIIME2. IBM SPSS Statistics 24 was used to test the significance of the Shannon index, Chao1 index, relative abundance of bacterial composition of different treatments, as well as the effect of root zone location, nitrogen application, and fertility stage, on the relative abundances of the bacterial Shannon index, Chao1 index, and bacterial composition, and the difference was statistically significant at p < 0.05. Principal coordinate analysis (PCoA) based on the relative abundance of ASVs was performed using the “vegan” package in R4.1.1 to visualize the differences in the rhizosphere and endosphere community structure. Permutational multivariate ANOVA (PERMANOVA) based on Bray–Curtis distance matrices was performed with 999 permutations using the “vegan” package in R4.1.1 to quantify the contribution of growth stages and N fertilization to community variation. Linear discriminant analysis (LDA) was performed using ImageGP to elucidate the biomarkers with a logarithmic LDA score > 3 and p < 0.05 in bacterial communities. The field experiment was conducted in ShantouTown, Yangzhou City, Jiangsu Province (119°52′N, 32°31′E), using the rice variety Huaidao No. 5. Two nitrogen (N) application treatments were established as the following: no N application (N0) and conventional N application (N1, 270 kg N ha −1 ). The experiment utilized a randomized block design with three replicates per treatment. Each plot measured 9 m 2 (3 m × 3 m), with a transplanting density of 30 cm × 16 cm, and two seedlings per hill. In addition to nitrogen, phosphate fertilizer (P 2 O 5 , 12%) at 1000 kg ha −1 and potash fertilizer (K 2 O, 60%) at 200 kg ha −1 were uniformly applied across all plots. The experimental field was characterized by conventional soil with the following baseline physicochemical properties: pH 7.42, ammonium nitrogen (NH 4 + ) 10.7 mg/kg, nitrate nitrogen (NO 3 − ) 9.53 mg/kg, total nitrogen (TN) 1.57 g/kg, organic matter (OM) 51.02 g/kg, and electrical conductivity (EC) 0.4 mS/cm. Other field management practices, such as weeding and pest control, were performed according to standard high-yield cultivation techniques. Rice rhizosphere soil was collected according to the method described by Edwards et al. . Roots with 1 mm soil attached were placed in a triangular vial containing 50 mL of sterile phosphate buffer and vortexed for 10 min to clean all soil from the root surface. The soil cleaned from the roots was freeze-dried and stored as rhizosphere soil in a −80 °C freezer. Root surface decontamination of the above washed rice roots was performed with reference to the method described by Wang et al. . First, washed rice roots were placed in sterile 50 mL centrifuge tubes, and 30 mL of sterile phosphate buffer (containing 0.02% Tween 20) was added to submerge the samples. The samples were processed in an ultrasonic cleaner (SB-5200) at 40 kHz for 1 min, and the procedure was repeated five times, followed by a 5 min immersion in 2% sodium hypochlorite solution, then washing again with sterile water. The final washing solution was used as a template for PCR amplification to check whether surface sterilization was complete. After surface sterilization, the plant samples were lyophilized, pulverized, sieved, and stored at −80° C until DNA extraction. Root and soil samples were extracted using the FastDNA ® SPIN Kit for soil (MP Biomedicals, Santa Ana, CA, USA), which was described in the kit manual. The DNA of the samples was detected by 2% agarose gel electrophoresis and stored at −20 °C for subsequent analysis. Primers 799F (5′-AACMGGATTAGATACCCKG-3′) and 1193R (5′-ACGTCATCCCCACCTTCC-3′), which contained index sequences, were used to amplify the V5–V7 region of the 16S rRNA gene of soil and root bacteria . PCR amplification products were purified and sent to Shanghai Tianhao Biotechnology Co., Ltd. (Shanghai, China), for sequencing using the NovaSeq 6000 platform (Illumina, San Diego, CA, USA) with the bipartite sequencing strategy of SP-Xp (PE250). Raw sequencing data were processed using QIIME2 software (2023.5) . Firstly, the primer fragments of the sequences were excised and the unmatched primers were removed. Then, according to the default settings, the DADA2 plug-in was invoked to perform quality control, denoising, splicing , and chimera removal on the sequences to form the characteristic sequences of Amplicon Sequence Variants (ASVs) and the ASV abundance table. Species annotation was performed using a Naive Bayes classifier pre-trained on the Silva132 database, according to default parameters (confidence level 0.8). To correct for differences in bacterial community diversity due to sequencing depth, a random leveling process was performed for each sample. The leveling depth was based on the lowest sequence count in the sequenced samples. The data were processed using Microsoft Excel 2016, and the bacterial Chao1 index and Shannon indices were calculated using QIIME2. IBM SPSS Statistics 24 was used to test the significance of the Shannon index, Chao1 index, relative abundance of bacterial composition of different treatments, as well as the effect of root zone location, nitrogen application, and fertility stage, on the relative abundances of the bacterial Shannon index, Chao1 index, and bacterial composition, and the difference was statistically significant at p < 0.05. Principal coordinate analysis (PCoA) based on the relative abundance of ASVs was performed using the “vegan” package in R4.1.1 to visualize the differences in the rhizosphere and endosphere community structure. Permutational multivariate ANOVA (PERMANOVA) based on Bray–Curtis distance matrices was performed with 999 permutations using the “vegan” package in R4.1.1 to quantify the contribution of growth stages and N fertilization to community variation. Linear discriminant analysis (LDA) was performed using ImageGP to elucidate the biomarkers with a logarithmic LDA score > 3 and p < 0.05 in bacterial communities. In this study, we analyzed the structure and diversity of rhizosphere and endosphere bacterial communities in Huaidao No. 5 under varying nitrogen application rates and growth stages, exploring the effects on the abundances of functional genes associated with nitrogen metabolism. Root zone location was identified as the key factor influencing bacterial diversity and composition in the rice root, with nitrogen application being the primary factor affecting rhizosphere bacterial diversity, while the growth stage more significantly impacted endosphere bacterial community diversity. Additionally, the growth stage was the main driver of variation in both the rhizosphere and endosphere bacterial communities in Huaidao No. 5. The dominant bacterial phyla in both the rhizosphere and endosphere of Huaidao No. 5 were Actinobacteria, Proteobacteria, Firmicutes, Bacteroidetes, and Acidobacteria. Rice at different growth stages recruited distinct beneficial bacteria within the root, which varied by stage. The relative abundance of nitrogen metabolism functional genes in the rhizosphere and endosphere was not significantly affected by nitrogen application at 270 kg N ha −1 . However, the types and relative abundances of these genes changed as the growth stage progressed, with glnB, glnK, ptsN, ntrBC, and ntrXY showing higher relative abundances, playing essential roles in regulating nitrogen metabolism at various growth stages. These results are valuable for furthering our understanding of the interactions among a plant growth stage, nitrogen fertilizer application, and root microbiome. Future research directions should involve selecting different rice varieties, setting different nitrogen application rates, and sampling them at whole growth stages. Using metagenomics and metaproteomics to explore functional shifts in microbial communities under different nitrogen regimes could help identify key genes and pathways involved in nitrogen fixation, nitrification, and denitrification. Functional analysis may also reveal potential biomarkers for nitrogen use efficiency (NUE) in rice systems. |
Changes in soft tissue dimensions following horizontal guided bone regeneration with a split-thickness flap design – evaluation of 8 cases with a digital method | 88ed74a2-d5c3-4c05-a24d-1a6051c5dfcd | 11438005 | Dentistry[mh] | Over the course of the past 20 years guided bone regeneration (GBR) has become well-established method for reconstructing lost alveolar structures . Literature data has also shown that long-term success of implants placed into augmented sites does not differ significantly to those placed into native bone . Even though various aspects of GBR are well researched, there is less information available regarding soft tissue changes following ridge augmentation procedures. In the clinical practice reduction of keratinized tissue width- and thickness is a well-known occurrence, consequently, soft tissue augmentation and re-establishment of keratinized mucosa width following alveolar ridge reconstructions is often required. Yet, there is only limited literature data available reporting on soft tissue alterations and the factors that influence morphological changes in keratinized tissues after ridge augmentation procedures. It can be hypothesized that the reduction of the keratinized mucosa and distortion of the vestibule following ridge augmentation procedures occur due to extensive flap mobilization. In order to avoid severe vestibular distortion, Windisch et al. have suggested the application of a split-thickness flap design - instead of a conventional full-thickness flap – to be used in conjunction with GBR. The lack of adequate peri-implant soft tissue dimensions – i.e., the supracrestal soft tissue thickness and the width of peri-implant keratinized mucosa (PIKM) – results in reduced long-term implant success rates . As early as 1996 , it was shown in a preclinical study that the reduction of supracrestal soft tissue thickness below 2 mm led to a marginal peri-implant bone loss. In 2015, Linkevicius et al. showed that thick (> 2 mm) supracrestal soft tissue dimensions resulted in significantly less marginal bone loss. While a narrow band of PIKM facilitates plaque accumulation, increasing the rate of peri-implant mucositis and eventually peri-implantitis . To avoid peri-implant soft tissue related complications, the surgical reconstruction of the keratinized mucosa at the edentulous ridge after augmentation is often necessary . Different methods can be found in the literature for the measurement of supracrestal soft tissues. Width of the supracrestal soft tissues can be easily assess clinically with the use of periodontal probes or calipers. Supracrestal soft tissue thickness on the other hand is most commonly measured by the means of transmucosal probing (bone sounding) which is a relatively invasive method . Alternatively, ultrasonographic devices or the superimposition of cone-beam computed tomography (CBCT) scans and intraoral scans (IOSs) can be utilized to assess supracrestal soft tissue thickness in a non-invasive way . A previous study by Di Raimondo et al. , analyzed soft tissue changes occurring after simultaneous horizontal GBR and implant placement utilizing digitalized casts. After superimposition of baseline, 4-month and 12-month casts authors performed horizontal cross sectional linear measurements. Authors have concluded that soft tissue contours have increased after horizontal GBR. However, measurements represent the cumulative change in alveolar ridge dimensions (hard tissue and soft tissue changes) rather than solely analyzing soft tissue alterations. In a previous study, hard tissue changes following horizontal GBR have been investigated utilizing a 3D methodology . However, concomitant soft tissue alterations were not analyzed. With the combination of 3D reconstructed CBCT scans and IOSs, digital hybrid models can be generated . Digital hybrid models depict all relevant anatomical structures –i.e., teeth, alveolar bone, soft tissues – separately, allowing to analyze hard and soft tissue changes independently from one another. Hence the aim of our study was to evaluate soft tissue dimensional changes following horizontal GBR utilizing 3D digital hybrid models.
Study design This prospective single-center case series included a total of 8 surgical sites in the posterior mandible. Data included in the current paper are derived from a single group of a larger ongoing randomized clinical trial. The current pilot study investigated the supracrestal soft tissue alterations following horizontal GBR, before second stage dental implant placement. The study followed the PROCESS guidelines checklist (originally published in 2016, revised in 2018) . The study protocol was approved by the Semmelweis University Regional and Institutional Committee of Science and Research Ethics (Approval Number: SE RKEB 145/2018) and the U.S. National Library of Medicine ( www.clinicaltrials.gov ; trial registration number: NCT05538715; registration date: 09/09/2022). The study was conducted in full accordance with the Declaration of Helsinki of 1975, revised in 2013 . Surgical interventions were performed with the understanding and written informed consent of every participant. Patient selection Participants enrolled in the study, were treated at the Department of Periodontology, Semmelweis University. In the included cases, horizontal ridge augmentation in the posterior mandible was necessary for a prosthetically driven implant placement. Baseline defect morphologies were classified according to the HVC (horizontal, vertical, combined) ridge deficiency classification . Exclusion criteria were: (i) presence of general medical conditions contraindicating surgical treatment; (ii) age < 20 years, (iii) smoking; (iv) untreated periodontitis with high levels of residual inflammation (full mouth bleeding score > 25%); (v) inadequate oral hygiene (full mouth plaque score > 25%) and (iv) vertical or combined alveolar ridge deficiencies. CBCT images were taken with a Planmeca ProMax 3D Plus and a Planmeca Viso G7 device (Planmeca Oy, Helsinki, Finland) (FOV: 10 × 10 cm, voxel size: 150 μm) prior to and 6 months following the augmentation procedure. Intraoral scans were acquired with Planmeca Emerald S (Planmeca Oy, Helsinki, Finland) at baseline and 6-month follow-up. Surgical procedure Steps of the surgical protocol have been described in detail elsewhere . Briefly, first a mid-crestal incision was made, thereafter a double layer (mucosa and periosteum), split-thickness flap was raised on the buccal aspect. While on the lingual aspect a full-thickness mucoperiosteal flap was raised. To maintain the blood supply of the periosteum, vertical releasing incisions were avoided. Autogenous bone chips were harvested by a single use bone collector device (Safescraper Twist, Meta, Reggio Emilia, Italy) locally without the preparation of a second surgical site and were mixed with a bovine-derived xenograft (Bio_Oss, Gesitlich, Wolhusen, Germany) in a 1:1 ratio. A resorbable collagen membrane (BioGide, Geistlich, Wolhusan, Germany) was shaped and fixated on the lingual aspect with titanium micro-screws (Pro-fix, Osteogenics, Lubbock, USA). Thereafter, the composite graft was compacted on the residual ridge and the collagen membrane was folded over and subsequently fixed with titanium pins (Ustomed, Tuttlingen, Germany). Double-layer wound closure was carried out, first the buccal periosteal layer was sutured to the lingual flap with a 3–0 expanded polytetrafluoroethylene (e-PTFE) suturing material (Cytoplast, Osteogenics, Lubbock, USA). The buccal mucosal layer was also sutured to the lingual flap with horizontal mattress sutures and single interrupted sutures using a 4–0 non-resorbable e-PTFE suturing material (Cytoplast, Osteogenics, Lubbock, USA). Baseline and 6-month follow-up of keratinized soft tissue situations are visible in Fig. . Digital data processing Baseline and 6-month follow-up CBCT scans were segmented in an open-source radiographic image processing software (3D Slicer, www.slicer.org ) using a dedicated semi-automatic image segmentation method . Anatomical structures (teeth, alveolar bone and nerves) were segmented separately, constituting individual components of the model. The output of image segmentation is a 3D virtual model of the dento-alveolar hard tissues. Following segmentation, standard tessellation language (.stl) files of IOSs were superimposed with segmented 3D models, using identical landmark registration. Corresponding mark-up points were placed on fixed anatomical landmarks (cusps or incisor edges of teeth) (Fig. ). Alignment of the two models were inspected by two individual investigators. Registration of baseline- and follow-up data Following digital data processing, baseline and 6-month follow-up data were spatially registered (both CBCT scans and IOS). Utilizing an intensity-based medical image registration algorithm (Elastix) baseline and follow-up CBCT scans were aligned with a linear transformation. The same transformation algorithm was applied to register the IOSs. Outcome measures Vertical- and horizontal supracrestal soft tissue changes The primary outcome of the study was to evaluate the vertical supracrestal soft tissue changes at three measurement planes. Coronal, sagittal and axial planes of CBCT scans were oriented in a manner that the sagittal plane became parallel and the coronal plane became perpendicular to the alveolar ridge. In the coronal view window, three planes were selected for measurements (mesial plane: at the line of the most mesial titanium pin used for GBR membrane fixation, distal plane: at the line of the most distal titanium pin, middle plane: halfway between the mesial and distal planes). Supracrestal soft tissue height was measured between the most coronal point of the alveolar crest and the most coronal point of the keratinized alveolar mucosa on both the baseline and the follow-up data. Additionally to vertical soft tissue dimensions, the bucco-lingual width of the supracrestal soft tissues was measured. Linear measurements were made perpendicular to the long axis of the alveolar crest at the level of the buccal MGJ (visible on IOS) (Fig. ). Horizontal and vertical shift of the supracrestal soft tissues A measuring grid with a 1 mm interval was overlayed on the previously mentioned measurement planes (mesial, middle, distal). Two reference points were placed midcrestally at the most coronal level of the keratinized tissue crest (KTC) both on the baseline and 6-month follow-up models (ST-pre; ST-post). Two additional reference points were placed at the top of baseline and follow-up edentulous alveolar crests (AC-pre; AC-post). Horizontal distances between the ST-pre; ST-post points and the corresponding AC reference point were measured to assess the occasional shift of KTC following horizontal GBR (Fig. ). Statistical analysis Descriptive statistics were used to describe the variables, data were expressed as means ± standard deviations. Statistical differences between baseline and 6-month follow-up supracrestal soft tissue dimensions were analyzed using inferential statistics with a significance level of α = 0.05. Normality of the previously mentioned variables was checked with the Shapiro-Wilk test. Levene’s test was used to check the homogeneity of the variances. Data were found to be normally distributed, and the homogeneity assumption of the variances were met. The continuous variables between subgroups were compared with parametric statistics. The paired sample t-test was utilized to evaluate statistical differences for each variable at different time points. The statistical analysis was performed using the STATA 18 software package (StataCorp LLC, College Station, TX, USA).
This prospective single-center case series included a total of 8 surgical sites in the posterior mandible. Data included in the current paper are derived from a single group of a larger ongoing randomized clinical trial. The current pilot study investigated the supracrestal soft tissue alterations following horizontal GBR, before second stage dental implant placement. The study followed the PROCESS guidelines checklist (originally published in 2016, revised in 2018) . The study protocol was approved by the Semmelweis University Regional and Institutional Committee of Science and Research Ethics (Approval Number: SE RKEB 145/2018) and the U.S. National Library of Medicine ( www.clinicaltrials.gov ; trial registration number: NCT05538715; registration date: 09/09/2022). The study was conducted in full accordance with the Declaration of Helsinki of 1975, revised in 2013 . Surgical interventions were performed with the understanding and written informed consent of every participant.
Participants enrolled in the study, were treated at the Department of Periodontology, Semmelweis University. In the included cases, horizontal ridge augmentation in the posterior mandible was necessary for a prosthetically driven implant placement. Baseline defect morphologies were classified according to the HVC (horizontal, vertical, combined) ridge deficiency classification . Exclusion criteria were: (i) presence of general medical conditions contraindicating surgical treatment; (ii) age < 20 years, (iii) smoking; (iv) untreated periodontitis with high levels of residual inflammation (full mouth bleeding score > 25%); (v) inadequate oral hygiene (full mouth plaque score > 25%) and (iv) vertical or combined alveolar ridge deficiencies. CBCT images were taken with a Planmeca ProMax 3D Plus and a Planmeca Viso G7 device (Planmeca Oy, Helsinki, Finland) (FOV: 10 × 10 cm, voxel size: 150 μm) prior to and 6 months following the augmentation procedure. Intraoral scans were acquired with Planmeca Emerald S (Planmeca Oy, Helsinki, Finland) at baseline and 6-month follow-up.
Steps of the surgical protocol have been described in detail elsewhere . Briefly, first a mid-crestal incision was made, thereafter a double layer (mucosa and periosteum), split-thickness flap was raised on the buccal aspect. While on the lingual aspect a full-thickness mucoperiosteal flap was raised. To maintain the blood supply of the periosteum, vertical releasing incisions were avoided. Autogenous bone chips were harvested by a single use bone collector device (Safescraper Twist, Meta, Reggio Emilia, Italy) locally without the preparation of a second surgical site and were mixed with a bovine-derived xenograft (Bio_Oss, Gesitlich, Wolhusen, Germany) in a 1:1 ratio. A resorbable collagen membrane (BioGide, Geistlich, Wolhusan, Germany) was shaped and fixated on the lingual aspect with titanium micro-screws (Pro-fix, Osteogenics, Lubbock, USA). Thereafter, the composite graft was compacted on the residual ridge and the collagen membrane was folded over and subsequently fixed with titanium pins (Ustomed, Tuttlingen, Germany). Double-layer wound closure was carried out, first the buccal periosteal layer was sutured to the lingual flap with a 3–0 expanded polytetrafluoroethylene (e-PTFE) suturing material (Cytoplast, Osteogenics, Lubbock, USA). The buccal mucosal layer was also sutured to the lingual flap with horizontal mattress sutures and single interrupted sutures using a 4–0 non-resorbable e-PTFE suturing material (Cytoplast, Osteogenics, Lubbock, USA). Baseline and 6-month follow-up of keratinized soft tissue situations are visible in Fig. .
Baseline and 6-month follow-up CBCT scans were segmented in an open-source radiographic image processing software (3D Slicer, www.slicer.org ) using a dedicated semi-automatic image segmentation method . Anatomical structures (teeth, alveolar bone and nerves) were segmented separately, constituting individual components of the model. The output of image segmentation is a 3D virtual model of the dento-alveolar hard tissues. Following segmentation, standard tessellation language (.stl) files of IOSs were superimposed with segmented 3D models, using identical landmark registration. Corresponding mark-up points were placed on fixed anatomical landmarks (cusps or incisor edges of teeth) (Fig. ). Alignment of the two models were inspected by two individual investigators.
Following digital data processing, baseline and 6-month follow-up data were spatially registered (both CBCT scans and IOS). Utilizing an intensity-based medical image registration algorithm (Elastix) baseline and follow-up CBCT scans were aligned with a linear transformation. The same transformation algorithm was applied to register the IOSs.
Vertical- and horizontal supracrestal soft tissue changes The primary outcome of the study was to evaluate the vertical supracrestal soft tissue changes at three measurement planes. Coronal, sagittal and axial planes of CBCT scans were oriented in a manner that the sagittal plane became parallel and the coronal plane became perpendicular to the alveolar ridge. In the coronal view window, three planes were selected for measurements (mesial plane: at the line of the most mesial titanium pin used for GBR membrane fixation, distal plane: at the line of the most distal titanium pin, middle plane: halfway between the mesial and distal planes). Supracrestal soft tissue height was measured between the most coronal point of the alveolar crest and the most coronal point of the keratinized alveolar mucosa on both the baseline and the follow-up data. Additionally to vertical soft tissue dimensions, the bucco-lingual width of the supracrestal soft tissues was measured. Linear measurements were made perpendicular to the long axis of the alveolar crest at the level of the buccal MGJ (visible on IOS) (Fig. ). Horizontal and vertical shift of the supracrestal soft tissues A measuring grid with a 1 mm interval was overlayed on the previously mentioned measurement planes (mesial, middle, distal). Two reference points were placed midcrestally at the most coronal level of the keratinized tissue crest (KTC) both on the baseline and 6-month follow-up models (ST-pre; ST-post). Two additional reference points were placed at the top of baseline and follow-up edentulous alveolar crests (AC-pre; AC-post). Horizontal distances between the ST-pre; ST-post points and the corresponding AC reference point were measured to assess the occasional shift of KTC following horizontal GBR (Fig. ).
The primary outcome of the study was to evaluate the vertical supracrestal soft tissue changes at three measurement planes. Coronal, sagittal and axial planes of CBCT scans were oriented in a manner that the sagittal plane became parallel and the coronal plane became perpendicular to the alveolar ridge. In the coronal view window, three planes were selected for measurements (mesial plane: at the line of the most mesial titanium pin used for GBR membrane fixation, distal plane: at the line of the most distal titanium pin, middle plane: halfway between the mesial and distal planes). Supracrestal soft tissue height was measured between the most coronal point of the alveolar crest and the most coronal point of the keratinized alveolar mucosa on both the baseline and the follow-up data. Additionally to vertical soft tissue dimensions, the bucco-lingual width of the supracrestal soft tissues was measured. Linear measurements were made perpendicular to the long axis of the alveolar crest at the level of the buccal MGJ (visible on IOS) (Fig. ).
A measuring grid with a 1 mm interval was overlayed on the previously mentioned measurement planes (mesial, middle, distal). Two reference points were placed midcrestally at the most coronal level of the keratinized tissue crest (KTC) both on the baseline and 6-month follow-up models (ST-pre; ST-post). Two additional reference points were placed at the top of baseline and follow-up edentulous alveolar crests (AC-pre; AC-post). Horizontal distances between the ST-pre; ST-post points and the corresponding AC reference point were measured to assess the occasional shift of KTC following horizontal GBR (Fig. ).
Descriptive statistics were used to describe the variables, data were expressed as means ± standard deviations. Statistical differences between baseline and 6-month follow-up supracrestal soft tissue dimensions were analyzed using inferential statistics with a significance level of α = 0.05. Normality of the previously mentioned variables was checked with the Shapiro-Wilk test. Levene’s test was used to check the homogeneity of the variances. Data were found to be normally distributed, and the homogeneity assumption of the variances were met. The continuous variables between subgroups were compared with parametric statistics. The paired sample t-test was utilized to evaluate statistical differences for each variable at different time points. The statistical analysis was performed using the STATA 18 software package (StataCorp LLC, College Station, TX, USA).
Patient demographics Seven systemically healthy patients (6 female, 1 male, aged between 40 and 75 years; mean age 54,7 years) with 8 surgical sites were included in the current case series. Every augmentation procedure was performed in the premolar-molar region of the mandible. Five of the included defects were classified as HL (horizontal-large) and three of the included defects were classified as HM (horizontal- medium). A slight bucco-lingual discrepancy could be detected between the top of the alveolar crest and the midline of the KTC. In the cross sectional planes the midline of the KTC was located buccally from the midline of the alveolar crest. Primary outcome – changes in supracrestal soft tissue height Baseline supracrestal soft tissue height was measured midcrestally at three measurement planes (Table ). In the mesial plane baseline supracrestal soft tissue height averaged at 2.37 mm ± 0.68 mm. At the mesial measurement plane vertical soft tissue dimensions showed no statistically significant difference ( p = 0.21) at 6-month follow-up, being 2.62 mm ± 0.72 mm on average. In the middle plane supracrestal soft tissue height changed from 2.37 mm ± 0.71 mm at baseline to 2.67 mm ± 0.67 mm at 6-month follow-up, however this difference was statistically not significant ( p = 0.21). Contrary in the distal measurement plane a statistically significant difference ( p = 0.0002) was detected between the baseline and follow-up vertical soft tissue values, being 2.64 mm ± 0.87 mm and 3.69 mm ± 1.02 mm respectively. Secondary outcome measures The soft tissue width was assessed at the same three measurement planes at the level of the MGJ (Table ). At the mesial plane supracrestal soft tissue width was measured to be an average of 2.14 mm ± 0.72 mm at baseline compared to 2.47 mm ± 0.46 mm at follow-up. This difference was not found to be statistically significant ( p = 0.06). Also at the mesial plane, horizontal supracrestal soft tissue dimensions showed no statistically significant difference ( p = 0.21) between baseline and 6-month follow-up, being 1.72 mm ± 0.44 mm and 2.07 mm ± 0.67 mm respectively. At the distal plane supracrestal soft tissue width change from 2.15 mm ± 0.36 mm at baseline to 2.36 mm ± 0.59 mm at follow-up, which difference was stistically not significant ( p = 0.28). The midline of the KTC showed a horizontal substantial shift in the buccal direction (Table ). The horizontal distance between ST-pre and ST-post showed an average of 1.61 mm ± 0.72 mm, 2.15 mm ± 1.02 mm and 2.07 mm ± 0.62 mm at the mesial, middle and distal planes. Simultaneously, the vertical distance between ST-pre and ST-post showed an average of -0.45 mm ± 1.16 mm, -0.18 mm ± 1.35 mm and 0.27 mm ± 0.99 mm at the mesial, middle and distal planes.
Seven systemically healthy patients (6 female, 1 male, aged between 40 and 75 years; mean age 54,7 years) with 8 surgical sites were included in the current case series. Every augmentation procedure was performed in the premolar-molar region of the mandible. Five of the included defects were classified as HL (horizontal-large) and three of the included defects were classified as HM (horizontal- medium). A slight bucco-lingual discrepancy could be detected between the top of the alveolar crest and the midline of the KTC. In the cross sectional planes the midline of the KTC was located buccally from the midline of the alveolar crest.
Baseline supracrestal soft tissue height was measured midcrestally at three measurement planes (Table ). In the mesial plane baseline supracrestal soft tissue height averaged at 2.37 mm ± 0.68 mm. At the mesial measurement plane vertical soft tissue dimensions showed no statistically significant difference ( p = 0.21) at 6-month follow-up, being 2.62 mm ± 0.72 mm on average. In the middle plane supracrestal soft tissue height changed from 2.37 mm ± 0.71 mm at baseline to 2.67 mm ± 0.67 mm at 6-month follow-up, however this difference was statistically not significant ( p = 0.21). Contrary in the distal measurement plane a statistically significant difference ( p = 0.0002) was detected between the baseline and follow-up vertical soft tissue values, being 2.64 mm ± 0.87 mm and 3.69 mm ± 1.02 mm respectively.
The soft tissue width was assessed at the same three measurement planes at the level of the MGJ (Table ). At the mesial plane supracrestal soft tissue width was measured to be an average of 2.14 mm ± 0.72 mm at baseline compared to 2.47 mm ± 0.46 mm at follow-up. This difference was not found to be statistically significant ( p = 0.06). Also at the mesial plane, horizontal supracrestal soft tissue dimensions showed no statistically significant difference ( p = 0.21) between baseline and 6-month follow-up, being 1.72 mm ± 0.44 mm and 2.07 mm ± 0.67 mm respectively. At the distal plane supracrestal soft tissue width change from 2.15 mm ± 0.36 mm at baseline to 2.36 mm ± 0.59 mm at follow-up, which difference was stistically not significant ( p = 0.28). The midline of the KTC showed a horizontal substantial shift in the buccal direction (Table ). The horizontal distance between ST-pre and ST-post showed an average of 1.61 mm ± 0.72 mm, 2.15 mm ± 1.02 mm and 2.07 mm ± 0.62 mm at the mesial, middle and distal planes. Simultaneously, the vertical distance between ST-pre and ST-post showed an average of -0.45 mm ± 1.16 mm, -0.18 mm ± 1.35 mm and 0.27 mm ± 0.99 mm at the mesial, middle and distal planes.
In the current case report, soft tissue alterations following horizontal GBR were investigated utilizing digital models acquired with the combination of segmented CBCT models and IOSs. In the literature the information regarding soft tissue alterations following GBR is scarce, even though soft tissue dimensions at future implant sites contribute substantially to long-term implant success . Compared to previous articles , our investigation aimed to measure the dimensions of the supracrestal soft tissues before and after hard tissue augmentation and to observe the effects of split thickness flap mobilization during GBR on soft tissue dimensional changes. Measurements taken at three different measurement planes showed approximately 0.3 mm increase of both supracrestal soft tissue height and width, however none of the differences were found to be statistically significant. Except, the vertical soft tissue increase at the distal aspect of the surgical sites were found to be statistically significant, which may be due to the distortion of the retromolar trigone and the mandibular tuberosity following the repositioning of the buccal flap. These findings contradict those previous clinical observations that keratinized soft tissue dimensions are inevitably reduced following augmentation procedures. In a previous study of our group a slight crestal/ lingual hard tissue resorption following GBR was detected . Another aspect that can influence the vertical increase of keratinized tissues is the applied suturing technique. Due to the double layer suturing, supracrestal keratinized tissues were repositioned coronally. The observation of a horizontal increase is well in line with the horizontal increase of the underlying alveolar ridge, however compared to previous articles the extent of the horizontal increase is substantially less. This occurrence is more likely, due to the fact that horizontal measurement in our investigation were made in a more coronal level. Additionally, the buccal horizontal shift of supracrestal keratinized tissues was observed in the current study. It can be emphasized that this horizontal shift may be caused by the buccal displacement of the lingual flap due to the flap mobilization and the horizontal increase of hard tissues as a result of GBR. This buccal shift results in the discrepancy of midlines of the alveolar ridge and the supracrestal soft tissues, requiring a soft tissue shift or occasionally augmentation of the keratinized mucosa during implant uncovery to avoid potentially unfavorable peri-implant soft tissue conditions. Although our observations on soft tissue dimensional changes following horizontal GBR are unique in the literature, the current study has some drawbacks that have to be addressed. The greatest limitation of the study is the low sample size, therefore, soft tissue changes following horizontal GBR must be investigated on a much larger scale in the future. Another limitation of the current approach is the relatively high possibility of human error during the landmark-based registration of IOSs and CBCT models. In the future this hinderance can be overcome with the automation of the registration process .
The current study did not report significant supracrestal soft tissue reduction following horizontal GBR with a split-thickness flap. Even though there was a slight increase in both vertical and horizontal dimensions, differences are clinically negligible. Additionally, the buccal horizontal shift of supracrestal keratinized tissues was observed, which might be caused by the buccal displacement of the lingual flap due to flap mobilization and the horizontal increase of hard tissues. To derive further conclusions on soft tissue changes following ridge augmentation, a study on a larger population has to be conducted.
|
Psychophysiological and behavioral responses to descriptive labels in modern art museums | 1e6bb838-5d78-45a0-8b05-c635887c3f50 | 10155981 | Physiology[mh] | Over the last few years, museums have seen a significant increase in specific attention to the quality of visitors’ experience . Understanding the behavior of the public, their needs, expectations, and learning processes, is now a prerequisite for the development of any project addressing the enhancement and communication of heritage. In this context, the pandemic crisis has made even more evident the need to pursue research and experimentation initiatives aimed at identifying tools and conditions useful for improving the quality of the cultural and aesthetic experience. The beneficial and soothing effect of contact with artworks has been recognized by the World Health Organisation (WHO) and recently reaffirmed by the Organisation for Economic Co-operation and Development itself , as an important factor in the prevention of diseases and in increasing the state of well-being of the population. For these reasons, museums should focus their attention not only on “what” is exhibited but also on “how” works are exhibited and explained and try to adopt policies for reaching the large public of non-expert visitors. In this framework, basing strategic choices only on qualitative data rather than scientific evidence may not ensure reliable results. In the last few years, studies have focused specifically on the quality of the visitor experience in terms of psychological and cognitive satisfaction . Most studies on empirical aesthetics have been conducted in laboratories, assessing the experience with questionnaires . For example, Nadal and colleagues (2010) explored the influence of complexity, degree of abstraction, and artistry on beauty appreciation of artistic stimuli using multiple subjective rating scales . Other studies also measured psychophysiological parameters in response to pieces of art, such as skin conductance, heart rate, eye movements, and pupillary response . These parameters are known to reflect emotional and cognitive processes and they could be considered measures of individual reactions to artworks. For instance, skin conductance is a sensitive marker of individual meaningful events related to emotion, novelty, or attention, therefore it can be considered a “particularly pertinent window on the mind, when subjectively reported experience is not possible” . In laboratory settings, it has also been found that gazing behavior and pupillary responses reflect the internal state of the observer in terms of attention, pleasure, understanding, familiarity, imagination, cognitive effort, and subjective interpretation of complex visual stimuli . Some laboratory studies, focused on the effect of artworks’ title and labels on the aesthetic experience, found that elaborative titles congruent with the content of the paintings, as well as descriptive information, facilitate the comprehension of the artworks and increase aesthetic appreciation . Although the importance of these studies is largely recognized, recent research on art perception showed that when moving from the lab to the museum, looking at art becomes far more engaging and satisfying . For example, original artworks in museums were liked more, viewed longer, and found more arousing compared to their digital reproductions in the laboratory . Also, according to Mastandrea and colleagues (2009), one aspect that characterizes visitor experience and expectation for museums of ancient and modern art was to see the work in person . Therefore, research should be conducted in the real context where art is exhibited because the originality of the artworks together with the exhibition display contribute to the complexity of the aesthetic experience. For these reasons, more recently, some studies have been conducted inside museums, mainly through observing visitors’ behavior in free-choice setting conditions and administering questionnaires after the visit, but also recording psychophysiological parameters thanks to advanced psycho-physiological portable devices . Recent studies have also analyzed visitors’ pathways and experiences in relation to the arrangement of the exhibition . For example, Reitstätter and colleagues (2020) investigate how the rearrangement of a museum influences the way people see and experience art, combining mobile eye tracking, subjective mapping, and a questionnaire . In particular, regarding the introduction of interpretive labels in the museum setting, they wonder how visitors combine looking at art and reading labels, finding that the introduction of new labels provides benefits to artworks’ viewing time and that visitors’ engagement with the artworks was deeper, as assessed by post-visit exhibition verbal reflections . A more recent study investigated the role of the presence and consistency of titles influences visual exploration of artworks, finding that consistent titles produce longer saccade durations and amplitudes than untitled artworks . Although the evidence suggests that the educational tools in museums may be crucial to improve the process of understanding, appreciation, and promoting individual well-being , their role has been challenged and some museums have chosen to reduce or even eliminate explanations and labels in the attempt to make the experience more emotional and less cultural-driven . Scientifically evaluating the impact of labels on the perception and understanding of artworks can thus contribute to enhancing the engagement of museums in developing the quality of visitors’ experience and the efficacy of their educational offer . This is particularly relevant for modern/contemporary art museums and visitors with poor art training. Non-expert people usually prefer figurative paintings compared to abstract ones , since their content is very often ambiguous and indefinite, compared to figurative art, where the objects represented are clearly recognizable. Indeed, appreciation is correlated to the understanding of artworks , and incomplete comprehension may lead to visitors’ disappointment and potentially discourage further museum visits . The exploratory studies described above have delivered remarkable results and present the advantage of large sample sizes due to working with regular museum visitors. However, they do not allow measuring, with accuracy and reproducibility, the very specific cognitive and emotional processes that occur in the observer while looking at artworks as a function of specific variables, such as the labels provided by museums. Also, none of the studies investigating the impact of different labels have recorded multiple physiological and behavioral parameters in the context of a real art exhibition. Therefore, here we aim to conduct a comprehensive study to test whether descriptive labels improve the aesthetic experience, by combining multiple objective and subjective measurements in a structured experimental protocol in the very context of a modern art museum. To this purpose, we specifically tested the impact of essential and more descriptive written labels on the fruition of XX-XXI century paintings, for which the lay public expresses a lot of difficulty and perplexity in understanding and appreciating the content. We measured psychophysiological (skin conductance, heart rate, pupillary response, eye movements) and behavioral (viewing time, questionnaires) parameters in a group of art-naïve participants while looking at the artworks with different types of labels. Participants assigned to the experimental condition experienced the artworks with essential labels during a first visit and with descriptive labels during a second visit (intra-subject design). To control that the effects can be actually attributed to descriptive labels and not to the double exposure to paintings and essential labels, which could lead to familiarity effects, we introduced a control condition, in which essential labels were shown to an additional sample of participants during both sessions. We hypothesize that descriptive labels can influence both aesthetic emotional reactions and cognitive judgments . Indeed, we expect increased skin conductance, heart rate, and pupillary dilation, due to changes in physiological arousal and emotional response . Furthermore, we expect that descriptive labels yield a more detailed visual inspection and prolonged viewing of paintings, leading to a better understanding of artworks revealed by higher questionnaire scores. The outcome of this study could be of interest to museum operators, which can receive useful insight to offer more educational, descriptive, and interesting visits to a wider public. Participants Thirty healthy volunteers participated in the present study (aged 21–30 years, M = 23.60, SD = 0.44); randomly assigned to the experimental (twenty observers) or the control condition (ten observers). Prior to the experiment, we collected information about participants’ personal data, art historical background, and art expertise. All selected participants had normal or corrected-to-normal visual acuity, did not take any type of medication, did not present any brain damage, and were free of cognitive disorders. All participants were university students (not art students) naive to the purpose of the experiment, with high-school level art history background. None of them were painters. On average, they had visited museums or art exhibitions only 1 or 2 times in the last year and they did not read art-related blogs, magazines, or books. To measure participants’ artistic preferences for different art types, items (e.g., “how much do you like abstract art?”) were rated with a 5-point Likert scale. On average, they like “figurative” art significantly more than “abstract” art (mean score = 3.5 ± 0.2 vs. 2.7 ± 0.1; t (29) = 1.8, p <0.05). Participants were mostly unfamiliar with the paintings and their authors: they only knew the author Mirò (14 over 30), and only two of them were familiar with the painting used. All participants were covid-free. Ethics Experimental procedures were approved by the local ethics committee (Comitato Etico Pediatrico Regionale–Azienda Ospedaliero-Universitaria Meyer–Firenze FI) and are in line with the Declaration of Helsinki. Written informed consent was obtained from each participant prior to their inclusion in the study. Setup Pupil and gaze data were recorded by means of a wearable eye tracking headset (Pupil Core from Pupil Labs, Berlin, Germany), composed of two eye cameras (200Hz) and a world camera (60Hz). The device was USB-connected to a MacBook Pro running a dedicated software (Pupil Capture, version 3.5.7) that enabled real-time data capture, camera recording, and calibration routines for natural conditions. A wearable wireless device equipped with high-quality data sensors (E4 wristband from Empatica Inc, Boston, USA) was used to acquire electrodermal activity (EDA, 4Hz) and heart rate (HR, computed in spans of 10 seconds) measures. The internal memory of E4 allowed us to record data continuously during the daily session (about 30 minutes per participant). The E4 device also included the possibility to press a central button during the session to mark the times of our events of interest (“tags”). Stimuli The present study was conducted in the “Roberto Casamonti Collection” ( https://collezionerobertocasamonti.com ), a modern and contemporary art (XX-XXI century) private museum, hosted in Palazzo Bartolini Salimbeni in Florence. During each session, participants were required to follow the visit path indicated by the experimenter and stop in front of eight selected paintings. The presentation order was the same for all participants, following the position of the artwork in the exhibition. Experimental paintings were selected before data collection, excluding those representing human figures, those totally black or white, and those that were too small or too big to be framed by the Pupil Core world camera. For each selected painting, we set an adequate distance at which observers had to stop for observation, such that each one subtended a visual angle of about 21°x15°. The paintings’ physical luminance was measured at five different points of the canvas (top-left and top-right, lower-left and lower-right, and in the center) that were averaged in a single value. Five paintings resulted in the range between 11–31 cd/m2 (16 cd/m 2 on average), two were darker (5 and 10 cd/m 2 , 7.5 cd/m 2 on average) and one lighter (116 cd/m 2 ). We then created three luminance control stimuli (30x42cm uniform-colour canvas) for the three different levels of luminance, in order to measure baseline individual pupil diameter to those luminances. To see all selected paintings, see . Conditions Participants were divided into two groups: twenty of them participated in the experimental condition and ten participated in the control condition. Both conditions consisted of two sessions at the museum on two different days, at least one month apart (on average, the second session was carried out five/six weeks after the first). In the first session, all participants were presented with essential labels (i.e., author, title, year, and technique) before seeing the paintings. In the second session, experimental participants were provided with descriptive labels (i.e., author, title, year, technique, and description of the painting’s content and the technique), whereas control participants were shown the same essential label as in the first session. See for all essential and descriptive labels. Procedure At the beginning of each session, participants wore the instruments and familiarized with them in a dedicated room. Before starting the visit, they were positioned in front of the three luminance-control stimuli and asked to look at each for ten seconds. They were then instructed to follow the experimenter from one painting to the next and to press the timestamp button on the Empatica wristband (“tag”) every time they started and stopped reading a label and looked at the painting. A two-second red light was displayed on the wristband after each button press; therefore, the experimenter could check that participants correctly pressed the tag when needed (and promptly reminded him/her to press it in case of occasional forgetfulness). After the pre-session measurements, observers reached the first painting indicated by the experimenter and stood in front at the preset distance. Once the eye tracker was calibrated (we used an 8-points natural-features calibration routine), participants could read the label. The label, written on a sheet of paper, was shown by one experimenter standing in front of the participant. Then participants looked at the painting for as long as they wanted, pressing a “tag” when they started and stopped reading and observing. After they finished observing the painting, the experimenter asked them some questions about the artwork and reported the answers on a notepad. The questionnaire required the participants to score on a 5-points Likert scale the following items: complexity, comprehensibility, title informativeness, positive emotions, negative emotions, appreciation, interest, and curiosity for other works of the same author. Participants were also asked to report if the paintings and the authors were familiar or unfamiliar. Then participants continued the visit to the next selected painting. For a schematic representation of the experimental procedure see . Data processing and statistical analysis Physiological parameters were recorded from the start to the end of each museum session, so that, for each participant, we obtained a continuous recording of about 25–30 minutes per session. Raw data from the wristband and the eye tracker were extracted in.csv format and synchronized through an ad-hoc procedure in Matlab (R2020b version; Natick, Massachusetts: The MathWorks Inc.). The timestamps (“tags”) were converted to real times and used to delimitate our events of interest. The participants’ artistic preferences for different art types, rated with a 5-point Likert scale, were calculated. We performed paired sample t -tests across subjects to assess significant differences between art types. The reading time of each label was calculated as the difference in seconds between the two tags indicating the start and the end of the observer’s reading. The time of viewing of each painting was calculated as the difference in seconds between the two tags indicating the start and the end of the observer’s visualization. Thus, the artworks’ viewing time does not include the time spent reading the labels. For each participant, the viewing times of all the paintings were averaged together. Then, the times of all participants were averaged. To compare the average viewing time between the essential and the descriptive label sessions, and between the two control sessions with essential labels, a two-way ANOVA analysis with within-subjects factor session (two levels: first vs. second session) and between-subjects factor condition (two levels: experimental vs. control condition) was done. P -values obtained from post hoc analyses were adjusted using the Bonferroni correction. Effect sizes of the differences were estimated by eta-squared statistics ( η 2 ) with 95% confidence intervals. The average viewing time of each participant was also correlated (Pearson linear-correlation coefficient) to the information collected prior to the experiment about their art expertise and artistic preferences. For each questionnaire item, administered during the experiment after viewing each painting, the scores assigned by each participant to each painting were averaged together. To compare average scores between the essential and the descriptive label conditions, and between the two control sessions with essential labels, paired-sample Wilcoxon signed-rank tests across paintings were performed. The effect size of differences between conditions was estimated by Rank-Biserial correlation (r rb ) with 95% confidence intervals. Also, individual scores for each painting were related to the corresponding EDA, HR, and pupil responses to calculate the Pearson linear-correlation coefficient. For measuring the changes in EDA and HR responses induced by the painting, we normalized each trace considering as baseline the average EDA/HR value in the last three seconds before looking at the artwork. Pupil diameter was converted from pixel to millimeters by measuring the eye tracker recording of a 4 mm artificial pupil, positioned at the location of the observer’s eyes. For measuring pupil size variations induced by paintings, each trace was normalized, considering as baseline the average pupil diameter in response to the luminance-matched control stimulus presented before each session . To produce plots as a function of time, for each painting, normalized traces were averaged for each recorded time across participants. Since viewing time changes across participants, only means including at least five participants were considered. This process has led to average recordings where the initial values include all participants, whereas the last values include only participants with long viewing times. Finally, average traces for each painting were averaged together. To perform statistical analysis, for each normalized trace of each participant for each painting, the average value and the root mean square error (RMSE) during the whole viewing time were calculated. The means and RMSE of all parameters were compared with two-way ANOVA analyses with within-subjects factor session (two levels: first vs. second session) and between-subjects factor condition (two levels: experimental vs. control condition). P -values obtained from post hoc analyses were adjusted using the Bonferroni correction. Effect sizes of the differences were estimated by eta squared statistics ( η 2 ) with 95% confidence intervals. Since the gaze is recorded through a head-centred camera, and thus subjected to head movements, to analyze the gaze pattern we adopted a manual procedure . We subdivided each painting into 25 equally sized areas so that each area subtended a visual angle of about 4°x3° in each painting. All video recordings were extracted by using the Pupil Player software and each position of the gaze shown in the videos was manually converted to a position in one of the 25 areas. Then we counted how many times each area had been watched by each participant. To compare the difference of fixations in the descriptive vs. essential label or between the two control sessions for each of the 25 areas a heat map for each painting was calculated as follows. Since the number of fixations in the two sessions is different, the proportion of fixations in each area (with respect to the total number of fixations in that condition) in the different sessions was calculated and then their difference was computed. For representational purposes, the distribution of differences between the second and first session was binned into five density levels: one (the middle) corresponding to the median of the distribution (equal density of fixations), the others corresponding to quartiles of the distribution. To study the distribution of fixations as a function of eccentricity, three eccentricities were considered for the whole canvas area: central area 0°-7°, nearby periphery 7°-14°, and periphery 14°-21°. Then the number of fixations for each eccentricity and for all observers was calculated. Finally, for each eccentricity, the difference between fixations in the descriptive label and essential label sessions and between the two control sessions was calculated. Comparisons of these values between different eccentricities were done with paired-sample two-tailed t -tests. The effect size of differences between conditions was estimated by Cohen’s d statistics with 95% confidence intervals. Thirty healthy volunteers participated in the present study (aged 21–30 years, M = 23.60, SD = 0.44); randomly assigned to the experimental (twenty observers) or the control condition (ten observers). Prior to the experiment, we collected information about participants’ personal data, art historical background, and art expertise. All selected participants had normal or corrected-to-normal visual acuity, did not take any type of medication, did not present any brain damage, and were free of cognitive disorders. All participants were university students (not art students) naive to the purpose of the experiment, with high-school level art history background. None of them were painters. On average, they had visited museums or art exhibitions only 1 or 2 times in the last year and they did not read art-related blogs, magazines, or books. To measure participants’ artistic preferences for different art types, items (e.g., “how much do you like abstract art?”) were rated with a 5-point Likert scale. On average, they like “figurative” art significantly more than “abstract” art (mean score = 3.5 ± 0.2 vs. 2.7 ± 0.1; t (29) = 1.8, p <0.05). Participants were mostly unfamiliar with the paintings and their authors: they only knew the author Mirò (14 over 30), and only two of them were familiar with the painting used. All participants were covid-free. Experimental procedures were approved by the local ethics committee (Comitato Etico Pediatrico Regionale–Azienda Ospedaliero-Universitaria Meyer–Firenze FI) and are in line with the Declaration of Helsinki. Written informed consent was obtained from each participant prior to their inclusion in the study. Pupil and gaze data were recorded by means of a wearable eye tracking headset (Pupil Core from Pupil Labs, Berlin, Germany), composed of two eye cameras (200Hz) and a world camera (60Hz). The device was USB-connected to a MacBook Pro running a dedicated software (Pupil Capture, version 3.5.7) that enabled real-time data capture, camera recording, and calibration routines for natural conditions. A wearable wireless device equipped with high-quality data sensors (E4 wristband from Empatica Inc, Boston, USA) was used to acquire electrodermal activity (EDA, 4Hz) and heart rate (HR, computed in spans of 10 seconds) measures. The internal memory of E4 allowed us to record data continuously during the daily session (about 30 minutes per participant). The E4 device also included the possibility to press a central button during the session to mark the times of our events of interest (“tags”). The present study was conducted in the “Roberto Casamonti Collection” ( https://collezionerobertocasamonti.com ), a modern and contemporary art (XX-XXI century) private museum, hosted in Palazzo Bartolini Salimbeni in Florence. During each session, participants were required to follow the visit path indicated by the experimenter and stop in front of eight selected paintings. The presentation order was the same for all participants, following the position of the artwork in the exhibition. Experimental paintings were selected before data collection, excluding those representing human figures, those totally black or white, and those that were too small or too big to be framed by the Pupil Core world camera. For each selected painting, we set an adequate distance at which observers had to stop for observation, such that each one subtended a visual angle of about 21°x15°. The paintings’ physical luminance was measured at five different points of the canvas (top-left and top-right, lower-left and lower-right, and in the center) that were averaged in a single value. Five paintings resulted in the range between 11–31 cd/m2 (16 cd/m 2 on average), two were darker (5 and 10 cd/m 2 , 7.5 cd/m 2 on average) and one lighter (116 cd/m 2 ). We then created three luminance control stimuli (30x42cm uniform-colour canvas) for the three different levels of luminance, in order to measure baseline individual pupil diameter to those luminances. To see all selected paintings, see . Participants were divided into two groups: twenty of them participated in the experimental condition and ten participated in the control condition. Both conditions consisted of two sessions at the museum on two different days, at least one month apart (on average, the second session was carried out five/six weeks after the first). In the first session, all participants were presented with essential labels (i.e., author, title, year, and technique) before seeing the paintings. In the second session, experimental participants were provided with descriptive labels (i.e., author, title, year, technique, and description of the painting’s content and the technique), whereas control participants were shown the same essential label as in the first session. See for all essential and descriptive labels. At the beginning of each session, participants wore the instruments and familiarized with them in a dedicated room. Before starting the visit, they were positioned in front of the three luminance-control stimuli and asked to look at each for ten seconds. They were then instructed to follow the experimenter from one painting to the next and to press the timestamp button on the Empatica wristband (“tag”) every time they started and stopped reading a label and looked at the painting. A two-second red light was displayed on the wristband after each button press; therefore, the experimenter could check that participants correctly pressed the tag when needed (and promptly reminded him/her to press it in case of occasional forgetfulness). After the pre-session measurements, observers reached the first painting indicated by the experimenter and stood in front at the preset distance. Once the eye tracker was calibrated (we used an 8-points natural-features calibration routine), participants could read the label. The label, written on a sheet of paper, was shown by one experimenter standing in front of the participant. Then participants looked at the painting for as long as they wanted, pressing a “tag” when they started and stopped reading and observing. After they finished observing the painting, the experimenter asked them some questions about the artwork and reported the answers on a notepad. The questionnaire required the participants to score on a 5-points Likert scale the following items: complexity, comprehensibility, title informativeness, positive emotions, negative emotions, appreciation, interest, and curiosity for other works of the same author. Participants were also asked to report if the paintings and the authors were familiar or unfamiliar. Then participants continued the visit to the next selected painting. For a schematic representation of the experimental procedure see . Physiological parameters were recorded from the start to the end of each museum session, so that, for each participant, we obtained a continuous recording of about 25–30 minutes per session. Raw data from the wristband and the eye tracker were extracted in.csv format and synchronized through an ad-hoc procedure in Matlab (R2020b version; Natick, Massachusetts: The MathWorks Inc.). The timestamps (“tags”) were converted to real times and used to delimitate our events of interest. The participants’ artistic preferences for different art types, rated with a 5-point Likert scale, were calculated. We performed paired sample t -tests across subjects to assess significant differences between art types. The reading time of each label was calculated as the difference in seconds between the two tags indicating the start and the end of the observer’s reading. The time of viewing of each painting was calculated as the difference in seconds between the two tags indicating the start and the end of the observer’s visualization. Thus, the artworks’ viewing time does not include the time spent reading the labels. For each participant, the viewing times of all the paintings were averaged together. Then, the times of all participants were averaged. To compare the average viewing time between the essential and the descriptive label sessions, and between the two control sessions with essential labels, a two-way ANOVA analysis with within-subjects factor session (two levels: first vs. second session) and between-subjects factor condition (two levels: experimental vs. control condition) was done. P -values obtained from post hoc analyses were adjusted using the Bonferroni correction. Effect sizes of the differences were estimated by eta-squared statistics ( η 2 ) with 95% confidence intervals. The average viewing time of each participant was also correlated (Pearson linear-correlation coefficient) to the information collected prior to the experiment about their art expertise and artistic preferences. For each questionnaire item, administered during the experiment after viewing each painting, the scores assigned by each participant to each painting were averaged together. To compare average scores between the essential and the descriptive label conditions, and between the two control sessions with essential labels, paired-sample Wilcoxon signed-rank tests across paintings were performed. The effect size of differences between conditions was estimated by Rank-Biserial correlation (r rb ) with 95% confidence intervals. Also, individual scores for each painting were related to the corresponding EDA, HR, and pupil responses to calculate the Pearson linear-correlation coefficient. For measuring the changes in EDA and HR responses induced by the painting, we normalized each trace considering as baseline the average EDA/HR value in the last three seconds before looking at the artwork. Pupil diameter was converted from pixel to millimeters by measuring the eye tracker recording of a 4 mm artificial pupil, positioned at the location of the observer’s eyes. For measuring pupil size variations induced by paintings, each trace was normalized, considering as baseline the average pupil diameter in response to the luminance-matched control stimulus presented before each session . To produce plots as a function of time, for each painting, normalized traces were averaged for each recorded time across participants. Since viewing time changes across participants, only means including at least five participants were considered. This process has led to average recordings where the initial values include all participants, whereas the last values include only participants with long viewing times. Finally, average traces for each painting were averaged together. To perform statistical analysis, for each normalized trace of each participant for each painting, the average value and the root mean square error (RMSE) during the whole viewing time were calculated. The means and RMSE of all parameters were compared with two-way ANOVA analyses with within-subjects factor session (two levels: first vs. second session) and between-subjects factor condition (two levels: experimental vs. control condition). P -values obtained from post hoc analyses were adjusted using the Bonferroni correction. Effect sizes of the differences were estimated by eta squared statistics ( η 2 ) with 95% confidence intervals. Since the gaze is recorded through a head-centred camera, and thus subjected to head movements, to analyze the gaze pattern we adopted a manual procedure . We subdivided each painting into 25 equally sized areas so that each area subtended a visual angle of about 4°x3° in each painting. All video recordings were extracted by using the Pupil Player software and each position of the gaze shown in the videos was manually converted to a position in one of the 25 areas. Then we counted how many times each area had been watched by each participant. To compare the difference of fixations in the descriptive vs. essential label or between the two control sessions for each of the 25 areas a heat map for each painting was calculated as follows. Since the number of fixations in the two sessions is different, the proportion of fixations in each area (with respect to the total number of fixations in that condition) in the different sessions was calculated and then their difference was computed. For representational purposes, the distribution of differences between the second and first session was binned into five density levels: one (the middle) corresponding to the median of the distribution (equal density of fixations), the others corresponding to quartiles of the distribution. To study the distribution of fixations as a function of eccentricity, three eccentricities were considered for the whole canvas area: central area 0°-7°, nearby periphery 7°-14°, and periphery 14°-21°. Then the number of fixations for each eccentricity and for all observers was calculated. Finally, for each eccentricity, the difference between fixations in the descriptive label and essential label sessions and between the two control sessions was calculated. Comparisons of these values between different eccentricities were done with paired-sample two-tailed t -tests. The effect size of differences between conditions was estimated by Cohen’s d statistics with 95% confidence intervals. Considering average time viewing, ANOVA analysis reveals no significant difference between sessions (F(1,24) = 1.04, p >0.05, η 2 = 0.003). On the other hand, there is a significant effect of conditions (F(1,28) = 5.2, p <0.01, η 2 = 0.1) and of the interaction between factors (F(1,28) = 40.9, p <0.001, η 2 = 0.1). Indeed, in the experimental condition, the average time spent viewing the paintings is significantly lower in the first than in the second session, thus observers’ viewing time is significantly longer after reading a descriptive label compared to an essential label (post-hoc comparisons; t = -4.6, p <0.001; ) . On the contrary, in the control condition, the average viewing time is significantly longer in the first session than in the second session with essential labels (post-hoc comparisons; t = 4.5, p <0.001; ) . Viewing times in the first sessions of the experimental and control conditions are comparable s ince both groups read essential labels during the first visit (post-hoc comparisons; t = -0.013, p >0.05). Average viewing times of each painting in the experimental and control condition are shown in . Average reading times of labels of each painting in the experimental and control condition are reported in . A positive correlation emerges between participants’ preference for abstract art (rated with a 5-point Likert scale before the experiment; see section) and the average viewing time of paintings during the first visit (Pearson linear-correlation; r(28) = 0.56, p <0.01). For the experimental condition, questionnaire’ scoring reveals that descriptive labels, compared to essential labels, influence with very strong effect sizes several dimensions : perception of artwork’s complexity decreases (Paired-sample Wilcoxon signed-rank test; W(7) = 36.01, p <0.05, r rb = 1.00, 95% CI [1.00, 1.00]), contents’ comprehensibility increases (W(7) = 0.01, p <0.01, r rb = 1.00, CI [-1.00, -1.00]), the title looks more informative (W(7) = 2.01, p <0.05, r rb = 0.89, CI [-0.97, -0.56]), positive emotions increase (W(7) = 0.01, p <0.05, r rb = 1.00), while negative emotions decrease (W(7) = 35.01, p <0.05, r rb = 0.94, CI [0.76, 0.99]). Aesthetic appreciation, interest, and curiosity for other artworks of the same authors do not change significantly between different labels. No significant questionnaire differences were found between sessions in the control condition . Regarding the average EDA response, ANOVA analysis reveals a significant difference between conditions (F(1,14) = 12.3, p <0.01, η 2 = 0.2) and sessions (F(1,14) = 27.7, p <0.001, η 2 = 0.3). Indeed, EDA in the second session increases during the first seconds of painting viewing more than with essential labels and remains higher during the whole viewing time (line graph in ), as well as in the control condition (see line graph in ). The average EDA response is significantly higher both for experimental (see bar graph in —left panel) and control condition (see bar graph in —left panel). No significant interactions emerge between factors (F(1,14) = 0.3, p >0.05, η 2 = 0.004). For RMSE, ANOVA shows a significant effect of the session (F(1,14) = 26.7, p <0.001, η 2 = 0.3), no effect of the condition (F(1,14) = 0.07, p >0.05, η 2 = 0.002), and a significant interaction between the two factors (F(1,14) = 10.9, p <0.01, η 2 = 0. 1). Particularly, post-hoc comparisons reveal a statistical difference between the first and the second session during the experimental condition (t = -6.1, p <0.001; —right panel ). However, no statistical differences in the average RMSE are found in the control condition (see –right panel ). Average EDA responses to each painting in the experimental and control condition are shown in . Neither session (F(1,14) = 0.0, p >0.05, η 2 = 0.0), condition (F(1,14) = 0.07, p >0.05, η 2 = 0.002) or their interaction (F(1,14) = 1.3, p >0.05, η 2 = 0.05) affect heart rate response. Considering pupillary responses, ANOVA reveals no statistical differences between conditions (F(1,14) = 0.003, p >0.05, η 2 = 0.0). On the contrary, there is a significant effect of the session (F(1,14) = 5.9, p <0.05, η 2 = 0.02) and of the interaction between factors (F(1,14) = 8.4, p <0.05, η 2 = 0.02). Indeed, in the experimental condition, pupillary responses differ between the two sessions: the pupil is always more dilated with descriptive than with essential labels (line graph in ). Average pupil variation is positive and statistically higher with descriptive labels than with essential labels (post-hoc comparisons; t = -6.1, p <0.001; bar graph in ), but no differences are found in the control condition . Average pupillary responses to each painting in the experimental and control condition are shown in . There are no correlations (Pearson linear-correlation) between individual psychophysiological responses to paintings and corresponding questionnaire scores (all p >0.05). Also, the responses are not affected by familiarity with the paintings (i.e., the responses to the most familiar painting— Femme , Miró, 1977–1978 –are the same as those found for all the other unknown stimuli). Eye movements analysis during painting viewing highlights some differences across label conditions (see ). First, gaze patterns result to be related to the painting’s description. For example, eye movements in the Miró painting ( Femme; 1977–1978) are directed toward the elements depicting feminine body parts, as described in the descriptive label of the experimental condition (see and the heatmap in ). Instead, in the control condition without a description, the eyes are less directed towards salient paintings’ elements (see the heat map in ). Since participants in the experimental condition spent more time looking at the painting with descriptive labels, the number of fixations in this condition is higher (275±30 vs 220±37, on average). On average, after reading the description, participants tend to fixate more in the closer periphery (7°-14°) than the painting’s centre (0°-7°; Paired sample t -test; t(7) = -4.05, p < .01, d = -1.43, 95% CI [-2.42, -0.41]; see the bar graph in ). On the other hand, in the control condition, the number of fixations is lower in the second session, as expected by less time spent observing the painting. Observers mainly fixate the centre of the paintings: the number of fixations in the near (7–14°) and far periphery (14–21°) is much lower than in the centre (0–7°) between the two sessions (t(7) = 5.78, p <0.001, d = 2.05, CI [0.77, 3.28] and t(7) = 5.84, p <0.001, d = 2.07, CI [0.78, 3.31] respectively; see bar graph ). In the present study, we compared the impact of essential and descriptive labels on the cognitive and emotional experience of naïve visitors, through multiple objective and subjective measurements, focusing on controversial modern paintings. Indeed, socio-cultural bias and stereotypes tend to twist the answers of people, often worried to be judged for their lack of expertise in art history, and this process becomes particularly relevant in the context of modern art museums, where the sense of self-efficacy and ease of visitors tend to decrease and–at the same time–the need for educational support, from the majority of the public, becomes unavoidable. People can feel a sense of frustration in front of artists and movements which they merely know and, most of all, hardly understand . Our decision to work with a collection of modern art derives from these considerations and from the widely diffused commonplace that ancient art is “easier” than modern art . Coherently with these premises, we have decided to select a group of people who lacked any particular artistic experience and art history background: a characteristic which made our sample quite homogenous from the art-expertise point of view. Our results show that objective and subjective responses while inspecting modern artworks change depending on the information received before experiencing the paintings. A detailed description encourages participants to spend more time observing artworks, following the information provided. It is difficult for non-experts to catch the meaning of modern artworks . For example, the Miró painting may appear as a series of wide black brushstrokes with small, coloured spots. But when participants come to know that the spots outline the shape of a female body, their eyes perform a greater number of fixations on the parts depicting the figure. This suggests that the explanation provides a key to cognitive and emotive comprehension, confirmed by the subjective perception of increased positive feelings and comprehension. The increase and variability of electrodermal activity and pupil dilation, suggest an increase in physiological arousal . This could be due to a deeper understanding of artworks and a higher cognitive load , as well as to an increase in emotional load . Subjective judgments after viewing the paintings cannot further shed light on the relative contribution of the two dimensions. The increased EDA in the control condition suggests that this psychophysiological response might be modulated by the familiarity for the stimuli, which has been linked to faster processing and higher preference for familiar stimuli compared to novel ones . However, the phasic EDA activity changes when subjects have read the painting description, maybe related to focusing on different elements, while it does not change in the control condition . Our results also show that participants who appreciate more abstract art are the ones spending more time in front of the paintings. However, aesthetic appreciation for the specific paintings presented during the experiment does not change upon explanation. This suggests that, although labels can facilitate comprehension, this is not enough to cause an increased appreciation. We can speculate that specific art training is needed to appreciate modern artworks. Indeed, expertise in art facilitates the so-called aesthetic fluency ; a process that could lead people to better grasp the meaning of an artwork and to its aesthetic appreciation. Also, this result may be interpreted by the fact that modern/contemporary art does not have as its main objective to be "beautiful", rather to be interesting, activating, provocative, ambiguous, and meaningful. Overall, our findings show that visitors do receive important benefits from reading detailed information about artworks. On a more general ground, art description leads to changes in aesthetic judgment and aesthetic emotion outputs, described in the Leder and colleagues’ aesthetic experience model . Since descriptive labels are used when paintings are also seen for the second time, it could be that label-based effects may be conditional on paintings that participants are already familiar with. On the other hand, participants who visited the museum twice receiving the same essential information do not show increased satisfaction. They spent less time observing the painting and assigned the same scores to the questionnaire for understanding and appreciation of the artworks. Also, observers fixate and explore less the artworks, as expected. Except for a slight increase in electrodermal activity, variability of EDA and pupil dilation do not change in the second session. Overall, these results suggest that familiarity with the stimuli without additional information does not improve the museum experience in terms of aesthetic judgments and emotional reactions in naïve visitors, but rather causes them to pay less attention to the artworks . Some of our findings can be compared to those of previous studies. The average viewing times found here with essential and descriptive labels are in line with those found in previous works using unfamiliar artworks . Viewing times are generally longer for well-known paintings (e.g., “The Kiss” by Klimt; ); however, here the only painting familiar to our participants ( Femme , Miró, 1977–1978) received the same amount of time. Mastandrea and colleagues (2019) measured blood pressure and heart rate before and after museum visits, finding that visits to art museums decreased the level of systolic blood pressure but did not influence the heart rate . This is in line with our absence of effects of labels on the HR. It has been found that the display influences the way people experience art, causing different viewing times, levels of engagement and different patterns of fixations , as we found with different types of labels. In contrast, interests in specific artworks and art style preferences proved to be robust and independent of presentation modes. This confirms our results on aesthetic appreciation, which does not increase by introducing descriptive labels. They also found that when labels are more complex (with more text), visitors’ interpretation differs according to the information received, and so do we. Tschacher and colleagues (2012) found that artworks’ understanding was correlated with higher skin conductance variability . We also found higher EDA variability with descriptive labels, even if we did not find any correlation with specific cognitive or emotional domains of the questionnaire. Increased pupil dilation, which we found after presenting a description of the artwork, has been found in some studies conducted outside the museum context, where it was associated with aspects of aesthetic emotions . Note however that, none of this research has measured psychophysiological and behavioral responses during the museum visit as a function of descriptive material, that could influence the aesthetic experience. We cannot rule out that our findings depend on the particular type of artworks involved. Further research might be undertaken in order to make comparisons between “ancient” art and “modern” art, trying to explore possible differences in reactions of visitors in front of visual languages which are perceived as more familiar or simply closer to a general “common taste”. Indeed, robust findings show that non-expert observers usually prefer figurative paintings compared to abstract or conceptual artworks , and that art appreciation correlates with educational level . Ancient art is more easily recognizable and can result to be less anxiety-trigger, especially for non-expert visitors; nevertheless, we should not underestimate the emotional effect (in terms of involvement and gratification) that “ancient masters” can activate on the general public. During real museum visits, visitors can go back to view some paintings they particularly like while ignoring others, they usually go back and forth between reading labels and viewing artworks, sometimes they may be in the company of other people, and they generally experience many artworks, facing the problem of museum fatigue (for a review, see ). In our study, we gave up these situations occurring in an optimal ecological condition in favor of reproducible and accurate measurements, with the aim of avoiding confounding variables. Also, we cannot exclude that the outcome might have been different having as participants a sample of regular museum-goers. In the future, it would be interesting to test art-expert participants; the knowledge they already possess should be enough to understand the meaning of the artwork and have a satisfying experience without the need for informative materials. Expertise is known to lead to higher aesthetic appreciation of artworks and differences in viewing strategies, gaze patterns, fixation distributions, and even in electrophysiological correlates . Fixations should be more focused on salient parts of artworks because the meaning could be grasped even without a descriptive label. More complex is the question of aesthetic appreciation, which experts tend to underestimate in comparison to the complexity of meaning (while for naïve visitors may result as the leading value). We expect that a naïve group would receive more beneficial effects from the explanation of artworks than experts would do, as they might be influenced by their personal evaluation of the quality and amount of information received. Overall, our work suggests that elaborating effective labels, based on scientific evidence rather than on qualitative observations, should be a primary goal for museums. Indeed, if museums aim to attract a wider public, they need to focus their attention on didactic tools provided by panels and captions, with the hope to fill the gap generated by the lack of art knowledge. This is particularly relevant for modern art which is less known and harder to understand and appreciate by non-art experts, that can otherwise perceive the museum visit as a frustrating experience due to their little art education. Aesthetic experience is a psychophysiological process which arises during artworks fruition and involving a variety of emotional and cognitive responses. Here we use multiple psychophysiological and behavioral tools to measure rigorously, in the very context of a modern art museum, the effects of explanatory texts and labels on modern art experience. Our findings show that people receive important benefits, in terms of cognitive and emotional involvement, by reading detailed descriptions of modern artworks. The outcome of our studies could be of interest to museum operators and may become instrumental for improving exhibition, website information content, and advertising material, and for achieving optimal fruition, satisfaction, and thus contribute to well-being of naive visitors as well as experts. On more general grounds our results indicate that psychophysiological changes can be an effective probe into artworks processing and interpretation, making them useful tools for the study of museum aesthetic experience. S1 Table Paintings and their essential and descriptive labels. First column: web link to the paintings used as stimuli (in the same order as presented to the visitors at the Casamonti collection). Second column: essential label read by the participants in the first experimental session and by control participants in the two control sessions (regular style: Italian version; italics style: English version). Third column: descriptive label read by the participants in the second experimental session (regular style: Italian version; italics style: English version). (DOCX) Click here for additional data file. S2 Table Reading time of labels of each painting in the experimental and control condition. Data in the table are reading time in seconds averaged across participants. (DOCX) Click here for additional data file. S1 Fig Viewing time of each painting. (A) Experimental condition. (B) Control condition. The bars show the viewing time of each painting (from painting 1 to painting 8, in the same order as presented to the visitors at the Casamonti collection), averaged across participants. Errors are SE across participants. (TIF) Click here for additional data file. S2 Fig Electrodermal response to each painting. (A) Experimental condition. (B) Control condition. The bars show the EDA response to each painting (from painting 1 to painting 8, in the same order as presented to the visitors at the Casamonti collection), averaged across participants. Errors are SE across participants. (TIF) Click here for additional data file. S3 Fig Pupillary response to each painting. (A) Experimental condition. (B) Control condition. The lines show pupil response over time to each painting (from painting 1 to painting 8, in order as presented to the visitors at the Casamonti collection), averaged across participants. Errors are SE across participants. (TIF) Click here for additional data file. |
Evaluation of aortic arch calcification to predict prognosis after transcatheter aortic valve replacement | 5dff6821-6621-48b5-b037-d8a5b8a950b6 | 11845605 | Surgical Procedures, Operative[mh] | The number of transcatheter aortic valve replacements (TAVR) has rapidly increased during the last decade , . Randomized controlled studies have provided favorable evidence for TAVR; therefore, it is recommended as the first-line therapy for severe aortic stenosis (AS) in older patients . However, patients with extensively calcified aortic arch are generally excluded from randomized controlled trials because of embolism and vascular complications associated with the passing of delivery system through the heavily calcified aortic arch , . In fact, aortic arch calcification (AAC) can provide lots of information. It can reflect arterial stiffness, the magnitude of calcified change in the whole aorta, and even the frailty of the patients – . A large prospective population-based study demonstrated that a more severe AAC was associated with a higher risk of all-cause mortality and cardiovascular mortality in middle-aged and elderly persons . Besides, AAC was found in over 60% of patients who underwent TAVR . Therefore, AAC should be carefully evaluated in TAVR candidates. Computed tomography (CT) is an essential examination to assess the anatomy of patients before TAVR and is routinely performed in most centers. Theoretically, AAC is easily evaluated; however, few centers routinely report AAC in their CT analysis because of the lack of an easy and reliable method. Accordingly, a limited number of studies focus on AAC. The association between AAC and the clinical prognosis of patients who undergo TAVR is also unclear. Recording AAC information and images is very important since these are necessary for further machine learning or deep learning study, which can provide valuable insights in prediction in cardiovascular contexts , . Therefore, we envisaged this study to propose a simple and rapid method to assess AAC and to assess the impact of AAC on the prognosis after TAVR. Study population Consecutive severe AS patients who underwent TAVR between March 2016 and December 2020 were included in this study. The majority of patients were enrolled from the TORCH registry (Transcatheter Aortic Valve Replacement Single Center Registry in Chinese Population, NCT02803294); this is a prospective, ongoing, real-world registry, launched in June 2016. Patients who had a history of aortic valve replacement or did not have analyzable pre-operative multidetector CT (MDCT) images were excluded. Consecutive severe AS patients who underwent TAVR between March 2016 and December 2020 were included in this study. The majority of patients were enrolled from the TORCH registry (Transcatheter Aortic Valve Replacement Single Center Registry in Chinese Population, NCT02803294); this is a prospective, ongoing, real-world registry, launched in June 2016. Patients who had a history of aortic valve replacement or did not have analyzable pre-operative multidetector CT (MDCT) images were excluded. All TAVR procedures were decided by our multidisciplinary heart team. Self-expanding valves from VenusA (Venus Medtech, China), VitaFlow (Microport, China), TaurusOne (Peijia Medical, China), J-valve (Jiecheng, China) series were used in the majority of patients. The remaining patients were implanted with balloon-expandable (SAPIEN 3/XT, Edwards Lifesciences, California, USA) or mechanically-expandable (Lotus valve, Boston Scientific, Massachusetts, USA) valves. Patients with bicuspid aortic valve underwent TAVR using Hangzhou Solution strategy , . The detailed procedure has been previously described – . Our professional team subsequently followed up with patients at 30 days, 1 year, and yearly thereafter through face-to-face or telephonic consultation. We collected data during the 3-year follow-up to reduce the influence of limited sample size in the fourth/fifth-year follow-up (many patients did not reach the fourth or fifth follow-up window). The clinical events were defined according to the Valve Academic Research Consortium-2 consensus document . Contrast-enhanced MDCT was routinely performed on the second-generation dual-energy CT (SOMATOM Definition Flash, Siemens Medical Solutions, Germany) following our scan protocol. The anatomy of patients was assessed using 3mensio software (3mensio Medical Imaging BV, Bilthoven, The Netherlands). Aortic root anatomy was evaluated in a double-oblique reconstruction image using the best systolic phase. The severity of valve calcification was classified into no, mild, moderate, and severe, as previously described . Maximal intensity projection image was used to assess AAC (Fig. ). The aortic arch was defined as the segment between the ascending and descending aorta (between the innominate artery and the aortic isthmus) , and AAC was visible as high-density portions in this region. Notably, the calcification in the innominate, left common carotid, and left subclavian arteries was distinguished and was not included in the AAC. According to our protocol, the optimized image for aortic arch length visualization was used to assess the calcification involving the ratio of the aortic arch length. Then, an image that visually overlapped the vascular lumen of the aortic arch was assessed to identify the calcification involving the ratio of the vessel circumference. In this way, AAC was graded as no (free of calcification), mild (calcification involving ≤ 1/3rd of aortic arch length or circumference), moderate (calcification involving > 1/3rd of aortic arch length and circumference, but not both > 2/3rd of aortic arch length and circumference), and severe calcification (calcification involving > 2/3rd of aortic arch length and circumference). The typical images are shown in Fig. . The analyses of AAC were performed by two authors (D.Z and H.Y.D) experienced in TAVR pre-operative CT imaging analyses. First, the two observers assessed the AAC in 20 patients together. Thereafter, CT images of 40 randomly selected consecutive patients were separately evaluated to measure interobserver variations. The two observers reassessed these cases independently at an interval of > 3 months to measure intraobserver variations in AAC analyses. The reproducibility of this method was confirmed, and the remaining images are evaluated independently. The two experts who analyzed the MDCT images were blinded to the procedural data and clinical outcomes. Quantitative variables were expressed as means with standard deviation or median with interquartile ranges (IQR) according to the distribution as determined by Shapiro–Wilk test. Student’s t-test or Mann–Whitney U test was performed for continuous variables. All categorical data were presented as frequencies with percentages and were compared using a chi-square test or Fisher’s exact test, as appropriate. Cohen’s kappa coefficient was calculated to determine interobserver and intraobserver variations in the measurements of AAC grade. A kappa value of 0.81–1, 0.61–0.80, 0.41–0.60, 0.21–0.40, 0.01–0.20, and < 0 indicated almost perfect, substantial, fair, slight, and no agreements, respectively , . The cumulative all-cause mortality and cardiovascular/non-cardiovascular mortality rates were determined using the Kaplan–Meier survival analysis and were compared using the log-rank test. COX regression analyses were performed to evaluate the correlation of baseline characteristics and estimated 3-year all-cause mortality. Univariate COX regression analyses were performed for all baseline variables; variables with p-values < 0.05 in univariate COX regression analysis were entered in multivariate COX regression models using a forward likelihood-ratio method. A two-tailed p-value < 0.05 was considered to be statistically significant. All data analyses were performed using SPSS software version 20.0 (IBM, New York, USA). A total of 464 consecutive patients with AS who underwent TAVR were enrolled in our study, and 12.1%, 58.2%, 18.5%, and 11.2% of patients had no, mild, moderate, and severe AAC, respectively. Furthermore, we compared 138 (29.7%) patients with moderate or severe AAC and 326 (70.3%) patients with no or mild AAC. The baseline characteristics of the patients are shown in Table . Individuals in the moderate/severe AAC group were older (80.0 years [IQR: 73.8 to 84.0 years] vs.75.0 years [IQR: 70.0 to 79.0 years]; p < 0.001) and had higher Society of Thoracic Surgeons Predicted Risk of Mortality (STS-PROM) scores (7.7 [IQR:4.4 to 10.5] vs.4.8 [IQR:3.2 to 8.3]; p < 0.001) than those in no/mild AAC group. Higher prevalence of hypertension (69.6% vs. 48.8%; p < 0.001), chronic kidney disease (18.1% vs. 7.4%; p = 0.001), and history of stroke (8.7% vs. 3.4%; p = 0.016) were found in moderate/severe AAC group. The proportion of sex and smokers was not significantly different between the two groups. Further, the proportion of patients with hyperlipidemia, diabetes, chronic obstructive pulmonary diseases, prior myocardial infarction, prior percutaneous coronary intervention, or prior pacemaker implantation was not significantly different between the two groups. The echocardiography and CT data are displayed in Table . The left ventricular ejection fraction, maximum velocity, mean gradient, and aortic valve area were comparable between the two groups. Approximately 18.1% and 15.1% of patients had low-gradient AS (severe AS with mean gradient lower than 40 mm Hg) in the moderate/severe AAC and no/mild AAC groups ( p = 0.457), respectively. MDCT data revealed that patients with moderate/severe AAC had a smaller sinotubular junction, smaller ascending aorta diameter, and lower left main coronary height. The proportion of patients with moderate/severe aortic valve calcification was similar between the moderate/severe AAC and no/mild AAC groups (76.1% vs. 81.0%, p = 0.232). The reproducibility of the proposed method was determined by repeated assessment of randomly selected consecutive 40 AAC cases independently; the kappa value for interobserver variability was 0.807. The kappa values for intraobserver variation were 0.962 (D.Z) and 0.922 (H.Y.D). Kappa values can be interpreted as: ≤ 0 (no agreement), 0.01–0.20 (none to slight), 0.21–0.40 (fair), 0.41–0.60 (moderate), 0.61–0.80 (substantial), and 0.81–1.00 (almost perfect agreement). Therefore, we achieved almost perfect interobserver and intraobserver agreements. The procedural characteristics and 30-day clinical outcomes are mentioned in Table . Self-expandable valves were the most commonly used prosthesis in moderate/severe AAC (83.3%) and no/mild AAC (88.3%) groups. The transfemoral approach was the most frequently adopted access approach in the two groups (92.0% vs. 95.7%; p = 0.109). The proportion of patients who underwent balloon predilatation was similar between the two groups (97.8% vs. 97.2%, p = 0.965), while the proportion of those who underwent post-dilation was higher in the no/mild AAC group (47.8% vs. 59.8%; p = 0.017). All-cause mortality during 30 days of follow-up after TAVR was numerically more frequent in the moderate/severe AAC group than in the no/mild AAC group (3.6% vs. 1.5%); however, the difference was not statistically significant ( p = 0.284). Moreover, the rates of myocardial infarction, stroke, bleeding, and permanent pacemaker implantation did not significantly differ between the two groups. Individuals with moderate or severe AAC had a statistically higher incidence of all-cause death than those with no or mild AAC (12.3% vs. 4.9%; p = 0.005) at 1-year follow-up; non-cardiovascular deaths were the major contributory factors (6.5% vs. 0.6%; p < 0.001; Fig. ). The median follow-up time was 25.1 months (IQR: 12.4–36.0). The time-to-event analysis determined that the patients with severe, moderate, mild, and no AAC had an estimated 3-year all-cause mortality of 39.6%, 20.8%, 13.4%, and 6.7% (log-rank p < 0.001, Figure ). Figures and show that the patients with moderate/severe AAC had significantly higher 3-year all-cause mortality (27.7% vs. 12.3%, log-rank p = 0.001) and non-cardiovascular mortality (14.3% vs. 5.0%, log-rank p = 0.001) than those with no/mild AAC. Nonetheless, cardiovascular mortality was not significantly different between the two groups (15.6% vs. 7.7%, log-rank p = 0.123). Further, we evaluated patients with moderate or severe AAC separately and observed that patients with severe AAC had significantly higher incidences of cardiovascular and non-cardiovascular mortality compared with patients with no/mild AAC (cardiovascular: 31.5% vs. 7.7%, log-rank p = 0.002; non-cardiovascular: 13.1% vs. 5.0%, log-rank p = 0.009). In addition, patients with severe AAC had a higher rate of estimated 3-year cardiovascular mortality than those with moderate AAC (31.5% vs. 7.2%, log-rank p = 0.010). Patients with moderate AAC had a higher incidence of non-cardiovascular mortality but had a similar incidence of cardiovascular mortality compared to patients with no/mild AAC (cardiovascular: 7.2% vs. 7.7%, log-rank p = 0.726; non-cardiovascular: 14.7% vs. 5.0%, log-rank p = 0.003). Table shows that moderate/severe AAC was a risk factor for 3-year all-cause mortality in univariate (hazard ratio [HR]: 2.39, 95% confidence interval [CI]: 1.41–4.03; p = 0.001) and multivariate COX regression analyses (HR: 1.78, 95% CI: 1.04–3.06; p = 0.037). Higher STS-PROM strata ( p = 0.011), stage 4 or 5 chronic kidney disease ( p = 0.027), and low gradient AS ( p = 0.018) were suggested as independent predictors in multivariate regression analyses. Furthermore, in another COX regression analyses model in which AAC was classified according to the grade of arch calcification, a more severe AAC grade was also an independent predictor of all-cause mortality (univariate: p < 0.001; multivariate: p = 0.025) and the above-mentioned independent predictors were still valid (Table ). The main findings of our studies are as follows: (1) we have proposed a rapid and reproducible method to assess AAC; (2) among patients who underwent TAVR in our centers, 87.9% patients had AAC; of these, 18.5% and 11.2% had moderate and severe AAC, respectively; (3) patients with moderate/severe AAC had a higher incidence of 1-year all-cause mortality and non-cardiovascular mortality; (4) moderate/severe AAC was an independent predictor of 3-year all-cause mortality; (5) compared with no/mild AAC, severe AAC was associated with a higher incidence of 3-year cardiovascular and non-cardiovascular mortalities, whereas moderate AAC was associated with a higher incidence of non-cardiovascular mortality. The assessment of aortic arch calcification is important for the operator since many manipulations during the procedure involve the aortic arch. The current main limiting factor in evaluating AAC is the lack of an easy, effective, and repeatable method. Although calcium volume score can provide a quantitative assessment of calcification, calculating the aortic arch calcium volume score is complicated. The analysts need to decide centerline from the aortic root to descending aorta, make a reconstruction, place a region of interest to measure average Hounsfield units, and manually adjust the best cut-off of Hounsfield units to optimize calcification detection. Few centers routinely do this analysis since clinicians usually focus more on the anatomy of the aortic root and access. In addition, simply reporting arch calcium volume score misses information regarding calcification distribution. AAC can be evaluated using another semiquantitative assessment method. Sagittal and reconstructed axial CT images are analyzed slice-by-slice, and the severity of arch calcification is determined by estimating the degree of calcification involving the vessel length and circumference . This method is also relatively complicated and may have potentially large intraobserver and interobserver differences due to the slice-by-slice analyses. Therefore, we designed this study to provide a rapid and useful method to assess AAC. Using our method, the AAC can be easily assessed during the analysis of access anatomy. During routinely analyzing aortic valve root structure before TAVR, it takes only 20 s to 1 min to finish the analyses and grading of AAC. Besides, adding typical imaging to the report of the anatomy of the patient is easy. The clinicians can also visualize the information of the AAC in patients just by reading the report. Severe AAC and porcelain aorta can be quickly identified, and then the operators can decide whether the patient should receive a transapical-TAVR rather than transfemoral-TAVR . In addition, the application of cerebral protection devices has increased rapidly in recent years , . Some cerebral protection devices are placed in the aortic arch and information on the anatomy of the aortic arch is required. The routine use of this rapid assessment method can also provide information for evaluating the availability of cerebral protection devices. In addition to providing important information for making clinical decision, using this approach to grade AAC and record images can also provide important data and information for conducting further research using machine learning techniques. Currently, more and more studies use machine learning techniques to enhance diagnostic accuracy and prediction by optimizing feature extraction and model performance. Related work in areas such as structural health monitoring and advanced imaging analysis has demonstrated the potential of these techniques , , . Besides, novel algorithms have been proposed to enhance training performance and have been proven their predictive power in recent years , . However, the lack of an easy evaluate method of AAC results in this information being ignored in most center, restricting application of machine learning techniques to assess the predictive value of AAC in TAVR patients. Therefore, this method might help to conduct further machine learning studies. Currently, the knowledge of AAC in old patients with severe AS is scarce. In our study, a considerable number of patients (29.7%) had moderate or severe AAC, suggesting that the occurrence of AAC should be critically considered. Intimal injury caused by hypertension and calcium and phosphorus metabolism disorder caused by renal insufficiency can lead to AAC. Thus, patients with moderate or severe AAC were older and had a higher proportion of hypertension and renal insufficiency, which were consistent with those of previous studies conducted in other populations , , . Moreover, previous studies have indicated that prevalent stroke was associated with AAC , . Similarly, in our study, a history of stroke was more common in patients with moderate or severe AAC. In a population-based cohort study including 2,408 participants from the Rotterdam study, AAC was found to be the strongest indicator of cardiovascular and non-cardiovascular mortalities among major vessel beds . However, to the best of our knowledge, our study is the first one to illustrate the relationship between AAC and mid-to-long term clinical outcomes in patients who underwent TAVR. We also found that the patients with moderate or severe AAC had a higher incidence of estimated 3-year all-cause and non-cardiovascular mortalities, whereas patients with severe AAC had a higher incidence of 3-year all-cause, cardiovascular, and non-cardiovascular mortalities. Therefore, it is plausible that a better management of comorbidity after TAVR is needed in patients with moderate/severe AAC, especially in patients with severe AAC. For example, over 50% of patients can have an increased blood pressure after TAVR – , and management of hypertension in such patients may help delay further progression of aortic arch atherosclerosis and calcification. We also found that higher STS-PROM risk strata, chronic kidney disease stage 4 or 5, and low gradient AS were independent predictors of the 3-year all-cause mortality. STS score includes several clinical characteristics (such as age, comorbidity, and medication) and is widely used in clinical practice. Low-gradient AS was associated with worse left heart function. All these three factors were classic strong risk factors for worse clinical outcomes. Although patients with moderate/severe AAC had higher STS scores and a higher proportion of chronic kidney disease stage 4 or 5 than those with no/mild AAC, moderate/severe AAC was still an independent predictor in multivariate COX regression analysis. Our study has some limitations. First, this method cannot identify low-density atherosclerosis. However, this rapid method can be combined with other assessment methods. Therefore, if obvious low-density atherosclerosis plaque or ulcers are present, we recommend that the analysts should further evaluate low-density atherosclerosis plaque or ulcers in the aortic arch in addition to the routine use of our method. Second, this method is only focused on assessing AAC. Nevertheless, the AAC can reflect the total burden of vascular calcification given its central location in the arterial system , . Third, since patients who had a history of aortic valve replacement or did not have analyzable MDCT images were excluded, patient selection biases may exist in this study. Fourth, although almost perfect agreement was achieved in our reproducibility test, there still existed some observer variability (especially interobserver variations). The variability were mainly found in analyzing AAC images with boundary values (such as 1/3rd and 2/3rd). Therefore, we suggest that clinicians should be trained using typical cases to reduce interobserver variability in clinical practice. Besides, some baseline characteristics differed between patients with moderate/severe and no/mild AAC (such as age, STS score, and renal function). Although moderate/severe AAC was an independent predictor of 3-year all-cause mortality in multivariate regression analysis, a well-designed study matching baseline characteristics well between different groups is needed to further verify our results. Finally, the sample size was relatively small and the follow-up was not long enough. Future large-scale studies with longer follow-ups are needed to validate our findings. We developed a semi-quantitative method for evaluating patients undergoing TAVR and classified AAC into no, mild, moderate, and severe. The method had good reproducibility and can be routinely used to rapidly assess AAC in clinical practice. We found that approximately 30% of patients undergoing TAVR had a moderate/severe AAC, which was associated with a higher incidence of 1-year and 3-year all-cause mortalities. In addition, patients with moderate or severe AAC had an increased incidence of 3-year non-cardiovascular mortality, whereas only patients with severe AAC had a higher risk of 3-year cardiovascular mortality. Below is the link to the electronic supplementary material. Supplementary Material 1 |
673e498a-2c6e-4c5c-ae30-eaa34a9368e7 | 11935578 | Cardiovascular System[mh] | Phosphorylation of α1 C at serine 1928 (S1928) controls Ca V 1.2 channels spatiotemporal remodeling during activation of angiotensin II signaling and hypertension. Phosphorylation of α1 C at S1928 underlies increased arterial myocyte intracellular calcium, vascular reactivity, and blood pressure during angiotensin II signaling and hypertension.
Phosphorylation of α1 C at S1928 may be a rheostat of vascular function in health and disease. Increased phosphorylation of α1 C at S1928 may be a risk factor underlying vascular complications during hypertension. Results may help develop therapeutics with single amino acid accuracy to reduce α1 C S1928 phosphorylation, which may ameliorate hypertension‐related vascular complications.
Data Availability Statement All data are included in the article, and detailed methodology, (Table ), (Figures through ), full unedited blots, and source Matlab codes are included in Data . Mouse Study Approval Given the well‐known sex‐dependent differences in Ang II‐induced hypertension, male wild‐type (WT) C57BL/6J and S1928A mice and male BP normal (BPN) and BPH mice were used to properly power experiments to resolve statistical differences between data sets. To assess potential sex differences, a set of properly powered experiments were done using female mice/samples, although this was not a primary goal of the study. All studies conform with the US National Institutes of Health Guide for the Care and Use of Laboratory Animals and were carried out in strict accordance with the protocols and guidelines approved by the Institutional Animal Care and Use Committee of the University of California, Davis. In Vivo Ang II Infusion Chronic Ang II infusion was performed using minipumps subcutaneously implanted in WT and S1928A mice. , Dissection of Mesenteric Arteries and Isolation of Arterial Myocytes Mice were euthanized by a lethal intraperitoneal injection of sodium pentobarbital. Mesenteric arteries were isolated, and arterial myocytes were dissociated using enzymatic/mechanical dissociation. Electrophysiology Whole‐cell and cell‐attached electrophysiology was performed in isolated mesenteric arterial myocytes using an Axopatch 200B amplifier. Data were analyzed using pClamp. Computational Modeling Simulations were performed using an established mathematical model of membrane electrophysiology and Ca 2+ cycling. , Imaging of Cell Ca 2+ and Contraction Ca 2+ imaging in mesenteric cells was performed using a spinning disk confocal microscope. The fluorescence signal was also used to calculate contractility. Superresolution and PLA α 1C protein distribution and clustering were determined using a direct stochastic optical reconstruction microscopy superresolution microscope. PLA was used to define close association of α 1C subunits. , Laser Speckle Blood flow in anesthetized animals was measured using laser speckle imaging. Edge detection was used to calculate arterial diameter from the images. , Pressure Myography Myogenic/vascular tone was calculated in pressurized mesenteric arteries. , , Blood Pressure and Echocardiography BP was measured in freely moving mice using DSI telemetry. Echocardiography was performed in anesthetized mice using a Vevo 2100 system. Statistical Analysis Data were analyzed using GraphPad Prism v10 or R Studio software and expressed as mean±SEM. The Shapiro–Wilk normality test was used to assess whether a data set deviated significantly from a normal (eg, Gaussian) distribution. The comparisons between groups were performed using hierarchical “nested” analyses to account for the number of mice and replicates, with appropriate t test, 1‐way ANOVA, 2‐way ANOVA, and 3‐way ANOVA with Bonferroni post hoc test. Parametric or nonparametric t tests and ANOVAs were applied when nested analyses were not implemented. Unless otherwise indicated, P <0.05 was considered statistically significant.
All data are included in the article, and detailed methodology, (Table ), (Figures through ), full unedited blots, and source Matlab codes are included in Data .
Given the well‐known sex‐dependent differences in Ang II‐induced hypertension, male wild‐type (WT) C57BL/6J and S1928A mice and male BP normal (BPN) and BPH mice were used to properly power experiments to resolve statistical differences between data sets. To assess potential sex differences, a set of properly powered experiments were done using female mice/samples, although this was not a primary goal of the study. All studies conform with the US National Institutes of Health Guide for the Care and Use of Laboratory Animals and were carried out in strict accordance with the protocols and guidelines approved by the Institutional Animal Care and Use Committee of the University of California, Davis.
Ang II Infusion Chronic Ang II infusion was performed using minipumps subcutaneously implanted in WT and S1928A mice. ,
Mice were euthanized by a lethal intraperitoneal injection of sodium pentobarbital. Mesenteric arteries were isolated, and arterial myocytes were dissociated using enzymatic/mechanical dissociation.
Whole‐cell and cell‐attached electrophysiology was performed in isolated mesenteric arterial myocytes using an Axopatch 200B amplifier. Data were analyzed using pClamp.
Simulations were performed using an established mathematical model of membrane electrophysiology and Ca 2+ cycling. ,
2+ and Contraction Ca 2+ imaging in mesenteric cells was performed using a spinning disk confocal microscope. The fluorescence signal was also used to calculate contractility.
PLA α 1C protein distribution and clustering were determined using a direct stochastic optical reconstruction microscopy superresolution microscope. PLA was used to define close association of α 1C subunits. ,
Blood flow in anesthetized animals was measured using laser speckle imaging. Edge detection was used to calculate arterial diameter from the images. ,
Myogenic/vascular tone was calculated in pressurized mesenteric arteries. , ,
BP was measured in freely moving mice using DSI telemetry. Echocardiography was performed in anesthetized mice using a Vevo 2100 system.
Data were analyzed using GraphPad Prism v10 or R Studio software and expressed as mean±SEM. The Shapiro–Wilk normality test was used to assess whether a data set deviated significantly from a normal (eg, Gaussian) distribution. The comparisons between groups were performed using hierarchical “nested” analyses to account for the number of mice and replicates, with appropriate t test, 1‐way ANOVA, 2‐way ANOVA, and 3‐way ANOVA with Bonferroni post hoc test. Parametric or nonparametric t tests and ANOVAs were applied when nested analyses were not implemented. Unless otherwise indicated, P <0.05 was considered statistically significant.
All experiments were performed using freshly isolated mouse mesenteric arteries and arterial myocytes (ie, immediate use after isolation/dissociation). The rationale for this is that mesenteric arteries and arterial myocytes play a key role in regulating BP, and their use immediately after isolation/dissociation may better reflect the native environment. Ang II Increases Vascular Ca V 1 .2 Function Via pS1928 Acute Ang II exposure has been shown to increase Ca V 1.2 channel activity in arterial myocytes. , , , Using whole‐cell patch clamp with barium as the charge carrier and a single voltage step from −70 to +10 mV (maximal current), , , , , we confirmed that Ang II exposure increased Ca V 1.2 current density (whole‐cell barium current) in WT arterial myocytes (Figure ). In stark contrast, Ang II failed to augment whole‐cell barium current in arterial myocytes from a genetically modified mouse in which the α1 C serine 1928 position was mutated to alanine to prevent its phosphorylation (S1928A mouse ; Figure ). Accordingly, the Ang II‐induced change in whole‐cell barium current was 3.3±0.6 pA/pF in WT cells versus 0.3±0.2 pA/pF in S1928A cells ( P =0.0028 with unpaired t test; Shapiro–Wilk normality test P =0.6173). Note that α1 C , and PKCα (Figure ), as well as BKα, BKβ1, and K V 2.1 protein abundance in arterial lysates and basal voltage dependency of activation and inactivation of Ca V 1.2 channels, are similar in WT and S1928A male arterial myocytes, suggesting that changes in the expression of key proteins and Ca V 1.2 biophysical properties do not account for the Ang II effects. These results indicate that pS1928 is necessary for Ang II‐induced potentiation of vascular whole‐cell barium current in male arterial myocytes. Activation of Ang II signaling has been shown to promote cooperative gating of vascular Ca V 1.2 channels. , , To examine if pS1928 is important for this gating mode, cell‐attached electrophysiology with Ca 2+ as the charge carrier was done in freshly isolated WT and S1928A male arterial myocytes. Ang II significantly increased Ca V 1.2 channel nPo (ie, n is the number of channels and Po is the channel open probability; Figure and ) and availability (ie, likelihood of at least one event per sweep; Figure and ) in WT cells (nested t test between WT controls and WT Ang II). In addition, the frequency of Ca V 1.2 cooperative events (ie, number of traces showing openings with 2 or more channels; see insets in Figure ) and the coupling strength (ie, κ, which was quantified using a Markov chain model , ) were higher in WT cells after Ang II (Figure and ). Conversely, Ang II failed to increase Ca V 1.2 channel nPo, availability, coupling frequency, or coupling strength in S1928A cells (Figure through ). Results in control and Ang II‐treated S1928A male cells were comparable to WT male controls. Overall, these results suggest that pS1928 is necessary for Ang II to increase Ca V 1.2 channel activity and cooperative gating behavior. Similar single‐channel experiments performed in WT female arterial myocytes revealed no differences in nPo and coupling frequency but significant differences in availability and coupling strength compared with WT male cells (Figure through ). Ang II increased coupling frequency, and trends toward higher values were found in nPo and availability properties but not coupling strength in female arterial myocytes (Figure and through ). These results suggest sex‐dependent differences in basal conditions and mechanisms by which Ang II modulates vascular Ca V 1.2 channel activity and cooperative gating. pS1928 Mediates Ang II ‐Induced Vascular α1 C Clustering The Ang II‐induced elevations in Ca V 1.2 cooperative gating prompted us to examine if this was due to increased α1 C clustering in the sarcolemma of arterial myocytes. For this, we performed superresolution imaging of α1 C in freshly dissociated male mesenteric arterial myocytes (Figure ). , We used pair‐correlation analysis to characterize the physical properties of α1 C clusters and objectively quantify spatial scales of density fluctuations (Figure ). , This analysis is not influenced by the multiple appearances of a single fluorescent molecule detected multiple times across all frames before being irreversibly photobleached. , Data revealed a significant increase in cluster size (Figure ) and the estimated number of molecules per cluster (N clusters ; Figure ) but not cluster density (Figure ) in WT arterial myocytes after Ang II. Changes in α1 C cluster size and N clusters were a direct consequence of pS1928, as cluster size (Figure ), cluster density (Figure ), and N clusters (Figure ) were similar in S1928A cells in control and Ang II conditions and comparable to WT control cells. Note that PKA inhibition does not prevent the increase in Ang II‐induced α1 C clustering properties (Figure through ). Thus, PKA is not involved in the pathway triggering the α1 C clustering in response to Ang II. Intriguingly, superresolution experiments in female mesenteric arterial myocytes found an increase in basal α1 C cluster size and density, but not N clusters , compared with male mesenteric myocytes (Figure and through ). Moreover, Ang II did not elicit significant changes in α1 C cluster size, cluster density, and N clusters in female WT and S1928A myocytes (Figure , , through ). These results reveal sex‐dependent differences in α1 C clustering mechanisms at the basal and Ang II‐stimulated levels. The superresolution results in male arterial myocytes were confirmed using a modified PLA approach with an α1 C monoclonal antibody labeled with the PLUS or MINUS probe that has been extensively validated to test α1 C oligomerization/clustering (Figure ). , PLA showed that Ang II increased the number of α1 C PLA puncta per cell area in WT male arterial myocytes but not WT cells pretreated with a broad PKC inhibitor or S1928A cells (Figure through ). PKC inhibition in WT cells did not change the α1 C PLA puncta per cell area compared with WT control and PLA puncta were never observed if 1 of the primary antibody probes was omitted (Figure and ). These results suggest Ang II/PKC‐dependent pS1928 increases α1 C clustering in male arterial myocytes. Ang II Increases Arterial Myocyte Ca 2+ and Contractility Via pS1928 Ca V 1.2 cooperative gating amplifies Ca 2+ influx into cells, accounting for ~50% of the total Ca 2+ influx in arterial myocytes. , This is important because Ang II promotes this Ca V 1.2 gating modality, influencing arterial myocyte Ca 2+ and contractility. Thus, the role of pS1928 in modulating arterial myocyte excitability in response to Ang II signaling was initially investigated using an in silico model. We did this to prevent oversimplification and consider the complex cascade of ionic conductances that may be altered and influence cell electrophysiology and Ca 2+ properties during Ang II signaling activation. , The model was fitted with experimental data showing Ang II‐induced alterations in the activity of Ca V 1.2 channels, several K + channels, and transient receptor potential channels. , , , , , We also modeled the effect of preventing pS1928 (as in S1928A cells) upon Ang II to assess the role of pS1928 on arterial myocyte membrane potential and [Ca 2+ ] i . The model predicted that Ang II could promote arterial myocyte membrane depolarization in WT and S1928A cells, but pS1928 was still necessary for elevating [Ca 2+ ] i in arterial myocytes (Figure ). To test this prediction, we first measured the membrane potential of WT and S1928A freshly isolated arterial myocytes upon Ang II using the perforated whole‐cell patch‐clamp. , Basal membrane potential was similar in WT (−56±2 mV) and S1928A (−57±2 mV) cells (Figure and ). Ang II depolarized membrane potential to about the same magnitude in both cell types (−35±3 mV in WT and −32±3 mV in S1928A; Figure and ). These results are consistent with the in silico prediction and suggest that Ang II effects on membrane potential in arterial myocytes are not reliant on pS1928. To test the prediction that pS1928 is necessary for Ang II‐induced elevations in [Ca 2+ ] i , male arterial myocytes loaded with the Ca 2+ indicator fluo 4‐AM were used. This fluorescent signal was also used to track cell length and assess contractility (see Methods for details). Data showed that Ang II increased peak [Ca 2+ ] i with a concomitant contraction in WT arterial myocytes (Figure ). WT cells pretreated with the Ca V 1.2 channel blocker nifedipine or the broad PKC inhibitor calphostin C did not show the Ang II‐induced elevation in [Ca 2+ ] i and cell contraction, indicating that Ca V 1.2 channel activity and PKC are required for the Ang II effects. Moreover, Ang II failed to elevate [Ca 2+ ] i and induce contraction in S1928A cells. These results are consistent with the in silico model prediction and suggest that pS1928 is necessary for the Ang II‐induced, PKC‐dependent elevation of [Ca 2+ ] i leading to contraction of male arterial myocytes. pS1928 Regulates Arterial Diameter and Blood Flow We measured vascular tone ex vivo in response to Ang II in WT and S1928A mesenteric arteries to corroborate pS1928 physiological relevance. Mesenteric arteries from WT and S1928A male mice were freshly dissected and pressurized to 80 mm Hg. Basal myogenic tone (ie, pressure‐induced constriction) was lower in S1928A mesenteric arteries compared with WT (Figure and ). When exposed to Ang II, both WT and S1928A arteries constricted to about the same magnitude (Figure and ). However, in the continuous presence of Ang II, WT arteries reached a new steady‐state plateau level with higher vascular tone (ie, agonist‐induced constriction) compared with S1928A arteries (Figure and ). Similar Ang II‐induced responses were observed in pressurized cerebral pial arteries (Figure through ), suggesting conserved mechanisms in different vascular beds. These data indicate that Ang II‐induced elevation in steady‐state tone in male arteries requires pS1928. To confirm the in vivo physiological significance of pS1928 upon Ang II signaling activation, arterial diameter and blood flow were measured in response to different Ang II concentrations in exposed mesenteric arteries of anesthetized WT and S1928A male mice using laser speckle imaging. , , A solution containing vasodilatory drugs was used to promote maximal dilation to normalize arterial diameter and blood flow measurements. Under control conditions, basal tone was higher and corresponding blood flow was lower in mesenteric arteries of WT male mice compared with S1928A (Figure , and ), consistent with the ex vivo basal tone data. Ang II induced a concentration‐dependent decrease in arterial diameter and concomitant reduction in blood flow (ie, flux) in mesenteric arteries from WT and S1928A mice (Figure and ). Peak constriction and corresponding flux were relatively similar at most points examined immediately after Ang II (Figure , and ; Figure and ). However, 5 minutes after Ang II, arteries from the S1928A mice showed larger diameter and flux compared with arteries from WT male mice (Figure , and through ). The Ca V 1.2 channel blocker nifedipine prevented the Ang II effects on arterial diameter and flux (Figure and Figure and ). Altogether, these results suggest a key role for pS1928 and L‐type Ca 2+ channels in the in vivo regulation of male arterial diameter (and vascular tone) and blood flow upon Ang II. Laser speckle experiments in mesenteric arteries from WT female mice revealed lower basal tone and Ang II‐induced changes in diameter and flux compared with male mice (Figure through ), consistent with reported sex differences in these properties. Moreover, while Ang II induced a concentration‐dependent decrease in arterial diameter and concomitant reduction in blood flow (ie, flux) in mesenteric arteries from WT and S1928A female mice, preventing pS1928 did not lessen the Ang II response (Figure through ). These results suggest sex‐dependent differences in mechanisms by which Ang II modulates vascular function in vivo. pS1928 Enhances Vascular Ca V 1 .2 Activity and Cooperativity in Hypertension Considering that chronic activation of Ang II signaling contributes to enhanced Ca V 1.2 channel activity and the development of hypertension, , , , , , , , we examined if pS1928 underlies the alterations in channel properties. Unitary data showed that channel nPo and availability were increased in WT hypertensive arterial myocytes compared with WT sham cells (Figure and and Figure ). The frequency and strength (κ) of cooperative Ca V 1.2 events showed a robust trend toward higher values in WT hypertensive myocytes compared with WT sham cells (Figure and ). The changes in Ca V 1.2 channel properties in WT hypertensive cells were not observed in S1928A hypertension compared with S1928A sham cells (Figure through ; Figure ). Moreover, channel nPo and availability, as well as coupling frequency and strength, were significantly higher in WT hypertension compared with S1928A hypertensive myocytes. These results suggest that pS1928 triggers increased vascular Ca V 1.2 channel function in male arterial myocytes during hypertension. pS1928 Mediates Increased Vascular α1 C Clustering in Hypertension We next examined if α1 C clustering was increased in freshly isolated male arterial myocytes during hypertension and if this process required pS1928. Superresolution imaging with pair‐correlation analysis (Figure and ) found an increase in α1 C cluster size (Figure ), no significant differences in cluster density (Figure ), and trends toward higher N clusters (Figure ) in WT hypertensive arterial myocytes compared with WT sham and S1928A HTN cells. These changes were not observed in S1928A hypertensive cells compared with S1928A sham, which had similar α1 C cluster properties as WT sham (Figure through ). The modified PLA assay confirmed the increase in α1 C clustering in WT hypertension compared with WT sham, and S1928A sham and hypertensive cells (Figure ; Figure ). No PLA signal was detected when one of the primary antibody‐probe complexes was omitted from the preparation (Figure ; Figure ). Intriguingly, no change in total α1 C protein abundance was detected by Western blot in WT and S1928A hypertensive lysates compared with sham conditions (Figure ). These results were unexpected as prior studies have shown increased total α1 C protein abundance in arterial lysates during hypertension. Results may point to offsetting changes, possibly associated with the hypertension model. Regardless, these results indicate that α1 C clustering properties in male arterial myocytes are elevated during hypertension, and this spatial remodeling is mediated by pS1928. To examine if similar increases in α1 C clustering were observed in different models of hypertension, superresolution experiments were performed using male arterial myocytes from the BPH compared to BPN mice. Hypertension in BPH mice is thought to be mediated by sympathetic overactivity, which acting through the α‐adrenergic receptors/PKC axis, may modulate α1 C clustering. Data show that α1 C clustering is increased in BPH arterial myocytes compared with BPN cells (Figure through ). The α1 C superclustering in BPH arterial myocytes correlated with increased Ca V 1.2 nPo and cooperativity previously reported in these cells. Moreover, in pressurized arteries, noradrenaline caused an initial peak constriction of similar magnitude in WT and S1928A vessels (Figure ). Yet, in the continuous presence of noradrenaline, WT arteries reached a steady‐state level (eg, plateau level) with higher vascular tone compared with S1928A arteries (Figure ). These results are consistent with the Ang II observations here and prior studies indicating that noradrenaline increases L‐type Ca 2+ channel activity leading to enhanced vascular tone and suggest that noradrenaline‐induced elevation in steady‐state tone in mesenteric arteries requires α1 C S1928 phosphorylation. Results further indicate that enhanced α1 C clustering may be a general feature in hypertensive arterial myocytes to increase Ca V 1.2 activity and cooperativity, and vascular tone. pS1928 Underlies Increased Myogenic Tone and Blood Pressure in Hypertension To establish the physiological relevance of pS1928 during hypertension, we measured myogenic tone in pressurized male mesenteric arteries. We found that WT hypertensive arteries developed higher myogenic tone at physiologically relevant intravascular pressures (ie, >60 mm Hg) than WT sham arteries (Figure ). The hypertension‐induced elevations in myogenic tone were not observed when comparing tone levels in S1928A sham and hypertensive arteries (Figure ). These results suggest that elevations in myogenic tone in male mesenteric arteries during hypertension require pS1928. Lastly, we examined the role of pS1928 in BP regulation in freely moving WT and S1928A male mice implanted with radiotelemetry devices and osmotic minipumps eluting either saline (ie, sham) or Ang II (ie, hypertension). BP measurements were analyzed at day 0 and day 7 after minipump implantation. Telemetry data showed that resting (Day 0) systolic pressure (Figure ), diastolic pressure (Figure ), and mean arterial pressure (MAP; Figure ) were similar in WT and S1928A mice. However, a significant increase in systolic pressure (Figure ), diastolic pressure (Figure ), and MAP (Figure ) was observed in both WT and S1928A mice after 7 days with a minipump eluting Ang II compared with saline. The increase in MAP in S1928A hypertensive mice was not surprising as BP is regulated by many factors in addition to peripheral vascular resistance, such as alterations in sympathetic activity, baroreflex dysfunction, or renal abnormalities. Importantly, the change in MAP (ΔMAP) during hypertension was significantly larger in WT mice compared with S1928A mice (Figure ). These results are remarkable in that preventing phosphorylation of a single amino acid in α1 C significantly ameliorates Ang II‐induced hypertension. Note that, in our hands, pulse pressure (an indicator of arterial stiffness and cardiac contraction; Figures ), heart rate (Figure ), and cardiac hemodynamics (Figure through ) were similar in WT and S1928A sham and hypertensive mice. Overall, these results suggest that pS1928 contributes to increasing BP without a major impact on cardiac function in an Ang II‐induced hypertensive male mouse model.
Increases Vascular Ca V 1 .2 Function Via pS1928 Acute Ang II exposure has been shown to increase Ca V 1.2 channel activity in arterial myocytes. , , , Using whole‐cell patch clamp with barium as the charge carrier and a single voltage step from −70 to +10 mV (maximal current), , , , , we confirmed that Ang II exposure increased Ca V 1.2 current density (whole‐cell barium current) in WT arterial myocytes (Figure ). In stark contrast, Ang II failed to augment whole‐cell barium current in arterial myocytes from a genetically modified mouse in which the α1 C serine 1928 position was mutated to alanine to prevent its phosphorylation (S1928A mouse ; Figure ). Accordingly, the Ang II‐induced change in whole‐cell barium current was 3.3±0.6 pA/pF in WT cells versus 0.3±0.2 pA/pF in S1928A cells ( P =0.0028 with unpaired t test; Shapiro–Wilk normality test P =0.6173). Note that α1 C , and PKCα (Figure ), as well as BKα, BKβ1, and K V 2.1 protein abundance in arterial lysates and basal voltage dependency of activation and inactivation of Ca V 1.2 channels, are similar in WT and S1928A male arterial myocytes, suggesting that changes in the expression of key proteins and Ca V 1.2 biophysical properties do not account for the Ang II effects. These results indicate that pS1928 is necessary for Ang II‐induced potentiation of vascular whole‐cell barium current in male arterial myocytes. Activation of Ang II signaling has been shown to promote cooperative gating of vascular Ca V 1.2 channels. , , To examine if pS1928 is important for this gating mode, cell‐attached electrophysiology with Ca 2+ as the charge carrier was done in freshly isolated WT and S1928A male arterial myocytes. Ang II significantly increased Ca V 1.2 channel nPo (ie, n is the number of channels and Po is the channel open probability; Figure and ) and availability (ie, likelihood of at least one event per sweep; Figure and ) in WT cells (nested t test between WT controls and WT Ang II). In addition, the frequency of Ca V 1.2 cooperative events (ie, number of traces showing openings with 2 or more channels; see insets in Figure ) and the coupling strength (ie, κ, which was quantified using a Markov chain model , ) were higher in WT cells after Ang II (Figure and ). Conversely, Ang II failed to increase Ca V 1.2 channel nPo, availability, coupling frequency, or coupling strength in S1928A cells (Figure through ). Results in control and Ang II‐treated S1928A male cells were comparable to WT male controls. Overall, these results suggest that pS1928 is necessary for Ang II to increase Ca V 1.2 channel activity and cooperative gating behavior. Similar single‐channel experiments performed in WT female arterial myocytes revealed no differences in nPo and coupling frequency but significant differences in availability and coupling strength compared with WT male cells (Figure through ). Ang II increased coupling frequency, and trends toward higher values were found in nPo and availability properties but not coupling strength in female arterial myocytes (Figure and through ). These results suggest sex‐dependent differences in basal conditions and mechanisms by which Ang II modulates vascular Ca V 1.2 channel activity and cooperative gating.
Mediates Ang II ‐Induced Vascular α1 C Clustering The Ang II‐induced elevations in Ca V 1.2 cooperative gating prompted us to examine if this was due to increased α1 C clustering in the sarcolemma of arterial myocytes. For this, we performed superresolution imaging of α1 C in freshly dissociated male mesenteric arterial myocytes (Figure ). , We used pair‐correlation analysis to characterize the physical properties of α1 C clusters and objectively quantify spatial scales of density fluctuations (Figure ). , This analysis is not influenced by the multiple appearances of a single fluorescent molecule detected multiple times across all frames before being irreversibly photobleached. , Data revealed a significant increase in cluster size (Figure ) and the estimated number of molecules per cluster (N clusters ; Figure ) but not cluster density (Figure ) in WT arterial myocytes after Ang II. Changes in α1 C cluster size and N clusters were a direct consequence of pS1928, as cluster size (Figure ), cluster density (Figure ), and N clusters (Figure ) were similar in S1928A cells in control and Ang II conditions and comparable to WT control cells. Note that PKA inhibition does not prevent the increase in Ang II‐induced α1 C clustering properties (Figure through ). Thus, PKA is not involved in the pathway triggering the α1 C clustering in response to Ang II. Intriguingly, superresolution experiments in female mesenteric arterial myocytes found an increase in basal α1 C cluster size and density, but not N clusters , compared with male mesenteric myocytes (Figure and through ). Moreover, Ang II did not elicit significant changes in α1 C cluster size, cluster density, and N clusters in female WT and S1928A myocytes (Figure , , through ). These results reveal sex‐dependent differences in α1 C clustering mechanisms at the basal and Ang II‐stimulated levels. The superresolution results in male arterial myocytes were confirmed using a modified PLA approach with an α1 C monoclonal antibody labeled with the PLUS or MINUS probe that has been extensively validated to test α1 C oligomerization/clustering (Figure ). , PLA showed that Ang II increased the number of α1 C PLA puncta per cell area in WT male arterial myocytes but not WT cells pretreated with a broad PKC inhibitor or S1928A cells (Figure through ). PKC inhibition in WT cells did not change the α1 C PLA puncta per cell area compared with WT control and PLA puncta were never observed if 1 of the primary antibody probes was omitted (Figure and ). These results suggest Ang II/PKC‐dependent pS1928 increases α1 C clustering in male arterial myocytes.
Increases Arterial Myocyte Ca 2+ and Contractility Via pS1928 Ca V 1.2 cooperative gating amplifies Ca 2+ influx into cells, accounting for ~50% of the total Ca 2+ influx in arterial myocytes. , This is important because Ang II promotes this Ca V 1.2 gating modality, influencing arterial myocyte Ca 2+ and contractility. Thus, the role of pS1928 in modulating arterial myocyte excitability in response to Ang II signaling was initially investigated using an in silico model. We did this to prevent oversimplification and consider the complex cascade of ionic conductances that may be altered and influence cell electrophysiology and Ca 2+ properties during Ang II signaling activation. , The model was fitted with experimental data showing Ang II‐induced alterations in the activity of Ca V 1.2 channels, several K + channels, and transient receptor potential channels. , , , , , We also modeled the effect of preventing pS1928 (as in S1928A cells) upon Ang II to assess the role of pS1928 on arterial myocyte membrane potential and [Ca 2+ ] i . The model predicted that Ang II could promote arterial myocyte membrane depolarization in WT and S1928A cells, but pS1928 was still necessary for elevating [Ca 2+ ] i in arterial myocytes (Figure ). To test this prediction, we first measured the membrane potential of WT and S1928A freshly isolated arterial myocytes upon Ang II using the perforated whole‐cell patch‐clamp. , Basal membrane potential was similar in WT (−56±2 mV) and S1928A (−57±2 mV) cells (Figure and ). Ang II depolarized membrane potential to about the same magnitude in both cell types (−35±3 mV in WT and −32±3 mV in S1928A; Figure and ). These results are consistent with the in silico prediction and suggest that Ang II effects on membrane potential in arterial myocytes are not reliant on pS1928. To test the prediction that pS1928 is necessary for Ang II‐induced elevations in [Ca 2+ ] i , male arterial myocytes loaded with the Ca 2+ indicator fluo 4‐AM were used. This fluorescent signal was also used to track cell length and assess contractility (see Methods for details). Data showed that Ang II increased peak [Ca 2+ ] i with a concomitant contraction in WT arterial myocytes (Figure ). WT cells pretreated with the Ca V 1.2 channel blocker nifedipine or the broad PKC inhibitor calphostin C did not show the Ang II‐induced elevation in [Ca 2+ ] i and cell contraction, indicating that Ca V 1.2 channel activity and PKC are required for the Ang II effects. Moreover, Ang II failed to elevate [Ca 2+ ] i and induce contraction in S1928A cells. These results are consistent with the in silico model prediction and suggest that pS1928 is necessary for the Ang II‐induced, PKC‐dependent elevation of [Ca 2+ ] i leading to contraction of male arterial myocytes.
Regulates Arterial Diameter and Blood Flow We measured vascular tone ex vivo in response to Ang II in WT and S1928A mesenteric arteries to corroborate pS1928 physiological relevance. Mesenteric arteries from WT and S1928A male mice were freshly dissected and pressurized to 80 mm Hg. Basal myogenic tone (ie, pressure‐induced constriction) was lower in S1928A mesenteric arteries compared with WT (Figure and ). When exposed to Ang II, both WT and S1928A arteries constricted to about the same magnitude (Figure and ). However, in the continuous presence of Ang II, WT arteries reached a new steady‐state plateau level with higher vascular tone (ie, agonist‐induced constriction) compared with S1928A arteries (Figure and ). Similar Ang II‐induced responses were observed in pressurized cerebral pial arteries (Figure through ), suggesting conserved mechanisms in different vascular beds. These data indicate that Ang II‐induced elevation in steady‐state tone in male arteries requires pS1928. To confirm the in vivo physiological significance of pS1928 upon Ang II signaling activation, arterial diameter and blood flow were measured in response to different Ang II concentrations in exposed mesenteric arteries of anesthetized WT and S1928A male mice using laser speckle imaging. , , A solution containing vasodilatory drugs was used to promote maximal dilation to normalize arterial diameter and blood flow measurements. Under control conditions, basal tone was higher and corresponding blood flow was lower in mesenteric arteries of WT male mice compared with S1928A (Figure , and ), consistent with the ex vivo basal tone data. Ang II induced a concentration‐dependent decrease in arterial diameter and concomitant reduction in blood flow (ie, flux) in mesenteric arteries from WT and S1928A mice (Figure and ). Peak constriction and corresponding flux were relatively similar at most points examined immediately after Ang II (Figure , and ; Figure and ). However, 5 minutes after Ang II, arteries from the S1928A mice showed larger diameter and flux compared with arteries from WT male mice (Figure , and through ). The Ca V 1.2 channel blocker nifedipine prevented the Ang II effects on arterial diameter and flux (Figure and Figure and ). Altogether, these results suggest a key role for pS1928 and L‐type Ca 2+ channels in the in vivo regulation of male arterial diameter (and vascular tone) and blood flow upon Ang II. Laser speckle experiments in mesenteric arteries from WT female mice revealed lower basal tone and Ang II‐induced changes in diameter and flux compared with male mice (Figure through ), consistent with reported sex differences in these properties. Moreover, while Ang II induced a concentration‐dependent decrease in arterial diameter and concomitant reduction in blood flow (ie, flux) in mesenteric arteries from WT and S1928A female mice, preventing pS1928 did not lessen the Ang II response (Figure through ). These results suggest sex‐dependent differences in mechanisms by which Ang II modulates vascular function in vivo.
Enhances Vascular Ca V 1 .2 Activity and Cooperativity in Hypertension Considering that chronic activation of Ang II signaling contributes to enhanced Ca V 1.2 channel activity and the development of hypertension, , , , , , , , we examined if pS1928 underlies the alterations in channel properties. Unitary data showed that channel nPo and availability were increased in WT hypertensive arterial myocytes compared with WT sham cells (Figure and and Figure ). The frequency and strength (κ) of cooperative Ca V 1.2 events showed a robust trend toward higher values in WT hypertensive myocytes compared with WT sham cells (Figure and ). The changes in Ca V 1.2 channel properties in WT hypertensive cells were not observed in S1928A hypertension compared with S1928A sham cells (Figure through ; Figure ). Moreover, channel nPo and availability, as well as coupling frequency and strength, were significantly higher in WT hypertension compared with S1928A hypertensive myocytes. These results suggest that pS1928 triggers increased vascular Ca V 1.2 channel function in male arterial myocytes during hypertension.
Mediates Increased Vascular α1 C Clustering in Hypertension We next examined if α1 C clustering was increased in freshly isolated male arterial myocytes during hypertension and if this process required pS1928. Superresolution imaging with pair‐correlation analysis (Figure and ) found an increase in α1 C cluster size (Figure ), no significant differences in cluster density (Figure ), and trends toward higher N clusters (Figure ) in WT hypertensive arterial myocytes compared with WT sham and S1928A HTN cells. These changes were not observed in S1928A hypertensive cells compared with S1928A sham, which had similar α1 C cluster properties as WT sham (Figure through ). The modified PLA assay confirmed the increase in α1 C clustering in WT hypertension compared with WT sham, and S1928A sham and hypertensive cells (Figure ; Figure ). No PLA signal was detected when one of the primary antibody‐probe complexes was omitted from the preparation (Figure ; Figure ). Intriguingly, no change in total α1 C protein abundance was detected by Western blot in WT and S1928A hypertensive lysates compared with sham conditions (Figure ). These results were unexpected as prior studies have shown increased total α1 C protein abundance in arterial lysates during hypertension. Results may point to offsetting changes, possibly associated with the hypertension model. Regardless, these results indicate that α1 C clustering properties in male arterial myocytes are elevated during hypertension, and this spatial remodeling is mediated by pS1928. To examine if similar increases in α1 C clustering were observed in different models of hypertension, superresolution experiments were performed using male arterial myocytes from the BPH compared to BPN mice. Hypertension in BPH mice is thought to be mediated by sympathetic overactivity, which acting through the α‐adrenergic receptors/PKC axis, may modulate α1 C clustering. Data show that α1 C clustering is increased in BPH arterial myocytes compared with BPN cells (Figure through ). The α1 C superclustering in BPH arterial myocytes correlated with increased Ca V 1.2 nPo and cooperativity previously reported in these cells. Moreover, in pressurized arteries, noradrenaline caused an initial peak constriction of similar magnitude in WT and S1928A vessels (Figure ). Yet, in the continuous presence of noradrenaline, WT arteries reached a steady‐state level (eg, plateau level) with higher vascular tone compared with S1928A arteries (Figure ). These results are consistent with the Ang II observations here and prior studies indicating that noradrenaline increases L‐type Ca 2+ channel activity leading to enhanced vascular tone and suggest that noradrenaline‐induced elevation in steady‐state tone in mesenteric arteries requires α1 C S1928 phosphorylation. Results further indicate that enhanced α1 C clustering may be a general feature in hypertensive arterial myocytes to increase Ca V 1.2 activity and cooperativity, and vascular tone.
Underlies Increased Myogenic Tone and Blood Pressure in Hypertension To establish the physiological relevance of pS1928 during hypertension, we measured myogenic tone in pressurized male mesenteric arteries. We found that WT hypertensive arteries developed higher myogenic tone at physiologically relevant intravascular pressures (ie, >60 mm Hg) than WT sham arteries (Figure ). The hypertension‐induced elevations in myogenic tone were not observed when comparing tone levels in S1928A sham and hypertensive arteries (Figure ). These results suggest that elevations in myogenic tone in male mesenteric arteries during hypertension require pS1928. Lastly, we examined the role of pS1928 in BP regulation in freely moving WT and S1928A male mice implanted with radiotelemetry devices and osmotic minipumps eluting either saline (ie, sham) or Ang II (ie, hypertension). BP measurements were analyzed at day 0 and day 7 after minipump implantation. Telemetry data showed that resting (Day 0) systolic pressure (Figure ), diastolic pressure (Figure ), and mean arterial pressure (MAP; Figure ) were similar in WT and S1928A mice. However, a significant increase in systolic pressure (Figure ), diastolic pressure (Figure ), and MAP (Figure ) was observed in both WT and S1928A mice after 7 days with a minipump eluting Ang II compared with saline. The increase in MAP in S1928A hypertensive mice was not surprising as BP is regulated by many factors in addition to peripheral vascular resistance, such as alterations in sympathetic activity, baroreflex dysfunction, or renal abnormalities. Importantly, the change in MAP (ΔMAP) during hypertension was significantly larger in WT mice compared with S1928A mice (Figure ). These results are remarkable in that preventing phosphorylation of a single amino acid in α1 C significantly ameliorates Ang II‐induced hypertension. Note that, in our hands, pulse pressure (an indicator of arterial stiffness and cardiac contraction; Figures ), heart rate (Figure ), and cardiac hemodynamics (Figure through ) were similar in WT and S1928A sham and hypertensive mice. Overall, these results suggest that pS1928 contributes to increasing BP without a major impact on cardiac function in an Ang II‐induced hypertensive male mouse model.
We report 5 key findings (Figure ). First, the α1 C subunit of the Ca V 1.2 channel in WT male arterial myocytes exposed to Ang II or from the Ang II‐induced hypertension, as well as genetically induced hypertensive (eg, BPH) mice reorganizes into larger clusters (ie, α1 C superclusters), and this requires PKC. Second, the α1 C superclustering leads to an increase in Ca V 1.2 channel current density and the frequency and strength of Ca V 1.2 cooperative events. Third, α1 C S1928 phosphorylation mediates the spatial reorganization of vascular α1 C into superclusters, as well as the increase in Ca V 1.2 current density and cooperativity during Ang II exposure and hypertension as concluded using cells/tissue from S1928A mice. Fourth, α1 C S1928 phosphorylation underlies a PKC‐dependent increase in male arterial myocyte [Ca 2+ ] i and contractility, leading to enhanced myogenic tone in arteries exposed to Ang II and from WT male hypertensive mice (also concluded using cells/tissue from S1928A mice). Fifth, α1 C S1928 phosphorylation contributes to hypertension as determined by a reduced change in mean arterial pressure when comparing WT and S1928A mice. These results uncovered a previously unappreciated role for pS1928 in mediating a PKC‐dependent spatiotemporal regulation of vascular Ca V 1.2 channels to alter vascular reactivity and BP during activation of Ang II signaling and hypertension. We propose that pS1928 is a rheostat of vascular Ca V 1.2 function and vascular reactivity and a risk factor for hypertension. Prior work in heterologous expression systems suggested the possibility that Ca V 1.2 channels could be in complex with PKC to promote phosphorylation of α1 C at S1928 to alter channel function. , Our findings support this premise, as they revealed a key role for pS1928 in mediating changes in α1 C /Ca V 1.2 spatiotemporal properties in response to Ang II/PKC signaling and hypertension. These results are significant because they uncover previously unappreciated mechanisms by which Ang II/PKC signaling may alter vascular α1 C /Ca V 1.2 function and vascular reactivity in health and hypertension. Given that diabetic hyperglycemia has been found to elevate pS1928 levels, , , it raises the intriguing possibility that this α 1C amino acid may serve as a “pathological hub” mediating vascular dysfunction in various diseases. The current general model of Ca V 1.2 cooperativity suggests the stochastic, Ca 2+ ‐dependent self‐assembly of α1 C in clusters of various sizes at the plasma membrane. , Our findings have important implications as they provide evidence that pS1928 is necessary for the PKC (but not PKA)‐dependent spatial reorganization of α1 C subunits into superclusters during acute Ang II and hypertension (Figures and ; Figure ). This α1 C spatial remodeling upon Ang II exposure and hypertension was correlated with increased Ca V 1.2 function (Figures and ). Supporting a central role for PKC and its anchoring near Ca V 1.2 in this process, few to no changes in Ca V 1.2 properties were observed in PKCα knockout (PKCα −/− ) or AKAP5 knockout (AKAP5 −/− ) arterial myocytes acutely exposed to Ang II or from Ang II‐infused PKCα −/− and AKAP5 −/− mice. , , Caveolin may also play a structural or signaling role in regulating Ca V 1.2 function, which remains to be determined. The results highlight the key involvement of α1 C pS1928 in the spatiotemporal regulation (ie, clustering and cooperativity) of vascular Ca V 1.2 channels upon activation of Ang II/PKC signaling and hypertension. Moreover, considering that vascular α1 C S1928 is a substrate for both PKA and PKC phosphorylation, we propose that this site is a master regulator of vascular Ca V 1.2 channel function. Future studies should assess how PKA and PKC may differentially regulate α1 C /Ca V 1.2, which may involve spatial segregation of G S versus G q protein‐coupled receptors with specific subpopulations of the channel. An emerging and important concept is that Ca V 1.2 channels can be distinctively regulated in different cell types. Intriguingly, a recent study found that acute Ang II exposure of freshly isolated cardiomyocytes resulted in reduced surface expression and cluster size of cardiac α1 C . This was correlated with a decrease in Ca V 1.2 current density and Ca 2+ transients in cardiomyocytes. In stark contrast, our study shows that acute and chronic Ang II exposure increased vascular α1 C clustering and Ca V 1.2 function, which was associated with increased arterial myocyte [Ca 2+ ] i and contractility. In addition, whereas Ang II regulation of α1 C /Ca V 1.2 was linked to the depletion of the membrane phospholipid phosphatidylinositol 4,5 bisphosphate in cardiomyocytes, pS1928 was required in arterial myocytes. These contrasting Ang II effects on cardiac versus vascular α1 C /Ca V 1.2 (and underlying functional impact) provide additional evidence supporting the tissue‐specific regulation of this important ion channel. Although the mechanisms mediating α1 C /Ca V 1.2 tissue‐specific regulation are unclear, they could involve the expression of different α1 C splice variants or distinct lipid microenvironments in cardiac versus vascular tissue. Regardless, these fundamental differences may be exploited to treat hypertension without affecting cardiac function. A significant outcome of pS1928‐mediated α1 C superclustering and increased Ca V 1.2 function in Ang II/PKC signaling activation and hypertension is the underlying amplification of Ca 2+ influx, which may alter cellular and tissue responses. , Accordingly, this study linked the pS1928‐induced α1 C superclustering and Ca V 1.2 channel function to elevations in arterial myocyte [Ca 2+ ] i and contraction in response to acute Ang II (Figure ) and hypertension (Figure ). Inhibiting PKC in WT cells and blocking pS1928 (as in S1928A cells) prevented the Ang II‐induced elevation of arterial myocyte [Ca 2+ ] i and contraction. The results are significant because they suggest a direct link between PKC and pS1928 in modulating arterial myocyte [Ca 2+ ] i and contraction during Ang II signaling. At the tissue level, however, acute Ang II induced an initial constriction of similar magnitude in both WT and S1928A arteries in ex vivo and in vivo preparations (Figure ). By contrast, the cellular data in S1928A arterial myocytes showed no significant changes in cell length (ie, contraction) to Ang II exposure. This unexpected result may reflect the complex interplay between different cells and signaling pathways that could be engaged by Ang II exposure in intact tissue versus isolated arterial myocytes. Nonetheless, the vascular and myogenic response was elevated in WT compared with S1928A tissue (both ex vivo and in vivo) after continued exposure to Ang II and during hypertension, which correlated with concomitant yet contrasting changes in blood flow and BP (Figures and ). The results are comparable with prior work from our group showing that genetic ablation of PKCα or AKAP5 ameliorates Ang II‐induced elevation in BP, , thus providing a link between activation of AKAP5‐anchored PKC near vascular Ca V 1.2 and pS1928 in control of BP. The differences in ΔMAP between WT and S1928A hypertensive mice were not mediated by alterations in heart function, as heart rate and cardiac hemodynamics were similar between these cohorts of mice (Figure through ). Overall, data here suggest a key role and strong connection between the spatiotemporal regulation of vascular α1 C /Ca V 1.2 and the underlying cellular/tissue responses to control blood flow and BP in hypertension, which is mediated by pS1928. Given that preventing pS1928 ameliorates hypertension, this α1 C amino acid may be a new therapeutic target that could help correct Ca V 1.2 dysfunction and ameliorate vascular complications. Increased Ca V 1.2 function leading to Ca 2+ influx amplification in arterial myocytes may also regulate excitation‐transcription coupling. Thus, increased Ca V 1.2 channel function upon Ang II signaling activation during hypertension may engage the prohypertensive transcriptional cascade involving calcineurin and NFATc3 (nuclear factor of activated T cells 3). , , Consistent with this, calcineurin and NFATc3 activation are enhanced in arterial myocytes from Ang II‐induced hypertensive mice. NFATc3 activation in hypertensive arterial myocytes led to the selective downregulation of several K + channel subunits, including the BKβ1 and K V 2.1 subunits. , The NFATc3 activation and downregulation of K + channel subunits in hypertensive arterial myocytes was blocked by the Ca V 1.2 channel blocker nifedipine, providing a link between Ca V 1.2 function, NFATc3 signaling, and transcriptional regulation. It is intriguing to speculate that blocking pS1928 will prevent not only enhanced Ca V 1.2 channel function but also activation of the calcineurin/NFATc3 pathway and underlying transcriptional changes in K + channel functional expression. Future studies should examine these possibilities. Although we confirmed an increase in α1 C superclustering in 2 models of hypertension (Ang II‐induced hypertension and BPH), all other experimental series were examined using the Ang II‐induced hypertension model. Additional experiments in samples from other hypertension models and patients with hypertension will be useful to ascertain the general impact of pS1928 in modulating Ca V 1.2 activity, vascular function, and BP. While this study provides strong functional data highlighting the importance of pS1928, we were not able to provide direct evidence of changes in the phosphorylation state of the site due to the unavailability of well‐validated antibodies, which should be addressed in future studies. The role of pS1928 in α1 C trafficking, which may contribute to α1 C superclustering, , will have to be investigated. Our data also highlight distinct basal and sex‐dependent responses to Ang II. These results suggest that distinctive mechanisms may be engaged in the regulation of α1 C /Ca V 1.2 spatiotemporal properties and vascular function in males versus females, which may contribute to the observed sex differences in Ang II‐induced hypertension. Accordingly, recent studies suggest that α1 C superclusters in female arterial myocytes are sustained by a concomitant self‐assembly superclustering of K V 2.1 channels due to increased phosphorylation of S590 in K V 2.1 channels. , These sex‐dependent changes in K V 2.1‐dependent α1 C clustering may be driven by sex hormones and affected by age, which can then influence vascular function and BP/flow control. , These observations and results here may inspire future studies to comprehensively compare the sex‐ and age‐dependent role of α1 C pS1928 and K V 2.1 pS590 in modulating α1 C /Ca V 1.2, vascular function, and BP during basal conditions and hypertension.
In summary, the findings here identify α1 C S1928 as a target for Ang II/PKC signaling in native vascular tissue. Data also indicate that pS1928 mediates a spatiotemporal remodeling of vascular α1 C /Ca V 1.2 to control arterial myocyte [Ca 2+ ] i and contraction, leading to alterations in vascular reactivity, blood flow, and BP during hypertension. Results are highly significant, as they reveal a previously unappreciated mechanism contributing to pathological changes in a crucial ion channel central for regulating cardiovascular function, thus providing a potential new target for developing novel therapeutics aimed at controlling hypertension.
This work was supported by National Institutes of Health grants R01HL149127, R01HL121059, R01HL171014 and R01HL161872 (to Manuel F. Navedo), and R00HL138160 (to Stefano Morotti), and American Heart Association Postdoctoral Fellowship 830629 (to Miguel Martín‐Aragón Baudel) and 1022782 (to Jade L. Taylor) as well as an American Heart Association Career Development Award 852984 (to Madeline Nieves‐Cintrón). Madeline Nieves‐Cintrón is a University of California Davis CAMPOS Fellow.
None.
Data S1 Table S1 Figures S1–S12
|
|
A Brief Review of Anatomy Education in Korea, Encompassing Its Past, Present, and Future Direction | e95e74b6-a6cc-4f9d-86f9-2b1b11cd64f0 | 11136677 | Anatomy[mh] | The history of anatomy in ancient times was the history of medicine. Anatomy has been described as the cornerstone of good medical practice and the foundation for clinical studies. The first person known to have dissected the human body in the western world to understand its structure and function was Alcmaeon of Croton, who lived in ancient Greece around 500 BC. Hippocrates (BC 460–377) had the belief that medicine should be practiced as a scientific discipline based on the natural sciences, diagnosing and preventing diseases as well as treating them. Also, he believed that physicians should study anatomy. During a brief period in ancient history when human dissection was permitted, Herophilus (BC 330–280) and Erasistratos (BC 310–250) revealed the structure and function of the human body through dissection. Galen (AD 131–200) advanced anatomical knowledge inherited from his predecessors by supplementing it with insights gained through patient care and animal dissection. He also offered scientific insights into pathophysiology, linking it with disease. For approximately 1,300 years, Galen remained a prominent figure in western medicine. Human dissection was prohibited during the Middle Ages. However, in the 13th century, Mondino (1276–1326) published “Anathomia” in 1316, which solidified the importance of anatomy in medical education. Leonardo da Vinci (1452–1519) is considered the most outstanding figure among Renaissance artists who studied anatomy. He attempted to understand the human form and function by correlating them, leaving behind over 750 anatomical drawings that are both artistically beautiful and scientifically accurate. Later, Vesalius (1514–1564), often regarded as the father of modern anatomy, significantly contributed to the field with the publication of the detailed “De Humani Corporis Fabrica” in 1543, establishing anatomy as the cornerstone of modern medicine. In the early 19th century (1828), there was an explosive population growth in Britain, leading to a significant increase in students aspiring to become doctors. This surge resulted in the establishment and popularization of anatomy and medical schools. Many of these students, after their studies, played a crucial role in founding the Department of Anatomy in universities and spreading the teaching of anatomy worldwide. At that time, there was a high demand for cadavers for students' dissection practices in medical schools, but the supply was critically low. Cadavers were clandestinely traded in illegal markets. However, body snatchers or grave robbers significantly contributed to anatomical teaching and played a vital role in advancing anatomy. Due to murders associated with the illicit trade of cadavers, the Anatomy Act was enacted by the British Parliament on August 1, 1832, catalyzing to recognition of the need for anatomical dissection in medical training. Before 1825, the medical curriculum in Britain consisted of basic sciences and clinical subjects. According to Rothstein, anatomy was fundamental and the most popular and important subject in medical schools. Modern anatomy education includes gross anatomy, histology, neuroanatomy, and embryology. The advent of advanced imaging methods like CT and MRI has coincided with the advancement of minimally invasive treatments aimed at particular organs or locations within them. Consequently, a solid grasp of gross anatomy has become more crucial, not just for interpreting the images generated by these advanced techniques, but also for comprehending the route for directing therapy to a precise location. In addition, gross anatomy education helps students develop medical professionalism at the initial stage of entering medicine. It is also recognized to have a significant impact on students' career paths, with those who have experience performing cadaver dissection being more inclined to specialize in surgery, which is an essential field of medical practice. In Korea, though it is known that the concept of anatomy was introduced in the Age of Three Kingdoms, anatomy in the modern sense meaning was introduced in the late Joseon Dynasty by western missionary doctors. Since then, anatomy education has played an important role in helping the poorest countries grow to the point where they can provide world-class medical services after colonial rule and the Korean War. Recently, the Korean government announced a plan to increase the medical college seats as a response to the medical demand triggered by public healthcare, essential healthcare, regional imbalances in medical services, and an aging society. This brief review focuses on the past and present status of gross anatomy education in Korea, which is the basis of medical education, which holds the greatest contact hours and huge infrastructure for laboratories. In preparation for this review, we conducted a brief telephone survey to determine the current state of anatomy education in Korea and present the results here. In addition, we will simulate the demand for gross anatomy education expected when the government's planned increase in the number of medical college seats progresses. Finally, we will discuss the possibility of solving the problem triggered by the abrupt increase in medical students. While the history of anatomy in Korea is quite long, the actual teaching of anatomy to train doctors began with the introduction of western medicine during the late Joseon Dynasty. Although anatomy education began with the establishment of medical schools, hands-on dissection started in 1910. From the 1920s onwards, not only dissection practices but also active research on the bones and organs of Koreans through actual dissections became more prevalent. Before Korea’s liberation on August 15, 1945, anatomy education in Korea adopted the German-style teaching methods used by Japanese anatomists. After liberation, anatomy education continued to be conducted by anatomists trained during the Japanese colonial period. At that time, anatomy education utilized textbooks from the Japanese colonial era, and there was a severe shortage of professors specialized in anatomy. The teaching methods for anatomy were carried out through verbal lectures and educational charts, and the lectures were mainly focused on note-writing. In the 1950s, overcoming the painful scars of the Korean War, anatomy education in some medical colleges was conducted as printed-based lectures, and gradually changed to film slide-based lectures. In the early 1970s, with the revision of the medical college curriculum, the hours dedicated to anatomy education were significantly reduced. In the 1980s, with the establishment of many new medical colleges, the total number increased to 28. Each newly established medical college had its own unique curriculum. During this period, the number of anatomy professors in each medical college was one (4.4%, one college), two (21.7%, five colleges), and three (34.8%, eight colleges), with most of the three or fewer (60.9%) professors in charge of anatomy education. In the 1990s, the number of medical colleges in Korea sharply increased to a total of 37. In most medical colleges (78.3%), the number of professors responsible for anatomy ranged from 2 to 4. With the reduction in anatomy education hours, the teaching method shifted from systematic anatomy to regional anatomy. Some medical schools introduced integrated education (integration of basic medical subjects or integration of basic medical and clinical subjects), leading to changes in traditional anatomy teaching methods. In the 1990s, the total anatomy education hours (including lectures and practice) varied significantly by medical colleges, but the average was 234.5 hours. Since the introduction of the medical graduate school system in the 2000s, most medical colleges have increased clinical practice hours. As a result, not only anatomy but also other basic medical subjects have seen a significant reduction in educational hours. From the 1970s, as society stabilized, the supply of cadavers for anatomy education decreased. With the increase in newly established medical colleges, anatomy dissection faced many challenges. Especially after the ‘Busan Hyungje Welfare Center Incident’ and the ‘Daejeon Seongji Center Incident’ in 1987, coupled with the non-cooperation of officials from related agencies and changes in social welfare systems, the procurement of cadavers became extremely difficult. At that time, there were medical colleges where 30 to 40 students dissected one cadaver, reflecting the serious state of anatomy education. However, with the activation of the organ donation campaign that began in the 1980s, the number of cadaver donors for anatomy education increased. As a result, the situation regarding cadaver procurement began to improve gradually in the late 1990s. Since the 2000s, cadavers donated for anatomy dissection in Korea have been legally provided to medical colleges. Most medical colleges rigorously manage and respect the cadavers, ensuring that remains are treated with dignity, either through respectful cremation or burial. As a result, the widely promoted cadaver donation programs are operating successfully, and public awareness of cadaver donation and dissection has been improving. Currently, most of the cadavers used for anatomy dissection are donated. A survey was conducted targeting 40 medical colleges nationwide (3 groups according to their locations, ) to investigate the number of students attending anatomy lectures and practical sessions, the number of anatomy professors, the number of professors responsible for gross anatomy lectures and practical sessions, the number of staff members responsible for cadaver management, the number of teaching assistants responsible for gross anatomy education, and the number of cadavers used in gross anatomy practical sessions. The survey was conducted via telephone calls with the anatomy education coordinator of each college in April 2024, and the data were collected based on the average numbers over the past five years. The number of students enrolled in anatomy courses exceeded the official capacity due to additional enrollees, re-takers of classes, etc. (Official capacity: 3,058 students, actual number of students taking anatomy courses: approximately 3,246 students). The average number of anatomy professors per medical college was 4.5, with Group A having an average of 6.5 professors, which was higher than the average of 3.9 professors in Groups B and C ( , P = 0.000, Student’s t -test, two-tailed). The average number of professors responsible for gross anatomy lectures and practical sessions was 3.3 per college, indicating that approximately 86.7% of all anatomy professors are involved in teaching gross anatomy lectures and practical sessions . Thus, most anatomy professors seem to be actively engaged in both gross anatomy lectures and practical sessions. On the other hand, among the eight medical colleges in Group A, the average number of professors responsible for gross anatomy lectures and practical sessions was 5.3 per college, higher than the average of 3.3 professors in the other Groups ( , P = 0.001, Student’s t -test, two-tailed). Regarding the number of teaching assistants (TAs) responsible for gross anatomy education, it was found that nationally, each college had fewer than one TA. However, the eight medical colleges in Group A had an average of 2.0 TAs per college, which was more than four times the average of 0.5 TAs in the other Groups ( , P = 0.000, Student’s t -test, two-tailed). In terms of the cadavers essential for gross anatomy practical sessions, approximately 450 cadavers are utilized annually for medical students’ anatomy education. While there are variations depending on the Group, the ratio of students to cadavers stands at 7.4 ( , 3,246 students/438 cadavers) in Korea. This indicates a challenging educational environment compared to the United States, where the ratio is 5.1 students per cadaver. Nationally, each medical college utilizes an average of 11.0 cadavers for anatomy practical sessions. In the eight medical colleges in Group A, the average stands at 16.9 cadavers per college, which is 77.9% higher than the average of 9.5 cadavers in the other Groups ( , P = 0.000, Student’s t -test, two-tailed). Regarding the staff responsible for managing the cadavers, the survey found an average of 1.3 staff members per medical college. In the eight universities in Group A, the average was 2.4 staff members per college, while in the medical colleges of Groups B and C, the average was 1.1 staff members per college. This indicates that medical colleges of Group A have more than twice the number of staff compared to the other Groups ( , P = 0.000, Student’s t -test, two-tailed). There were no statistically significant differences observed in any of the metrics between Group B and Group C. Given that the eight medical colleges in Group A have more professors, teaching assistants, cadaver management staff, and cadavers compared to the others in Group B and C it is evident that there is a significant disparity in educational conditions for gross anatomy lectures and practical sessions across the Groups. Since the content and volume of anatomy lectures and practical sessions are expected to be relatively consistent across colleges, it suggests that professors and education assistants in medical colleges in Groups B and C are handling more anatomy-related educational tasks compared to those in the eight medical colleges in Group A. On the other hand, while there was a slight difference in the student-to-professor ratio for anatomy education across Groups, with an average of 24.4 students per professor nationwide, 20.9 in Group A, and 26.0 in Group B and C, this difference was not statistically significant . Similarly, the student-to-cadaver ratio was slightly different across Groups, with an average of 7.4 students per cadaver nationwide, 6.5 in Group A, and 7.8 in Groups B and C, but again, this difference was not statistically significant . The reason the student-to-professor and student-to-cadaver ratios in Groups B and C are similar to those in Group A, despite having fewer professors and cadavers, could be attributed to the significantly higher student number in the eight medical colleges in Group A. The average student number in Group A’s eight medical colleges is 109.5 students, which is statistically significantly higher than the average of 74.1 students in the other Groups ( , P = 0.020, Student’s t -test, two-tailed). The recruitment of professors capable of conducting gross anatomy lectures and practical sessions has become increasingly challenging in recent years. Primarily, there is a limited number of graduate students specializing in gross anatomy due to its nature as a niche field. Moreover, not many professionals meet the expectations of universities in terms of both teaching and research capabilities. As a result, many universities have been unable to fill vacant anatomy professor positions despite posting job openings for several years. Consequently, some universities have resorted to appointing professionals from other disciplines with strong research capabilities as anatomy professors. These individuals either handle other subfields of anatomy, such as neuroanatomy, histology, and embryology or deliver gross anatomy lectures without participating in practical sessions, merely transmitting textbook content without verifying their teaching skills (personal communication). This ongoing cycle makes it increasingly difficult to attract students who aspire to specialize in gross anatomy or become future anatomy professors, perpetuating a vicious cycle of limited recruitment and specialized training in the field. In summary, as of 2023, the 32 medical colleges of Groups B and C fall significantly short in terms of anatomy lectures and practical session professors, teaching assistants, cadavers, and cadaver management staff compared to the eight medical colleges in Group A. However, due to a smaller student population, these colleges maintain a gross anatomy educational environment that is somewhat comparable to that of medical colleges in Group A. Nevertheless, the workload per professor and per teaching assistant is likely higher in the 32 medical colleges of Groups B and C than in those of Group A. The current study focused primarily on gross anatomy among various subjects taught in anatomy classrooms due to time constraints. This study focused on examining the human resources, cadaver status, and infrastructure related to education. Subsequent research will be necessary to investigate other subjects and explore aspects such as the number of educational hours, practical session hours, and teaching methods for these subjects. If the number of medical college students continues to increase, it is anticipated that the anatomy education environments, which have been barely maintained despite limited educational resources, will significantly deteriorate. Particularly, an increase in the student population at medical colleges of Groups B and C is expected to further worsen the anatomy education environment. predicts the situation at anatomy education sites based on the scale of the increase in medical college admissions. The current nationwide student-to-anatomy professor ratio is 24.4 students per professor, and the ratio of students per cadaver is 7.4 . To maintain these ratios with an increase in student numbers by 500, 1,000, and 2,000, we calculated the additional professors and cadavers needed . For an increase of 500 students, approximately 20 more anatomy professors and 68 additional cadavers would be required. For an increase of 1,000 students, about 41 more professors and 135 additional cadavers would be needed. For an increase of 2,000 students, roughly 82 more professors and 270 additional cadavers would be necessary. Considering the current number of anatomy professors nationwide is 92 (25 in Group A and 67 in Groups B and C), and there are 30 TAs responsible for anatomy education nationwide, it appears to be extremely challenging to secure the additional anatomy professors and TAs needed in the short term. In the case of an increase of 2,000 students, even if all the 30 anatomy TAs nationwide are promoted to professors, there will still be a shortage of 52 professors, leaving no assistants remaining. A more significant challenge is that many of the current professors responsible for anatomy education, around 23, are expected to retire within the next five years. In the cases of University A in Gyeonggi-do and University B in Gyeongsang-do, two anatomy teaching professors from each institution are set to retire within the next two years. To prepare additional cadavers, the number of staff members responsible for cadavers must also increase. At present, each anatomy professor in Korea handles 24.4 students, and even when including TAs who participate in practical training as capable individuals, the number is 17.8 students per faculty member. In contrast, the United Kingdom (UK) has a ratio of 13.3 students per educator, indicating a severe shortage of qualified educators for gross anatomy education in Korea. Given this situation, an increase in medical college enrolment would undoubtedly exacerbate the inadequacies in gross anatomy education. Considering the number of current gross anatomy professors who are expected to retire within the next five years, it’s not difficult to anticipate a further shortage of instructors capable of teaching gross anatomy, even if there is no increase in medical student enrolment. This pressing issue needs to be addressed promptly to prevent a decline in the quality of anatomy education. Finally, to support anatomy education, infrastructural elements such as lecture halls, labs, and multimedia environments are essential. While this review did not delve into the specific numbers for each university's educational infrastructure, most institutions only have infrastructure tailored to their current student numbers. Therefore, an increase in medical college seats requires not only securing human and cadaver resources for anatomy education but also investing in educational infrastructure. Just as University C in the Jeolla-do which saw an increase in enrolment following the closure of Seonam University, took approximately five years to establish educational infrastructure aligned with the increased numbers showing that the planning, design, and construction processes require not only financial resources but also considerable time. Thus, significant challenges are anticipated until educational environments adapted to the increased student body are fully prepared. Given the current state and conditions of gross anatomy education in Korea, along with the societal demand for an increase in healthcare professionals, we aim to address the anticipated challenges in gross anatomy education due to this expansion. Specifically, we propose the following recommendations concerning the training of personnel capable of teaching gross anatomy and the supply of cadavers, which serve as the material basis for education. Faculty recruitment strategy Establishment of an anatomy educator national fellow (basic medical educator fellow) program The most significant challenge in the current gross anatomy faculty recruitment lies in the shortage of qualified faculty. Basic medical subjects, including gross anatomy, are essential components of clinical practice. To cultivate faculty capable of teaching gross anatomy, a system akin to providing subsidies to pediatricians and thoracic surgeons, who specialize in essential medical fields, should be instituted for graduate students majoring in basic medical sciences like anatomy. This program should be designed not only to attract doctors who have graduated from medical colleges but also to bring in other highly qualified talents from various disciplines. Currently, the government’s ongoing physician-scientist training program focuses on enhancing the research capabilities of clinical doctors. Expanding this concept to include the training of educational experts in essential basic medical fields could foster a pool of faculty in basic medical sciences. Anatomy educator matching program The current landscape of gross anatomy education is characterized by a significant shortage of faculty. To address this issue in the short term, leveraging doctoral students, postdoctoral researchers, and retired anatomists can be a viable option. However, connecting the demand in the field with the potential educational workforce has proven challenging. To address this gap, the establishment of an “Anatomy Educator Matching Program” within the Korean Association of Anatomists (KAA) can be proposed. This program would aim to build a pool of potential educators and facilitate connections between educational institutions and available anatomy educators. By matching institutions with verified anatomy educators, this program can provide temporary relief to the shortage of anatomy faculty and also enable training for the next generation of anatomy educators. Qualified anatomy educator certificate To maintain the quality of gross anatomy education, it is essential to certify professors and instructors who can effectively teach and guide practical sessions in gross anatomy. Monitoring should also be in place to ensure that proper education and practical training are consistently provided. This should be integrated into medical college accreditation evaluations to encourage universities to prioritize effective gross anatomy education. Similar to the “Supervising Specialist” system implemented by the Korean Hospital Association (KHA) in clinical medicine, a system could be developed for gross anatomy education. This system would certify educators with over four years of experience, including practical training, in gross anatomy and who possess research capabilities. The Korean Association of Anatomists could certify these individuals as specialized gross anatomy instructors and maintain continuous educational oversight. By utilizing these certified educators to deliver gross anatomy lectures and practical sessions to medical students, the quality of gross anatomy education can be maintained and even improved. Maintaining diverse hiring tracks for anatomy education specialists According to 2019 statistics from the UK, among 760 anatomy educators across 39 medical schools, 103 are professors who combine both anatomy research and teaching. In contrast, 143 are professors dedicated solely to anatomy education, supported by 514 anatomy demonstrators. Among these, the 143 professors focused solely on anatomy education deserve attention. As mentioned earlier, due to the nature of gross anatomy, it can be challenging to find professors who excel equally in both research and teaching. Instead of hiring professors based solely on their research capabilities, there seems to be a trend in employing dedicated anatomy education specialists. This suggests that medical colleges in Korea should consider utilizing diverse hiring tracks, similar to those observed in the UK. (Note: Anatomy demonstrators are doctors who have completed a two-year internship and are in the stage of refining surgical skills and enhancing competencies through assisting in anatomy classrooms before entering the surgical field.) Securing cadavers for anatomy education Recently, efforts have been made to replace anatomy labs with technologies such as augmented reality (AR) and virtual reality (VR) for anatomy education. AR and VR have been proven effective in understanding human organs and structures in three dimensions, aiding in the perception of their spatial relationships. Consequently, their use as supplementary educational materials and tools in gross anatomy education, particularly in practical training, has increased. Despite these advancements, certain essential contents that medical professionals must acquire, such as anatomical techniques, understanding of human dignity, and medical professionalism, cannot be learned solely through simulations. Thus, the importance of cadaveric training in graduate medical education (GME) in surgical disciplines has grown. In fact, a lack of cadaveric education has resulted in a significant decline in the number of doctors choosing surgical fields as their careers. Consequently, many medical education institutions continue to prioritize anatomy practical sessions, and the situation is no different in South Korea. Therefore, securing an adequate number of cadavers is crucial for anatomy training. Currently, South Korea receives approximately 1,100–1,200 body donations annually, which are used for medical student anatomy education, GME for clinical physicians, and gross anatomy research in life sciences. With the increasing emphasis on cadaver-based practical training in GME, coupled with legal revisions regarding body dissection and preservation for life science research, the shortage of cadavers will inevitably worsen with an increase in medical student enrolments. To address this, efforts should focus on improving awareness and promoting cadaver donation, extending respect and support to cadaver donors and their families, and supporting universities and practitioners responsible for cadaver collection, utilization, and disposal. Considering the difficulties in cadaver supply faced by some medical schools in the United States, establishing and operating regional cadaver centers could also be considered. Moving forward, collaboration among medical colleges, university hospitals, societies, local governments, and government departments will be essential to devise solutions for securing educational cadavers. Infrastructure support Securing adequate infrastructure is essential for providing quality anatomy education. This encompasses space allocation, facility construction, and acquiring the necessary equipment. Ultimately, addressing these challenges requires substantial financial resources and both short-term and long-term development plans. To resolve these issues, research projects and national financial support are essential for providing the necessary backing and resources. Establishment of an anatomy educator national fellow (basic medical educator fellow) program The most significant challenge in the current gross anatomy faculty recruitment lies in the shortage of qualified faculty. Basic medical subjects, including gross anatomy, are essential components of clinical practice. To cultivate faculty capable of teaching gross anatomy, a system akin to providing subsidies to pediatricians and thoracic surgeons, who specialize in essential medical fields, should be instituted for graduate students majoring in basic medical sciences like anatomy. This program should be designed not only to attract doctors who have graduated from medical colleges but also to bring in other highly qualified talents from various disciplines. Currently, the government’s ongoing physician-scientist training program focuses on enhancing the research capabilities of clinical doctors. Expanding this concept to include the training of educational experts in essential basic medical fields could foster a pool of faculty in basic medical sciences. Anatomy educator matching program The current landscape of gross anatomy education is characterized by a significant shortage of faculty. To address this issue in the short term, leveraging doctoral students, postdoctoral researchers, and retired anatomists can be a viable option. However, connecting the demand in the field with the potential educational workforce has proven challenging. To address this gap, the establishment of an “Anatomy Educator Matching Program” within the Korean Association of Anatomists (KAA) can be proposed. This program would aim to build a pool of potential educators and facilitate connections between educational institutions and available anatomy educators. By matching institutions with verified anatomy educators, this program can provide temporary relief to the shortage of anatomy faculty and also enable training for the next generation of anatomy educators. The most significant challenge in the current gross anatomy faculty recruitment lies in the shortage of qualified faculty. Basic medical subjects, including gross anatomy, are essential components of clinical practice. To cultivate faculty capable of teaching gross anatomy, a system akin to providing subsidies to pediatricians and thoracic surgeons, who specialize in essential medical fields, should be instituted for graduate students majoring in basic medical sciences like anatomy. This program should be designed not only to attract doctors who have graduated from medical colleges but also to bring in other highly qualified talents from various disciplines. Currently, the government’s ongoing physician-scientist training program focuses on enhancing the research capabilities of clinical doctors. Expanding this concept to include the training of educational experts in essential basic medical fields could foster a pool of faculty in basic medical sciences. The current landscape of gross anatomy education is characterized by a significant shortage of faculty. To address this issue in the short term, leveraging doctoral students, postdoctoral researchers, and retired anatomists can be a viable option. However, connecting the demand in the field with the potential educational workforce has proven challenging. To address this gap, the establishment of an “Anatomy Educator Matching Program” within the Korean Association of Anatomists (KAA) can be proposed. This program would aim to build a pool of potential educators and facilitate connections between educational institutions and available anatomy educators. By matching institutions with verified anatomy educators, this program can provide temporary relief to the shortage of anatomy faculty and also enable training for the next generation of anatomy educators. To maintain the quality of gross anatomy education, it is essential to certify professors and instructors who can effectively teach and guide practical sessions in gross anatomy. Monitoring should also be in place to ensure that proper education and practical training are consistently provided. This should be integrated into medical college accreditation evaluations to encourage universities to prioritize effective gross anatomy education. Similar to the “Supervising Specialist” system implemented by the Korean Hospital Association (KHA) in clinical medicine, a system could be developed for gross anatomy education. This system would certify educators with over four years of experience, including practical training, in gross anatomy and who possess research capabilities. The Korean Association of Anatomists could certify these individuals as specialized gross anatomy instructors and maintain continuous educational oversight. By utilizing these certified educators to deliver gross anatomy lectures and practical sessions to medical students, the quality of gross anatomy education can be maintained and even improved. According to 2019 statistics from the UK, among 760 anatomy educators across 39 medical schools, 103 are professors who combine both anatomy research and teaching. In contrast, 143 are professors dedicated solely to anatomy education, supported by 514 anatomy demonstrators. Among these, the 143 professors focused solely on anatomy education deserve attention. As mentioned earlier, due to the nature of gross anatomy, it can be challenging to find professors who excel equally in both research and teaching. Instead of hiring professors based solely on their research capabilities, there seems to be a trend in employing dedicated anatomy education specialists. This suggests that medical colleges in Korea should consider utilizing diverse hiring tracks, similar to those observed in the UK. (Note: Anatomy demonstrators are doctors who have completed a two-year internship and are in the stage of refining surgical skills and enhancing competencies through assisting in anatomy classrooms before entering the surgical field.) Recently, efforts have been made to replace anatomy labs with technologies such as augmented reality (AR) and virtual reality (VR) for anatomy education. AR and VR have been proven effective in understanding human organs and structures in three dimensions, aiding in the perception of their spatial relationships. Consequently, their use as supplementary educational materials and tools in gross anatomy education, particularly in practical training, has increased. Despite these advancements, certain essential contents that medical professionals must acquire, such as anatomical techniques, understanding of human dignity, and medical professionalism, cannot be learned solely through simulations. Thus, the importance of cadaveric training in graduate medical education (GME) in surgical disciplines has grown. In fact, a lack of cadaveric education has resulted in a significant decline in the number of doctors choosing surgical fields as their careers. Consequently, many medical education institutions continue to prioritize anatomy practical sessions, and the situation is no different in South Korea. Therefore, securing an adequate number of cadavers is crucial for anatomy training. Currently, South Korea receives approximately 1,100–1,200 body donations annually, which are used for medical student anatomy education, GME for clinical physicians, and gross anatomy research in life sciences. With the increasing emphasis on cadaver-based practical training in GME, coupled with legal revisions regarding body dissection and preservation for life science research, the shortage of cadavers will inevitably worsen with an increase in medical student enrolments. To address this, efforts should focus on improving awareness and promoting cadaver donation, extending respect and support to cadaver donors and their families, and supporting universities and practitioners responsible for cadaver collection, utilization, and disposal. Considering the difficulties in cadaver supply faced by some medical schools in the United States, establishing and operating regional cadaver centers could also be considered. Moving forward, collaboration among medical colleges, university hospitals, societies, local governments, and government departments will be essential to devise solutions for securing educational cadavers. Securing adequate infrastructure is essential for providing quality anatomy education. This encompasses space allocation, facility construction, and acquiring the necessary equipment. Ultimately, addressing these challenges requires substantial financial resources and both short-term and long-term development plans. To resolve these issues, research projects and national financial support are essential for providing the necessary backing and resources. In the relatively brief period since the introduction of western medicine to Korea, anatomy education in Korea has played a significant role in shaping proficient medical professionals. This review of the current state of anatomy education reveals variations in educational conditions across medical colleges. Although each institution strives to maintain a high standard of education despite challenging circumstances, the field faces fragility and numerous challenges, particularly in aspects like the availability of anatomy educators, cadaver supply, and infrastructure. It is evident that an abrupt increase in the number of medical students amid these challenges could jeopardize the quality of anatomy education. To address this, there is a pressing need to establish a comprehensive mid- to long-term development strategy. Such a plan requires sufficient time and financial investment to tackle issues including the recruitment of qualified medical educators, ensuring an adequate supply of cadavers, and improving infrastructure. Only through these concerted efforts can the integrity and quality of anatomy education be upheld in the face of evolving demands . |
Inter-observer reproducibility of classical lobular neoplasia (B3 lesions) in preoperative breast biopsies: a study of the Swiss Working Group of breast and gynecopathologists | de2b0900-e2b0-48c8-9c03-8dc8ec67747c | 7230045 | Gynaecology[mh] | Lobular neoplasia of the breast comprises a large variation in atypical epithelial proliferation within the acinar breast structures (Foote and Stewart ; Haagensen et al. ; King et al. ; King and Reis-Filho ; Lakhani et al. ; Tavassoli ; WHO ; Wen and Brogi ). In low-grade lesions, several existing alternative terminologies such as classical lobular neoplasia (LN classical type) including both atypical lobular hyperplasia (ALH) and lobular carcinoma in situ (classical LCIS) and the so-called Lobular Intraepithelial Neoplasia (LIN) covering LIN-1, LIN-2, LIN-3 (Lakhani et al. ; Tavassoli ; WHO ; Rageth et al. , ) allow to classify the same lobular breast lesion with different diagnostic terms. Although clinical management of these alternative low-grade terminologies is quite similar and all are considered as risk factors and non-obligate precursor for breast cancer, there is still no single pathological factor to predict upgrade, progression, and/or local recurrence (Foote and Stewart ; Haagensen et al. ; King et al. ; King and Reis-Filho ; Lakhani et al. ; Tavassoli ; WHO ; Wen and Brogi ; Rageth et al. , ; AGO ). On the contrary, high-grade lobular in situ lesions such as pleomorphic or florid LCIS/LN exhibit a biologically similar behavior and require the same management as their ductal carcinoma in situ (DCIS) counterpart (WHO ; Wen and Brogi ; AGO ; Shamir et al. ). Inter-observer agreement data on different terminologies are sparse and these data point to improved agreement when favoring one category to more than one descriptive subgroup (AGO ; Choi et al. ; Fitzgibbons ; Gomes et al. ; Singh et al. ). In our study, we addressed the question on inter-observer agreement on six existing non-pleomorphic LN terminologies using 40 diagnostic LN breast core and vacuum biopsy cases with eight participating pathologists specialized in breast pathology.
40 cases of breast core- and vacuum biopsies with the diagnosis B3 lesion and lobular neoplasia were retrieved from the Institute of Pathology and Molecular Pathology, University Hospital Zurich Switzerland, in the years 2012–2013. All cases were diagnostic cases from routine histological diagnostics. The diagnosis of lobular neoplasia and B3 category was made on conventional hematoxylin–eosin (H&E) stains and confirmed with immunohistochemistry (E-Cadherin loss and/or catenin p120 cytoplasmic staining) in all cases at the time of the routine diagnostics. The study was conducted within the project approved by the cantonal committee of the Canton Zurich (KEK-2012-554). Informed consent was not necessary as all cases were analyzed in a fully anonymized way. Study design All eight participants of the study were members of the Working Group of Breast and Gynecopathology of the Swiss Society of Pathology. For the study, a digital link containing H&E images of the biopsies as well as an excel data sheet were sent to all participants. Participants were asked to assess the H&E images and to enter any of the following further diagnostic subcategory which they think would fit to the index case. These categories were named as follows: atypical lobular hyperplasia (ALH), lobular carcinoma in situ of classical type (LCIS classical type), lobular intraepithelial neoplasia I, II, III (LIN-I, LIN-II, LIN-III), lobular neoplasia of classical type (LN, classical type), focal or extensive classical LN (one or more than one focus of LN), and others (different from the mentioned category). A given case could be classified in multiple categories by the participant (Figs. and ). An agreement in classification of a lesion was defined if a category was entered by at least six out of participating eight pathologists. Statistical analyses Results were analyzed using the Chi-square statistics and Kappa Fleiss to compare agreed diagnostic categories.
All eight participants of the study were members of the Working Group of Breast and Gynecopathology of the Swiss Society of Pathology. For the study, a digital link containing H&E images of the biopsies as well as an excel data sheet were sent to all participants. Participants were asked to assess the H&E images and to enter any of the following further diagnostic subcategory which they think would fit to the index case. These categories were named as follows: atypical lobular hyperplasia (ALH), lobular carcinoma in situ of classical type (LCIS classical type), lobular intraepithelial neoplasia I, II, III (LIN-I, LIN-II, LIN-III), lobular neoplasia of classical type (LN, classical type), focal or extensive classical LN (one or more than one focus of LN), and others (different from the mentioned category). A given case could be classified in multiple categories by the participant (Figs. and ). An agreement in classification of a lesion was defined if a category was entered by at least six out of participating eight pathologists.
Results were analyzed using the Chi-square statistics and Kappa Fleiss to compare agreed diagnostic categories.
The highest agreement between eight pathologists was reached using the category lobular neoplasia (LN, classical), 26/40 (65%) cases were diagnosed as such. Agreements in other categories was low or poor: 12/40 (30%) (ALH), 9/40 (22%) (LCIS), 8/40 (20%) (LIN-1), 8/40 (20%) (focal B3), 3/40 (7.5%) (LIN-2), and 2/40 (5%) (extensive B3). Chi-square statistic was significant for the differences on agreement between classical LN versus the other nomenclatures ( p = 0.001137). Kappa Fleiss could not be applied due to multiple answers per case (Fig. ).
This is the first study to address reproducibility among existing established designations for classical B3 lobular lesions. In this study, we could show that different existing terminological categories all describing classical type lobular neoplasia of the breast are unequally reproduced among expert breast pathologists. Among the known diagnostic categories such as atypical lobular hyperplasia, lobular carcinoma in situ of classical type, lobular intraepithelial neoplasia I, II, III, and lobular neoplasia of classical type, the diagnostic agreement varies from 5–65%. The terminology ‘lobular neoplasia of classical type’ reached the highest agreement with 65% among breast pathologists. The term lobular neoplasia encompasses a spectrum of histological lesions with differences in extent and the degree of nuclear atypia (Foote and Stewart ; Haagensen et al. ; King et al. ; King and Reis-Filho ; Lakhani et al. ; Tavassoli ; WHO ; Wen and Brogi ; Jorns et al. ). The original paper by Foote and Stewart from described and defined the morphological criteria and differences between ALH and LCIS as both lesions exhibiting the same low-grade monotonous nuclear atypia but differing quantitatively in their acinar involvement (Foote and Stewart ; King et al. ; King and Reis-Filho ; Wen and Brogi ; Jorns et al. ). Alternative terminologies such as classical lobular neoplasia (LN classical type) including both ALH and LCIS as well as the term Lobular Intraepithelial Neoplasia (LIN) covering LIN-1, LIN-2, LIN-3 as consecutive morphological categories represent a further approach to classify the same lobular breast lesions (Lakhani et al. ; Tavassoli ; WHO ; Rageth et al. ; AGO ). These lesions are considered both as risk-factor and non-obligate precursor for breast cancer in terms of uncertain malignant potential also categorized as B3 lesions in some guidelines (Lakhani et al. ; Tavassoli ; WHO ; Wen and Brogi ; Rageth et al. , ). Long-term cumulative risk of classical LN for breast cancer is 1–2% per year, resulting in 8–10 × relative risk for LCIS and 4–5 × relative risk for ALH (King et al. ; King and Reis-Filho ; Lakhani et al. ; WHO ; Rageth et al. ). Morphological variants with high nuclear grade, with the presence of necrosis or with extensive involvement of the multiple acini, are considered as separate entities and are designated as pleomorphic LCIS/LN, florid LCIS/LN, or LIN-3, and are also categorized as B5a category (non-invasive pre-malignant lesion) (Lakhani et al. ; Tavassoli ; WHO ; Rageth et al. ; ). Pleomorphic LCIS/LN, florid LCIS/LN, or LIN-3 can morphologically mimic solid type of DCIS, but represent molecularly distinct entities (WHO ; Wen and Brogi ; Shamir et al. ). However, lobular lesions in the B5a category behave biologically similar as their DCIS counterpart, have higher risk for local recurrence and progression to invasive cancer, are more often Her2 positive, and therefore, their clinical management is very similar to DCIS (WHO ; Wen and Brogi ; Shamir et al. ). On the contrary, classical lobular neoplasia forms are known to have a different biological behavior in terms of local recurrence and development of synchronous or subsequent breast cancer than the high-grade variants (King et al. ; King and Reis-Filho ; Wen and Brogi ; Rageth et al. , ; Schmidt et al. ). Upgrade rate to in situ or invasive cancer in open excision specimens has been conflictingly reported in the literature varying from 0 to 25% in some papers up to 50% (Rageth et al. , ; Schmidt et al. ). No association with common clinical risk factors as positive family history or age can be linked to clinical behavior, and until now, no single histopathological factor could predict upgrade or development of concurrent or subsequent breast cancer (King et al. ; Rageth et al. , ). However, clinical management of classical lobular neoplasia has undergone relevant modifications during the last decade, including the identification of imaging target lesions as visible lesions and the histological association to mammographic calcifications into the management workflow (Rageth et al. , ; AGO ). Current therapeutic guidelines recommend open excision for classical LN forms in breast core biopsies if there is a target lesion on imaging and in case of any inconsistency between imaging modalities and pathological assessment (Rageth et al. , ; AGO ). In all other classical LN cases, a conservation approach with a high-risk senological follow-up is acceptable, especially in diagnoses made by breast vacuum biopsy and if the radiological target has been removed (Rageth et al. , ; AGO ). Although this therapeutic approach has been the standard for all classical LN forms, until now, guidelines do not consider different subgroups of classical lobular neoplasia as ALH vs LCIS or LIN-1 vs LIN-2 (Rageth et al. , ; AGO ). Classical forms of B3 LN lesions are mainly subjected to a very similar therapeutic workflow (Rageth et al. , ; AGO ). The AGO guidelines specifically do not recommend the distinction between LIN-1 and LIN-2, because prognostic differences have not adequately been documented and proven until now, even though absolute risk for breast cancer development differs between ALH and classical LCIS (AGO ). Reproducibility issues concerning a wide spectrum of pre-malignant breast lesions, biomarkers, or degree of atypia have been the subject of several previous papers (Shamir et al. ; Choi et al. ; Fitzgibbons ; Gomes et al. ; Allison et al. ; Carney et al. ; Elmore et al. , , ; O'Malley et al. ; Onega et al. ; Schuh et al. ; Sloane et al. ; Tan et al. ; Wells et al. ). The use of immunohistochemistry with aberrant E-Cadherin staining combined with morphological criteria led to an excellent agreement (86.9%) of correctly classifying in situ or invasive lobular carcinomas and rule out morphological differential diagnoses of duct lesions such as solid-type DCIS or invasive ductal carcinomas (Choi et al. ). Gomes et al. reported differential inter-observer variability among pre-malignant breast lesions including atypical ductal hyperplasia, columnar cell lesions, lobular neoplasia, and DCIS in a large series of second opinions. In this paper, ALH and LCIS had both had a substantial inter-observer agreement after external review (Kappa 0.62 vs 0.66) (Gomes et al. ). Similar data were observed in a study by Fitzgibbons, where ALH and LCIS were inadequately classified when considered as separate entities (17% and 58% correct diagnoses); however, diagnostic accuracy improved to 74% when both lesions were categorized as one entity (Fitzgibbons ). Our results corroborate with these observations, single entities such as ALH, classical LCIS, or LIN-1 or LIN-2 did not result in satisfactory agreement (10–30% agreement), only using one category as classical lobular neoplasia including all B3 entities had an improved agreement (65%), which was also statistically significant. Singh et al. reported on a similar trend on improved reproducibility when pleomorphic and florid lobular carcinoma in situ were grouped into one diagnostic category. As was also suggested by Haagensen et al. more than 4 decades ago and also supported by the current study, insufficient reproducibility between slightly different histological entities can be improved using one category as classical LN. Similar issues were addressed in DCIS in several previous studies (Onega et al. ; Schuh et al. ; Sloane et al. ; Wells et al. ). Comparing three DCIS classification systems, the van Nuys system resulted in the highest diagnostic agreement in the Sloane project and by Shuh et al. (Kappa 0.42 and 0.37), although the final histological grading of DCIS was better reproducible using the Holland classification in the other studies (Kappa 0.53) (Schuh et al. ; Sloane et al. ; Wells et al. ). Applying a two-tiered grading system in DCIS (as low vs high grade) as opposed with reference diagnoses, high-grade DCIS proved to be more robust than low grade (83% vs 46% agreement with reference diagnoses) (Onega et al. ). Reproducibility issues in atypical ductal breast lesions, such as columnar cell lesions, flat epithelial atypia (FEA), atypical ductal hyperplasia (ADH), or DCIS show a similarly unequal trend (Allison et al. ; Carney et al. ; O'Malley et al. ; Tan et al. ). Agreement for FEA varies in the literature from poor (Kappa 0.27) to excellent (Kappa 0.83) (O'Malley et al. ; Tan et al. ). Regarding ADH, solid or micropapillary pattern with borderline cytological atypia was shown to be associated with lower agreement than those with cribriform pattern and clearly monotonous atypia (Allison et al. ). Differences and agreements in pathologist’s opinions in a broader range of breast surgical specimens were documented in several earlier papers (Carney et al. ; Elmore et al. ; ; ). Under- and overestimation of atypia and consistency in overall agreement with diagnostic standards were found between non-academic and academic pathologists (77.6% vs. 46%) (Carney et al. ). Elmore et al. reported misinterpretation in terms of atypia as highest after one single evaluation (52.2%) and the level of diagnostic concordance as highest in invasive carcinoma and lowest for DCIS and atypia (Elmore et al. , , ). In summary, our results show that existing different terminologies on classical form of LN in general have a poor-to-substantial agreement among expert breast pathologists on the same lesion, except when using a single category of classical lobular neoplasia. Regarding therapeutic approaches, until now, there is no difference in management between ALH, LCIS, LIN-1, and LIN-2 or classical LN (Rageth et al. , ; AGO ). Decisions for open surgery currently require discordant lesions between histology and imaging, a suspicious mass lesion in imaging or inadequately removed target lesions by vacuum-assisted biopsies (Rageth et al. , ; AGO ). Although until now no single histopathological factor of classical LN diagnosis could be identified to predict upgrade or local recurrence, helpful morphological ancillary tools such as information on associated calcifications in LN and a rough LN extension in breast core and vacuum-assisted biopsies can contribute to management decisions and possibly enable image-based senological follow-up in larger subset of LN cases.
|
Health literacy development of Primary Health Care patients: qualitative research | 149a90fc-a308-471d-a232-8f11b6f358e6 | 11654553 | Health Literacy[mh] | Health literacy proposes empowering patients over their health condition, enabling shared decision-making with the health team. Gaps and difficulties in health literacy processes have been associated with an increase in unfavorable clinical outcomes . In this direction, the World Health Organization (WHO) has developed four manuals on the subject. In the first volume, some dimensions are highlighted as important in health literacy development, such as identifying how patients acquire knowledge about health, how they put this knowledge into practice, how political and social contexts influence this development, among others. With this, the WHO warns of the need to know how people acquire health knowledge before developing strategies for health promotion and prevention . At the moment, studies on the subject have been carried out predominantly in developed countries, and have prioritized quantitative analyses to measure health literacy based on the development of scales, such as the Test of Functional Health Literacy In Adults (TOFHLA) and Rapid Estimate of Adult Literacy in Medicine (REALM) . In the Brazilian context, validity studies of scales such as the Short Assessment of Health Literacy for Portuguese-speaking Adults (SAHLPA) and the Short Test of Functional Health Literacy In Adults (STOFHLA) also predominate . In 2018, a Brazilian study investigated the relationship between health literacy and sociodemographic factors, self-perception of health and quality of life in Primary Health Care (PHC) users, finding a significant association between low education and inadequate health literacy . In qualitative research in PHC in Brazil, a study on health literacy among older adults that showed satisfaction with the information received in PHC units and another that assessed users’ perceptions of health promotion and prevention activities carried out by students in these units stand out . However, since health literacy is a concept that encompasses complex skills and highlighting the importance highlighted by the WHO of assessing how people acquire health knowledge, there is a gap in analyzing how health literacy is developed in the context of the Brazilian population, of which three out of ten Brazilians are considered functionally illiterate . Added to this is the importance of understanding this phenomenon in PHC, which is the main place of care for patients with chronic non-communicable diseases. Thus, the current study aimed to identify the process of developing patients’ health literacy, relating it to their self-care practices.
To identify how patients develop health literacy and relate these findings to their self-care actions in the PHC context.
Ethical aspects The study complied with national and international ethical precepts for research involving human beings, and was approved by the Hospital de Câncer de Barretos - Fundação Pio XII Research Ethics Committee. All participants signed the Informed Consent Form (ICF) and were instructed regarding their anonymity and their freedom to withdraw their data from the research at any time. To preserve participant anonymity, they were identified with the letter P, followed by a number, referring to the order in which they were interviewed and, finally, the initial letter of the health unit belonging to patients. Theoretical framework Analyses on health literacy used in this study were carried out from the perspective of health literacy manuals developed by the WHO . Study design This is a qualitative, prospective study that followed the COnsolidating criteria for REporting Qualitative research (COREQ) guidelines . Methodological procedures The interviews were collected by the main authors of the study who, during the collection period, were residents in family and community medicine and, according to the organization of the residency, spent two years working in Family Health Strategy (FHS) units. This longitudinal work allowed them to form significant bonds with the population covered by the units and to perceive common weaknesses and potentialities within their respective health units, culminating in a curiosity about the development of patients’ health literacy, which is so important for health care, especially in primary care. Two instruments were used to collect data: a questionnaire with patient sociodemographic data; and a semi-structured interview script constructed by the researchers, with ten guiding questions. The interviews were recorded individually with each participant after their consent, using the researchers’ personal recorder. Afterwards, the audios were stored in REdCap . Study setting The study was conducted in two FHS units, one of which was made up of a FHS team, covering a population of approximately 4,500 inhabitants, and the other, made up of three health teams, covering 9,684 individuals. Data source Study participants are users of two FHS units, who were included because they were from the area covered by the FHS units of the study, were between 18 and 70 years old and had at least one chronic disease. Patients with severe neurological or psychiatric conditions and patients under exclusive home care were excluded from the study. Firstly, the researchers identified the three most frequent chronic diseases in each reference team of the FHS units in the study, which became the chronic diseases investigated in this study. Selection was carried out according to Flick’s (2009) recommendations for sample selection in qualitative research. Among the various possibilities for recruitment in qualitative research indicated by the author, the researchers used formal sampling, chosen due to limited time to complete the research . According to the author, in qualitative research, it is crucial to select participants based on relevant criteria to obtain homogeneous content, in order to demonstrate cohesion in the analysis of reports provided by users as well as highlight the diversity between the different attributes of research participants . To achieve this, participants with extreme or deviant cases for each health condition were initially chosen, followed by typical cases. Within each chronic health condition, at least two participants were selected with distinct variables, such as age, biological sex, years of education and whether or not they had a companion. Patients who met the inclusion criteria were selected based on the researchers’ experience and knowledge of the population being monitored and through case discussions in team meetings, until the criteria were in accordance with the author’s recommendations for formal sampling . As a result, 22 patients were recruited between August and September 2024. Data collection and organization Firstly, a script was prepared with personal and sociodemographic data of each participant and then a semi-structured interview was conducted with the following guiding questions: what do you understand about your disease? How do you take care of your health? How did you learn the information about your disease? Can you understand all the information that health professionals give you during consultation? What difficulties do you have in taking care of your health? When the doctor writes a prescription, do you understand it? How do you feel about the decisions for your treatment? When you receive information through another means, other than the health team, do you talk to the team about it? How do you see access to our health unit? The interviews took place between 10/01/2023 and 12/01/2023, and were conducted in person at the FHS of reference for each participant, lasting 12 to 38 minutes. Each interview was recorded on the researchers’ own recorder and stored in REDcap, with the presence of the researcher and the participant. Data analysis Participant sociodemographic characterization was presented through descriptive statistics generated by REDCap , presented in graphs and tables. The interviews were analyzed based on the theory of thematic content analysis proposed by Bardin and interpreted in light of the manual on health literacy developed by the WHO . This process began with pre-analysis, in which we sought, through skimming, to familiarize ourselves with the data, choosing and highlighting passages with similar meanings. In parallel, we resumed an in-depth reading of the manuals, noting the dialogue between the results obtained in the interviews and the dimensions presented in the first volume of the WHO manual on health literacy, which explains how users acquire health knowledge and how they put it into practice . Hence, it was possible to identify two thematic categories, which were submitted to the authors’ interpretation. illustrates the stages contemplated in the analysis.
The study complied with national and international ethical precepts for research involving human beings, and was approved by the Hospital de Câncer de Barretos - Fundação Pio XII Research Ethics Committee. All participants signed the Informed Consent Form (ICF) and were instructed regarding their anonymity and their freedom to withdraw their data from the research at any time. To preserve participant anonymity, they were identified with the letter P, followed by a number, referring to the order in which they were interviewed and, finally, the initial letter of the health unit belonging to patients.
Analyses on health literacy used in this study were carried out from the perspective of health literacy manuals developed by the WHO .
This is a qualitative, prospective study that followed the COnsolidating criteria for REporting Qualitative research (COREQ) guidelines .
The interviews were collected by the main authors of the study who, during the collection period, were residents in family and community medicine and, according to the organization of the residency, spent two years working in Family Health Strategy (FHS) units. This longitudinal work allowed them to form significant bonds with the population covered by the units and to perceive common weaknesses and potentialities within their respective health units, culminating in a curiosity about the development of patients’ health literacy, which is so important for health care, especially in primary care. Two instruments were used to collect data: a questionnaire with patient sociodemographic data; and a semi-structured interview script constructed by the researchers, with ten guiding questions. The interviews were recorded individually with each participant after their consent, using the researchers’ personal recorder. Afterwards, the audios were stored in REdCap .
The study was conducted in two FHS units, one of which was made up of a FHS team, covering a population of approximately 4,500 inhabitants, and the other, made up of three health teams, covering 9,684 individuals.
Study participants are users of two FHS units, who were included because they were from the area covered by the FHS units of the study, were between 18 and 70 years old and had at least one chronic disease. Patients with severe neurological or psychiatric conditions and patients under exclusive home care were excluded from the study. Firstly, the researchers identified the three most frequent chronic diseases in each reference team of the FHS units in the study, which became the chronic diseases investigated in this study. Selection was carried out according to Flick’s (2009) recommendations for sample selection in qualitative research. Among the various possibilities for recruitment in qualitative research indicated by the author, the researchers used formal sampling, chosen due to limited time to complete the research . According to the author, in qualitative research, it is crucial to select participants based on relevant criteria to obtain homogeneous content, in order to demonstrate cohesion in the analysis of reports provided by users as well as highlight the diversity between the different attributes of research participants . To achieve this, participants with extreme or deviant cases for each health condition were initially chosen, followed by typical cases. Within each chronic health condition, at least two participants were selected with distinct variables, such as age, biological sex, years of education and whether or not they had a companion. Patients who met the inclusion criteria were selected based on the researchers’ experience and knowledge of the population being monitored and through case discussions in team meetings, until the criteria were in accordance with the author’s recommendations for formal sampling . As a result, 22 patients were recruited between August and September 2024.
Firstly, a script was prepared with personal and sociodemographic data of each participant and then a semi-structured interview was conducted with the following guiding questions: what do you understand about your disease? How do you take care of your health? How did you learn the information about your disease? Can you understand all the information that health professionals give you during consultation? What difficulties do you have in taking care of your health? When the doctor writes a prescription, do you understand it? How do you feel about the decisions for your treatment? When you receive information through another means, other than the health team, do you talk to the team about it? How do you see access to our health unit? The interviews took place between 10/01/2023 and 12/01/2023, and were conducted in person at the FHS of reference for each participant, lasting 12 to 38 minutes. Each interview was recorded on the researchers’ own recorder and stored in REDcap, with the presence of the researcher and the participant.
Participant sociodemographic characterization was presented through descriptive statistics generated by REDCap , presented in graphs and tables. The interviews were analyzed based on the theory of thematic content analysis proposed by Bardin and interpreted in light of the manual on health literacy developed by the WHO . This process began with pre-analysis, in which we sought, through skimming, to familiarize ourselves with the data, choosing and highlighting passages with similar meanings. In parallel, we resumed an in-depth reading of the manuals, noting the dialogue between the results obtained in the interviews and the dimensions presented in the first volume of the WHO manual on health literacy, which explains how users acquire health knowledge and how they put it into practice . Hence, it was possible to identify two thematic categories, which were submitted to the authors’ interpretation. illustrates the stages contemplated in the analysis.
Participant characteristics shows research participant sociodemographic characteristics. shows the average number of years lived with the underlying disease. Thematic categories of interviews Two categories emerged from the interviews conducted: 1) Health knowledge construction; and 2) Dialogue between health knowledge construction and patient care actions. Health knowledge construction According to the reports of research participants, knowledge construction for health literacy is established in four ways: through interpersonal relationships; through social media; through bodily sensations; and through relationships with health professionals. Among them, the most evident way of constructing health literacy was through personal experiences with family, friends or acquaintances and the beliefs that emerge from these relationships: I was always dealing with it, because I took care of my elderly mother, for many years, right? I took care of everything for ten years . (P3D) I had a lot of experience, because my husband had heart surgery twice and that rush, always with him too, and with my mother, taking care of everything so we deal with a lot of health problems, right? (P3D) Because at work there are a lot of heart problems. I talked to everyone but why did you show up with your problem? So, we talk, even about medicine, which doctor are you going to? So, you’ll gain a lot of knowledge. (P5D) My colleagues, I know a lot of people with diabetes, they talk so much, then things start to add up, check mate, you know? The eyesight, the tired legs, the tired body. (P8D) Even when accompanied by guidance from a health professional, knowledge construction for health literacy is facilitated when the professional is a family member or someone they know: Friends from work, from the Basic Health Unit and from the family unit also have family members who are doctors. (P7D) My niece is a secretary. She is now a nurse. She works at the reception desk at Santa Casa. My sister is now a social worker. She also teaches me a lot. I have doubts, but at home I have an introduction, is that what I say? An interpreter, who is Joana. She has already studied nursing, worked in the health area . (P4D) My daughter is a nursing technician, today she works in the laboratory, but she worked more in the administrative area. So, I ask her to ask the doctor everything, there is always guidance. (P3D) Sometimes, this construction, through family relationships, generates doubts and can generate a certain disbelief in relation to guidelines on chronic diseases: My mother died at 94 years old, eating only pork fat. The only thing that happened was that she lost her sight, she couldn’t see anymore. She was lucid, she didn’t have diabetes. My sister had it, my brother had it and my sister’s daughters already have it. So, I don’t know, is it sedentary or what else can it be. (P3D) I don’t even know what to say about my illness, because my mother never drank, never smoked, never did anything, she has diabetes and even high blood pressure so, I don’t know. (P9D) The internet also appears as an ally in knowledge construction for participants’ health literacy: Because I also do a lot of research on Google on Google, because you have to research. Don’t be so ignorant about health, your own health, right? (P4D) On Google or YouTube. There are always reports of people explaining it. But even though the disease is the same, it’s just that each person is different. (P8I) On the internet, I see the lesions that diabetes causes and I tell the doctor. I was a little scared. (P11I) However, patients often compare information on the internet with the advice of health professionals: I only feel trustworthy when I see a health center or some health unit, because I even look for information like that on the internet. The other day it came up that diabetes is caused by a worm in the pancreas, and then you’re like, “Oh my God”. (P1I) I look for it because I don’t like to do things on my own, I try to find out, because there are things on the internet that you can’t . (P4I) Guidance and connections with health professionals also appear to be important in building knowledge for participants’ health literacy: Without the help of a doctor, without the help of a professional in this area, you can’t do anything, you’ll just read, but you won’t do anything, you need to have someone to help you, to show you the right path. (P3I) I go to the pharmacist and always mark exactly how to take it, stick a little piece of paper, time, how to take it, whether it’s before a meal, whether it’s after. (P3D) Look, doctor, I’m not one to research things on the internet, no. I have the information I have, it’s from my doctors who follow me, right? (P4I) I’ve gotten so used to the doctor that I try to come on the day she’s there to solve it. I feel like: if I have a problem, I can come here tomorrow morning and it will be solved. I’m very happy. (P6D) Finally, another way to build health literacy is through the perception of one’s own body: My head gets cloudy sometimes and my head feels bad, then I know it’s a little high. (P2D) Sometimes, a little dizziness, nausea, shortness of breath. Sometimes, tachycardia until I discovered that it was anxiety. (P5I) Because when it goes up uncontrollably, it makes you feel bad, it makes you nauseous. And when it goes down, it makes me desperately hungry, I even eat things I don’t like. And I feel very weak, sometimes it even feels like I’m going to faint from dizziness. (P8I) Dialogue between health knowledge construction and patient care actions When health literacy construction occurs mainly through interpersonal relationships, health care is linked to these relationships: Because my neighbor, she was like that. I’m not feeling anything, I’m not going to take it. I was like that too but then, she lay down, had a facial hemorrhage. So, from now on, oh, I’m not without my medicine under any circumstances. (P6D) I want them to be proud of me, because they say they’re there thinking about me here. my son said, “Hey, come here so we can buy the medicine for you” I took it and said, “Well, I’m not going to disappoint him, right? I’m going to do it right”. (P6D) she found out she had cancer, right? It was too late, you know? So, even because I’ve been with her through all this, I try to take good care of my health, like, as much as I can, you know? (P6I) Because my neighbor, she went to do it because I didn’t have anyone to go with me, right? So, my neighbor is doing it because she’s pre-diabetic and needs to do it and she’s taking me. Because there I am alone, I am afraid, I will not go . (P2D) Then, I was talking to a friend of mine, and she said, “No, Regina, there’s no danger, they won’t let you drown.” It gave me confidence she said, “You can go with me, we’ll come together” . (P2D) My daughter is studying medicine, she’s always on our case, she keeps a close eye on us, you know? She sees how we’re doing, she sees all the tests, you know? She sees what we’re eating, what we’re not eating, she keeps an eye on us. She keeps a close eye on us too. (P6I) I hear at the bar, you know? That there are a lot of people who have diabetes there, that’s why I don’t understand the diabetes thing. I can’t eat sweets, I can’t eat anything. There are others who go there, who drink pinga [Brazilian beverage] with coke and are diabetic too. Why is one different from the other? (P6D) Although social media appears to be a contributor to health knowledge construction, it does not appear to be an aid to the care process. However, the relationship between knowledge acquired by health professionals appears to be an important contributor, and is permeated by factors such as bonding, listening, seeing professionals as an authority figure, and professionals’ body and verbal language: She only asked for the tests. She said, “You have to do this thing”. I get really down, worried it makes me afraid to do the test and find out so I say, “I’m not going to do this test” . (P4D) a decision I made together was made by me and my psychologist, we were discussing the next steps, and I realized that the place I was living was not good. I went there and moved, I changed the activities I was doing during the day. Over time, I started waking up earlier again, going for walks, exercising. (P1I) and I was working normally, and they called me to do a regular check-up the doctor took my blood pressure, listened to my heartbeat, and then told me I had to wait, because my blood pressure had changed another person quickly attended to me and called an ambulance, and said I had to go straight to the emergency room, and I thought I was dying, I froze . (P5D) I feel fine, he’s a specialist, he knows, he’s studied, right? So, he advised me to change, I have to change, right? (P3D) because he’s there making himself available and seeing part of the process I’m going through. So, I believe he’s prescribing it to me because it’s going to do me good, right? . (P11D) In this regard, the professional-patient relationship often initiates the motivation for health care, and the sensations perceived in the body, as well as interpersonal relationships, function as facilitators or hinders in maintaining this care: Now, I’m participating in the group, right? A women’s group. It’s something that’s really helping me learn techniques, right? Meditation, breathing, we talk. So, it’s doing me good. I’m also having individual consultations with her. (P11D) Oh, I try to spend time with my family on the weekends, to pay more attention to my children because I realize that the emotional part, anxiety, is a factor that changes a lot at least on the weekends, I try to do some things differently go for a walk, knowing what we have to do to feel better. (P7D) Yes, I’m going for walks and my psychiatrist even advised me to go to the gym, because I was very quiet. But otherwise, I’m going to therapy. I can’t do without it. We have groups, a wonderful support group, so I don’t stop going. If I stopped going, I wouldn’t be here anymore . (P4I) Because if there is a treatment, a care for this, why not do it? Walk a little, eat more fruit and be calm. Try to stay calm too, because that contributes a lot to these feelings. (P7I) I stopped on my own, when I see that I’m getting really bad, I cry, and then it goes away a little and the medicine improves one thing, but hinders the other. As for dating, I had no desire at all. So, I think the medicine helped at that time when I was crying too much, all the time, but I think that now there’s no need for it . (P10I) Unlike social media, which appear as contributors to health knowledge construction, but not to care, faith and religion appear as factors that help in health care: I can tell you that every day we have to seek our faith because without faith we don’t get anywhere we read the Bible, we read the word, looking for a place to go. I think that’s the fundamental point. Not only for this, but for all the issues in our lives. (P7I) Thank God, I got rid of all that, because I sought it in my faith of course there is psychological and psychiatric help from doctors who are qualified for that. But fundamentally it’s this search, it’s a set of things that you have to do. (P7I)
shows research participant sociodemographic characteristics. shows the average number of years lived with the underlying disease.
Two categories emerged from the interviews conducted: 1) Health knowledge construction; and 2) Dialogue between health knowledge construction and patient care actions.
According to the reports of research participants, knowledge construction for health literacy is established in four ways: through interpersonal relationships; through social media; through bodily sensations; and through relationships with health professionals. Among them, the most evident way of constructing health literacy was through personal experiences with family, friends or acquaintances and the beliefs that emerge from these relationships: I was always dealing with it, because I took care of my elderly mother, for many years, right? I took care of everything for ten years . (P3D) I had a lot of experience, because my husband had heart surgery twice and that rush, always with him too, and with my mother, taking care of everything so we deal with a lot of health problems, right? (P3D) Because at work there are a lot of heart problems. I talked to everyone but why did you show up with your problem? So, we talk, even about medicine, which doctor are you going to? So, you’ll gain a lot of knowledge. (P5D) My colleagues, I know a lot of people with diabetes, they talk so much, then things start to add up, check mate, you know? The eyesight, the tired legs, the tired body. (P8D) Even when accompanied by guidance from a health professional, knowledge construction for health literacy is facilitated when the professional is a family member or someone they know: Friends from work, from the Basic Health Unit and from the family unit also have family members who are doctors. (P7D) My niece is a secretary. She is now a nurse. She works at the reception desk at Santa Casa. My sister is now a social worker. She also teaches me a lot. I have doubts, but at home I have an introduction, is that what I say? An interpreter, who is Joana. She has already studied nursing, worked in the health area . (P4D) My daughter is a nursing technician, today she works in the laboratory, but she worked more in the administrative area. So, I ask her to ask the doctor everything, there is always guidance. (P3D) Sometimes, this construction, through family relationships, generates doubts and can generate a certain disbelief in relation to guidelines on chronic diseases: My mother died at 94 years old, eating only pork fat. The only thing that happened was that she lost her sight, she couldn’t see anymore. She was lucid, she didn’t have diabetes. My sister had it, my brother had it and my sister’s daughters already have it. So, I don’t know, is it sedentary or what else can it be. (P3D) I don’t even know what to say about my illness, because my mother never drank, never smoked, never did anything, she has diabetes and even high blood pressure so, I don’t know. (P9D) The internet also appears as an ally in knowledge construction for participants’ health literacy: Because I also do a lot of research on Google on Google, because you have to research. Don’t be so ignorant about health, your own health, right? (P4D) On Google or YouTube. There are always reports of people explaining it. But even though the disease is the same, it’s just that each person is different. (P8I) On the internet, I see the lesions that diabetes causes and I tell the doctor. I was a little scared. (P11I) However, patients often compare information on the internet with the advice of health professionals: I only feel trustworthy when I see a health center or some health unit, because I even look for information like that on the internet. The other day it came up that diabetes is caused by a worm in the pancreas, and then you’re like, “Oh my God”. (P1I) I look for it because I don’t like to do things on my own, I try to find out, because there are things on the internet that you can’t . (P4I) Guidance and connections with health professionals also appear to be important in building knowledge for participants’ health literacy: Without the help of a doctor, without the help of a professional in this area, you can’t do anything, you’ll just read, but you won’t do anything, you need to have someone to help you, to show you the right path. (P3I) I go to the pharmacist and always mark exactly how to take it, stick a little piece of paper, time, how to take it, whether it’s before a meal, whether it’s after. (P3D) Look, doctor, I’m not one to research things on the internet, no. I have the information I have, it’s from my doctors who follow me, right? (P4I) I’ve gotten so used to the doctor that I try to come on the day she’s there to solve it. I feel like: if I have a problem, I can come here tomorrow morning and it will be solved. I’m very happy. (P6D) Finally, another way to build health literacy is through the perception of one’s own body: My head gets cloudy sometimes and my head feels bad, then I know it’s a little high. (P2D) Sometimes, a little dizziness, nausea, shortness of breath. Sometimes, tachycardia until I discovered that it was anxiety. (P5I) Because when it goes up uncontrollably, it makes you feel bad, it makes you nauseous. And when it goes down, it makes me desperately hungry, I even eat things I don’t like. And I feel very weak, sometimes it even feels like I’m going to faint from dizziness. (P8I)
When health literacy construction occurs mainly through interpersonal relationships, health care is linked to these relationships: Because my neighbor, she was like that. I’m not feeling anything, I’m not going to take it. I was like that too but then, she lay down, had a facial hemorrhage. So, from now on, oh, I’m not without my medicine under any circumstances. (P6D) I want them to be proud of me, because they say they’re there thinking about me here. my son said, “Hey, come here so we can buy the medicine for you” I took it and said, “Well, I’m not going to disappoint him, right? I’m going to do it right”. (P6D) she found out she had cancer, right? It was too late, you know? So, even because I’ve been with her through all this, I try to take good care of my health, like, as much as I can, you know? (P6I) Because my neighbor, she went to do it because I didn’t have anyone to go with me, right? So, my neighbor is doing it because she’s pre-diabetic and needs to do it and she’s taking me. Because there I am alone, I am afraid, I will not go . (P2D) Then, I was talking to a friend of mine, and she said, “No, Regina, there’s no danger, they won’t let you drown.” It gave me confidence she said, “You can go with me, we’ll come together” . (P2D) My daughter is studying medicine, she’s always on our case, she keeps a close eye on us, you know? She sees how we’re doing, she sees all the tests, you know? She sees what we’re eating, what we’re not eating, she keeps an eye on us. She keeps a close eye on us too. (P6I) I hear at the bar, you know? That there are a lot of people who have diabetes there, that’s why I don’t understand the diabetes thing. I can’t eat sweets, I can’t eat anything. There are others who go there, who drink pinga [Brazilian beverage] with coke and are diabetic too. Why is one different from the other? (P6D) Although social media appears to be a contributor to health knowledge construction, it does not appear to be an aid to the care process. However, the relationship between knowledge acquired by health professionals appears to be an important contributor, and is permeated by factors such as bonding, listening, seeing professionals as an authority figure, and professionals’ body and verbal language: She only asked for the tests. She said, “You have to do this thing”. I get really down, worried it makes me afraid to do the test and find out so I say, “I’m not going to do this test” . (P4D) a decision I made together was made by me and my psychologist, we were discussing the next steps, and I realized that the place I was living was not good. I went there and moved, I changed the activities I was doing during the day. Over time, I started waking up earlier again, going for walks, exercising. (P1I) and I was working normally, and they called me to do a regular check-up the doctor took my blood pressure, listened to my heartbeat, and then told me I had to wait, because my blood pressure had changed another person quickly attended to me and called an ambulance, and said I had to go straight to the emergency room, and I thought I was dying, I froze . (P5D) I feel fine, he’s a specialist, he knows, he’s studied, right? So, he advised me to change, I have to change, right? (P3D) because he’s there making himself available and seeing part of the process I’m going through. So, I believe he’s prescribing it to me because it’s going to do me good, right? . (P11D) In this regard, the professional-patient relationship often initiates the motivation for health care, and the sensations perceived in the body, as well as interpersonal relationships, function as facilitators or hinders in maintaining this care: Now, I’m participating in the group, right? A women’s group. It’s something that’s really helping me learn techniques, right? Meditation, breathing, we talk. So, it’s doing me good. I’m also having individual consultations with her. (P11D) Oh, I try to spend time with my family on the weekends, to pay more attention to my children because I realize that the emotional part, anxiety, is a factor that changes a lot at least on the weekends, I try to do some things differently go for a walk, knowing what we have to do to feel better. (P7D) Yes, I’m going for walks and my psychiatrist even advised me to go to the gym, because I was very quiet. But otherwise, I’m going to therapy. I can’t do without it. We have groups, a wonderful support group, so I don’t stop going. If I stopped going, I wouldn’t be here anymore . (P4I) Because if there is a treatment, a care for this, why not do it? Walk a little, eat more fruit and be calm. Try to stay calm too, because that contributes a lot to these feelings. (P7I) I stopped on my own, when I see that I’m getting really bad, I cry, and then it goes away a little and the medicine improves one thing, but hinders the other. As for dating, I had no desire at all. So, I think the medicine helped at that time when I was crying too much, all the time, but I think that now there’s no need for it . (P10I) Unlike social media, which appear as contributors to health knowledge construction, but not to care, faith and religion appear as factors that help in health care: I can tell you that every day we have to seek our faith because without faith we don’t get anywhere we read the Bible, we read the word, looking for a place to go. I think that’s the fundamental point. Not only for this, but for all the issues in our lives. (P7I) Thank God, I got rid of all that, because I sought it in my faith of course there is psychological and psychiatric help from doctors who are qualified for that. But fundamentally it’s this search, it’s a set of things that you have to do. (P7I)
In this study, we can see the alignment with the WHO manual regarding the dimensions of how health service users acquire knowledge and apply it in self-care practices. We realize that knowledge acquisition occurs predominantly through closer social interactions, including friends and family, especially if the latter are health professionals or if this learning emerges from the experience of individuals as caregivers of a family member with a health condition similar to their own. Dialogue allows individuals to share their experiences, concerns, seek emotional support and learn from each other’s experiences, demonstrating, according to WHO manuals, that health literacy development is a social practice . Knowledge construction for health literacy through interpersonal relationships motivated most of the interviewees to self-care. When they observed complications from chronic diseases in people close to them, they started to take better care of their diet or practice physical activity. However, for some, this relationship generated skepticism regarding the information provided by health professionals, comparing it with the reality experienced by their family members. One participant in this study mentioned a family member with unhealthy eating habits who had a long life without developing diabetes, questioning the relationship between healthy habits and chronic diseases. This result is in line with studies aimed at understanding individuals’ health literacy . The concept of health literacy mediators describes people who make their health literacy available to others, formally or informally, and highlights social support as one of the most important mediators of health literacy , highlighting what the WHO manual on health literacy warns: the fact that health literacy is not an individual task and that friends and family communicate with health professionals on behalf of and in collaboration with the patients they assist . The proximity of interpersonal relationships was also in line with those found in this study, with the nuclear family being the main mediator in health literacy construction, followed by friends and, finally, coworkers and members of support groups . These findings highlight the importance of initiatives such as the Patientand Family-Centered Care Model (PFCCM) . However, recent studies show the difficulty of health professionals in including patients and family members in care decisions. In Brazil, research indicates the difficulty of integrating PFCCM into patient care , in addition to having an insufficient number of studies in the area. Another important point in the relationship between health knowledge construction and interpersonal relationships is the importance of strategies within the patient community. Patients who participated in health education groups learned more about their conditions than those who did not participate . Interventions that promote group support are also emphasized by the WHO health literacy manual . This study, like others on health literacy, showed that digital technology and social networks are fundamental tools in the dissemination of public health information . Some patients interviewed consider the media to be one of the main sources of information acquisition, due to the speed and accessibility of content. However, there is a duality, since the population of this study sought to compare digital information with that provided by trusted health professionals, perceiving it as confusing, contradictory or even false. Therefore, the results of this study support previous research, by identifying social media as an important factor in health literacy development. However, in this study, the media are viewed with suspicion and appear as secondary to interpersonal relationships. However, as the use of social media appears in participants’ statements, the importance of using it as an ally in health care is evident. National and international studies show that social media can provide reliable information, form support communities among users with the same health conditions, facilitate behavioral changes necessary to manage chronic conditions, train health professionals, and monitor and track patients’ habits . In line with studies on health literacy, participants in this study highlight the role of health professionals as facilitators of the learning process, with a good doctor-patient relationship being crucial for individuals to trust the information received. The quality of the health professional-patient relationship is related to trust, empathy, communication, listening and information sharing. However, studies show that not all professionals support health literacy development or encourage patients to interact with information before choosing treatment. Some create barriers that prevent the development of self-care skills . In this study, patients highlighted that observing health professionals’ body and verbal language is essential for decision-making, generating trust or distrust. Patients with less trust and connection with health professionals have higher levels of glycated hemoglobin and more cardiovascular events . The main barriers to health literacy are poor communication skills among health professionals, which makes patients feel that they have not received enough information and have not been listened to. Some patients report that their information was not taken into account by professionals during consultation. By not taking into account the needs of patients themselves, behavior change becomes more difficult. This fact appears in the reports of participants in this study. It is clear that participants are more likely to change their behavior when they realize that the information provided by health professionals is constructed jointly, based on the bond and their own needs. In contrast, when professionals are seen only as an authority figure, reports of behavior change do not accompany the discourse. In this regard, communication strategies, such as motivational interviewing (MI), are important allies in promoting these changes. In Brazil, MI is applied, in short, in the context of the use of psychoactive substances, and is still little used in the context of PHC . Another way of constructing knowledge for health literacy evidenced in this study was their own body perceptions, which were correlated to a given health condition. It is up to health professionals to encourage patients to develop self-awareness and guide them in interpreting such signs and symptoms so that they can differentiate between common symptoms and serious symptoms. The WHO states that activities that encourage moments of self-awareness favor health literacy construction and self-care promotion, such as Mindfulness and Yoga, which are considered important tools in self-care construction . Finally, faith, based mainly on the Christian religion, appears as a mediator of motivation for self-care in health, a fact not found in other qualitative studies on health literacy. It is known that religious practices and groups have a great influence on the community structure and, at times, participate in the promotion of preventive programs, such as rehabilitation centers, physical activity practices and health education . Hence, religious places can be designed to distribute educational pamphlets, disseminate health information and even serve as a stage for health professionals to give educational lectures. The current study highlights the need for trained professionals to explore and utilize patients’ prior knowledge about their chronic health conditions, especially their personal and interpersonal experiences. It is crucial to implement communication actions that reach communities and disseminate health information broadly, beyond individual consultations. Continuing education in communication techniques and motivation for behavior change can be an effective strategy. Complementary therapies, especially mind-body practices, are important allies, as they promote body self-knowledge and can be applied in groups. Study limitations This study identified some difficulties. Firstly, conducting the interviews was hampered by the limitation of physical space, as one of the health units had inadequate infrastructure and a small room, without adequate ventilation, for conducting the interviews. Finally, the deadline for submitting the Residency Conclusion Work, combined with the practical workload required by the residency program, prevented the exploration of other aspects of health literacy, such as comparing health literacy construction indicated by participants with the opinions of health professionals who assist this population, in order to investigate discrepancies and similarities in the discourses as well as to carry out a content analysis before and after health education interventions in the same participants. These notes may constitute future fields of research needed in Brazil. Contributions to nursing, health or public policy This study was the first to qualitatively investigate knowledge acquisition for health literacy in PHC patients in Brazil. Studies with similar objectives have been conducted outside Brazil, analyzing populations with different cultural habits, ethnicities, economies, geographies and social practices, without considering Brazilian specificities. This may explain why this study was the only one to report the contribution of religion to health literacy, highlighting the need for more Brazilian research. Although quantitative studies on the topic are important, qualitative research is crucial to capture the subjectivities of discourses, essential to understanding complex social phenomena such as health literacy.
This study identified some difficulties. Firstly, conducting the interviews was hampered by the limitation of physical space, as one of the health units had inadequate infrastructure and a small room, without adequate ventilation, for conducting the interviews. Finally, the deadline for submitting the Residency Conclusion Work, combined with the practical workload required by the residency program, prevented the exploration of other aspects of health literacy, such as comparing health literacy construction indicated by participants with the opinions of health professionals who assist this population, in order to investigate discrepancies and similarities in the discourses as well as to carry out a content analysis before and after health education interventions in the same participants. These notes may constitute future fields of research needed in Brazil.
This study was the first to qualitatively investigate knowledge acquisition for health literacy in PHC patients in Brazil. Studies with similar objectives have been conducted outside Brazil, analyzing populations with different cultural habits, ethnicities, economies, geographies and social practices, without considering Brazilian specificities. This may explain why this study was the only one to report the contribution of religion to health literacy, highlighting the need for more Brazilian research. Although quantitative studies on the topic are important, qualitative research is crucial to capture the subjectivities of discourses, essential to understanding complex social phenomena such as health literacy.
The results indicate that the analysis of how health literacy is constructed in chronic non-communicable disease management is a continuous process of self-knowledge and self-management of one’s own health condition. Family members, friends and co-workers are also the main mediating agents in health literacy development and practice, sharing knowledge, experiences, skills and forming a support network in decision-making. Therefore, understanding family dynamics also allows us to understand the health literacy construct. The health professional-patient relationship is an important pillar for the information received by patients to be transformed into motivation for self-care. Thus, PHC, based on longitudinality and the person-centered approach (PFCCM), contributes so that the health professional-patient relationship is not a barrier, but rather a dyad alliance for health literacy construction.
|
Concept of defensive medicine and litigation among Sudanese doctors working in obstetrics and gynecology | 1ec25837-a407-4381-a20e-20abea685a23 | 4748468 | Gynaecology[mh] | Defensive medicine is defined as a doctor’s deviation from the usual practice in order to reduce or prevent criticism and complaints by patients or their relatives . Some would claim that it is a legitimate phenomenon, while others consider it immoral . In addition to this definition the United States Congress also include the action of ordering tests, procedures and visits, or avoidance of high risk patients or procedures with the primary (but not sole) aim of reducing mal-practice liability as a part of defensive medicine . A genuine difficulty exists when trying to identify and quantify the extent of defensive medicine practices. This is partially because there is a grey area between proper and overly self-protective treatment. It may be difficult to recognize medical actions that are more likely to result in legal action. Obstetricians and Gynaecologists like other health care professional have a legal obligation to adhere to reasonable standards of care while acting in their professional capacity, they always has reputation for being a highly litigious . Their field is surrounded by different circumstances that stimulate them to practice defensive medicine. About 5–7.4 % of physicians in USA faced a malpractice claim annually . Gynaecology alone had the 12th highest average annual proportion of physicians with a claim, with the highest payment rate (38 %) . Obstetrics and general surgery are regarded as high risk specialties . As a result, the rising cost of malpractice insurance in obstetrics and genecology has led to a reality where doctors may refrain from treating high risk patients . Again medical litigation represents a real threat for the doctors and may be a direct cause to leave the profession. Medical law is the aspect of the law which governs the relationship between the healthcare provider and patient . The medical practitioner is bound by certain laws depending on the circumstances of his practice. Law and ethics may overlap since obtaining patient permission is both legally required and the “right thing to do” . The Sudan Medical Council (SMC) standing disciplinary committee investigates any complaint that come to its notice or violation to medical ethics. The SMC has the power to erase doctors from its register or withhold the license of medical, dental or pharmacy institution or facility. Many studies were done worldwide concerning the medical litigation especially against obstetrics and gynaecology however none was carried out in Sudan thus this study was directed to assess the concept of defensive medicine (in term of knowledge and prevalence) and to determine any experience of medical litigations and their sources among different grades of Sudanese doctors working in obstetrics and gynaecology.
Hypothesis While we did not adopt a formal hypothesis for this study, our working hypothesis/assumption was that defensive medicine affects daily doctor’s clinical judgement and practice. Study design and data collection This study was directed to assess the extent and the possible effect of defensive medicine phenomenon (in term of knowledge and prevalence) on medical decision making (development of tools that can guide physicians to make good decisions in practice) among different grades of obstetric and gynaecologic doctors, and to determine any experience of medical litigations with respect to sources and factors associated with it (in term of area of work, characteristics of the area at which the doctors worked, professionalism, hospitals systems…ect). Using a self administered questionnaire (Additional file ) and after obtaining informed written consent the data was collected from the different certified OBGYN professionals (Registrars, Specialists and Consultants) working in obstetrics and gynaecology and who attended the 27th congress of obstetrical and gynaecological society of the Sudan held from (20th -23rd February 2015) in Khartoum. The survey included only Sudanese doctors who were practicing obstetrics and gynaecology in Sudan. Visiting Doctors who are practicing in a different context abroad were excluded from the study. We used a questionnaire which was constructed by the authors to consider different forms of defensive medicine and medical litigation in obstetrics and gynaecology. Information sought by the questionnaire included: socio-demographic characteristics (age, grade, gender, area of work, duration of work, health insurance coverage), information on the area of work (blame culture: which defined as no one accepts medical errors as being all right), information on the hospital where the respondent worked (hospital guidelines and protocol, auditing system, committees…ect) daily experience (informed consent, high risk consent, documentation), whether the respondents knew the concept of defensive medicine or not? and questions on different examples of positive and negative defensive medicine (prescription of unnecessary medication, experience unnecessary refer, refuse to manage high risk patient, request for unnecessary investigation, experience unnecessary surgical procedure and avoidance high risk surgical procedure because of fear of criticism or litigation). Other information obtained from the respondents included: is the litigation in OBGYN increasing?, whether the respondents experienced litigations during their daily practice and the source of the litigations (fetal distress, misdiagnosis, injury to the viscera, shoulder dystocia, death…ect). The definition of defensive medicine was framed by the investigators according the definition in the literature ; defined as a doctor’s deviation from their usual behaviour or that considered good practice, to reduce or prevent complaints or criticism by patients or their families. This definition was not set out in the questionnaire for the respondents however in the questionnaire we asked the respondents whether they know the defensive medicine or not. Defensive medical practices were further subcategorized into positive and negative practice. When extra tests and procedures are performed primarily to reduce malpractice liability, this is a positive defensive medicine. Negative defensive medicine consists of avoidance of certain patients and procedures, thereby withdrawing medical services, and can deny patients productive care . In the questionnaire we explained the situation of high risk consent which is taken in case of serious / complicated / risky / new - surgeries or procedures; for removing any organ; in high risk patients; for proceeding with a surgery / procedure in spite of any abnormal parameters of the patient. This list is indicative not exhaustive and in case of a dilemma it is always advisable to take this high-risk consent and not a general consent.) Statistics Data were entered into a computer database and SPSS software (SPSS Inc., Chicago, IL, USA, version 16.0) and double checked before analysis. Chi-squire test was used and P < 0.05 was considered significant. Univariate and multivariate analyses were performed. Defensive medicine was the dependent variable and other variables were independent factors. Confidence intervals of 95 % were calculated and P < 0.05 was considered significant. In case of discrepancy between the results of univariate and multivariate analyses, the later was taken as final. Ethics The study received ethical clearance from the Health Research Board at Ministry of Health, Kassala State, and Sudan Medical Specialization Board (SMSB), Obstetrics and Gynaecology Department, Sudan.
While we did not adopt a formal hypothesis for this study, our working hypothesis/assumption was that defensive medicine affects daily doctor’s clinical judgement and practice.
This study was directed to assess the extent and the possible effect of defensive medicine phenomenon (in term of knowledge and prevalence) on medical decision making (development of tools that can guide physicians to make good decisions in practice) among different grades of obstetric and gynaecologic doctors, and to determine any experience of medical litigations with respect to sources and factors associated with it (in term of area of work, characteristics of the area at which the doctors worked, professionalism, hospitals systems…ect). Using a self administered questionnaire (Additional file ) and after obtaining informed written consent the data was collected from the different certified OBGYN professionals (Registrars, Specialists and Consultants) working in obstetrics and gynaecology and who attended the 27th congress of obstetrical and gynaecological society of the Sudan held from (20th -23rd February 2015) in Khartoum. The survey included only Sudanese doctors who were practicing obstetrics and gynaecology in Sudan. Visiting Doctors who are practicing in a different context abroad were excluded from the study. We used a questionnaire which was constructed by the authors to consider different forms of defensive medicine and medical litigation in obstetrics and gynaecology. Information sought by the questionnaire included: socio-demographic characteristics (age, grade, gender, area of work, duration of work, health insurance coverage), information on the area of work (blame culture: which defined as no one accepts medical errors as being all right), information on the hospital where the respondent worked (hospital guidelines and protocol, auditing system, committees…ect) daily experience (informed consent, high risk consent, documentation), whether the respondents knew the concept of defensive medicine or not? and questions on different examples of positive and negative defensive medicine (prescription of unnecessary medication, experience unnecessary refer, refuse to manage high risk patient, request for unnecessary investigation, experience unnecessary surgical procedure and avoidance high risk surgical procedure because of fear of criticism or litigation). Other information obtained from the respondents included: is the litigation in OBGYN increasing?, whether the respondents experienced litigations during their daily practice and the source of the litigations (fetal distress, misdiagnosis, injury to the viscera, shoulder dystocia, death…ect). The definition of defensive medicine was framed by the investigators according the definition in the literature ; defined as a doctor’s deviation from their usual behaviour or that considered good practice, to reduce or prevent complaints or criticism by patients or their families. This definition was not set out in the questionnaire for the respondents however in the questionnaire we asked the respondents whether they know the defensive medicine or not. Defensive medical practices were further subcategorized into positive and negative practice. When extra tests and procedures are performed primarily to reduce malpractice liability, this is a positive defensive medicine. Negative defensive medicine consists of avoidance of certain patients and procedures, thereby withdrawing medical services, and can deny patients productive care . In the questionnaire we explained the situation of high risk consent which is taken in case of serious / complicated / risky / new - surgeries or procedures; for removing any organ; in high risk patients; for proceeding with a surgery / procedure in spite of any abnormal parameters of the patient. This list is indicative not exhaustive and in case of a dilemma it is always advisable to take this high-risk consent and not a general consent.)
Data were entered into a computer database and SPSS software (SPSS Inc., Chicago, IL, USA, version 16.0) and double checked before analysis. Chi-squire test was used and P < 0.05 was considered significant. Univariate and multivariate analyses were performed. Defensive medicine was the dependent variable and other variables were independent factors. Confidence intervals of 95 % were calculated and P < 0.05 was considered significant. In case of discrepancy between the results of univariate and multivariate analyses, the later was taken as final.
The study received ethical clearance from the Health Research Board at Ministry of Health, Kassala State, and Sudan Medical Specialization Board (SMSB), Obstetrics and Gynaecology Department, Sudan.
Characteristics of the respondents and area of work A total of 117 doctors were approached, their age ranged from 26 to 73 years. Their distribution according to job description was as follow: consultants (42.7 %, 50\117) registrars (34.2 %, 40\117) and specialists (23.1 %, 27\117), Table . Of them 106 (90 %) worked in teaching hospital, 11 (9.4 %) in rural hospitals, again 39 (33.3 %) of the respondents claimed that they worked in blame free culture while 78 (66.7 %) believed the opposite. More than half of the participants were female 60\117 (51.3 %) and the majority (76\117, 65 %) were not covered by health insurance. With regard to duration of experience in obstetrics and gynaecology 14.5 % had an experience of less than 5 years, (41.9 %) were of 5–10 years and 43.6 % were more than 10 years of experience. Again with respect to the area of work the vast majority of the investigated doctors mentioned that their hospitals having guidelines and protocol (58.1 %), auditing system (72.6 %) however only 45.3 % and 39.3 % reported having high risk and ethical committees respectively. Respondents reports on daily experience and practices with regards to documentations/and communications was quite variable: they always (61.5 %), usually (29.1 %) and sometimes (9.4 %) applied informed consent, always (55.6 %) usually (25.6 %) and sometimes (18.8) applied high risk consent and they always (39.7 %), usually (43.6 %) and sometimes (17.1 %) documented their findings and intervention. Medical litigation The majority 89.7 % ( n = 105) had the impression that litigation against doctors are increasing and 27.6 % ( n = 32) had a direct experience of litigation. The different sources of the litigations reported by the doctors included: maternal death ( n = 15), perinatal death ( n = 5), other {misdiagnosis, intra-uterine fetal death, uterine perforation, rupture uterus} ( n = 4), fetal distress ( n = 3), injury to viscera ( n = 3) and shoulder dystocia ( n = 2). Defensive medicine Less than one half (50\117, 42.7 %) of the surveyed doctors knew the concept of defensive medicine and 71.8 % ( n = 84) reported practicing one or another form of defensive medicine. With further classification of defensive medicine; 48 (41 %) reported practicing positive defensive medicine while 36 (30.8 %) reported practicing negative one. Arranging un-necessary refer was the most common form of defensive medicine practiced by the investigated doctors ( n = 27, 23.1 %) followed by avoiding high risk procedure ( n = 24, 20.5 %) and ordering unnecessary investigations ( n = 14, 12 %). Among our respondents 7 (6 %) prescribed un-necessary medication to avoid litigation and criticism, 6 (5.1 %) refused to manage high risk patient because of fear from litigation and 6 (5.1 %) performed un-necessary surgery (caesarean section) to avoid litigation and criticism. Factors associated with medical litigation and defensive medicine In this study the experience of medical litigation was significantly observed among those who worked in area of blame culture (90.6 % Vs 56.5 %, P < 0.001), Table while in logistic regression model the different variables (duration of work, qualification, place of work and area of blame culture) were not associated with the concept of defence medicine, Table .
A total of 117 doctors were approached, their age ranged from 26 to 73 years. Their distribution according to job description was as follow: consultants (42.7 %, 50\117) registrars (34.2 %, 40\117) and specialists (23.1 %, 27\117), Table . Of them 106 (90 %) worked in teaching hospital, 11 (9.4 %) in rural hospitals, again 39 (33.3 %) of the respondents claimed that they worked in blame free culture while 78 (66.7 %) believed the opposite. More than half of the participants were female 60\117 (51.3 %) and the majority (76\117, 65 %) were not covered by health insurance. With regard to duration of experience in obstetrics and gynaecology 14.5 % had an experience of less than 5 years, (41.9 %) were of 5–10 years and 43.6 % were more than 10 years of experience. Again with respect to the area of work the vast majority of the investigated doctors mentioned that their hospitals having guidelines and protocol (58.1 %), auditing system (72.6 %) however only 45.3 % and 39.3 % reported having high risk and ethical committees respectively. Respondents reports on daily experience and practices with regards to documentations/and communications was quite variable: they always (61.5 %), usually (29.1 %) and sometimes (9.4 %) applied informed consent, always (55.6 %) usually (25.6 %) and sometimes (18.8) applied high risk consent and they always (39.7 %), usually (43.6 %) and sometimes (17.1 %) documented their findings and intervention.
The majority 89.7 % ( n = 105) had the impression that litigation against doctors are increasing and 27.6 % ( n = 32) had a direct experience of litigation. The different sources of the litigations reported by the doctors included: maternal death ( n = 15), perinatal death ( n = 5), other {misdiagnosis, intra-uterine fetal death, uterine perforation, rupture uterus} ( n = 4), fetal distress ( n = 3), injury to viscera ( n = 3) and shoulder dystocia ( n = 2).
Less than one half (50\117, 42.7 %) of the surveyed doctors knew the concept of defensive medicine and 71.8 % ( n = 84) reported practicing one or another form of defensive medicine. With further classification of defensive medicine; 48 (41 %) reported practicing positive defensive medicine while 36 (30.8 %) reported practicing negative one. Arranging un-necessary refer was the most common form of defensive medicine practiced by the investigated doctors ( n = 27, 23.1 %) followed by avoiding high risk procedure ( n = 24, 20.5 %) and ordering unnecessary investigations ( n = 14, 12 %). Among our respondents 7 (6 %) prescribed un-necessary medication to avoid litigation and criticism, 6 (5.1 %) refused to manage high risk patient because of fear from litigation and 6 (5.1 %) performed un-necessary surgery (caesarean section) to avoid litigation and criticism.
In this study the experience of medical litigation was significantly observed among those who worked in area of blame culture (90.6 % Vs 56.5 %, P < 0.001), Table while in logistic regression model the different variables (duration of work, qualification, place of work and area of blame culture) were not associated with the concept of defence medicine, Table .
To our knowledge this is the first published data on concept of defensive medicine and medical litigation among Sudanese doctors. The majority 89.7 % had the impression that litigation against doctors are increasing and 27.6 % had a direct experience of litigation. In this study less than one half (42.7 %) of the surveyed doctors knew the concept of defensive medicine and 71.8 % reported practicing one or another form of defensive medicine. Worldwide there is a growing awareness of the need for more effective communication among caregivers, patients, and their families . With the increasing rates of negligence, patients are beginning to seek redress and are being enlightened by legal practitioners. Health care practitioners are thus confronted with the problem and risk of being sued. This is believed to have influence various aspects of gynecological and obstetrical practice . Thus the National Heath Care system in Sudan should insures medical staff employees, providing compensation to victims of alleged malpractice, including reasonable court fees. This will allow the health practitioners to exercise their minds individually and jointly to effectively give better service to patients. Less than one half of the investigated doctors in this study practiced defensive medicine. Defensive medicine was reported in 96 % among the USA neurosurgeons and in Europe 94 % of the gastroenterologists reported practicing defensive medicine . The practice of defensive medicine has also spread to Italy where 83 % of the surgeons and anesthetists reported practicing defensive medicine . In Japan 98 % of the gastroenterologists also reported practicing at least one or another form of defensive medicine . This discrepancy might be attributed to the culture of area and people motivation and awareness. In 1991 Ennis et al . investigated the members and fellows of the Royal College of Obstetricians and gynecologists and found that the majority of the surveyed doctors were using some of tests which were known to them as unnecessary . The most frequent explanations given for this practice were that such tests were an aid to clinical judgment and were necessary for medicolegal reasons. However we don’t believe this is an excellent explanation to practice defensive medicine. In the literature many studies suggested that there is significant association between medical litigation and specialty, however and inconsistent with Ortashi et al . our study did not show any significant different in the practice of defensive medicine among different specialties . Also in line with Ortashi et al . our study revealed no significant correlation between litigation and different investigated variables. Defensive medicine brings with it exponential increases in the costs associated with clinical practice. This is explained by poor communication and other causes of medical litigation such as poor decision making. Doctors’ decisions for their patients are strongly affected by concerns of possible legal consequences. Doctors therefore practice defensive medical decision making aiming to protect themselves from blame and litigation and some fears can be healthy and can lead to adaptive responses. Good process should help create trust, rapport and alliance by showing respect for the patient. Lack of ethical issues as well as hospitals guidelines will lead to increase in the medical litigation and thus defensive medicine; this is obviously observed among our respondents since only 45.3 % and 39.3 % reported having high risk and ethical committees respectively.
The majority of Sudanese doctors who are working in Obstetrics and Gynaecology had the impression that litigation against doctors are increasing, almost one third had a direct experience of litigation and more than two thirds reported practicing one or another form of defensive medicine. There should be strategic plan to reduce the practice of defensive medicine and medical litigation against doctors.
|
Research Progress and Hot Topics in Telerehabilitation for Hip or Knee Arthroplasty | 135344c5-c731-4b0f-b3c0-ed88d20a45f2 | 11872367 | Surgical Procedures, Operative[mh] | Introduction Approximately 250 million people worldwide suffer from osteoarthritis of the hip and knee, and this number is estimated to increase over time . Several studies have shown that total hip arthroplasty (THA) and total knee arthroplasty (TKA) are effective procedures for treating osteoarthritis of the hip and knee, as they can remedy deformities, improve function, maintain mobility, and reduce pain in the joint . Notably, after lower limb joint replacement, physical rehabilitation also plays a crucial role in improving functional outcomes, optimizing physical activity, enhancing clinical and social benefits, and promoting patients' return to normal activities . However, the incidence of total joint arthroplasty (TJA) is increasing globally because of the aging population and the prevalence of osteoarthritis . The growing demand for postoperative rehabilitation services has resulted in increased costs and longer waiting lists, presenting a great challenge for the sustainability of face‐to‐face rehabilitation services . Nonetheless, the spread of coronavirus disease 2019 (COVID‐19) since March 2020 has had a significant global impact . The strict social distancing measures between individuals and stringent infection control protocols within hospitals and across communities have impeded the ability of health care practitioners to provide face‐to‐face care . Therefore, using telerehabilitation services may be one approach to address the current challenges in managing THA/TKA patients. Telerehabilitation is a form of rehabilitation in which physical rehabilitation services, such as assessment, monitoring, intervention, supervision, education, consultation, and counseling, can be delivered remotely via information and communication technology . This provides a novel strategy as an alternative or complementary therapy for overcoming the barriers associated with face‐to‐face interventions. Telerehabilitation can assist patients in returning home and to the community after an acute illness to reduce hospitalization times and healthcare costs . Furthermore, when compared with conventional in‐person rehabilitation, telerehabilitation reduces travel time, follows social distancing measures, and is more accessible and feasible . Telerehabilitation technologies allow patients experiencing more pain and limited mobility to receive health care services at home or in other remote environments , which helps reduce patients' stress and increases their satisfaction by allowing them to exercise in a familiar environment after joint replacement . Consequently, telerehabilitation has attracted a great deal of attention in different clinical conditions, such as stroke, musculoskeletal conditions, cardiopulmonary rehabilitation, and breast cancer . These studies have found that telerehabilitation appears to be effective and comparable to conventional face‐to‐face approaches in improving physical function and reducing pain in a variety of musculoskeletal disorders . For stroke patients, telerehabilitation interventions have equal or better efficacy for motor, higher cortical, and mood disorders compared to traditional face‐to‐face treatment . Many recent systematic reviews have also examined the effectiveness of telerehabilitation among patients after TJA . As reported in these studies, compared with face‐to‐face rehabilitation, telerehabilitation significantly reduces pain and improves range of motion (ROM) and physical function . According to these reviews, telerehabilitation appears to be promising for both preventive and therapeutic activities. There are numerous publications related to telerehabilitation for TJA, which makes it imperative to provide a comprehensive overview of the field by analyzing these research studies. Bibliometric analysis is regarded as an established and rigorous statistical approach for investigating and evaluating voluminous scientific literature . This method examines the knowledge structure and developing trends in a certain field to generate quantifiable, repeatable, and reliable data . By analyzing citations, co‐citations, geographic distribution, and word frequency, bibliometric approaches enable researchers to investigate specific research fields and provide valuable findings . Furthermore, bibliometric analyses promote interdisciplinary cooperation by assisting academics, physicians, and health care policy makers in collecting information to gain insight into a specific field of research and its applications . There is some bibliometric research related to telerehabilitation, and these studies have performed systematic bibliometrics. However, their fundamental ideas are restricted to the global use of telerehabilitation, with a focus on post‐stroke patients . Telerehabilitation after joint replacement has not received much attention. To the best of our knowledge, no specific studies have attempted to characterize research hotspots, global research collaborations, or trends related to telerehabilitation after joint replacement. Thus, the aim of this bibliometric analysis was to provide an overview of the current status of research and map the research landscape on telerehabilitation for TJA to understand current trends, identify research gaps, and guide future research directions.
Methods 2.1 Data Source and Search Strategy In this study, data were collected from the Science Citation Index‐Expanded (SCI‐E) of the Web of Science Core Collection (WoSCC) and PubMed. Using the retrieval strategy shown in the (Tables ), we searched for publications related to telerehabilitation for knee or hip arthroplasty. From 2003 to June 7, 2024, a range of publication dates were chosen. 2.2 Literature Screening and Data Extraction There were two stages of screening: preliminary screening based on title and abstract, and full‐text screening based on the inclusion and exclusion criteria (Table ). We included published records, including articles and reviews related to knee or hip arthroplasty and telerehabilitation. We excluded duplicates, studies not involving knee or hip arthroplasty, studies not involving telerehabilitation intervention programs, and manuscripts published in any language other than English. Other document types, such as letters, news, editorial materials, proceedings papers, short reports, and conference abstracts, were also excluded. Following standard selection procedures, two reviewers independently screened the article titles and abstracts. Full texts of the articles were retrieved if necessary. Disagreements were discussed and resolved between the two reviewers. Figure provides an overview of the comprehensive search strategy and inclusion criteria implemented in this study. 2.3 Data Analysis Tool and Analysis Methodologies VOSviewer was developed by Van Eck and Waltman and provides a powerful bibliometric mapping function for viewing networks of keywords, literature, authors, and so on. We used VOSviewer 1.6.18 (Leiden University, Van Eck N. J.) to draw diagrams for keyword co‐occurrence and the full‐counting bibliographic coupling analysis of journals. A number of parameters were chosen: counting method (full counting), type of analysis (co‐occurrence), unit of analysis (all keywords), minimum number of occurrences (one), and number of keywords (1000). Meanwhile, a density map was created to show the frequencies of keywords appearing in the publications. The yellow area represents keywords or clusters with a high frequency count, whereas the green area indicates keywords with a low frequency count. CiteSpace, a Java application developed by Chen , was employed to develop co‐occurrence and co‐citation networks and analyze their characteristics to identify research hotspots, frontiers, and foundations in the primary areas of knee or hip arthroplasty and telerehabilitation. The following parameters were established: the time period ranged from 2003 to 2024 with 1 year per slice, and the selection criterion for top N was set to 25. Several co‐occurrence node types, including keywords and references, were chosen for the co‐citation knowledge graph. The size of a node represents the number or frequency of documents. The relationship between nodes is determined by their connections, such as co‐existence, co‐occurrence, or co‐citation . Centrality is represented by a purple outer circle; the wider the circle, the greater the centrality. The red node indicates “citation emergence,” which means that the number of citations in this node has suddenly increased over time. In addition, to generate reference co‐citation clusters, the log‐likelihood ratio (LLR) strategy was adopted, and module values ( Q values) and weighted mean silhouette values ( S values) were chosen as clustering criteria.
Data Source and Search Strategy In this study, data were collected from the Science Citation Index‐Expanded (SCI‐E) of the Web of Science Core Collection (WoSCC) and PubMed. Using the retrieval strategy shown in the (Tables ), we searched for publications related to telerehabilitation for knee or hip arthroplasty. From 2003 to June 7, 2024, a range of publication dates were chosen.
Literature Screening and Data Extraction There were two stages of screening: preliminary screening based on title and abstract, and full‐text screening based on the inclusion and exclusion criteria (Table ). We included published records, including articles and reviews related to knee or hip arthroplasty and telerehabilitation. We excluded duplicates, studies not involving knee or hip arthroplasty, studies not involving telerehabilitation intervention programs, and manuscripts published in any language other than English. Other document types, such as letters, news, editorial materials, proceedings papers, short reports, and conference abstracts, were also excluded. Following standard selection procedures, two reviewers independently screened the article titles and abstracts. Full texts of the articles were retrieved if necessary. Disagreements were discussed and resolved between the two reviewers. Figure provides an overview of the comprehensive search strategy and inclusion criteria implemented in this study.
Data Analysis Tool and Analysis Methodologies VOSviewer was developed by Van Eck and Waltman and provides a powerful bibliometric mapping function for viewing networks of keywords, literature, authors, and so on. We used VOSviewer 1.6.18 (Leiden University, Van Eck N. J.) to draw diagrams for keyword co‐occurrence and the full‐counting bibliographic coupling analysis of journals. A number of parameters were chosen: counting method (full counting), type of analysis (co‐occurrence), unit of analysis (all keywords), minimum number of occurrences (one), and number of keywords (1000). Meanwhile, a density map was created to show the frequencies of keywords appearing in the publications. The yellow area represents keywords or clusters with a high frequency count, whereas the green area indicates keywords with a low frequency count. CiteSpace, a Java application developed by Chen , was employed to develop co‐occurrence and co‐citation networks and analyze their characteristics to identify research hotspots, frontiers, and foundations in the primary areas of knee or hip arthroplasty and telerehabilitation. The following parameters were established: the time period ranged from 2003 to 2024 with 1 year per slice, and the selection criterion for top N was set to 25. Several co‐occurrence node types, including keywords and references, were chosen for the co‐citation knowledge graph. The size of a node represents the number or frequency of documents. The relationship between nodes is determined by their connections, such as co‐existence, co‐occurrence, or co‐citation . Centrality is represented by a purple outer circle; the wider the circle, the greater the centrality. The red node indicates “citation emergence,” which means that the number of citations in this node has suddenly increased over time. In addition, to generate reference co‐citation clusters, the log‐likelihood ratio (LLR) strategy was adopted, and module values ( Q values) and weighted mean silhouette values ( S values) were chosen as clustering criteria.
Results The search yielded 229 publications published between 2003 and 2024. An overview of the number of papers published each year and the development trend in the field of knee or hip arthroplasty for telerehabilitation is provided in Figure . The overall trend of publications can be divided into three stages: the initial stage (2003–2014), the second stage (2015–2018), and the third stage (2019–2024). During the first stage (2003–2014), the number of publications was limited to no more than five per year. After that, the number of publications gradually increased and surpassed 14 for the first time in 2017. Publications increased considerably between 2019 and 2023. Considering that only 5 months were included in 2024, the numbers decreased sharply. 3.1 Analysis of Countries and Institutions As shown in Table , the top 15 countries according to the number of publications are as follows: The United States ranked first with 80 published articles, which accounted for 34.93% of all 229 publications. Australia was second (27/229, 11.79%), followed by the United Kingdom (24/229, 10.48%), China (17/229, 7.42%), Canada (104/229, 5.67%), and Germany (12/229, 5.24%). Furthermore, international cooperation among different nations is illustrated in Figure . According to the data, the United States had the highest centrality (0.41), followed by Denmark (0.22), Italy (0.16), and the United Kingdom (0.15). Based on the definition of centrality, it is estimated that these countries represented tremendous academic influence by cooperating closely with other countries. A combination of publication and centrality analysis revealed the United States, the United Kingdom, Denmark, and Italy to be the most dominant countries. The top 15 research institutions by number of publications are displayed in Table . The institution that published the most papers, with 8 publications, was the University of Sherbrooke, accounting for 3.49% ( n = 229) of all publications. Sinai Hospital of Baltimore, Laval University, Jefferson University, and Ingham Institute for Applied Medical Research were second, with 6 (2.62%) publications. The University of New South Wales Sydney came in third, with 2.18% ( n = 229) of all publications. Collaborations among various institutions are shown in Figure . Cleveland Clinic Foundation and Research Libraries of the United Kingdom demonstrated the highest centrality (0.02), followed by Sinai Hospital of Baltimore and Harvard University (0.01). According to an analysis of publication numbers and centrality, these institutions have strong academic influence. 3.2 Analysis of Authors and Co‐Cited Authors The top 15 authors on telerehabilitation of knee or hip arthroplasty research are shown in Table . Among the top 15 authors, Boissy P. ranked first concerning publication output (7 publications), followed by Cabana F. and Corriveau H. with 6 publications. As far as co‐cited authors are concerned, Tousignant M. had the most cited times (65 cited times), followed by Moffet H. (61 cited times) and Russell T. G. (59 cited times). These authors were actively involved in research related to telerehabilitation of knee or hip arthroplasty. Furthermore, we also analyzed the cooperation between authors and co‐cited authors to identify potential collaborations. Cooperation relationships are represented by connections, and the thickness of the connections indicates the degree of cooperation between nodes. According to Figure , the merged network of authors consists of 385 nodes and 631 links. The author's cooperation network is further divided into four smaller clusters, each led by a prominent author. The merged network of co‐cited authors in Figure is composed of 606 nodes and 2517 links. Using knowledge maps of authors and co‐cited authors, researchers can find influential research groups and potential collaboration partners. 3.3 Analysis of Journals A total of 90 scholarly journals have published articles in the field of knee or hip arthroplasty telerehabilitation research. The top 15 journals accounted for nearly a third of all publications, as shown in Figure and Table . The Journal of Arthroplasty (IF 2019 = 3.709) published the highest number of articles (29 publications, 12.66%), followed by Journal of Telemedicine and Telecare and Clinical Orthopaedics and Related Research . Compared to other journals, the top three journals contributed for at least 51 publications, representing 22.27% of the total, indicating their special position in the field of knee or hip arthroplasty telerehabilitation. According to the full counting co‐citation analysis of cited journals, the most frequently cited journal was the Journal of Arthroplasty (cited 477 times). The second was the Journal of Bone and Joint Surgery American Volume (cited 434 times). The third was the Journal of Clinical Orthopaedics and Related Research (cited 335 times). The bibliographic coupling analysis of highly cited journals is displayed in Figure . 3.4 Analysis of Reference Co‐Citation The strongest citation bursts in references represent the research basics for future frontiers. As shown in Figure , the red line denotes the time period of emergence, the light blue line indicates that the node has not yet begun to appear, and the dark blue line represents that the node has started to appear. The top 25 articles related to knee or hip arthroplasty telerehabilitation exhibited the strongest citation bursts. The 15 articles with the highest citation frequency and burst strength are listed in Table . Among these references, the article written by Moffet H. has the strongest burst (9.59), which supports the effectiveness of in‐home telerehabilitation for hospital discharge of patients following TKA, demonstrating its comparability to face‐to‐face service delivery. The article written by Tousignant M was second (5.76), which led to a cost analysis of in‐home telerehabilitation for post‐knee arthroplasty in Australia. In addition, three articles written by Bettger et al., Wang et al., and Dias et al. were the most popular from 2022 to 2024. Bettger et al. analyzed and compared the effects of virtual exercise rehabilitation in‐home therapy and traditional in‐person rehabilitation for TKA. Wang et al. investigated the effectiveness of internet‐based telerehabilitation among patients after TJA. Dias et al. assessed the feasibility of a novel artificial intelligence‐powered digital biofeedback system following THA and compared the clinical outcomes against supervised conventional rehabilitation. 3.5 Analysis of Keywords We can discover trends and changes in research topics by analyzing co‐occurrence networks of keywords, which is crucial to understanding the development of these topics. An analysis of the co‐occurrence network of keywords was conducted using VOSviewer software. Based on the bibliographic data, we created the map using a full counting strategy, limiting the minimum number of occurrences of a keyword to one. Figure shows an overlay visualization of 689 keyword co‐occurrences after the data were cleaned and purified using a thesaurus. As shown in Figure , color represents the average occurrence time, the node size indicates the occurrence frequency, and the link thickness denotes the co‐occurrence strength. Yellow denotes a more recent occurrence, whereas dark blue indicates an earlier appearance. According to the average time of occurrence, the earliest keywords were “health care,” “physical activity intervention,” “randomized controlled trials,” “knee range of motion,” “internet,” “range,” “Osteoarthritis Index,” “validation,” and “functional,” followed by “telemedicine,” “arthroplasty,” “telehealth,” “satisfiction,” “smartphone,” “technology,” “ehealth,” “virtual reality,” and “old audults,” and then gradually evolved into “telerehabilitation,” “total hip arthroplasty,” and “total knee arthroplasty.” These keywords were then expanded to include subfields such as “COVID‐19,” “mobile application,” “self‐efficacy,” “internet‐based interventions,” “Cost benefit analysis,” “anxiety,” “home exercise,” “mHealth intervention,” “smartphone appication,” “app‐based rehabilitation,” “exercise program,” and “smartphone application.” As shown in Figure , the largest nodes emerged from 2018 to 2022, including “telemedicine,” “telerehabilitation,” “total hip arthroplasty,” “total knee arthroplasty,” “osteoarthritis,” and “physiotherapy.” Subsequently, the frequency of many keywords, including “mobile application, self‐efficacy,” “home exercise,” and “smartphone application,” continued to increase, and these keywords gained rapid traction between 2022 and 2024. Additionally, Table shows the top 25 keywords according to occurrence. Additionally, we analyzed keyword co‐occurrence via density visualizations, as shown in Figure . Major keyword cluster maps were generated in Figure and Table using the LLR strategy, with the selection type adopted being the “keyword option.” The Q value and S value are considered evaluation strategies for clustering, and a Q value > 0.4 and S value > 0.7 indicate that the cluster is credible. Research on the telerehabilitation of knee or hip arthroplasty was grouped into different clusters. The modularity Q is 0.4611 (> 0.4), suggesting that the network clustering is reasonable, and the S value is 0.7528 (> 0.5), indicating that the homogeneity of the clustering is acceptable. The two largest clusters were composed of 49 and 42 members, respectively. References #6 and #9 were the three oldest clusters, which presented the knowledge base of knee or hip arthroplasty for telerehabilitation. The other areas were relatively new, especially #3 and #7, which were the latest, indicating future research trends and directions. Furthermore, the top 25 keywords with the strongest occurrence bursts are displayed in Figure . In terms of burst intensity, the five keywords with the greatest citation bursts in knee or hip arthroplasty telerehabilitation were “mobile application” (2.89), “randomized controlled trial” (2.64), “pain” (2.55), “telemedicine” (2.21), and “patient satisfaction” (2.04). Importantly, the values in brackets reflect the strength of the burst. An occurrence burst, which indicates a sharp rise in the frequency of a keyword over time, can reveal new research trends and developments in frontier topics. Additionally, “patient satisfaction,” “older adults,” “physical function,” “pain,” and “telehealth” were the most popular keywords from 2020 to 2024.
Analysis of Countries and Institutions As shown in Table , the top 15 countries according to the number of publications are as follows: The United States ranked first with 80 published articles, which accounted for 34.93% of all 229 publications. Australia was second (27/229, 11.79%), followed by the United Kingdom (24/229, 10.48%), China (17/229, 7.42%), Canada (104/229, 5.67%), and Germany (12/229, 5.24%). Furthermore, international cooperation among different nations is illustrated in Figure . According to the data, the United States had the highest centrality (0.41), followed by Denmark (0.22), Italy (0.16), and the United Kingdom (0.15). Based on the definition of centrality, it is estimated that these countries represented tremendous academic influence by cooperating closely with other countries. A combination of publication and centrality analysis revealed the United States, the United Kingdom, Denmark, and Italy to be the most dominant countries. The top 15 research institutions by number of publications are displayed in Table . The institution that published the most papers, with 8 publications, was the University of Sherbrooke, accounting for 3.49% ( n = 229) of all publications. Sinai Hospital of Baltimore, Laval University, Jefferson University, and Ingham Institute for Applied Medical Research were second, with 6 (2.62%) publications. The University of New South Wales Sydney came in third, with 2.18% ( n = 229) of all publications. Collaborations among various institutions are shown in Figure . Cleveland Clinic Foundation and Research Libraries of the United Kingdom demonstrated the highest centrality (0.02), followed by Sinai Hospital of Baltimore and Harvard University (0.01). According to an analysis of publication numbers and centrality, these institutions have strong academic influence.
Analysis of Authors and Co‐Cited Authors The top 15 authors on telerehabilitation of knee or hip arthroplasty research are shown in Table . Among the top 15 authors, Boissy P. ranked first concerning publication output (7 publications), followed by Cabana F. and Corriveau H. with 6 publications. As far as co‐cited authors are concerned, Tousignant M. had the most cited times (65 cited times), followed by Moffet H. (61 cited times) and Russell T. G. (59 cited times). These authors were actively involved in research related to telerehabilitation of knee or hip arthroplasty. Furthermore, we also analyzed the cooperation between authors and co‐cited authors to identify potential collaborations. Cooperation relationships are represented by connections, and the thickness of the connections indicates the degree of cooperation between nodes. According to Figure , the merged network of authors consists of 385 nodes and 631 links. The author's cooperation network is further divided into four smaller clusters, each led by a prominent author. The merged network of co‐cited authors in Figure is composed of 606 nodes and 2517 links. Using knowledge maps of authors and co‐cited authors, researchers can find influential research groups and potential collaboration partners.
Analysis of Journals A total of 90 scholarly journals have published articles in the field of knee or hip arthroplasty telerehabilitation research. The top 15 journals accounted for nearly a third of all publications, as shown in Figure and Table . The Journal of Arthroplasty (IF 2019 = 3.709) published the highest number of articles (29 publications, 12.66%), followed by Journal of Telemedicine and Telecare and Clinical Orthopaedics and Related Research . Compared to other journals, the top three journals contributed for at least 51 publications, representing 22.27% of the total, indicating their special position in the field of knee or hip arthroplasty telerehabilitation. According to the full counting co‐citation analysis of cited journals, the most frequently cited journal was the Journal of Arthroplasty (cited 477 times). The second was the Journal of Bone and Joint Surgery American Volume (cited 434 times). The third was the Journal of Clinical Orthopaedics and Related Research (cited 335 times). The bibliographic coupling analysis of highly cited journals is displayed in Figure .
Analysis of Reference Co‐Citation The strongest citation bursts in references represent the research basics for future frontiers. As shown in Figure , the red line denotes the time period of emergence, the light blue line indicates that the node has not yet begun to appear, and the dark blue line represents that the node has started to appear. The top 25 articles related to knee or hip arthroplasty telerehabilitation exhibited the strongest citation bursts. The 15 articles with the highest citation frequency and burst strength are listed in Table . Among these references, the article written by Moffet H. has the strongest burst (9.59), which supports the effectiveness of in‐home telerehabilitation for hospital discharge of patients following TKA, demonstrating its comparability to face‐to‐face service delivery. The article written by Tousignant M was second (5.76), which led to a cost analysis of in‐home telerehabilitation for post‐knee arthroplasty in Australia. In addition, three articles written by Bettger et al., Wang et al., and Dias et al. were the most popular from 2022 to 2024. Bettger et al. analyzed and compared the effects of virtual exercise rehabilitation in‐home therapy and traditional in‐person rehabilitation for TKA. Wang et al. investigated the effectiveness of internet‐based telerehabilitation among patients after TJA. Dias et al. assessed the feasibility of a novel artificial intelligence‐powered digital biofeedback system following THA and compared the clinical outcomes against supervised conventional rehabilitation.
Analysis of Keywords We can discover trends and changes in research topics by analyzing co‐occurrence networks of keywords, which is crucial to understanding the development of these topics. An analysis of the co‐occurrence network of keywords was conducted using VOSviewer software. Based on the bibliographic data, we created the map using a full counting strategy, limiting the minimum number of occurrences of a keyword to one. Figure shows an overlay visualization of 689 keyword co‐occurrences after the data were cleaned and purified using a thesaurus. As shown in Figure , color represents the average occurrence time, the node size indicates the occurrence frequency, and the link thickness denotes the co‐occurrence strength. Yellow denotes a more recent occurrence, whereas dark blue indicates an earlier appearance. According to the average time of occurrence, the earliest keywords were “health care,” “physical activity intervention,” “randomized controlled trials,” “knee range of motion,” “internet,” “range,” “Osteoarthritis Index,” “validation,” and “functional,” followed by “telemedicine,” “arthroplasty,” “telehealth,” “satisfiction,” “smartphone,” “technology,” “ehealth,” “virtual reality,” and “old audults,” and then gradually evolved into “telerehabilitation,” “total hip arthroplasty,” and “total knee arthroplasty.” These keywords were then expanded to include subfields such as “COVID‐19,” “mobile application,” “self‐efficacy,” “internet‐based interventions,” “Cost benefit analysis,” “anxiety,” “home exercise,” “mHealth intervention,” “smartphone appication,” “app‐based rehabilitation,” “exercise program,” and “smartphone application.” As shown in Figure , the largest nodes emerged from 2018 to 2022, including “telemedicine,” “telerehabilitation,” “total hip arthroplasty,” “total knee arthroplasty,” “osteoarthritis,” and “physiotherapy.” Subsequently, the frequency of many keywords, including “mobile application, self‐efficacy,” “home exercise,” and “smartphone application,” continued to increase, and these keywords gained rapid traction between 2022 and 2024. Additionally, Table shows the top 25 keywords according to occurrence. Additionally, we analyzed keyword co‐occurrence via density visualizations, as shown in Figure . Major keyword cluster maps were generated in Figure and Table using the LLR strategy, with the selection type adopted being the “keyword option.” The Q value and S value are considered evaluation strategies for clustering, and a Q value > 0.4 and S value > 0.7 indicate that the cluster is credible. Research on the telerehabilitation of knee or hip arthroplasty was grouped into different clusters. The modularity Q is 0.4611 (> 0.4), suggesting that the network clustering is reasonable, and the S value is 0.7528 (> 0.5), indicating that the homogeneity of the clustering is acceptable. The two largest clusters were composed of 49 and 42 members, respectively. References #6 and #9 were the three oldest clusters, which presented the knowledge base of knee or hip arthroplasty for telerehabilitation. The other areas were relatively new, especially #3 and #7, which were the latest, indicating future research trends and directions. Furthermore, the top 25 keywords with the strongest occurrence bursts are displayed in Figure . In terms of burst intensity, the five keywords with the greatest citation bursts in knee or hip arthroplasty telerehabilitation were “mobile application” (2.89), “randomized controlled trial” (2.64), “pain” (2.55), “telemedicine” (2.21), and “patient satisfaction” (2.04). Importantly, the values in brackets reflect the strength of the burst. An occurrence burst, which indicates a sharp rise in the frequency of a keyword over time, can reveal new research trends and developments in frontier topics. Additionally, “patient satisfaction,” “older adults,” “physical function,” “pain,” and “telehealth” were the most popular keywords from 2020 to 2024.
Discussion 4.1 Principal Findings In this study, a bibliometric analysis was conducted on publications related to telerehabilitation for hip or knee arthroplasty from 2003 to 2024. One study investigated the knowledge structure of the global use of telerehabilitation, and one study focused on post‐stroke patients . This research presents the first bibliometric analysis of telerehabilitation for hip or knee arthroplasty. Telerehabilitation for knee or hip arthroplasty research could be divided into three phases: the initial stage (2003–2014), the second stage (2015–2018), and the third stage (2019–2024). Publications increased considerably between 2019 and 2023. The possible reasons for the rapid growth of this research area include technological advancements in telerehabilitation, the growing demand in the aging population, and the outbreak of COVID‐19. Based on an analysis by country, the United States has dominated the field of telerehabilitation for hip or knee arthroplasty research. The United States contributed approximately one‐third of all documents, exhibiting its strong research capacity and dedication to the advancement of the field. The United States is followed by other developed nations, such as the United Kingdom and Australia, which indicates a high level of telerehabilitation technology development in those countries. In terms of the number of research articles published, the highest number was published at the University of Sherbrooke, which can be explained by the strong academic environment and scientific foundations of universities. Among several publication venues, the Journal of Arthroplasty was the most productive publication source in this field, followed by the Journal of Telemedicine and Telecare , Clinical Orthopaedics and Related Research . For individual researchers, influential authors based on contributions and co‐citations include Boissy P. and Tousignant M. Since the above‐mentioned countries, institutions, and writers have contributed significantly to the field of telerehabilitation for hip or knee arthroplasty research, it could be recommended that more attention be paid to their research. 4.2 A Major Concern at Present Is Telerehabilitation After Total Hip or Knee Arthroplasty Depending on the type of information and communication technology, therapists and patients can communicate synchronously or asynchronously. Synchronous telerehabilitation involves real‐time interaction between therapists and patients, with instant transmission and display of information, with telephone calls and videoconferencing as the most common methods . However, an asynchronous telerehabilitation model is characterized by a delay in information exchange among all participants. Most commonly, asynchronous technologies include email and messaging services . Some recent systematic reviews and meta‐analyses have shown that, compared with conventional in‐person rehabilitation, telerehabilitation results in comparable improvements in various clinical outcomes for patients who underwent joint arthroplasty, such as knee extension range, pain, muscle strength, swelling of the knee joints, WOMAC, and KOOS . Additionally, telerehabilitation has proven to be more economical than conventional face‐to‐face rehabilitation when the distance between the patient and physical therapist is > 30 km . Despite this, 13% of patients dropped out of asynchronous telerehabilitation; the authors noted that asynchronous telerehabilitation lacks a personal relationship between the therapist and patient . In addition, several factors may contribute to the effectiveness of telerehabilitation in the treatment of arthroplasty. Patients reported that telerehabilitation therapy had the greatest benefit by eliminating travel and waiting time for both patients and therapists . It is much easier for patients to continue their therapy more effectively at home because of easy access to equipment. As a result, the telerehabilitation group would have had more exercise sessions than the conventional group did . In the absence of in‐person supervision, telerehabilitation provides patients with more convenient options and stimulates their desire to return to their daily lives .
Principal Findings In this study, a bibliometric analysis was conducted on publications related to telerehabilitation for hip or knee arthroplasty from 2003 to 2024. One study investigated the knowledge structure of the global use of telerehabilitation, and one study focused on post‐stroke patients . This research presents the first bibliometric analysis of telerehabilitation for hip or knee arthroplasty. Telerehabilitation for knee or hip arthroplasty research could be divided into three phases: the initial stage (2003–2014), the second stage (2015–2018), and the third stage (2019–2024). Publications increased considerably between 2019 and 2023. The possible reasons for the rapid growth of this research area include technological advancements in telerehabilitation, the growing demand in the aging population, and the outbreak of COVID‐19. Based on an analysis by country, the United States has dominated the field of telerehabilitation for hip or knee arthroplasty research. The United States contributed approximately one‐third of all documents, exhibiting its strong research capacity and dedication to the advancement of the field. The United States is followed by other developed nations, such as the United Kingdom and Australia, which indicates a high level of telerehabilitation technology development in those countries. In terms of the number of research articles published, the highest number was published at the University of Sherbrooke, which can be explained by the strong academic environment and scientific foundations of universities. Among several publication venues, the Journal of Arthroplasty was the most productive publication source in this field, followed by the Journal of Telemedicine and Telecare , Clinical Orthopaedics and Related Research . For individual researchers, influential authors based on contributions and co‐citations include Boissy P. and Tousignant M. Since the above‐mentioned countries, institutions, and writers have contributed significantly to the field of telerehabilitation for hip or knee arthroplasty research, it could be recommended that more attention be paid to their research.
A Major Concern at Present Is Telerehabilitation After Total Hip or Knee Arthroplasty Depending on the type of information and communication technology, therapists and patients can communicate synchronously or asynchronously. Synchronous telerehabilitation involves real‐time interaction between therapists and patients, with instant transmission and display of information, with telephone calls and videoconferencing as the most common methods . However, an asynchronous telerehabilitation model is characterized by a delay in information exchange among all participants. Most commonly, asynchronous technologies include email and messaging services . Some recent systematic reviews and meta‐analyses have shown that, compared with conventional in‐person rehabilitation, telerehabilitation results in comparable improvements in various clinical outcomes for patients who underwent joint arthroplasty, such as knee extension range, pain, muscle strength, swelling of the knee joints, WOMAC, and KOOS . Additionally, telerehabilitation has proven to be more economical than conventional face‐to‐face rehabilitation when the distance between the patient and physical therapist is > 30 km . Despite this, 13% of patients dropped out of asynchronous telerehabilitation; the authors noted that asynchronous telerehabilitation lacks a personal relationship between the therapist and patient . In addition, several factors may contribute to the effectiveness of telerehabilitation in the treatment of arthroplasty. Patients reported that telerehabilitation therapy had the greatest benefit by eliminating travel and waiting time for both patients and therapists . It is much easier for patients to continue their therapy more effectively at home because of easy access to equipment. As a result, the telerehabilitation group would have had more exercise sessions than the conventional group did . In the absence of in‐person supervision, telerehabilitation provides patients with more convenient options and stimulates their desire to return to their daily lives .
Emerging Trends 5.1 Mobile Application‐Based Program for Telerehabilitation After Total Hip or Knee Arthroplasty The delivery modes of telerehabilitation interventions for arthroplasty were labeled as synchronous or asynchronous on the basis of whether health care providers communicated with patients in real time . As part of the synchronized telerehabilitation program, physical therapy assessments, supervised exercises, and prescriptions for home activities were performed. However, the system has the following limitations when it is implemented: patients and providers must sit in front of an excellent screen with a reliable internet connection at the same time; technicians are required to visit patients' homes to install or repair the equipment; and training is necessary because of the complexity of the technology . Moreover, the asynchronous telerehabilitation method provided information on diseases, pain management, cognitive behavioral therapy, and complication prevention. In particular, mobile applications enable health care providers to present more easily accessible and flexible telehealth support. Patients can review the rehabilitation program whenever and wherever it is most expedient for them . Thus, from 2003 to 2017, the synchronous mode was most prevalent, and videoconferencing or Skype was used for individual consultations with participants. However, studies using mobile apps, virtual exercise rehabilitation programs (VERAs), and interactive exercises have increasingly switched to asynchronous modes since 2017. This trend reflects the advancement of technology in telehealth: from telephone to videoconferencing and then to mobile applications . Owing to their convenience and accessibility, mobile applications are rapidly being used to deliver telerehabilitation services . These services include daily postoperative care education, exercise demonstrations, reminders, and progress monitoring . Consistent with this trend, our study also revealed that mobile applications presented a yellow color and larger nodes on the overlay visualization map, indicating that mobile applications have become a hot research topic in the last few years for telerehabilitation after total hip or knee arthroplasty. Numerous studies in recent years have demonstrated the efficacy of mobile app–based telerehabilitation programs for arthroplasty rehabilitation. Compared with conventional rehabilitation, application‐based telerehabilitation yields superior outcomes in terms of function, pain, and ROM . Recent studies have demonstrated that rehabilitation utilizing mobile applications may have beneficial impacts on patients' self‐efficacy, patient‐reported physical function, health‐related quality of life (HRQoL), and anxiety and depression levels . In this study, the telerehabilitation programs are offered through the mobile application WeChat, where all exercises are delivered via video demonstrations and uploaded to the mobile application. In addition, a randomized controlled trial showed that smartphone app–based remote rehabilitation worked better than home‐based rehabilitation with outpatient guidance in terms of short‐term results in ROM, the Five Times Sit‐to‐Stand Test (5xSST), and the Single‐Leg Stance Test (SLST) . The app named the Vital Health Remote Rehabilitation System consists of three parts: the patient‐side app, wearable sensors, and surgeon‐side websites. The app can offer real‐time guidance and assessment of patient movement according to monitoring data . However, most of the evidence supporting mobile app–based telerehabilitation in TJA has been derived from quantitative non‐randomized studies. Earlier studies focused on physical function and rarely examined psychological parameters, such as kinesiophobia, fear of falling, depression, and anxiety. Moreover, it is the elderly who are most likely to undergo joint replacements, and despite technological literacy steadily increasing among the general population, older adults rarely use mobile apps. Therefore, to confirm the effectiveness of mobile app–based telerehabilitation in TJA, robust RCTs are warranted . Moreover, researchers should consider incorporating qualitative components into future research designs and be aware of physical (sensory, motor, and cognitive) declines in older people to gain a better understanding of telerehabilitation and how it contributes to the recovery of patients . 5.2 “Patient Satisfaction,” “Self‐efficacy,” and “Cost Benefit Analysis” May Be Major Rehabilitation Outcomes for Hip or Knee Telerehabilitation Research Based on the co‐occurrence analysis of keywords via overlay visualization, the earliest keywords to appear were “motion” and “range.” This implies that the outcomes for the remote rehabilitation of arthroplasty in this period are related mainly to physical functions. In addition, in 2020, words such as “patient satisfaction,” “self‐efficacy,” and “cost benefit analysis” are displayed in yellow, implying that these aspects of research will be hotspots and research frontiers. Consistent with our findings, one study also reported that previous studies systematically emphasized pain, ROM, and physical function as outcomes of rehabilitation. The expert consensus on best practices in rehabilitation after TJA suggests that rehabilitation outcomes should include not only body structure/function (e.g., pain and ROM) and activity/participation (e.g., physical function) but also other outcomes, such as HRQoL and patient satisfaction with rehabilitation outcomes/processes . Psychological factors have been proven to affect recovery in many areas of surgery, including orthopedic, abdominal, breast, and oncological surgery . Psychological factors such as preoperative anxiety, depression, catastrophizing state, poor self‐efficacy, and poor coping skills in patients may be preoperative predictors of unsatisfactory outcomes in patients with arthroplasty . Therefore, psychological well‐being outcomes have been recommended because many patients after arthroplasty experience psychological problems such as anxiety and depression . Many patients who undergo arthroplasty suffer from poor psychological well‐being or psychological issues . These issues, such as pain and joint function, have adverse effects on patient rehabilitation outcomes . The outcomes of HRQoL indicate patients' general health and can be used to support clinical judgment for patients following TJA . Therefore, incorporating HRQoL as an outcome measure for assessing the recovery of patients after TJA is encouraged . However, the present study revealed that psychological consequences have rarely been investigated, and the duration and magnitude of the impact of psychological factors on postoperative outcomes are not clear . Only four out of 22 studies have examined psychological well‐being/problems ; therefore, psychological outcome measures should be included in future studies to assess the effects of telerehabilitation on psychological well‐being/problems among patients after TJA . Another essential healthcare outcome to consider when examining telerehabilitation is patient satisfaction . Both a healthcare recipient's cognitive assessment and the emotional response to their experience with healthcare can be characterized as satisfactory in healthcare management . The level of satisfaction has been cited as a key indicator of quality in healthcare because it is correlated with both adherence to treatment plans and improved clinical outcomes . Despite this, Kairy et al. have noted that there are no robust and standardized measures of satisfaction in articles discussing telerehabilitation for physical disabilities . Large‐scale clinical trials have not yet examined whether patients are satisfied with in‐home telerehabilitation after TKA. Few studies have examined the various factors affecting users' satisfaction with the service or technology employed. These factors include satisfaction with the care provided, perception of telehealth, satisfaction with telemedicine services, and individual appreciation of the quality of the technical system . The information on cost‐effectiveness is critical for guiding resource allocation decisions in the context of value‐based health systems . There have been encouraging results from cost‐effectiveness studies conducted in the orthopedic population. Telerehabilitation was found to be more cost‐effective in a Canadian‐based cost‐analysis study comparing in‐person care with telerehabilitation (via home visits) for knee replacement patients . A systematic review conducted by de la Torre‐Dez et al. also supported that some cost‐effectiveness studies revealed that telemedicine may reduce costs. Additionally, one study found that telerehabilitation could cut healthcare spending mainly due to lower transportation costs . These findings suggest that telerehabilitation may be an economically feasible alternative to in‐person rehabilitation after joint replacement. However, there are still several drawbacks, including a lack of randomized controlled trials, limited sample sizes, and poor data quality . As a result, the cost‐effectiveness of telemedicine applications could not be determined because of methodological flaws in the analyses . Therefore, many further studies are needed to address the problems mentioned above. Self‐efficacy in rehabilitation plays a significant role in determining the efficacy of rehabilitation. According to self‐efficacy theory, direct mastery experiences, vicarious experiences, verbal persuasion, and arousal state are four factors that contribute to individuals' confidence in their ability to successfully complete a task . By incorporating these concepts into a rehabilitation strategy, rehabilitation programs can be optimized in terms of content and structure, as well as the motivation and confidence of patients in the rehabilitation process, which will lead to greater compliance with exercise and better clinical outcomes such as physical function and HRQoL . However, a review revealed that few studies have examined patients' self‐efficacy during telerehabilitation . Therefore, in addition to pain, ROM, and physical function, other outcomes such as HRQoL, satisfaction, cost‐effectiveness, self‐efficacy, and psychological well‐being should be included to gain a holistic understanding of the effectiveness of internet‐based telerehabilitation. 5.3 Strengths and Limitations This research has multiple strengths. It is one of the pioneering studies that focuses on the specific topic of telerehabilitation for hip or knee arthroplasty. The use of the visualization software CiteSpace and VOSviewer for bibliometric analysis can provide objective and comprehensive results. A relatively scientific and intuitive picture was provided. In addition, the study is highly pertinent to the scope of the journal, and it would be beneficial to a wide variety of audiences, including freshmen, scholars, doctors, and rehabilitation therapists. Despite our study providing a comprehensive picture of telerehabilitation for hip or knee arthroplasty, several limitations should be addressed. First, as we know, co‐occurrence, collaborative network, and co‐citation analysis are the most important analyses in bibliometric studies. However, CiteSpace may have limitations when analyzing data from different databases. For example, CiteSpace is unable to analyze co‐occurrence data from Scopus or cooperative networks from Derwent. In addition, the CiteSpace software does not support the processing of data from the Google Scholar database. Therefore, we specifically analyzed the data from Web of Science and PubMed. In the future, as bibliometric software is updated, it will be necessary to incorporate additional database resources, such as Scopus and Google Scholar, to facilitate further investigations. Second, the WoS mainly indexes papers written in English. However, there may be Chinese papers or papers published in Chinese journals that are not yet included in WoS; therefore, this language restriction may have prevented us from finding some publications, which might have resulted in selection bias. Third, even though we used the search terms chosen, we cannot guarantee that we have retrieved all the documents related to the topic. Fourth, it is also possible that publication and citation counts do not directly represent the scientific quality of the papers, which may be impacted by multiple factors, including sample size, study design, and standard of reporting. In addition, bibliometric analysis cannot determine the scientific quality of the content, including bias risks, effect sizes, and statistical significance of the conclusions and whether they are justified by the respective results. Fifth, only articles and reviews were ultimately evaluated, while gray articles or publicly available materials such as meeting abstracts, letters, and editorials were not included in the final analysis. Even so, we believe that our study can be utilized to provide an overview of the overall condition and emerging trends for telerehabilitation for hip or knee arthroplasty, as well as to make recommendations for further research.
Mobile Application‐Based Program for Telerehabilitation After Total Hip or Knee Arthroplasty The delivery modes of telerehabilitation interventions for arthroplasty were labeled as synchronous or asynchronous on the basis of whether health care providers communicated with patients in real time . As part of the synchronized telerehabilitation program, physical therapy assessments, supervised exercises, and prescriptions for home activities were performed. However, the system has the following limitations when it is implemented: patients and providers must sit in front of an excellent screen with a reliable internet connection at the same time; technicians are required to visit patients' homes to install or repair the equipment; and training is necessary because of the complexity of the technology . Moreover, the asynchronous telerehabilitation method provided information on diseases, pain management, cognitive behavioral therapy, and complication prevention. In particular, mobile applications enable health care providers to present more easily accessible and flexible telehealth support. Patients can review the rehabilitation program whenever and wherever it is most expedient for them . Thus, from 2003 to 2017, the synchronous mode was most prevalent, and videoconferencing or Skype was used for individual consultations with participants. However, studies using mobile apps, virtual exercise rehabilitation programs (VERAs), and interactive exercises have increasingly switched to asynchronous modes since 2017. This trend reflects the advancement of technology in telehealth: from telephone to videoconferencing and then to mobile applications . Owing to their convenience and accessibility, mobile applications are rapidly being used to deliver telerehabilitation services . These services include daily postoperative care education, exercise demonstrations, reminders, and progress monitoring . Consistent with this trend, our study also revealed that mobile applications presented a yellow color and larger nodes on the overlay visualization map, indicating that mobile applications have become a hot research topic in the last few years for telerehabilitation after total hip or knee arthroplasty. Numerous studies in recent years have demonstrated the efficacy of mobile app–based telerehabilitation programs for arthroplasty rehabilitation. Compared with conventional rehabilitation, application‐based telerehabilitation yields superior outcomes in terms of function, pain, and ROM . Recent studies have demonstrated that rehabilitation utilizing mobile applications may have beneficial impacts on patients' self‐efficacy, patient‐reported physical function, health‐related quality of life (HRQoL), and anxiety and depression levels . In this study, the telerehabilitation programs are offered through the mobile application WeChat, where all exercises are delivered via video demonstrations and uploaded to the mobile application. In addition, a randomized controlled trial showed that smartphone app–based remote rehabilitation worked better than home‐based rehabilitation with outpatient guidance in terms of short‐term results in ROM, the Five Times Sit‐to‐Stand Test (5xSST), and the Single‐Leg Stance Test (SLST) . The app named the Vital Health Remote Rehabilitation System consists of three parts: the patient‐side app, wearable sensors, and surgeon‐side websites. The app can offer real‐time guidance and assessment of patient movement according to monitoring data . However, most of the evidence supporting mobile app–based telerehabilitation in TJA has been derived from quantitative non‐randomized studies. Earlier studies focused on physical function and rarely examined psychological parameters, such as kinesiophobia, fear of falling, depression, and anxiety. Moreover, it is the elderly who are most likely to undergo joint replacements, and despite technological literacy steadily increasing among the general population, older adults rarely use mobile apps. Therefore, to confirm the effectiveness of mobile app–based telerehabilitation in TJA, robust RCTs are warranted . Moreover, researchers should consider incorporating qualitative components into future research designs and be aware of physical (sensory, motor, and cognitive) declines in older people to gain a better understanding of telerehabilitation and how it contributes to the recovery of patients .
“Patient Satisfaction,” “Self‐efficacy,” and “Cost Benefit Analysis” May Be Major Rehabilitation Outcomes for Hip or Knee Telerehabilitation Research Based on the co‐occurrence analysis of keywords via overlay visualization, the earliest keywords to appear were “motion” and “range.” This implies that the outcomes for the remote rehabilitation of arthroplasty in this period are related mainly to physical functions. In addition, in 2020, words such as “patient satisfaction,” “self‐efficacy,” and “cost benefit analysis” are displayed in yellow, implying that these aspects of research will be hotspots and research frontiers. Consistent with our findings, one study also reported that previous studies systematically emphasized pain, ROM, and physical function as outcomes of rehabilitation. The expert consensus on best practices in rehabilitation after TJA suggests that rehabilitation outcomes should include not only body structure/function (e.g., pain and ROM) and activity/participation (e.g., physical function) but also other outcomes, such as HRQoL and patient satisfaction with rehabilitation outcomes/processes . Psychological factors have been proven to affect recovery in many areas of surgery, including orthopedic, abdominal, breast, and oncological surgery . Psychological factors such as preoperative anxiety, depression, catastrophizing state, poor self‐efficacy, and poor coping skills in patients may be preoperative predictors of unsatisfactory outcomes in patients with arthroplasty . Therefore, psychological well‐being outcomes have been recommended because many patients after arthroplasty experience psychological problems such as anxiety and depression . Many patients who undergo arthroplasty suffer from poor psychological well‐being or psychological issues . These issues, such as pain and joint function, have adverse effects on patient rehabilitation outcomes . The outcomes of HRQoL indicate patients' general health and can be used to support clinical judgment for patients following TJA . Therefore, incorporating HRQoL as an outcome measure for assessing the recovery of patients after TJA is encouraged . However, the present study revealed that psychological consequences have rarely been investigated, and the duration and magnitude of the impact of psychological factors on postoperative outcomes are not clear . Only four out of 22 studies have examined psychological well‐being/problems ; therefore, psychological outcome measures should be included in future studies to assess the effects of telerehabilitation on psychological well‐being/problems among patients after TJA . Another essential healthcare outcome to consider when examining telerehabilitation is patient satisfaction . Both a healthcare recipient's cognitive assessment and the emotional response to their experience with healthcare can be characterized as satisfactory in healthcare management . The level of satisfaction has been cited as a key indicator of quality in healthcare because it is correlated with both adherence to treatment plans and improved clinical outcomes . Despite this, Kairy et al. have noted that there are no robust and standardized measures of satisfaction in articles discussing telerehabilitation for physical disabilities . Large‐scale clinical trials have not yet examined whether patients are satisfied with in‐home telerehabilitation after TKA. Few studies have examined the various factors affecting users' satisfaction with the service or technology employed. These factors include satisfaction with the care provided, perception of telehealth, satisfaction with telemedicine services, and individual appreciation of the quality of the technical system . The information on cost‐effectiveness is critical for guiding resource allocation decisions in the context of value‐based health systems . There have been encouraging results from cost‐effectiveness studies conducted in the orthopedic population. Telerehabilitation was found to be more cost‐effective in a Canadian‐based cost‐analysis study comparing in‐person care with telerehabilitation (via home visits) for knee replacement patients . A systematic review conducted by de la Torre‐Dez et al. also supported that some cost‐effectiveness studies revealed that telemedicine may reduce costs. Additionally, one study found that telerehabilitation could cut healthcare spending mainly due to lower transportation costs . These findings suggest that telerehabilitation may be an economically feasible alternative to in‐person rehabilitation after joint replacement. However, there are still several drawbacks, including a lack of randomized controlled trials, limited sample sizes, and poor data quality . As a result, the cost‐effectiveness of telemedicine applications could not be determined because of methodological flaws in the analyses . Therefore, many further studies are needed to address the problems mentioned above. Self‐efficacy in rehabilitation plays a significant role in determining the efficacy of rehabilitation. According to self‐efficacy theory, direct mastery experiences, vicarious experiences, verbal persuasion, and arousal state are four factors that contribute to individuals' confidence in their ability to successfully complete a task . By incorporating these concepts into a rehabilitation strategy, rehabilitation programs can be optimized in terms of content and structure, as well as the motivation and confidence of patients in the rehabilitation process, which will lead to greater compliance with exercise and better clinical outcomes such as physical function and HRQoL . However, a review revealed that few studies have examined patients' self‐efficacy during telerehabilitation . Therefore, in addition to pain, ROM, and physical function, other outcomes such as HRQoL, satisfaction, cost‐effectiveness, self‐efficacy, and psychological well‐being should be included to gain a holistic understanding of the effectiveness of internet‐based telerehabilitation.
Strengths and Limitations This research has multiple strengths. It is one of the pioneering studies that focuses on the specific topic of telerehabilitation for hip or knee arthroplasty. The use of the visualization software CiteSpace and VOSviewer for bibliometric analysis can provide objective and comprehensive results. A relatively scientific and intuitive picture was provided. In addition, the study is highly pertinent to the scope of the journal, and it would be beneficial to a wide variety of audiences, including freshmen, scholars, doctors, and rehabilitation therapists. Despite our study providing a comprehensive picture of telerehabilitation for hip or knee arthroplasty, several limitations should be addressed. First, as we know, co‐occurrence, collaborative network, and co‐citation analysis are the most important analyses in bibliometric studies. However, CiteSpace may have limitations when analyzing data from different databases. For example, CiteSpace is unable to analyze co‐occurrence data from Scopus or cooperative networks from Derwent. In addition, the CiteSpace software does not support the processing of data from the Google Scholar database. Therefore, we specifically analyzed the data from Web of Science and PubMed. In the future, as bibliometric software is updated, it will be necessary to incorporate additional database resources, such as Scopus and Google Scholar, to facilitate further investigations. Second, the WoS mainly indexes papers written in English. However, there may be Chinese papers or papers published in Chinese journals that are not yet included in WoS; therefore, this language restriction may have prevented us from finding some publications, which might have resulted in selection bias. Third, even though we used the search terms chosen, we cannot guarantee that we have retrieved all the documents related to the topic. Fourth, it is also possible that publication and citation counts do not directly represent the scientific quality of the papers, which may be impacted by multiple factors, including sample size, study design, and standard of reporting. In addition, bibliometric analysis cannot determine the scientific quality of the content, including bias risks, effect sizes, and statistical significance of the conclusions and whether they are justified by the respective results. Fifth, only articles and reviews were ultimately evaluated, while gray articles or publicly available materials such as meeting abstracts, letters, and editorials were not included in the final analysis. Even so, we believe that our study can be utilized to provide an overview of the overall condition and emerging trends for telerehabilitation for hip or knee arthroplasty, as well as to make recommendations for further research.
Conclusions In this study, bibliometric and visual analyses of telerehabilitation after total hip or knee arthroplasty were presented via the Web of Science database, CiteSpace software, and VOSviewer software, and a relatively scientific and intuitive picture was provided. Based on publishing data, the number of publications, influential nations and organizations, authors and co‐cited authors, published journals, and collaboration networks were all given a comprehensive assessment. Furthermore, we presented historical and prospective perspectives as well as major research and development hotspots, trends, and frontiers in hip or knee arthroplasty telerehabilitation. A major concern at present is physical therapy for home telerehabilitation in elderly individuals, and future research could focus on mobile application–based programs after total hip or knee arthroplasty for telerehabilitation. Moreover, as outcomes for hip or knee arthroplasty telerehabilitation research, a previous study focused on pain, ROM, and physical function. Patient satisfaction, self‐efficacy, and cost–benefit analysis are expected to serve as major rehabilitation outcomes in the future. Our work will serve as a valuable resource to provide fundamental reference and directional guidance for future research.
Project conceptualization: H.C.Q. and W.L.Q. study design: W.L.Q. data collection/validation: W.L.Q. and Z.L.M. data analysis: W.L.Q. and Z.L.M. result interpretation: W.L.Q. reporting and editing: W.L.Q. Final approval of the version to be submitted: H.C.Q., W.L.Q., and Z.L.M. project guarantor: H.C.Q. and W.L.Q. All authors have read and approved the manuscript.
The authors have nothing to report.
The authors declare no conflicts of interest.
Tables S1–S2.
|
Knowledge mapping and visualized analysis of research progress in onconephrology: a bibliometric analysis | 8c62450c-cbe0-4c09-a15d-fea55d223f9a | 11921167 | Internal Medicine[mh] | Introduction Onconephrology is a new subspecialty that combines nephrology and oncology, focusing on their complex interactions. In general, onconephrology refers to drug nephrotoxicity and end-organ damage from underlying malignancies beyond cancers of the kidney. Although the term ‘onconephrology’ was coined in the 2010s , the practice of involving nephrologists in the care of cancer patients dates back much further. Common and specific scenarios of onconephrology often include acute kidney injury (AKI), electrolyte disturbances, cancer-related glomerular diseases, paraproteinemia, and hematopoietic stem cell transplant-related kidney disorders. Regardless of the specific form, the mission of onconephrology is to provide multidisciplinary, coordinated, and collaborative care involving both nephrologists and oncologists for cancer patients. Bibliometric analysis uses mathematical and statistical approaches to analyze research hotspots and provide insight into future directions by evaluating the published literature in a specific field quantitatively or semiquantitatively . In contrast to traditional meta-analysis, which predominantly synthesizes new data, bibliometric analysis instead focuses on analyzing countries, institutions, authors, journals, documents, references, and keywords to accurately capture the hot topics and research prospects in a particular research area. Bibliometric analysis has been successfully applied to identify future research directions in various medical specialties . Research has dramatically improved our understanding of the underlying pathogenesis and management of onconephrology. Although several narrative reviews on onconephrology have been published recently , there is still a significant need for comprehensive bibliometric analyses to provide information on development trends and key research topics. This study aimed to analyze the current status of onconephrology and to provide nephrologists and oncologists with an overview of hotspots and clinical insights in the field of onconephrology.
Materials and methods 2.1. Data sources and literature inclusion and exclusion criteria This study, like earlier research , used the Science Citation Index Expanded of the Web of Science Core Collection (SCIE-WOS) database for literature searches. The emphasis was on documents focusing on both cancer and kidney diseases, excluding renal cancer and kidney transplant-related issues. Only English-language articles or reviews published between 1 January 2000 and 27 April 2024 (the date of the literature search), were included. Exclusions were: (1) non-article/review documents (e.g., editorials, meeting abstracts); (2) non-English publications; and (3) articles published before 1 January 2020. Two authors (YWW and SLF) screened titles and abstracts to remove irrelevant studies, with disputes resolved by a third author (WW). The search and selection process is outlined in . 2.2. Data visualization and analysis The included documents were first downloaded from SCIE-WOS and stored in plain text format with full records and complete references ( Supplementary Material ). The analyses of countries, authors, and the generation of co-occurrence and burst graphs were enabled by using CiteSpace (version 6.2.7R, Dr. Chaomei Chen, China) with time slices (2000–2024), years per slice (1), node types (country/author), and g indices and scale factors ( k = 25). The software VOSviewer (Leiden University Center for Science and Technology Studies) was used for the analysis of institutions, journals, cited references, and keywords. During the keyword analysis, terms with similar meanings, such as acute renal failure and acute kidney injury, were grouped together. The journal impact factor (IF) was obtained from the most recent Journal Citation Reports.
Data sources and literature inclusion and exclusion criteria This study, like earlier research , used the Science Citation Index Expanded of the Web of Science Core Collection (SCIE-WOS) database for literature searches. The emphasis was on documents focusing on both cancer and kidney diseases, excluding renal cancer and kidney transplant-related issues. Only English-language articles or reviews published between 1 January 2000 and 27 April 2024 (the date of the literature search), were included. Exclusions were: (1) non-article/review documents (e.g., editorials, meeting abstracts); (2) non-English publications; and (3) articles published before 1 January 2020. Two authors (YWW and SLF) screened titles and abstracts to remove irrelevant studies, with disputes resolved by a third author (WW). The search and selection process is outlined in .
Data visualization and analysis The included documents were first downloaded from SCIE-WOS and stored in plain text format with full records and complete references ( Supplementary Material ). The analyses of countries, authors, and the generation of co-occurrence and burst graphs were enabled by using CiteSpace (version 6.2.7R, Dr. Chaomei Chen, China) with time slices (2000–2024), years per slice (1), node types (country/author), and g indices and scale factors ( k = 25). The software VOSviewer (Leiden University Center for Science and Technology Studies) was used for the analysis of institutions, journals, cited references, and keywords. During the keyword analysis, terms with similar meanings, such as acute renal failure and acute kidney injury, were grouped together. The journal impact factor (IF) was obtained from the most recent Journal Citation Reports.
Results 3.1. Trends of annual publications and cumulative citations After applying the inclusion and exclusion criteria, a total of 1853 documents, including 1,647 articles and 206 reviews, were included in the final bibliometric analysis. As shown in , the number of annual publications generally follows a steadily increasing trend, ranging from a low of 25 documents in 2003 to a high of 161 documents in 2023. Similarly, the number of cumulative citations is also increasing, suggesting that the area of onconephrology is gaining momentum. 3.2. Country and institution analysis The results of the country and institution analyses are presented in . Overall, 2757 institutions from 73 countries contributed to these 1853 documents. The United States ( n = 464), China ( n = 287), Japan ( n = 270), England ( n = 116), and France ( n = 112) were the top 5 countries in terms of total scientific output, accounting for 67.40% of the total onconephrology literature. However, when analyzed with an emphasis on importance as indicated by betweenness centrality, the United States, England, France, Canada, and Germany were hub nodes with betweenness centralities of 0.36, 0.23, 0.19, 0.15, and 0.12, respectively. Institutional analysis revealed that the top 5 institutions with the highest publication volume were the University of Texas MD Anderson Cancer Center ( n = 39), Mayo Clinic ( n = 36), National and Kapodistrian University of Athens ( n = 39), Dana-Farber Cancer Institute ( n = 26), and Yale University ( n = 26). 3.3. Analysis of prolific authors and journals shows that the 1,853 documents published in 601 academic journals were contributed by 11,606 authors. The authors with ≥10 total publications were Meletios A. Dimopoulos ( n = 21, National and Kapodistrian University of Athens), Kenar D. Jhaveri ( n = 14, Zucker School of Medicine), Nelson Leung ( n = 13, Mayo Clinic), Efstathios Kastritis ( n = 11, National and Kapodistrian University of Athens), Evangelos Terpos ( n = 11, National and Kapodistrian University of Athens), Jolanta Malyszko ( n = 11, Medical University of Warsaw), and Paul Cockwell ( n = 10, Queen Elizabeth Hospital). Among the 16 journals that published ≥20 documents in this field, the top 5 with the most publications were Nephrology Dialysis Transplantation ( n = 35, IF = 6.1), Internal Medicine ( n = 32, IF = 1.2), Cancer Therapy and Pharmacology ( n = 31, IF = 3.0), Clinical Nephrology ( n = 28, IF = 1.1), and PLoS One ( n = 27, IF = 3.7). Among the top 10 most productive journals, five were nephrology journals, three were multidisciplinary journals, and two were cancer journals. 3.4. Influential documents and highly cited references The analyses of the most cited documents and references are summarized in and , respectively. The timeline view of the cited references, as displayed in , shows that the references cited between 2000 and 2015 were related to lenalidomide, dose adjustment, acute interstitial nephritis, ibandronate, interleukin-2, multiple myeloma, and iron. In comparison, the clusters identified after 2015 were immune checkpoint inhibitors, glomerular filtration rate, and cisplatin. The reference burst analysis ( ) revealed that the articles ‘Clinical Features and Outcomes of Immune Checkpoint Inhibitor-Associated AKI: A Multicenter Study’ by Cortazar et al. in 2020 and ‘The Incidence, Causes, and Risk Factors of Acute Kidney Injury in Patients Receiving Immune Checkpoint Inhibitors’ by Seethapathy et al. in 2019 were recent bursts. 3.5. Keyword analysis A total of 5,630 keywords were extracted, of which 316 and 54 keywords appeared ≥10 times and ≥50 times, respectively. As shown in , the top 10 most frequent keywords were acute kidney injury ( n = 537), cancer ( n = 340), multiple myeloma ( n = 281), renal insufficiency ( n = 277), nephrotoxicity ( n = 269), risk factors ( n = 249), chemotherapy ( n = 235), chronic kidney disease (CKD, n = 215), dialysis ( n = 193), and cisplatin ( n = 189). Cluster analysis ( ) revealed that these keywords can be grouped into eight clusters, including multiple myeloma (#0), cisplatin (#1), nephrotic syndrome (#2), chronic kidney disease (#3), immunotherapy (#4), oxidative stress (#5), zoledronic acid (#6), and hepatocellular carcinoma (#7). Trend analysis ( ) revealed that immune checkpoint inhibitors, oxaliplatin, calcium, open-label, and thrombotic microangiopathy emerged after 2020. The keyword burst analysis ( ) indicated that outcome, acute kidney injury, immunotherapy, risk factors, and chronic kidney diseases were bursts that still persist through 2024.
Trends of annual publications and cumulative citations After applying the inclusion and exclusion criteria, a total of 1853 documents, including 1,647 articles and 206 reviews, were included in the final bibliometric analysis. As shown in , the number of annual publications generally follows a steadily increasing trend, ranging from a low of 25 documents in 2003 to a high of 161 documents in 2023. Similarly, the number of cumulative citations is also increasing, suggesting that the area of onconephrology is gaining momentum.
Country and institution analysis The results of the country and institution analyses are presented in . Overall, 2757 institutions from 73 countries contributed to these 1853 documents. The United States ( n = 464), China ( n = 287), Japan ( n = 270), England ( n = 116), and France ( n = 112) were the top 5 countries in terms of total scientific output, accounting for 67.40% of the total onconephrology literature. However, when analyzed with an emphasis on importance as indicated by betweenness centrality, the United States, England, France, Canada, and Germany were hub nodes with betweenness centralities of 0.36, 0.23, 0.19, 0.15, and 0.12, respectively. Institutional analysis revealed that the top 5 institutions with the highest publication volume were the University of Texas MD Anderson Cancer Center ( n = 39), Mayo Clinic ( n = 36), National and Kapodistrian University of Athens ( n = 39), Dana-Farber Cancer Institute ( n = 26), and Yale University ( n = 26).
Analysis of prolific authors and journals shows that the 1,853 documents published in 601 academic journals were contributed by 11,606 authors. The authors with ≥10 total publications were Meletios A. Dimopoulos ( n = 21, National and Kapodistrian University of Athens), Kenar D. Jhaveri ( n = 14, Zucker School of Medicine), Nelson Leung ( n = 13, Mayo Clinic), Efstathios Kastritis ( n = 11, National and Kapodistrian University of Athens), Evangelos Terpos ( n = 11, National and Kapodistrian University of Athens), Jolanta Malyszko ( n = 11, Medical University of Warsaw), and Paul Cockwell ( n = 10, Queen Elizabeth Hospital). Among the 16 journals that published ≥20 documents in this field, the top 5 with the most publications were Nephrology Dialysis Transplantation ( n = 35, IF = 6.1), Internal Medicine ( n = 32, IF = 1.2), Cancer Therapy and Pharmacology ( n = 31, IF = 3.0), Clinical Nephrology ( n = 28, IF = 1.1), and PLoS One ( n = 27, IF = 3.7). Among the top 10 most productive journals, five were nephrology journals, three were multidisciplinary journals, and two were cancer journals.
Influential documents and highly cited references The analyses of the most cited documents and references are summarized in and , respectively. The timeline view of the cited references, as displayed in , shows that the references cited between 2000 and 2015 were related to lenalidomide, dose adjustment, acute interstitial nephritis, ibandronate, interleukin-2, multiple myeloma, and iron. In comparison, the clusters identified after 2015 were immune checkpoint inhibitors, glomerular filtration rate, and cisplatin. The reference burst analysis ( ) revealed that the articles ‘Clinical Features and Outcomes of Immune Checkpoint Inhibitor-Associated AKI: A Multicenter Study’ by Cortazar et al. in 2020 and ‘The Incidence, Causes, and Risk Factors of Acute Kidney Injury in Patients Receiving Immune Checkpoint Inhibitors’ by Seethapathy et al. in 2019 were recent bursts.
Keyword analysis A total of 5,630 keywords were extracted, of which 316 and 54 keywords appeared ≥10 times and ≥50 times, respectively. As shown in , the top 10 most frequent keywords were acute kidney injury ( n = 537), cancer ( n = 340), multiple myeloma ( n = 281), renal insufficiency ( n = 277), nephrotoxicity ( n = 269), risk factors ( n = 249), chemotherapy ( n = 235), chronic kidney disease (CKD, n = 215), dialysis ( n = 193), and cisplatin ( n = 189). Cluster analysis ( ) revealed that these keywords can be grouped into eight clusters, including multiple myeloma (#0), cisplatin (#1), nephrotic syndrome (#2), chronic kidney disease (#3), immunotherapy (#4), oxidative stress (#5), zoledronic acid (#6), and hepatocellular carcinoma (#7). Trend analysis ( ) revealed that immune checkpoint inhibitors, oxaliplatin, calcium, open-label, and thrombotic microangiopathy emerged after 2020. The keyword burst analysis ( ) indicated that outcome, acute kidney injury, immunotherapy, risk factors, and chronic kidney diseases were bursts that still persist through 2024.
Discussion This bibliometric analysis revealed that the field of onconephrology is steadily expanding. Moreover, although China and Japan have made significant contributions to this field, their impact on the field is suboptimal. Current research hotspots are diverse and include chemotherapeutic nephrotoxicity, dosing of chemotherapeutic agents in CKD, kidney function assessment, glomerular disease in cancer, immunotherapy, and electrolyte disturbances. Future directions in this field may encompass clinical trials and thrombotic microangiopathy (TMA). 4.1. General information Although the bidirectional relationship between kidney disease/function and cancer has long been well known, the term ‘onconephrology’ was coined in the early 2010s. In fact, it was not until 2019 that an international conference on onconephrology was held to address the key issues at the challenging clinical interface of onconephrology . Interestingly, our bibliometric analysis indicated that practice and multidisciplinary cooperation have already been present, as indicated by the number of publications in this field. Country, institution, and author analysis suggested that Western countries and researchers are the driving force in this challenging endeavor. Although China and Japan ranked 2nd and 3rd in terms of scientific output, none of the top 10 institutions or researchers were based in China or Japan, highlighting that these nations should cultivate more influential research with greater scientific impact. The journal analysis also indicated that the research activities in the field of onconephrology encompass related specialties, including nephrology, cancer, and multiple disciplines. Therefore, interested researchers can obtain the latest information from these journals and should contribute their studies to these journals. 4.2. Current research hotspots The timeline view analysis of the most highly cited references and the keyword analysis revealed that current hotspots in the field of onconephrology are broad and involve both solid and hematologic malignancies. In addition, chemotherapeutic nephrotoxicity, the dosing of chemotherapeutic agents in CKD patients, glomerular diseases in cancer patients, kidney function measurement, electrolyte disturbances, immunotherapy, and basic research have emerged as current hotspots. 4.2.1. Types of malignancies and treatment modalities involved Among the top 10 most cited documents and references, 6/10 and 6/10 were related to multiple myeloma, indicating that multiple myeloma-related renal insufficiency is still the most important and hottest topic in onconephrology. Approximately half of multiple myeloma patients are estimated to develop renal dysfunction during the course of treatment . More importantly, numerous studies have associated renal impairment with decreased overall survival and premature mortality in multiple myeloma patients . The main mechanism of renal injury in multiple myeloma is the overproduction of free light chains, which are nephrotoxic because they promote proximal tubule apoptosis and induce inflammation with subsequent interstitial fibrosis . In addition, other conditions associated with multiple myeloma, such as amyloidosis, monoclonal immunoglobulin deposition disease, light-chain proximal tubulopathy, and cryoglobulinaemia, may also cause renal impairment. Notably, other multiple myeloma factors, such as dehydration, hypercalcemia, and tumor lysis syndrome, and treatment-related factors, such as hematopoietic stem cell transplantation, are also essential etiologies. Our bibliometric results also suggest that solid tumors, represented by the cluster ‘hepatocellular carcinoma’, are also involved. In fact, many treatments for solid tumors, such as targeted therapy and immunotherapy, have also been associated with a variety of renal complications . For example, immune checkpoint inhibitors, such as programmed cell death (PD)-1/PD-L1, have been found to cause hypophosphatemia, proteinuria, acute interstitial nephritis, and acute tubular necrosis . Our study revealed that in addition to conventional cytotoxic chemotherapies, emerging novel cancer therapies, such as immunotherapies (immune checkpoint inhibitors and chimeric antigen receptor-T [CAR-T] cell therapy) and targeted therapy, also cause kidney injury. Kidney injury in CAR-T most commonly occurs as AKI, which is related to hypoperfusion and the inflammatory effects of released cytokines . In addition, agents for targeted therapies, including anaplastic lymphoma kinase inhibitors (e.g., brigatinib and ceritinib), cyclin-dependent kinase 4/6 inhibitors (e.g., abemaciclib and ribociclib), BRAF inhibitors (e.g., vemurafenib) and poly(ADP–ribose) polymerase inhibitors (e.g., olaparib), have been found to cause pseudo-AKI by reducing renal tubular secretion of creatinine, resulting in increased serum creatinine . 4.2.2. Chemotherapeutic nephrotoxicity Anticancer agents are increasingly recognized as major causes of acute and chronic kidney injury. Clinically, the manifestations of onconephrology take multiple forms, such as AKI, CKD, electrolyte disturbances, Fanconi’s syndrome, onco-hypertension, proteinuria/nephrotic syndrome, renal cysts, and TMA. Our bibliometric analysis of keywords suggests that AKI appears to receive the most attention. Epidemiologic studies have suggested that the incidence of AKI in cancer patients ranges from 24% (95% CI 17–30%) to 52% (95% CI 34–70%) , highlighting that the incidence of AKI in cancer patients may be closely related to coexisting comorbidities, cancer type and stage, and specific treatment regimens. A nationwide cohort study of 3,120 children, 16,310 adults, and 3,802 hospitalized elderly patients revealed that the overall incidence of AKI was 4.9%, most commonly observed in genitourinary, hematological, and neuro-musculoskeletal cancers and caused by purine analogs, folic acid analogs, and combinations of antineoplastic agents . The causes of AKI in cancer patients are heterogeneous and can be broadly categorized into cancer-related factors and therapy-related factors. Cancer-related factors include urinary tract obstruction, metabolic disorders, glomerular disease, and hemodynamic alterations, whereas therapy-related factors include radiation, immunotherapy, targeted therapy, chemotherapy, hematopoietic stem cell transplantation, tumor lysis syndrome, and TMA. A retrospective study of 67,986 cancer patients revealed that stages 1, 2, and 3 AKI were associated with 18.3% (95% CI 1.145–1.221), 71.0% (95% CI 1.629–1.796), and 100.0% (95% CI 1.910–2.095) increased risk of all-cause mortality, respectively . Therefore, sensitive biomarkers for the timely detection and diagnosis of AKI in cancer patients are urgently needed for early intervention to improve patient outcomes. 4.2.3. Kidney function assessment Accurate assessment of kidney function in cancer patients is important for risk stratification and appropriate dosing of chemotherapeutic agents, antibiotics, opioid analgesics, and other medications. Dose adjustment on the basis of kidney function is often required for a variety of common chemotherapeutic agents, such as alkylating agents, antimicrotubule agents, platinum agents, antimetabolites, immunomodulatory agents, and proteasome inhibitors. Although assessment with radiopharmaceuticals, such as chromium-51-ethylenediaminetetraacetic acid ( 51 Cr-EDTA, not available in the US) or technetium-99m-diethylenetriaminepentaacetic acid ( 99 mTc-DTPA) is considered the gold standard for kidney function, these methods are time-consuming, costly, and labor-intensive, making them impractical to perform on a regular basis. In comparison, estimation of the kidney glomerular filtration rate by serum creatinine and/or cystatin C-based formulas, such as the Cockcroft-Gault equation, the Modification of Diet in Renal Disease equation, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation, and the race-free formula calculated on the basis of both creatinine and cystatin C , appears to be more convenient and practical. Nonetheless, the clinical use of these equations may be biased by various confounding factors in cancer patients, such as anorexia, muscle wasting, and nutritional status. Moreover, validation of these equations in cancer patients is limited by the fact that most involve specific types of cancer in a limited number of patients. In the largest validation cohort of 2471 survivors with various types of cancer, Janowitz et al. validated these formulas against the 51 Cr-EDTA method and reported that the body surface area-adjusted CKD-EPI formula is the most accurate . More recently, cancer-specific formulas have been proposed and developed. For example, Williams and colleagues developed an eGFR formula called ‘CamGFR v2’ based on either standardized or nonstandardized serum creatinine, which outperformed currently available formulas in 7,240 cancer patients . Currently, there is no consensus regarding the best formula for cancer patients, and a prospective clinical trial is needed. In cases where significant discrepancies (differences >10% or >10 mL/min/1.73 m 2 ) were noted when different formulas were used, dose adjustment based on the lower estimated glomerular filtration rate result may be considered, especially for chemotherapeutic agents with dramatic nephrotoxicity with a narrow therapeutic range or in vulnerable patient populations. 4.2.4. CKD in cancer patients and chemotherapy dosing The reported prevalence of CKD in patients with cancer ranges from 12 to 25%, depending on the exact cancer type and associated demographic factors . In an analysis of 5,831 biopsy-proven cancer patients, Ciorcan et al. demonstrated that cancers of the kidney, urinary tract, and pancreas had the highest prevalence of CKD, defined as an estimated glomerular filtration rate <60 mL/min/1.73 m 2 . Although the majority of CKD cases in cancer patients are modest in degree, the presence of diabetes and hypertension also synergistically worsens patient survival. It has been unequivocally demonstrated that the presence of CKD, especially stage 3–5 CKD, is related to a 27% (95% CI 1.00–1.60) increased mortality rate compared with those with estimated glomerular filtration rate >60 mL/min/1.73 m 2 . A critical and practical clinical issue related to CKD is dose adjustment for most chemotherapeutic agents. The prerequisite is accurate measurement of the glomerular filtration rate, which can have a dramatic impact on chemotherapeutic pharmacokinetics and drug toxicity. Studies have shown that the incidences of AKI or AKI transition to CKD are significantly increased in those receiving higher doses of preconditioning chemotherapy . A comprehensive review of the exact dosing regimens of anticancer drugs is beyond the scope of this study, but a brief discussion of the dosage adjustment of chemotherapeutic medications with different renal functions is essential to highlight their nuanced relationship. For example, reducing the dose of cisplatin by 50–70% in patients with a creatinine clearance rate <50 mL/min/1.73 m 2 , and by 75% for capecitabine in patients with a creatinine clearance rate between 30 and 50 mL/min/1.73 m 2 , is recommended . Nonetheless, no dose adjustments are required for vascular endothelial growth factor inhibitors, epithelial growth factor receptor inhibitors, PD-1/PD-L1 antagonists, or cytotoxic T-lymphocyte antigen 4 antagonists. In addition, other strategies in addition to dose adjustment may be used to mitigate nephrotoxicity associated with anticancer agents. For example, teneligliptin, a dipeptidyl peptidase-4 inhibitor, has been shown in vivo to attenuate cisplatin-induced nephrotoxicity and improve renal function by accelerating tubular regeneration and reducing injury and fibrosis . 4.2.5. Glomerular diseases associated with cancer Glomerular disease in cancer refers to the concurrent or metachronous occurrence of glomerular disease, most commonly manifested as proteinuria with or without elevated serum creatinine, in cancer. Although various patterns of glomerular injury have been described in the literature, the prototypical example is cancer-associated membranous nephropathy, which is largely mediated by antibodies to thrombospondin type 1 domain-containing 7A . Epidemiologic studies suggest that the incidence of cancer in membranous nephropathy is 10% (95% CI 6.1–14.6%), with cancer detected before or at the time of membranous nephropathy diagnosis in ∼80% of cases . The mechanistic link is hypothesized to be subepithelial deposition of tumor antigens associated with an enhanced immune response targeting the tumor. In contrast to idiopathic membranous nephropathy, which is histologically characterized by the IgG4 subclass, cancer-associated membranous nephropathy is characterized predominantly by the IgG1 and IgG2 subclasses. Other reported patterns of glomerular disease in solid tumors include pauci-immune glomerulonephritis, renal vasculitis, podocytopathy, C3 glomerulopathy, rapidly progressive glomerulonephritis, IgA nephropathy, and membranoproliferative glomerulonephritis . 4.2.6. Electrolyte disorders Electrolyte abnormalities are common in both solid and hematopoietic malignancies, most commonly as chemotherapy-induced, tumor-related sequelae. Common electrolyte disturbances observed in cancer patients include hyponatremia, hypokalemia/hyperkalemia, hypocalcemia/hypercalcemia, hyperphosphatemia/hypophosphatemia, and hypomagnesemia. Hyponatremia is the most common type of electrolyte disturbance and is most commonly caused by decreased appetite, anorexia, and vomiting, and inappropriate secretion of antidiuretic hormone by many malignancies and chemotherapeutic regimens. Hyperkalemia is mostly caused by tumor lysis syndrome, an oncologic emergency in which tumors lyse spontaneously or secondary to anticancer therapies with sudden release of intracellular ions and metabolites into the systemic circulation. In fact, in addition to hyperkalemia, tumor lysis syndrome may also include hyperphosphatemia, hypocalcemia, hyperuricemia, and AKI. Hypercalcemia in cancer patients is often caused by extensive cytokine-mediated osteolysis or the release of parathyroid hormone-related peptides. The appearance of zoledronic acid as a keyword cluster suggests that attention has been given to the treatment of hypercalcemia, which can often be an onconephrology emergency. Hypophosphatemia is often associated with platinum-based chemotherapy, especially cisplatin. Hypomagnesemia is often associated with the administration of monoclonal antibodies targeting epidermal growth factor receptor, particularly cetuximab, due to increased urinary magnesium loss, as magnesium reabsorption in the distal tubule is partially dependent on the activity of epidermal growth factor receptor proteins located on the basolateral tubular membrane . 4.2.7. Immunotherapy CAR-T cell therapy is a type of immunotherapy that represents a revolutionary treatment for hematological malignancies using genetically engineered host T cells. Although CAR-T cell therapy been shown to induce durable remission with reduced mortality and prolonged survival , it can potentially cause cytokine release syndrome, which can be observed in up to 90% of patients receiving CD19 CAR-T cells for B-cell acute lymphoblastic leukemia and non-Hodgkin B-cell lymphoma . A recent meta-analysis indicated that the pooled estimated incidences of AKI and AKI requiring renal replacement therapy following CAR-T cell therapy were 18.6% (95% CI 14.3–23.8%) and 4.4% (95% CI 2.1–8.9%), respectively . The cumulative incidence at day 100 was 21.7% (95% CI 9.7–33.8%) for grade 1 AKI and 8.7% (95% CI 0.4–17%) for grade 2–3 AKI , suggesting that AKI in patients receiving CAR-T cell therapy is mostly transient, mild in severity, and associated with rapid recovery . Nevertheless, 75 and 67% of patients who experienced AKI after CAR-T cell therapy still had CKD at the 6- and 12-month follow-ups, respectively . Potential mechanisms for AKI after CAR-T cell infusion include capillary leakage with hypovolemia, tumor lysis syndrome, and immune effector cell-associated neurotoxicity syndrome (ICANS). Immune checkpoint inhibitors are monoclonal antibodies that selectively block intrinsic down-regulating receptors of the immune system to activate suppressed T cells to enhance anti-tumor-directed immune responses . As reflected in the keywords trend analysis, immune checkpoint inhibitors have become widely used and considered standard of care in the management of many advanced cancers. AKI has been increasingly recognized as an uncommon but important form of kidney toxicity following the use of immune checkpoint inhibitors. A multi-center analysis revealed that AKI typically manifests at a median of 16 weeks following the initiation of immune checkpoint inhibitors, and kidney recovery was observed in 64.3% of patients at a median of 7 weeks following the onset of AKI . The underlying pathophysiology and corresponding treatments for AKI associated with CAR-T therapies and immune checkpoint inhibitors were different. Kidney injury in CAR-T is mostly related to hemodynamic changes and is typically transient and mild in severity, with only a small percentage of patients requiring continuous renal replacement therapy . Interestingly, studies have shown that preexisting CKD does not appear to have a significant impact on the safety, efficacy, or patient outcomes of CAR-T cell therapy . In comparison, acute tubulointerstitial nephritis is the most common cause of immune checkpoint inhibitor-associated AKI. Manohar et al. reported that prompt withholding of immune checkpoint inhibitors along with steroid use resulted in a complete response in 63% of patients . In addition, rechallenge in 4 patients was successful in 3 patients and the other patient developed recurrent acute allergic interstitial nephritis and fatal pneumonitis . Electrolyte abnormalities are extremely common in patients receiving immunotherapy and most commonly present as hyponatremia, hypokalemia, hypercalcemia/hypocalcemia, hyperphosphatemia/hypophosphatemia, and hypomagnesemia. A meta-analysis of 48 clinical trials revealed a pooled risk ratio of 1.67% (95% CI 0.89–3.12) for electrolyte disorders in cancer patients receiving PD-1 inhibitors . In a recent real-world study of 2,458 patients, Seethapathy’s group reported that 62% of patients had hyponatremia and that 6% had severe hyponatremia with sodium <124 mmol/L . There have been an increasing number of case reports highlighting hypophysitis and secondary adrenal insufficiency as the underlying cause of hyponatremia in patients receiving immune checkpoint inhibitors . Another common electrolyte abnormality in cancer patients receiving immunotherapy is hypercalcemia, which may be caused by immune endocrinopathies, parathyroid hormone-related peptide release, cancer hyperprogression, and sarcoidosis-like granulomas . Notable side effects associated with CAR-T cell therapy include cytokine release syndrome, tumor lysis syndrome, and ICANS. Patients with cytokine release syndrome often have nonspecific symptoms of fever, arthralgia, malaise and fatigue, anorexia, and tachycardia. Therapeutic approaches for cytokine release syndrome include supportive care and pharmacological intervention with the monoclonal anti-interleukin-6 antibody tocilizumab. There are also agents currently under investigation for the prophylaxis of cytokine release syndrome, such as tocilizumab, anakinra, teclistamab, and duvelisib . Tumor lysis treatment is an oncological emergency resulting from the release of intracellular electrolytes and nucleic acids from malignant cells. Treatment and prophylaxis for tumor lysis syndrome are similar and include aggressive intravenous hydration, the use of xanthine oxidase inhibitors and rasburicase, and medical management of associated electrolyte abnormalities. ICANS is a constellation of neuropsychiatric symptoms, including headache, aphasia, seizures, or decreased consciousness, that are observed in 20–70% of CAR-T cell therapy recipients. Risk factors for the occurrence of ICANS include cytokine release syndrome, high tumor burden, advanced age, and a strong inflammatory response . Antiepileptic medications, steroids, and tocilizumab have been the most common treatments for ICANS, especially for those with concurrent cytokine release syndrome. Preliminary studies investigating the interleukin-1 receptor antagonist anakinra for the prophylaxis of ICANS have reported encouraging results . 4.2.8. Underlying pathophysiology The investigation of the molecular mechanisms underlying nephrotoxicity caused by various anticancer drugs or treatments is particularly important, as it may shed light on prevention and intervention strategies. Moreover, the exact underlying pathways for anticancer drug-induced renal injury may differ depending on the exact type of agent. For example, cisplatin, a classic and highly effective chemotherapeutic agent for a variety of malignancies, is well known for its potential risk of inducing nephrotoxicity. Various molecular mechanisms have been reported in the pathogenesis of cisplatin-induced nephrotoxicity, including tubular cell apoptosis under oxidative stress, renal interstitial inflammatory cell infiltration leading to AKI, and extensive production of proinflammatory cytokines . Several lines of evidence also suggest that various signaling pathways are involved, including the toll-like receptor pathway, the NF-κB pathway, and the poly-ADP–ribose polymerase-1 pathway . Currently, the mechanism of action for nephrotoxicity resulting from the majority of traditional chemotherapeutic agents has been described and is largely related to the side effects of the desired anticancer properties of these agents. Specifically, antimetabolites, such as gemcitabine have a vasoconstrictive effect on afferent renal arteries, resulting in a decreased glomerular filtration rate and AKI . However, the mechanisms of action of targeted therapy-induced nephrotoxicity may extend beyond pharmacological action. For example, epidermal growth factor receptor inhibitors, such as cetuximab, cause electrolyte disturbances that can be explained by inhibited signaling at the distal convoluted tubule that regulates transepithelial magnesium transport. However, epidermal growth factor receptor inhibitors may also cause AKI, although the mechanisms of action remain unclear. Similarly, the exact mechanisms by which Her-2 inhibitors, anaplastic lymphoma kinase inhibitors, and BRAF inhibitors cause AKI remain elusive and warrant additional investigations. 4.3. Future research directions 4.3.1. TMA TMA is a pathological term that is clinically characterized by microangiopathic hemolytic anemia and thrombocytopenia and is associated with elevated lactate dehydrogenase and end-organ failure . The emergency of TMA as a trend topic is consistent with increasing reports suggesting that TMA represents ∼5.4% of the findings in kidney biopsies from cancer patients . Unlike primary TMA–which is mostly caused by genetic mutations–cancers and related therapies, especially hematopoietic stem cell transplantation, are important causes of secondary TMA. On biopsy, the renal TMA is characterized by extensive formation of fibrin thrombi in the capillary loops and arterioles, intimal on ionization, fragmented red blood cells, and mesangiolysis. Common solid tumors that can induce TMA include gastrointestinal, lung, genitourinary, and hepatobiliary cancers. In a series of 168 cases of cancer-related TMA, Lechner et al. reported that lymphoma accounted for ∼8.3% of all cases . Cancer therapy-related TMA is also relatively common and can be observed in those treated with conventional chemotherapy, targeted therapy, and immunotherapy, especially in those with underlying genetic defects in the alternative complement cascade. Commonly used conventional chemotherapy drugs include mitomycin-C, gemcitabine, bleomycin, and platinum-based agents, whereas epidermal growth factor monoclonal antibody agents are the most common targeted therapy used to induce TMA . Compared with conventional chemotherapy-induced TMA, which is often fulminant, systemic, and lethal, TMA caused by targeted therapy is generally confined to the kidney and has more favorable outcomes . However, discontinuation of stimulating agents may not lead to renal recovery, and additional therapeutic maneuvers, such as plasma exchange and rituximab administration, may have variable clinical efficacy . However, currently, all available evidence comes from case reports or small case series. The American Society for Apheresis guidelines consider therapeutic plasma exchange to be ineffective for treating chemotherapy-related TMA . In the largest cohort of gemcitabine-associated TMA, Daviet et al. reported that 54.7% of patients eventually died and that treatment with therapeutic plasma exchange did not improve outcomes compared with glucocorticoids and was associated with more adverse events . Eculizumab, a monoclonal antibody targeting terminal complement C5, has been shown in several case reports to promote renal recovery in cancer therapy-associated TMA . In a retrospective analysis of 12 cases of gemcitabine-induced TMA, Grall et al. showed that 17 and 67% achieved complete and partial recovery of renal function, respectively, after treatment with eculizumab . These increasing results encourage the conduct of prospective studies in the future to conclusively determine the efficacy of eculizumab in the treatment of cancer-related TMA. 4.3.2. Clinical trials Our trend topic analysis revealed that clinical trials are likely to be a hot topic in onconephrology in the future. We believe this raises 2 considerations. First, this may imply the recruitment of patients with impaired renal function into cancer clinical trials. Clearly, obtaining accurate information on the safety and effective dosing of chemotherapeutic agents in cancer patients with CKD is critical for formulating appropriate treatment regimens. Unfortunately, the majority of cancer clinical trials have enrolled patients with normal or mildly impaired renal function, thus underrepresenting patients with severely impaired renal function. For example, in a recent meta-analysis of 11,066 participants from 32 clinical trials receiving combination therapy with vascular endothelial growth factor pathway inhibitors and immune checkpoint inhibitors, Elyan et al. reported that all trials excluded patients with advanced CKD, and few trials included people with proteinuria . Butrovich et al. reported that pharmacokinetic analyses in patients with CKD stages 4–5 and hemodialysis were performed for only 29 and 6% of the 55 drugs approved between 2015 and 2019, respectively . The reasons for the exclusion of this specific patient population are complex, including sponsor concerns, safety concerns, and the lack of a robust nephrology clinical trial infrastructure . Another consideration is the design and application of specific clinical trials for specific onconephrology conditions or issues. Eculizumab or other novel treatments for cancer-associated TMA appear to be good examples of such clinical trials. Another area of interest is the design of prospective and controlled trials to evaluate the efficacy of renoprotective agents during cancer treatment. Successful trials of renoprotective agents would not only provide effective chemotherapy with minimized risk of renal toxicity but also offer chemotherapeutic options for cancer patients with preexisting renal impairment. For example, sirtuins, which are NAD+-dependent deacetylases with important antioxidant activity, have been demonstrated in animal studies to exert cytoprotective and renoprotective effects in various forms of nephrotoxicity . Thus, clinical trials evaluating their ability to reduce nephrotoxicity during cancer treatment are very meaningful.
General information Although the bidirectional relationship between kidney disease/function and cancer has long been well known, the term ‘onconephrology’ was coined in the early 2010s. In fact, it was not until 2019 that an international conference on onconephrology was held to address the key issues at the challenging clinical interface of onconephrology . Interestingly, our bibliometric analysis indicated that practice and multidisciplinary cooperation have already been present, as indicated by the number of publications in this field. Country, institution, and author analysis suggested that Western countries and researchers are the driving force in this challenging endeavor. Although China and Japan ranked 2nd and 3rd in terms of scientific output, none of the top 10 institutions or researchers were based in China or Japan, highlighting that these nations should cultivate more influential research with greater scientific impact. The journal analysis also indicated that the research activities in the field of onconephrology encompass related specialties, including nephrology, cancer, and multiple disciplines. Therefore, interested researchers can obtain the latest information from these journals and should contribute their studies to these journals.
Current research hotspots The timeline view analysis of the most highly cited references and the keyword analysis revealed that current hotspots in the field of onconephrology are broad and involve both solid and hematologic malignancies. In addition, chemotherapeutic nephrotoxicity, the dosing of chemotherapeutic agents in CKD patients, glomerular diseases in cancer patients, kidney function measurement, electrolyte disturbances, immunotherapy, and basic research have emerged as current hotspots. 4.2.1. Types of malignancies and treatment modalities involved Among the top 10 most cited documents and references, 6/10 and 6/10 were related to multiple myeloma, indicating that multiple myeloma-related renal insufficiency is still the most important and hottest topic in onconephrology. Approximately half of multiple myeloma patients are estimated to develop renal dysfunction during the course of treatment . More importantly, numerous studies have associated renal impairment with decreased overall survival and premature mortality in multiple myeloma patients . The main mechanism of renal injury in multiple myeloma is the overproduction of free light chains, which are nephrotoxic because they promote proximal tubule apoptosis and induce inflammation with subsequent interstitial fibrosis . In addition, other conditions associated with multiple myeloma, such as amyloidosis, monoclonal immunoglobulin deposition disease, light-chain proximal tubulopathy, and cryoglobulinaemia, may also cause renal impairment. Notably, other multiple myeloma factors, such as dehydration, hypercalcemia, and tumor lysis syndrome, and treatment-related factors, such as hematopoietic stem cell transplantation, are also essential etiologies. Our bibliometric results also suggest that solid tumors, represented by the cluster ‘hepatocellular carcinoma’, are also involved. In fact, many treatments for solid tumors, such as targeted therapy and immunotherapy, have also been associated with a variety of renal complications . For example, immune checkpoint inhibitors, such as programmed cell death (PD)-1/PD-L1, have been found to cause hypophosphatemia, proteinuria, acute interstitial nephritis, and acute tubular necrosis . Our study revealed that in addition to conventional cytotoxic chemotherapies, emerging novel cancer therapies, such as immunotherapies (immune checkpoint inhibitors and chimeric antigen receptor-T [CAR-T] cell therapy) and targeted therapy, also cause kidney injury. Kidney injury in CAR-T most commonly occurs as AKI, which is related to hypoperfusion and the inflammatory effects of released cytokines . In addition, agents for targeted therapies, including anaplastic lymphoma kinase inhibitors (e.g., brigatinib and ceritinib), cyclin-dependent kinase 4/6 inhibitors (e.g., abemaciclib and ribociclib), BRAF inhibitors (e.g., vemurafenib) and poly(ADP–ribose) polymerase inhibitors (e.g., olaparib), have been found to cause pseudo-AKI by reducing renal tubular secretion of creatinine, resulting in increased serum creatinine . 4.2.2. Chemotherapeutic nephrotoxicity Anticancer agents are increasingly recognized as major causes of acute and chronic kidney injury. Clinically, the manifestations of onconephrology take multiple forms, such as AKI, CKD, electrolyte disturbances, Fanconi’s syndrome, onco-hypertension, proteinuria/nephrotic syndrome, renal cysts, and TMA. Our bibliometric analysis of keywords suggests that AKI appears to receive the most attention. Epidemiologic studies have suggested that the incidence of AKI in cancer patients ranges from 24% (95% CI 17–30%) to 52% (95% CI 34–70%) , highlighting that the incidence of AKI in cancer patients may be closely related to coexisting comorbidities, cancer type and stage, and specific treatment regimens. A nationwide cohort study of 3,120 children, 16,310 adults, and 3,802 hospitalized elderly patients revealed that the overall incidence of AKI was 4.9%, most commonly observed in genitourinary, hematological, and neuro-musculoskeletal cancers and caused by purine analogs, folic acid analogs, and combinations of antineoplastic agents . The causes of AKI in cancer patients are heterogeneous and can be broadly categorized into cancer-related factors and therapy-related factors. Cancer-related factors include urinary tract obstruction, metabolic disorders, glomerular disease, and hemodynamic alterations, whereas therapy-related factors include radiation, immunotherapy, targeted therapy, chemotherapy, hematopoietic stem cell transplantation, tumor lysis syndrome, and TMA. A retrospective study of 67,986 cancer patients revealed that stages 1, 2, and 3 AKI were associated with 18.3% (95% CI 1.145–1.221), 71.0% (95% CI 1.629–1.796), and 100.0% (95% CI 1.910–2.095) increased risk of all-cause mortality, respectively . Therefore, sensitive biomarkers for the timely detection and diagnosis of AKI in cancer patients are urgently needed for early intervention to improve patient outcomes. 4.2.3. Kidney function assessment Accurate assessment of kidney function in cancer patients is important for risk stratification and appropriate dosing of chemotherapeutic agents, antibiotics, opioid analgesics, and other medications. Dose adjustment on the basis of kidney function is often required for a variety of common chemotherapeutic agents, such as alkylating agents, antimicrotubule agents, platinum agents, antimetabolites, immunomodulatory agents, and proteasome inhibitors. Although assessment with radiopharmaceuticals, such as chromium-51-ethylenediaminetetraacetic acid ( 51 Cr-EDTA, not available in the US) or technetium-99m-diethylenetriaminepentaacetic acid ( 99 mTc-DTPA) is considered the gold standard for kidney function, these methods are time-consuming, costly, and labor-intensive, making them impractical to perform on a regular basis. In comparison, estimation of the kidney glomerular filtration rate by serum creatinine and/or cystatin C-based formulas, such as the Cockcroft-Gault equation, the Modification of Diet in Renal Disease equation, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation, and the race-free formula calculated on the basis of both creatinine and cystatin C , appears to be more convenient and practical. Nonetheless, the clinical use of these equations may be biased by various confounding factors in cancer patients, such as anorexia, muscle wasting, and nutritional status. Moreover, validation of these equations in cancer patients is limited by the fact that most involve specific types of cancer in a limited number of patients. In the largest validation cohort of 2471 survivors with various types of cancer, Janowitz et al. validated these formulas against the 51 Cr-EDTA method and reported that the body surface area-adjusted CKD-EPI formula is the most accurate . More recently, cancer-specific formulas have been proposed and developed. For example, Williams and colleagues developed an eGFR formula called ‘CamGFR v2’ based on either standardized or nonstandardized serum creatinine, which outperformed currently available formulas in 7,240 cancer patients . Currently, there is no consensus regarding the best formula for cancer patients, and a prospective clinical trial is needed. In cases where significant discrepancies (differences >10% or >10 mL/min/1.73 m 2 ) were noted when different formulas were used, dose adjustment based on the lower estimated glomerular filtration rate result may be considered, especially for chemotherapeutic agents with dramatic nephrotoxicity with a narrow therapeutic range or in vulnerable patient populations. 4.2.4. CKD in cancer patients and chemotherapy dosing The reported prevalence of CKD in patients with cancer ranges from 12 to 25%, depending on the exact cancer type and associated demographic factors . In an analysis of 5,831 biopsy-proven cancer patients, Ciorcan et al. demonstrated that cancers of the kidney, urinary tract, and pancreas had the highest prevalence of CKD, defined as an estimated glomerular filtration rate <60 mL/min/1.73 m 2 . Although the majority of CKD cases in cancer patients are modest in degree, the presence of diabetes and hypertension also synergistically worsens patient survival. It has been unequivocally demonstrated that the presence of CKD, especially stage 3–5 CKD, is related to a 27% (95% CI 1.00–1.60) increased mortality rate compared with those with estimated glomerular filtration rate >60 mL/min/1.73 m 2 . A critical and practical clinical issue related to CKD is dose adjustment for most chemotherapeutic agents. The prerequisite is accurate measurement of the glomerular filtration rate, which can have a dramatic impact on chemotherapeutic pharmacokinetics and drug toxicity. Studies have shown that the incidences of AKI or AKI transition to CKD are significantly increased in those receiving higher doses of preconditioning chemotherapy . A comprehensive review of the exact dosing regimens of anticancer drugs is beyond the scope of this study, but a brief discussion of the dosage adjustment of chemotherapeutic medications with different renal functions is essential to highlight their nuanced relationship. For example, reducing the dose of cisplatin by 50–70% in patients with a creatinine clearance rate <50 mL/min/1.73 m 2 , and by 75% for capecitabine in patients with a creatinine clearance rate between 30 and 50 mL/min/1.73 m 2 , is recommended . Nonetheless, no dose adjustments are required for vascular endothelial growth factor inhibitors, epithelial growth factor receptor inhibitors, PD-1/PD-L1 antagonists, or cytotoxic T-lymphocyte antigen 4 antagonists. In addition, other strategies in addition to dose adjustment may be used to mitigate nephrotoxicity associated with anticancer agents. For example, teneligliptin, a dipeptidyl peptidase-4 inhibitor, has been shown in vivo to attenuate cisplatin-induced nephrotoxicity and improve renal function by accelerating tubular regeneration and reducing injury and fibrosis . 4.2.5. Glomerular diseases associated with cancer Glomerular disease in cancer refers to the concurrent or metachronous occurrence of glomerular disease, most commonly manifested as proteinuria with or without elevated serum creatinine, in cancer. Although various patterns of glomerular injury have been described in the literature, the prototypical example is cancer-associated membranous nephropathy, which is largely mediated by antibodies to thrombospondin type 1 domain-containing 7A . Epidemiologic studies suggest that the incidence of cancer in membranous nephropathy is 10% (95% CI 6.1–14.6%), with cancer detected before or at the time of membranous nephropathy diagnosis in ∼80% of cases . The mechanistic link is hypothesized to be subepithelial deposition of tumor antigens associated with an enhanced immune response targeting the tumor. In contrast to idiopathic membranous nephropathy, which is histologically characterized by the IgG4 subclass, cancer-associated membranous nephropathy is characterized predominantly by the IgG1 and IgG2 subclasses. Other reported patterns of glomerular disease in solid tumors include pauci-immune glomerulonephritis, renal vasculitis, podocytopathy, C3 glomerulopathy, rapidly progressive glomerulonephritis, IgA nephropathy, and membranoproliferative glomerulonephritis . 4.2.6. Electrolyte disorders Electrolyte abnormalities are common in both solid and hematopoietic malignancies, most commonly as chemotherapy-induced, tumor-related sequelae. Common electrolyte disturbances observed in cancer patients include hyponatremia, hypokalemia/hyperkalemia, hypocalcemia/hypercalcemia, hyperphosphatemia/hypophosphatemia, and hypomagnesemia. Hyponatremia is the most common type of electrolyte disturbance and is most commonly caused by decreased appetite, anorexia, and vomiting, and inappropriate secretion of antidiuretic hormone by many malignancies and chemotherapeutic regimens. Hyperkalemia is mostly caused by tumor lysis syndrome, an oncologic emergency in which tumors lyse spontaneously or secondary to anticancer therapies with sudden release of intracellular ions and metabolites into the systemic circulation. In fact, in addition to hyperkalemia, tumor lysis syndrome may also include hyperphosphatemia, hypocalcemia, hyperuricemia, and AKI. Hypercalcemia in cancer patients is often caused by extensive cytokine-mediated osteolysis or the release of parathyroid hormone-related peptides. The appearance of zoledronic acid as a keyword cluster suggests that attention has been given to the treatment of hypercalcemia, which can often be an onconephrology emergency. Hypophosphatemia is often associated with platinum-based chemotherapy, especially cisplatin. Hypomagnesemia is often associated with the administration of monoclonal antibodies targeting epidermal growth factor receptor, particularly cetuximab, due to increased urinary magnesium loss, as magnesium reabsorption in the distal tubule is partially dependent on the activity of epidermal growth factor receptor proteins located on the basolateral tubular membrane . 4.2.7. Immunotherapy CAR-T cell therapy is a type of immunotherapy that represents a revolutionary treatment for hematological malignancies using genetically engineered host T cells. Although CAR-T cell therapy been shown to induce durable remission with reduced mortality and prolonged survival , it can potentially cause cytokine release syndrome, which can be observed in up to 90% of patients receiving CD19 CAR-T cells for B-cell acute lymphoblastic leukemia and non-Hodgkin B-cell lymphoma . A recent meta-analysis indicated that the pooled estimated incidences of AKI and AKI requiring renal replacement therapy following CAR-T cell therapy were 18.6% (95% CI 14.3–23.8%) and 4.4% (95% CI 2.1–8.9%), respectively . The cumulative incidence at day 100 was 21.7% (95% CI 9.7–33.8%) for grade 1 AKI and 8.7% (95% CI 0.4–17%) for grade 2–3 AKI , suggesting that AKI in patients receiving CAR-T cell therapy is mostly transient, mild in severity, and associated with rapid recovery . Nevertheless, 75 and 67% of patients who experienced AKI after CAR-T cell therapy still had CKD at the 6- and 12-month follow-ups, respectively . Potential mechanisms for AKI after CAR-T cell infusion include capillary leakage with hypovolemia, tumor lysis syndrome, and immune effector cell-associated neurotoxicity syndrome (ICANS). Immune checkpoint inhibitors are monoclonal antibodies that selectively block intrinsic down-regulating receptors of the immune system to activate suppressed T cells to enhance anti-tumor-directed immune responses . As reflected in the keywords trend analysis, immune checkpoint inhibitors have become widely used and considered standard of care in the management of many advanced cancers. AKI has been increasingly recognized as an uncommon but important form of kidney toxicity following the use of immune checkpoint inhibitors. A multi-center analysis revealed that AKI typically manifests at a median of 16 weeks following the initiation of immune checkpoint inhibitors, and kidney recovery was observed in 64.3% of patients at a median of 7 weeks following the onset of AKI . The underlying pathophysiology and corresponding treatments for AKI associated with CAR-T therapies and immune checkpoint inhibitors were different. Kidney injury in CAR-T is mostly related to hemodynamic changes and is typically transient and mild in severity, with only a small percentage of patients requiring continuous renal replacement therapy . Interestingly, studies have shown that preexisting CKD does not appear to have a significant impact on the safety, efficacy, or patient outcomes of CAR-T cell therapy . In comparison, acute tubulointerstitial nephritis is the most common cause of immune checkpoint inhibitor-associated AKI. Manohar et al. reported that prompt withholding of immune checkpoint inhibitors along with steroid use resulted in a complete response in 63% of patients . In addition, rechallenge in 4 patients was successful in 3 patients and the other patient developed recurrent acute allergic interstitial nephritis and fatal pneumonitis . Electrolyte abnormalities are extremely common in patients receiving immunotherapy and most commonly present as hyponatremia, hypokalemia, hypercalcemia/hypocalcemia, hyperphosphatemia/hypophosphatemia, and hypomagnesemia. A meta-analysis of 48 clinical trials revealed a pooled risk ratio of 1.67% (95% CI 0.89–3.12) for electrolyte disorders in cancer patients receiving PD-1 inhibitors . In a recent real-world study of 2,458 patients, Seethapathy’s group reported that 62% of patients had hyponatremia and that 6% had severe hyponatremia with sodium <124 mmol/L . There have been an increasing number of case reports highlighting hypophysitis and secondary adrenal insufficiency as the underlying cause of hyponatremia in patients receiving immune checkpoint inhibitors . Another common electrolyte abnormality in cancer patients receiving immunotherapy is hypercalcemia, which may be caused by immune endocrinopathies, parathyroid hormone-related peptide release, cancer hyperprogression, and sarcoidosis-like granulomas . Notable side effects associated with CAR-T cell therapy include cytokine release syndrome, tumor lysis syndrome, and ICANS. Patients with cytokine release syndrome often have nonspecific symptoms of fever, arthralgia, malaise and fatigue, anorexia, and tachycardia. Therapeutic approaches for cytokine release syndrome include supportive care and pharmacological intervention with the monoclonal anti-interleukin-6 antibody tocilizumab. There are also agents currently under investigation for the prophylaxis of cytokine release syndrome, such as tocilizumab, anakinra, teclistamab, and duvelisib . Tumor lysis treatment is an oncological emergency resulting from the release of intracellular electrolytes and nucleic acids from malignant cells. Treatment and prophylaxis for tumor lysis syndrome are similar and include aggressive intravenous hydration, the use of xanthine oxidase inhibitors and rasburicase, and medical management of associated electrolyte abnormalities. ICANS is a constellation of neuropsychiatric symptoms, including headache, aphasia, seizures, or decreased consciousness, that are observed in 20–70% of CAR-T cell therapy recipients. Risk factors for the occurrence of ICANS include cytokine release syndrome, high tumor burden, advanced age, and a strong inflammatory response . Antiepileptic medications, steroids, and tocilizumab have been the most common treatments for ICANS, especially for those with concurrent cytokine release syndrome. Preliminary studies investigating the interleukin-1 receptor antagonist anakinra for the prophylaxis of ICANS have reported encouraging results . 4.2.8. Underlying pathophysiology The investigation of the molecular mechanisms underlying nephrotoxicity caused by various anticancer drugs or treatments is particularly important, as it may shed light on prevention and intervention strategies. Moreover, the exact underlying pathways for anticancer drug-induced renal injury may differ depending on the exact type of agent. For example, cisplatin, a classic and highly effective chemotherapeutic agent for a variety of malignancies, is well known for its potential risk of inducing nephrotoxicity. Various molecular mechanisms have been reported in the pathogenesis of cisplatin-induced nephrotoxicity, including tubular cell apoptosis under oxidative stress, renal interstitial inflammatory cell infiltration leading to AKI, and extensive production of proinflammatory cytokines . Several lines of evidence also suggest that various signaling pathways are involved, including the toll-like receptor pathway, the NF-κB pathway, and the poly-ADP–ribose polymerase-1 pathway . Currently, the mechanism of action for nephrotoxicity resulting from the majority of traditional chemotherapeutic agents has been described and is largely related to the side effects of the desired anticancer properties of these agents. Specifically, antimetabolites, such as gemcitabine have a vasoconstrictive effect on afferent renal arteries, resulting in a decreased glomerular filtration rate and AKI . However, the mechanisms of action of targeted therapy-induced nephrotoxicity may extend beyond pharmacological action. For example, epidermal growth factor receptor inhibitors, such as cetuximab, cause electrolyte disturbances that can be explained by inhibited signaling at the distal convoluted tubule that regulates transepithelial magnesium transport. However, epidermal growth factor receptor inhibitors may also cause AKI, although the mechanisms of action remain unclear. Similarly, the exact mechanisms by which Her-2 inhibitors, anaplastic lymphoma kinase inhibitors, and BRAF inhibitors cause AKI remain elusive and warrant additional investigations.
Types of malignancies and treatment modalities involved Among the top 10 most cited documents and references, 6/10 and 6/10 were related to multiple myeloma, indicating that multiple myeloma-related renal insufficiency is still the most important and hottest topic in onconephrology. Approximately half of multiple myeloma patients are estimated to develop renal dysfunction during the course of treatment . More importantly, numerous studies have associated renal impairment with decreased overall survival and premature mortality in multiple myeloma patients . The main mechanism of renal injury in multiple myeloma is the overproduction of free light chains, which are nephrotoxic because they promote proximal tubule apoptosis and induce inflammation with subsequent interstitial fibrosis . In addition, other conditions associated with multiple myeloma, such as amyloidosis, monoclonal immunoglobulin deposition disease, light-chain proximal tubulopathy, and cryoglobulinaemia, may also cause renal impairment. Notably, other multiple myeloma factors, such as dehydration, hypercalcemia, and tumor lysis syndrome, and treatment-related factors, such as hematopoietic stem cell transplantation, are also essential etiologies. Our bibliometric results also suggest that solid tumors, represented by the cluster ‘hepatocellular carcinoma’, are also involved. In fact, many treatments for solid tumors, such as targeted therapy and immunotherapy, have also been associated with a variety of renal complications . For example, immune checkpoint inhibitors, such as programmed cell death (PD)-1/PD-L1, have been found to cause hypophosphatemia, proteinuria, acute interstitial nephritis, and acute tubular necrosis . Our study revealed that in addition to conventional cytotoxic chemotherapies, emerging novel cancer therapies, such as immunotherapies (immune checkpoint inhibitors and chimeric antigen receptor-T [CAR-T] cell therapy) and targeted therapy, also cause kidney injury. Kidney injury in CAR-T most commonly occurs as AKI, which is related to hypoperfusion and the inflammatory effects of released cytokines . In addition, agents for targeted therapies, including anaplastic lymphoma kinase inhibitors (e.g., brigatinib and ceritinib), cyclin-dependent kinase 4/6 inhibitors (e.g., abemaciclib and ribociclib), BRAF inhibitors (e.g., vemurafenib) and poly(ADP–ribose) polymerase inhibitors (e.g., olaparib), have been found to cause pseudo-AKI by reducing renal tubular secretion of creatinine, resulting in increased serum creatinine .
Chemotherapeutic nephrotoxicity Anticancer agents are increasingly recognized as major causes of acute and chronic kidney injury. Clinically, the manifestations of onconephrology take multiple forms, such as AKI, CKD, electrolyte disturbances, Fanconi’s syndrome, onco-hypertension, proteinuria/nephrotic syndrome, renal cysts, and TMA. Our bibliometric analysis of keywords suggests that AKI appears to receive the most attention. Epidemiologic studies have suggested that the incidence of AKI in cancer patients ranges from 24% (95% CI 17–30%) to 52% (95% CI 34–70%) , highlighting that the incidence of AKI in cancer patients may be closely related to coexisting comorbidities, cancer type and stage, and specific treatment regimens. A nationwide cohort study of 3,120 children, 16,310 adults, and 3,802 hospitalized elderly patients revealed that the overall incidence of AKI was 4.9%, most commonly observed in genitourinary, hematological, and neuro-musculoskeletal cancers and caused by purine analogs, folic acid analogs, and combinations of antineoplastic agents . The causes of AKI in cancer patients are heterogeneous and can be broadly categorized into cancer-related factors and therapy-related factors. Cancer-related factors include urinary tract obstruction, metabolic disorders, glomerular disease, and hemodynamic alterations, whereas therapy-related factors include radiation, immunotherapy, targeted therapy, chemotherapy, hematopoietic stem cell transplantation, tumor lysis syndrome, and TMA. A retrospective study of 67,986 cancer patients revealed that stages 1, 2, and 3 AKI were associated with 18.3% (95% CI 1.145–1.221), 71.0% (95% CI 1.629–1.796), and 100.0% (95% CI 1.910–2.095) increased risk of all-cause mortality, respectively . Therefore, sensitive biomarkers for the timely detection and diagnosis of AKI in cancer patients are urgently needed for early intervention to improve patient outcomes.
Kidney function assessment Accurate assessment of kidney function in cancer patients is important for risk stratification and appropriate dosing of chemotherapeutic agents, antibiotics, opioid analgesics, and other medications. Dose adjustment on the basis of kidney function is often required for a variety of common chemotherapeutic agents, such as alkylating agents, antimicrotubule agents, platinum agents, antimetabolites, immunomodulatory agents, and proteasome inhibitors. Although assessment with radiopharmaceuticals, such as chromium-51-ethylenediaminetetraacetic acid ( 51 Cr-EDTA, not available in the US) or technetium-99m-diethylenetriaminepentaacetic acid ( 99 mTc-DTPA) is considered the gold standard for kidney function, these methods are time-consuming, costly, and labor-intensive, making them impractical to perform on a regular basis. In comparison, estimation of the kidney glomerular filtration rate by serum creatinine and/or cystatin C-based formulas, such as the Cockcroft-Gault equation, the Modification of Diet in Renal Disease equation, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation, and the race-free formula calculated on the basis of both creatinine and cystatin C , appears to be more convenient and practical. Nonetheless, the clinical use of these equations may be biased by various confounding factors in cancer patients, such as anorexia, muscle wasting, and nutritional status. Moreover, validation of these equations in cancer patients is limited by the fact that most involve specific types of cancer in a limited number of patients. In the largest validation cohort of 2471 survivors with various types of cancer, Janowitz et al. validated these formulas against the 51 Cr-EDTA method and reported that the body surface area-adjusted CKD-EPI formula is the most accurate . More recently, cancer-specific formulas have been proposed and developed. For example, Williams and colleagues developed an eGFR formula called ‘CamGFR v2’ based on either standardized or nonstandardized serum creatinine, which outperformed currently available formulas in 7,240 cancer patients . Currently, there is no consensus regarding the best formula for cancer patients, and a prospective clinical trial is needed. In cases where significant discrepancies (differences >10% or >10 mL/min/1.73 m 2 ) were noted when different formulas were used, dose adjustment based on the lower estimated glomerular filtration rate result may be considered, especially for chemotherapeutic agents with dramatic nephrotoxicity with a narrow therapeutic range or in vulnerable patient populations.
CKD in cancer patients and chemotherapy dosing The reported prevalence of CKD in patients with cancer ranges from 12 to 25%, depending on the exact cancer type and associated demographic factors . In an analysis of 5,831 biopsy-proven cancer patients, Ciorcan et al. demonstrated that cancers of the kidney, urinary tract, and pancreas had the highest prevalence of CKD, defined as an estimated glomerular filtration rate <60 mL/min/1.73 m 2 . Although the majority of CKD cases in cancer patients are modest in degree, the presence of diabetes and hypertension also synergistically worsens patient survival. It has been unequivocally demonstrated that the presence of CKD, especially stage 3–5 CKD, is related to a 27% (95% CI 1.00–1.60) increased mortality rate compared with those with estimated glomerular filtration rate >60 mL/min/1.73 m 2 . A critical and practical clinical issue related to CKD is dose adjustment for most chemotherapeutic agents. The prerequisite is accurate measurement of the glomerular filtration rate, which can have a dramatic impact on chemotherapeutic pharmacokinetics and drug toxicity. Studies have shown that the incidences of AKI or AKI transition to CKD are significantly increased in those receiving higher doses of preconditioning chemotherapy . A comprehensive review of the exact dosing regimens of anticancer drugs is beyond the scope of this study, but a brief discussion of the dosage adjustment of chemotherapeutic medications with different renal functions is essential to highlight their nuanced relationship. For example, reducing the dose of cisplatin by 50–70% in patients with a creatinine clearance rate <50 mL/min/1.73 m 2 , and by 75% for capecitabine in patients with a creatinine clearance rate between 30 and 50 mL/min/1.73 m 2 , is recommended . Nonetheless, no dose adjustments are required for vascular endothelial growth factor inhibitors, epithelial growth factor receptor inhibitors, PD-1/PD-L1 antagonists, or cytotoxic T-lymphocyte antigen 4 antagonists. In addition, other strategies in addition to dose adjustment may be used to mitigate nephrotoxicity associated with anticancer agents. For example, teneligliptin, a dipeptidyl peptidase-4 inhibitor, has been shown in vivo to attenuate cisplatin-induced nephrotoxicity and improve renal function by accelerating tubular regeneration and reducing injury and fibrosis .
Glomerular diseases associated with cancer Glomerular disease in cancer refers to the concurrent or metachronous occurrence of glomerular disease, most commonly manifested as proteinuria with or without elevated serum creatinine, in cancer. Although various patterns of glomerular injury have been described in the literature, the prototypical example is cancer-associated membranous nephropathy, which is largely mediated by antibodies to thrombospondin type 1 domain-containing 7A . Epidemiologic studies suggest that the incidence of cancer in membranous nephropathy is 10% (95% CI 6.1–14.6%), with cancer detected before or at the time of membranous nephropathy diagnosis in ∼80% of cases . The mechanistic link is hypothesized to be subepithelial deposition of tumor antigens associated with an enhanced immune response targeting the tumor. In contrast to idiopathic membranous nephropathy, which is histologically characterized by the IgG4 subclass, cancer-associated membranous nephropathy is characterized predominantly by the IgG1 and IgG2 subclasses. Other reported patterns of glomerular disease in solid tumors include pauci-immune glomerulonephritis, renal vasculitis, podocytopathy, C3 glomerulopathy, rapidly progressive glomerulonephritis, IgA nephropathy, and membranoproliferative glomerulonephritis .
Electrolyte disorders Electrolyte abnormalities are common in both solid and hematopoietic malignancies, most commonly as chemotherapy-induced, tumor-related sequelae. Common electrolyte disturbances observed in cancer patients include hyponatremia, hypokalemia/hyperkalemia, hypocalcemia/hypercalcemia, hyperphosphatemia/hypophosphatemia, and hypomagnesemia. Hyponatremia is the most common type of electrolyte disturbance and is most commonly caused by decreased appetite, anorexia, and vomiting, and inappropriate secretion of antidiuretic hormone by many malignancies and chemotherapeutic regimens. Hyperkalemia is mostly caused by tumor lysis syndrome, an oncologic emergency in which tumors lyse spontaneously or secondary to anticancer therapies with sudden release of intracellular ions and metabolites into the systemic circulation. In fact, in addition to hyperkalemia, tumor lysis syndrome may also include hyperphosphatemia, hypocalcemia, hyperuricemia, and AKI. Hypercalcemia in cancer patients is often caused by extensive cytokine-mediated osteolysis or the release of parathyroid hormone-related peptides. The appearance of zoledronic acid as a keyword cluster suggests that attention has been given to the treatment of hypercalcemia, which can often be an onconephrology emergency. Hypophosphatemia is often associated with platinum-based chemotherapy, especially cisplatin. Hypomagnesemia is often associated with the administration of monoclonal antibodies targeting epidermal growth factor receptor, particularly cetuximab, due to increased urinary magnesium loss, as magnesium reabsorption in the distal tubule is partially dependent on the activity of epidermal growth factor receptor proteins located on the basolateral tubular membrane .
Immunotherapy CAR-T cell therapy is a type of immunotherapy that represents a revolutionary treatment for hematological malignancies using genetically engineered host T cells. Although CAR-T cell therapy been shown to induce durable remission with reduced mortality and prolonged survival , it can potentially cause cytokine release syndrome, which can be observed in up to 90% of patients receiving CD19 CAR-T cells for B-cell acute lymphoblastic leukemia and non-Hodgkin B-cell lymphoma . A recent meta-analysis indicated that the pooled estimated incidences of AKI and AKI requiring renal replacement therapy following CAR-T cell therapy were 18.6% (95% CI 14.3–23.8%) and 4.4% (95% CI 2.1–8.9%), respectively . The cumulative incidence at day 100 was 21.7% (95% CI 9.7–33.8%) for grade 1 AKI and 8.7% (95% CI 0.4–17%) for grade 2–3 AKI , suggesting that AKI in patients receiving CAR-T cell therapy is mostly transient, mild in severity, and associated with rapid recovery . Nevertheless, 75 and 67% of patients who experienced AKI after CAR-T cell therapy still had CKD at the 6- and 12-month follow-ups, respectively . Potential mechanisms for AKI after CAR-T cell infusion include capillary leakage with hypovolemia, tumor lysis syndrome, and immune effector cell-associated neurotoxicity syndrome (ICANS). Immune checkpoint inhibitors are monoclonal antibodies that selectively block intrinsic down-regulating receptors of the immune system to activate suppressed T cells to enhance anti-tumor-directed immune responses . As reflected in the keywords trend analysis, immune checkpoint inhibitors have become widely used and considered standard of care in the management of many advanced cancers. AKI has been increasingly recognized as an uncommon but important form of kidney toxicity following the use of immune checkpoint inhibitors. A multi-center analysis revealed that AKI typically manifests at a median of 16 weeks following the initiation of immune checkpoint inhibitors, and kidney recovery was observed in 64.3% of patients at a median of 7 weeks following the onset of AKI . The underlying pathophysiology and corresponding treatments for AKI associated with CAR-T therapies and immune checkpoint inhibitors were different. Kidney injury in CAR-T is mostly related to hemodynamic changes and is typically transient and mild in severity, with only a small percentage of patients requiring continuous renal replacement therapy . Interestingly, studies have shown that preexisting CKD does not appear to have a significant impact on the safety, efficacy, or patient outcomes of CAR-T cell therapy . In comparison, acute tubulointerstitial nephritis is the most common cause of immune checkpoint inhibitor-associated AKI. Manohar et al. reported that prompt withholding of immune checkpoint inhibitors along with steroid use resulted in a complete response in 63% of patients . In addition, rechallenge in 4 patients was successful in 3 patients and the other patient developed recurrent acute allergic interstitial nephritis and fatal pneumonitis . Electrolyte abnormalities are extremely common in patients receiving immunotherapy and most commonly present as hyponatremia, hypokalemia, hypercalcemia/hypocalcemia, hyperphosphatemia/hypophosphatemia, and hypomagnesemia. A meta-analysis of 48 clinical trials revealed a pooled risk ratio of 1.67% (95% CI 0.89–3.12) for electrolyte disorders in cancer patients receiving PD-1 inhibitors . In a recent real-world study of 2,458 patients, Seethapathy’s group reported that 62% of patients had hyponatremia and that 6% had severe hyponatremia with sodium <124 mmol/L . There have been an increasing number of case reports highlighting hypophysitis and secondary adrenal insufficiency as the underlying cause of hyponatremia in patients receiving immune checkpoint inhibitors . Another common electrolyte abnormality in cancer patients receiving immunotherapy is hypercalcemia, which may be caused by immune endocrinopathies, parathyroid hormone-related peptide release, cancer hyperprogression, and sarcoidosis-like granulomas . Notable side effects associated with CAR-T cell therapy include cytokine release syndrome, tumor lysis syndrome, and ICANS. Patients with cytokine release syndrome often have nonspecific symptoms of fever, arthralgia, malaise and fatigue, anorexia, and tachycardia. Therapeutic approaches for cytokine release syndrome include supportive care and pharmacological intervention with the monoclonal anti-interleukin-6 antibody tocilizumab. There are also agents currently under investigation for the prophylaxis of cytokine release syndrome, such as tocilizumab, anakinra, teclistamab, and duvelisib . Tumor lysis treatment is an oncological emergency resulting from the release of intracellular electrolytes and nucleic acids from malignant cells. Treatment and prophylaxis for tumor lysis syndrome are similar and include aggressive intravenous hydration, the use of xanthine oxidase inhibitors and rasburicase, and medical management of associated electrolyte abnormalities. ICANS is a constellation of neuropsychiatric symptoms, including headache, aphasia, seizures, or decreased consciousness, that are observed in 20–70% of CAR-T cell therapy recipients. Risk factors for the occurrence of ICANS include cytokine release syndrome, high tumor burden, advanced age, and a strong inflammatory response . Antiepileptic medications, steroids, and tocilizumab have been the most common treatments for ICANS, especially for those with concurrent cytokine release syndrome. Preliminary studies investigating the interleukin-1 receptor antagonist anakinra for the prophylaxis of ICANS have reported encouraging results .
Underlying pathophysiology The investigation of the molecular mechanisms underlying nephrotoxicity caused by various anticancer drugs or treatments is particularly important, as it may shed light on prevention and intervention strategies. Moreover, the exact underlying pathways for anticancer drug-induced renal injury may differ depending on the exact type of agent. For example, cisplatin, a classic and highly effective chemotherapeutic agent for a variety of malignancies, is well known for its potential risk of inducing nephrotoxicity. Various molecular mechanisms have been reported in the pathogenesis of cisplatin-induced nephrotoxicity, including tubular cell apoptosis under oxidative stress, renal interstitial inflammatory cell infiltration leading to AKI, and extensive production of proinflammatory cytokines . Several lines of evidence also suggest that various signaling pathways are involved, including the toll-like receptor pathway, the NF-κB pathway, and the poly-ADP–ribose polymerase-1 pathway . Currently, the mechanism of action for nephrotoxicity resulting from the majority of traditional chemotherapeutic agents has been described and is largely related to the side effects of the desired anticancer properties of these agents. Specifically, antimetabolites, such as gemcitabine have a vasoconstrictive effect on afferent renal arteries, resulting in a decreased glomerular filtration rate and AKI . However, the mechanisms of action of targeted therapy-induced nephrotoxicity may extend beyond pharmacological action. For example, epidermal growth factor receptor inhibitors, such as cetuximab, cause electrolyte disturbances that can be explained by inhibited signaling at the distal convoluted tubule that regulates transepithelial magnesium transport. However, epidermal growth factor receptor inhibitors may also cause AKI, although the mechanisms of action remain unclear. Similarly, the exact mechanisms by which Her-2 inhibitors, anaplastic lymphoma kinase inhibitors, and BRAF inhibitors cause AKI remain elusive and warrant additional investigations.
Future research directions 4.3.1. TMA TMA is a pathological term that is clinically characterized by microangiopathic hemolytic anemia and thrombocytopenia and is associated with elevated lactate dehydrogenase and end-organ failure . The emergency of TMA as a trend topic is consistent with increasing reports suggesting that TMA represents ∼5.4% of the findings in kidney biopsies from cancer patients . Unlike primary TMA–which is mostly caused by genetic mutations–cancers and related therapies, especially hematopoietic stem cell transplantation, are important causes of secondary TMA. On biopsy, the renal TMA is characterized by extensive formation of fibrin thrombi in the capillary loops and arterioles, intimal on ionization, fragmented red blood cells, and mesangiolysis. Common solid tumors that can induce TMA include gastrointestinal, lung, genitourinary, and hepatobiliary cancers. In a series of 168 cases of cancer-related TMA, Lechner et al. reported that lymphoma accounted for ∼8.3% of all cases . Cancer therapy-related TMA is also relatively common and can be observed in those treated with conventional chemotherapy, targeted therapy, and immunotherapy, especially in those with underlying genetic defects in the alternative complement cascade. Commonly used conventional chemotherapy drugs include mitomycin-C, gemcitabine, bleomycin, and platinum-based agents, whereas epidermal growth factor monoclonal antibody agents are the most common targeted therapy used to induce TMA . Compared with conventional chemotherapy-induced TMA, which is often fulminant, systemic, and lethal, TMA caused by targeted therapy is generally confined to the kidney and has more favorable outcomes . However, discontinuation of stimulating agents may not lead to renal recovery, and additional therapeutic maneuvers, such as plasma exchange and rituximab administration, may have variable clinical efficacy . However, currently, all available evidence comes from case reports or small case series. The American Society for Apheresis guidelines consider therapeutic plasma exchange to be ineffective for treating chemotherapy-related TMA . In the largest cohort of gemcitabine-associated TMA, Daviet et al. reported that 54.7% of patients eventually died and that treatment with therapeutic plasma exchange did not improve outcomes compared with glucocorticoids and was associated with more adverse events . Eculizumab, a monoclonal antibody targeting terminal complement C5, has been shown in several case reports to promote renal recovery in cancer therapy-associated TMA . In a retrospective analysis of 12 cases of gemcitabine-induced TMA, Grall et al. showed that 17 and 67% achieved complete and partial recovery of renal function, respectively, after treatment with eculizumab . These increasing results encourage the conduct of prospective studies in the future to conclusively determine the efficacy of eculizumab in the treatment of cancer-related TMA. 4.3.2. Clinical trials Our trend topic analysis revealed that clinical trials are likely to be a hot topic in onconephrology in the future. We believe this raises 2 considerations. First, this may imply the recruitment of patients with impaired renal function into cancer clinical trials. Clearly, obtaining accurate information on the safety and effective dosing of chemotherapeutic agents in cancer patients with CKD is critical for formulating appropriate treatment regimens. Unfortunately, the majority of cancer clinical trials have enrolled patients with normal or mildly impaired renal function, thus underrepresenting patients with severely impaired renal function. For example, in a recent meta-analysis of 11,066 participants from 32 clinical trials receiving combination therapy with vascular endothelial growth factor pathway inhibitors and immune checkpoint inhibitors, Elyan et al. reported that all trials excluded patients with advanced CKD, and few trials included people with proteinuria . Butrovich et al. reported that pharmacokinetic analyses in patients with CKD stages 4–5 and hemodialysis were performed for only 29 and 6% of the 55 drugs approved between 2015 and 2019, respectively . The reasons for the exclusion of this specific patient population are complex, including sponsor concerns, safety concerns, and the lack of a robust nephrology clinical trial infrastructure . Another consideration is the design and application of specific clinical trials for specific onconephrology conditions or issues. Eculizumab or other novel treatments for cancer-associated TMA appear to be good examples of such clinical trials. Another area of interest is the design of prospective and controlled trials to evaluate the efficacy of renoprotective agents during cancer treatment. Successful trials of renoprotective agents would not only provide effective chemotherapy with minimized risk of renal toxicity but also offer chemotherapeutic options for cancer patients with preexisting renal impairment. For example, sirtuins, which are NAD+-dependent deacetylases with important antioxidant activity, have been demonstrated in animal studies to exert cytoprotective and renoprotective effects in various forms of nephrotoxicity . Thus, clinical trials evaluating their ability to reduce nephrotoxicity during cancer treatment are very meaningful.
TMA TMA is a pathological term that is clinically characterized by microangiopathic hemolytic anemia and thrombocytopenia and is associated with elevated lactate dehydrogenase and end-organ failure . The emergency of TMA as a trend topic is consistent with increasing reports suggesting that TMA represents ∼5.4% of the findings in kidney biopsies from cancer patients . Unlike primary TMA–which is mostly caused by genetic mutations–cancers and related therapies, especially hematopoietic stem cell transplantation, are important causes of secondary TMA. On biopsy, the renal TMA is characterized by extensive formation of fibrin thrombi in the capillary loops and arterioles, intimal on ionization, fragmented red blood cells, and mesangiolysis. Common solid tumors that can induce TMA include gastrointestinal, lung, genitourinary, and hepatobiliary cancers. In a series of 168 cases of cancer-related TMA, Lechner et al. reported that lymphoma accounted for ∼8.3% of all cases . Cancer therapy-related TMA is also relatively common and can be observed in those treated with conventional chemotherapy, targeted therapy, and immunotherapy, especially in those with underlying genetic defects in the alternative complement cascade. Commonly used conventional chemotherapy drugs include mitomycin-C, gemcitabine, bleomycin, and platinum-based agents, whereas epidermal growth factor monoclonal antibody agents are the most common targeted therapy used to induce TMA . Compared with conventional chemotherapy-induced TMA, which is often fulminant, systemic, and lethal, TMA caused by targeted therapy is generally confined to the kidney and has more favorable outcomes . However, discontinuation of stimulating agents may not lead to renal recovery, and additional therapeutic maneuvers, such as plasma exchange and rituximab administration, may have variable clinical efficacy . However, currently, all available evidence comes from case reports or small case series. The American Society for Apheresis guidelines consider therapeutic plasma exchange to be ineffective for treating chemotherapy-related TMA . In the largest cohort of gemcitabine-associated TMA, Daviet et al. reported that 54.7% of patients eventually died and that treatment with therapeutic plasma exchange did not improve outcomes compared with glucocorticoids and was associated with more adverse events . Eculizumab, a monoclonal antibody targeting terminal complement C5, has been shown in several case reports to promote renal recovery in cancer therapy-associated TMA . In a retrospective analysis of 12 cases of gemcitabine-induced TMA, Grall et al. showed that 17 and 67% achieved complete and partial recovery of renal function, respectively, after treatment with eculizumab . These increasing results encourage the conduct of prospective studies in the future to conclusively determine the efficacy of eculizumab in the treatment of cancer-related TMA.
Clinical trials Our trend topic analysis revealed that clinical trials are likely to be a hot topic in onconephrology in the future. We believe this raises 2 considerations. First, this may imply the recruitment of patients with impaired renal function into cancer clinical trials. Clearly, obtaining accurate information on the safety and effective dosing of chemotherapeutic agents in cancer patients with CKD is critical for formulating appropriate treatment regimens. Unfortunately, the majority of cancer clinical trials have enrolled patients with normal or mildly impaired renal function, thus underrepresenting patients with severely impaired renal function. For example, in a recent meta-analysis of 11,066 participants from 32 clinical trials receiving combination therapy with vascular endothelial growth factor pathway inhibitors and immune checkpoint inhibitors, Elyan et al. reported that all trials excluded patients with advanced CKD, and few trials included people with proteinuria . Butrovich et al. reported that pharmacokinetic analyses in patients with CKD stages 4–5 and hemodialysis were performed for only 29 and 6% of the 55 drugs approved between 2015 and 2019, respectively . The reasons for the exclusion of this specific patient population are complex, including sponsor concerns, safety concerns, and the lack of a robust nephrology clinical trial infrastructure . Another consideration is the design and application of specific clinical trials for specific onconephrology conditions or issues. Eculizumab or other novel treatments for cancer-associated TMA appear to be good examples of such clinical trials. Another area of interest is the design of prospective and controlled trials to evaluate the efficacy of renoprotective agents during cancer treatment. Successful trials of renoprotective agents would not only provide effective chemotherapy with minimized risk of renal toxicity but also offer chemotherapeutic options for cancer patients with preexisting renal impairment. For example, sirtuins, which are NAD+-dependent deacetylases with important antioxidant activity, have been demonstrated in animal studies to exert cytoprotective and renoprotective effects in various forms of nephrotoxicity . Thus, clinical trials evaluating their ability to reduce nephrotoxicity during cancer treatment are very meaningful.
Study limitations The present study has several limitations that should be acknowledged. First, we only used the SCIE-WOS database for bibliometric analysis, which may have missed important studies indexed in other databases, thus limiting the comprehensiveness of the findings. Previous studies have indicated that the Scopus database has more citations per article and better coverage than the SCIE-WOS database . Extending the search for articles to other databases, such as Scopus and Google Scholar, may provide a more comprehensive overview of progress in the field of onconephrology. Second, the inclusion of English language-only documents obviously discriminated against countries, institutions, or authors publishing research in other languages. Third, we did not evaluate literature quality, which is often done in meta-analyses. Finally, the concept of onconephrology is constantly evolving, and some scholars have included an increased risk of cancer in kidney transplant patients in the field of onconephrology , a topic that was not covered in the present study.
Conclusions In summary, this bibliometric study revealed that the field of onconephrology is gradually gaining momentum in terms of both scientific output and the number of citations. The United States, University of Texas MD Anderson Cancer Center, Meletios A. Dimopoulos, and Nephrology Dialysis Transplantation were the most productive country, institution, author, and journal, respectively. Current research hotspots in onconephrology include chemotherapeutic nephrotoxicity, assessment of kidney function in cancer patients, CKD in cancer patients and chemotherapy dosing, glomerular diseases in cancer patients, electrolyte disorders, immunotherapy, and basic studies of the underlying molecular mechanisms of nephrotoxicity. Future directions in this field may include TMA and clinical trials.
Supplementary Material.doc
|
Meeting report from the 2020 Annual (virtual) Meeting of the American Society of Clinical Oncology | b3206b41-515c-4599-858a-d79226e15467 | 7405821 | Gynaecology[mh] | Introduction The 2020 56th annual meeting of the American Society of Clinical Oncology (ASCO) was held virtually for the first time due to the COVID-19 pandemic. Though you may not have missed the $10 muffins, we missed the opportunity to network with our colleagues. Fortunately, this year's scientific content was second to none. The meeting was led by ASCO president Howard Burris III, medical oncologist at Sarah Cannon Research Institute and Tennessee Oncology and the theme of the meeting was “Unite and Conquer; Accelerating Progress Together”. Dr. Burris recognized the impact of the pandemic and importance of continuing to provide high-level care to cancer patients, who are particularly vulnerable to COVID 19. He addressed the goal to reduce the global burden of cancer through drug development, clinical trials, and the use of technology to overcome barriers and improve care delivery and research. One major effort is to mandate that insurance carriers, including Medicare, cover the routine costs associated with clinical trial participation. This initiative would greatly enhance access and enrollment to trials for all patients and accelerate progress in cancer care. The meeting also presented many promising advances in gynecologic oncology. The focus was multifaceted and included targeted therapies, PARP inhibitors, immunotherapy, and surgery with developments in multiple disease sites. Below we present some of the highlights from this year's meeting.
Ovarian cancer 2.1 The revival of secondary cytoreductive surgery This year, two prospective trials evaluating secondary cytoreductive surgery (SCS) were presented; provides an overview of both trials. Du Bois and colleagues, after validation of their selection criteria and demonstrating a benefit in progression free survival (PFS), reported the final results of the phase 3, randomized DESKTOP III trial (#6000) evaluating the impact of SCS in recurrent ovarian cancer . The investigators reported an overall survival (OS) benefit in the surgery arm with a median survival of 53.7 months compared to 46.2 months in those who did not undergo surgery. The true benefit was seen in women who underwent complete resection at the time of surgery (60.7 months versus 46.2 months). It is important to note that patients who were not able to undergo complete resection had a worse outcome (median survival 28.8 months) when compared to the no surgery arm. In a separate randomized phase 3 study from China, SOC 1/SGOG-OV2 (#6001), SCS was also evaluated. In this study, there were designated selection criteria based on the iModel score in conjunction with PET-CT scan imaging to determine complete resection . Zang and colleagues reported that SCS was associated with a benefit in PFS of 17.4 versus 11.9 months, though OS outcomes are still maturing. Similarly to DESKTOP III, those with residual disease did worse than chemotherapy alone. Therefore, though these results differed from GOG 213 which reported no difference in OS outcomes with SCS, both groups concluded that SCS is of benefit when performed in carefully selected patients, and that SCS should be done at centers of excellence/experience in which the opportunity to resect disease completely is optimal. Lastly, another study of SCS presented was a randomized phase 2 trial of SCS with or without carboplatin hyperthermic intraperitoneal chemotherapy (HIPEC) in women with recurrent platinum sensitive ovarian cancer (#6016) . The study randomized 98 patients and reported 0% perioperative mortality and comparable outcomes to those undergoing SCS only. Though preliminary, PFS and OS rates were not significantly different and did not demonstrate that HIPEC was superior to SCS alone. 2.2 Further insight into the activity of PARP inhibitors (PARPi) and PARPi combinations This year, SOLO2 (#6002), a randomized Phase 3 trial of olaparib maintenance therapy in women with relapsed platinum-sensitive BRCA-mutated ovarian cancer following response to platinum-based therapy, became the first report of mature OS data from a Phase 3 setting . The study enrolled 295 patients and previously reported a 13.6 month improvement in the primary endpoint of PFS in 2017. In this final analysis of SOLO2, a median improvement in OS of 12.9 months was observed (38.8 versus 51.7 months), with a hazard ratio of 0.74 (95% CI 0.54–1.00; p = 0.0537). While this finding was just shy of the threshold for statistical significance, the magnitude of the observed OS benefit was striking. Additionally, it is important to note that this final analysis included patients who received subsequent PARPi therapy, including 38% of patients in the control arm and 10% of those in the olaparib arm. When adjusted for subsequent PARPi in the placebo group, the observed improvement in median OS was 16.3 months (35.4 versus 51.7 months), with a hazard ratio of 0.56 (95% CI 0.35–0.97). These results support the use of maintenance PARPi in patients with BRCA-mutated platinum sensitive recurrent ovarian cancer who are PARPi naïve. The primary analysis of NRG-GY004, a Phase 3 trial comparing the combination of cediranib and olaparib or olaparib monotherapy to platinum-based chemotherapy in relapsed platinum-sensitive ovarian cancer was also presented (#6003) . This study asked whether a non-platinum alternative could improve PFS over platinum-based chemotherapy in platinum-sensitive ovarian cancer. The combination cediranib and olaparib did not meet the primary endpoint of improving PFS compared to chemotherapy, with a median PFS of 10.4 months for the combination and 10.3 months for chemotherapy (hazard ratio 0.856, 95% CI 0.663–1.105, p = 0.077), although the observed activity was comparable for both PFS and objective response rates. Due to a hierarchical statistical design, the activity of olaparib monotherapy was not formally compared to chemotherapy, but a median PFS of 8.2 months was observed. Side effects with the combination of cediranib and olaparib were significant; while hematologic adverse events were higher with chemotherapy, rates of non-hematologic adverse events were higher with cediranib/olaparib, and the discontinuation rate due to adverse events was 21%. Pre-specified subgroup analyses were notable for the high response rates and significant activity seen with both olaparib monotherapy and cediranib/olaparib combination in patients with a germline BRCA mutation. Further investigation of the cediranib/olaparib combination is ongoing in NRG-GY005 and ICON9, and these studies will give us additional insight into the activity of this combination in these settings. Two Phase 2 trials reported this ASCO also provided additional insight into PARPi activity as monotherapy or in combination. NSGO-AVANOVA2 (#6012) was a randomized Phase 2 study compared combination niraparib and bevacizumab to niraparib monotherapy in relapsed platinum-sensitive ovarian cancer . In an updated analysis, there was continued improvement in PFS, with a hazard ratio of 0.34 (95% CI 0.21–0.54, p < 0.0001). With 52% event maturity, the hazard ratio for OS was non-significant at 0.75 (95% CI 0.44–1.28, p = 0.30). While these results support increased activity of anti-angiogenic/PARP inhibitor combinations over PARPi alone, the activity of this combination in comparison to standard of care platinum therapy has not been established. The LIGHT study (#6013) further characterized the activity of olaparib as primary therapy for relapsed PARPi naïve platinum-sensitive ovarian cancer . As expected, activity in BRCA-mutated tumors (germline or somatic) was highest, with response rates of 64 to 69% and median PFS of 10.8 to 11.0 months, while activity in BRCA-wild type HR deficient tumors (as assessed by the Myriad MyChoice assay) was more modest (response rate of 29%, median PFS 7.2 months), and activity in BRCA-wild type HR proficient tumors was lowest (response rate of 10%, median PFS 5.4 months). Two additional studies provided insight on the dosing of niraparib and raised questions about how best to follow patients on PARP inhibitor maintenance. Weight and platelet-based dosing of niraparib was incorporated into the PRIMA study after enrollment was approximately two-thirds complete. Mirza and colleagues (#6050) reported no impact on efficacy of this individualized starting dose, with a hazard ratio for PFS with maintenance niraparib of 0.59 (95% CI 0.46–0.76) in patients enrolled to PRIMA who received the fixed starting dose of 300 mg daily regardless of weight or platelets, and a hazard ratio of 0.69 (95% CI 0.48–0.98) for those patients who received the individualized starting dose . The interaction test p-value for efficacy based upon starting dose was 0.30. Finally, a study by Tjokrowidjaja and colleagues (#6014) utilizing data from the SOLO2 trial intriguingly suggested that nearly half of patients experiencing RECIST progression on trial did not meet GCIG CA125 progression criteria . Some of these patients still experienced a rising CA125; however, approximately one quarter of patients had a stable or falling CA125. This observation raises the question of whether patients on PARPi maintenance should have regular imaging as opposed to relying on CA-125 surveillance alone. However, the study did not report how many patients had symptoms, and it is possible that some patients without rising CA125 would have been identified due to symptomatic recurrence. 2.3 Antibody-drug conjugates and immunotherapy Activity from the combination of the folate-receptor alpha-targeting antibody-drug conjugate mirvetuximab soravtansine in combination with bevacizumab in tumors demonstrating medium or high FRα membrane staining was reported (#6004) . Overall, an objective response rate of 47% was observed, with a response rate of 64% in patients with high FRα expression. Response rates were high regardless of platinum status, with a response rate of 59% in platinum-resistant patients and 69% in platinum-sensitive cases. As reported at last year's European Society of Medical Oncology (ESMO) meeting, FORWARD1, the randomized Phase 3 trial of mirvetuximab soravtansine failed to meet its primary endpoint of improved activity compared to chemotherapy in platinum-resistant ovarian cancer; however, the selectivity of the assay used to determine FRα expression may also have contributed to this outcome. Results from the upcoming Phase 3 study MIRASOL of mirvetuximab soravtansine in true high FRα expressers will be of high interest in the further development of this interesting agent. Final results from the KEYNOTE100 study of pembrolizumab monotherapy in relapsed ovarian cancer were also reported (#6005) . Overall activity across the cohort remained low, with an overall response rate of 8.5%. However, there was a trend towards increased activity in patients with high PD-L1 expression (defined as a CPS score ≥ 10), with response rate ranging from 11.6 to 18.2% in these patients. These findings again highlight the limited activity of PD1/PD-L1-directed therapy in ovarian cancer and the need for biomarkers to identify the small percentage of patients who may derive benefit from these agents.
The revival of secondary cytoreductive surgery This year, two prospective trials evaluating secondary cytoreductive surgery (SCS) were presented; provides an overview of both trials. Du Bois and colleagues, after validation of their selection criteria and demonstrating a benefit in progression free survival (PFS), reported the final results of the phase 3, randomized DESKTOP III trial (#6000) evaluating the impact of SCS in recurrent ovarian cancer . The investigators reported an overall survival (OS) benefit in the surgery arm with a median survival of 53.7 months compared to 46.2 months in those who did not undergo surgery. The true benefit was seen in women who underwent complete resection at the time of surgery (60.7 months versus 46.2 months). It is important to note that patients who were not able to undergo complete resection had a worse outcome (median survival 28.8 months) when compared to the no surgery arm. In a separate randomized phase 3 study from China, SOC 1/SGOG-OV2 (#6001), SCS was also evaluated. In this study, there were designated selection criteria based on the iModel score in conjunction with PET-CT scan imaging to determine complete resection . Zang and colleagues reported that SCS was associated with a benefit in PFS of 17.4 versus 11.9 months, though OS outcomes are still maturing. Similarly to DESKTOP III, those with residual disease did worse than chemotherapy alone. Therefore, though these results differed from GOG 213 which reported no difference in OS outcomes with SCS, both groups concluded that SCS is of benefit when performed in carefully selected patients, and that SCS should be done at centers of excellence/experience in which the opportunity to resect disease completely is optimal. Lastly, another study of SCS presented was a randomized phase 2 trial of SCS with or without carboplatin hyperthermic intraperitoneal chemotherapy (HIPEC) in women with recurrent platinum sensitive ovarian cancer (#6016) . The study randomized 98 patients and reported 0% perioperative mortality and comparable outcomes to those undergoing SCS only. Though preliminary, PFS and OS rates were not significantly different and did not demonstrate that HIPEC was superior to SCS alone.
Further insight into the activity of PARP inhibitors (PARPi) and PARPi combinations This year, SOLO2 (#6002), a randomized Phase 3 trial of olaparib maintenance therapy in women with relapsed platinum-sensitive BRCA-mutated ovarian cancer following response to platinum-based therapy, became the first report of mature OS data from a Phase 3 setting . The study enrolled 295 patients and previously reported a 13.6 month improvement in the primary endpoint of PFS in 2017. In this final analysis of SOLO2, a median improvement in OS of 12.9 months was observed (38.8 versus 51.7 months), with a hazard ratio of 0.74 (95% CI 0.54–1.00; p = 0.0537). While this finding was just shy of the threshold for statistical significance, the magnitude of the observed OS benefit was striking. Additionally, it is important to note that this final analysis included patients who received subsequent PARPi therapy, including 38% of patients in the control arm and 10% of those in the olaparib arm. When adjusted for subsequent PARPi in the placebo group, the observed improvement in median OS was 16.3 months (35.4 versus 51.7 months), with a hazard ratio of 0.56 (95% CI 0.35–0.97). These results support the use of maintenance PARPi in patients with BRCA-mutated platinum sensitive recurrent ovarian cancer who are PARPi naïve. The primary analysis of NRG-GY004, a Phase 3 trial comparing the combination of cediranib and olaparib or olaparib monotherapy to platinum-based chemotherapy in relapsed platinum-sensitive ovarian cancer was also presented (#6003) . This study asked whether a non-platinum alternative could improve PFS over platinum-based chemotherapy in platinum-sensitive ovarian cancer. The combination cediranib and olaparib did not meet the primary endpoint of improving PFS compared to chemotherapy, with a median PFS of 10.4 months for the combination and 10.3 months for chemotherapy (hazard ratio 0.856, 95% CI 0.663–1.105, p = 0.077), although the observed activity was comparable for both PFS and objective response rates. Due to a hierarchical statistical design, the activity of olaparib monotherapy was not formally compared to chemotherapy, but a median PFS of 8.2 months was observed. Side effects with the combination of cediranib and olaparib were significant; while hematologic adverse events were higher with chemotherapy, rates of non-hematologic adverse events were higher with cediranib/olaparib, and the discontinuation rate due to adverse events was 21%. Pre-specified subgroup analyses were notable for the high response rates and significant activity seen with both olaparib monotherapy and cediranib/olaparib combination in patients with a germline BRCA mutation. Further investigation of the cediranib/olaparib combination is ongoing in NRG-GY005 and ICON9, and these studies will give us additional insight into the activity of this combination in these settings. Two Phase 2 trials reported this ASCO also provided additional insight into PARPi activity as monotherapy or in combination. NSGO-AVANOVA2 (#6012) was a randomized Phase 2 study compared combination niraparib and bevacizumab to niraparib monotherapy in relapsed platinum-sensitive ovarian cancer . In an updated analysis, there was continued improvement in PFS, with a hazard ratio of 0.34 (95% CI 0.21–0.54, p < 0.0001). With 52% event maturity, the hazard ratio for OS was non-significant at 0.75 (95% CI 0.44–1.28, p = 0.30). While these results support increased activity of anti-angiogenic/PARP inhibitor combinations over PARPi alone, the activity of this combination in comparison to standard of care platinum therapy has not been established. The LIGHT study (#6013) further characterized the activity of olaparib as primary therapy for relapsed PARPi naïve platinum-sensitive ovarian cancer . As expected, activity in BRCA-mutated tumors (germline or somatic) was highest, with response rates of 64 to 69% and median PFS of 10.8 to 11.0 months, while activity in BRCA-wild type HR deficient tumors (as assessed by the Myriad MyChoice assay) was more modest (response rate of 29%, median PFS 7.2 months), and activity in BRCA-wild type HR proficient tumors was lowest (response rate of 10%, median PFS 5.4 months). Two additional studies provided insight on the dosing of niraparib and raised questions about how best to follow patients on PARP inhibitor maintenance. Weight and platelet-based dosing of niraparib was incorporated into the PRIMA study after enrollment was approximately two-thirds complete. Mirza and colleagues (#6050) reported no impact on efficacy of this individualized starting dose, with a hazard ratio for PFS with maintenance niraparib of 0.59 (95% CI 0.46–0.76) in patients enrolled to PRIMA who received the fixed starting dose of 300 mg daily regardless of weight or platelets, and a hazard ratio of 0.69 (95% CI 0.48–0.98) for those patients who received the individualized starting dose . The interaction test p-value for efficacy based upon starting dose was 0.30. Finally, a study by Tjokrowidjaja and colleagues (#6014) utilizing data from the SOLO2 trial intriguingly suggested that nearly half of patients experiencing RECIST progression on trial did not meet GCIG CA125 progression criteria . Some of these patients still experienced a rising CA125; however, approximately one quarter of patients had a stable or falling CA125. This observation raises the question of whether patients on PARPi maintenance should have regular imaging as opposed to relying on CA-125 surveillance alone. However, the study did not report how many patients had symptoms, and it is possible that some patients without rising CA125 would have been identified due to symptomatic recurrence.
Antibody-drug conjugates and immunotherapy Activity from the combination of the folate-receptor alpha-targeting antibody-drug conjugate mirvetuximab soravtansine in combination with bevacizumab in tumors demonstrating medium or high FRα membrane staining was reported (#6004) . Overall, an objective response rate of 47% was observed, with a response rate of 64% in patients with high FRα expression. Response rates were high regardless of platinum status, with a response rate of 59% in platinum-resistant patients and 69% in platinum-sensitive cases. As reported at last year's European Society of Medical Oncology (ESMO) meeting, FORWARD1, the randomized Phase 3 trial of mirvetuximab soravtansine failed to meet its primary endpoint of improved activity compared to chemotherapy in platinum-resistant ovarian cancer; however, the selectivity of the assay used to determine FRα expression may also have contributed to this outcome. Results from the upcoming Phase 3 study MIRASOL of mirvetuximab soravtansine in true high FRα expressers will be of high interest in the further development of this interesting agent. Final results from the KEYNOTE100 study of pembrolizumab monotherapy in relapsed ovarian cancer were also reported (#6005) . Overall activity across the cohort remained low, with an overall response rate of 8.5%. However, there was a trend towards increased activity in patients with high PD-L1 expression (defined as a CPS score ≥ 10), with response rate ranging from 11.6 to 18.2% in these patients. These findings again highlight the limited activity of PD1/PD-L1-directed therapy in ovarian cancer and the need for biomarkers to identify the small percentage of patients who may derive benefit from these agents.
Uterine cancer A single arm Phase 2 trial reported on the activity of the Wee1 inhibitor adavosertib in uterine serous carcinomas (#6009) . A response rate of 29.4% was observed in this small trial of 34 evaluable patients and the clinical benefit rate (responses and stable disease for at least 6 months) was 50%. While the authors hypothesized that a combination of a TP53 mutation together with oncogenically-driven replication stress and additional cell cycle dysregulation could make these cells particular sensitive to the effects of Wee1 inhibition, no clear correlation was identified between single gene alterations and clinical outcomes in their sample set. Adavosertib is associated with some toxicities, and over 50% of patients required at least one dose reduction, although dose discontinuations due to adverse events were infrequent. Given the limited options for uterine serous cancers, targeting Wee1 may represent a novel therapeutic alternative in this challenging disease, although validation of these study results is needed. Lheureux and colleagues reported results from a trial comparing the combination of nivolumab and cabozantinib to nivolumab monotherapy in recurrent endometrial cancer (#6010) . The combination of nivolumab and bevacizumab had increased activity, with a median PFS of 5.3 months (compared to 1.9 months), and a response rate of 25% (compared to 16.7%). In an exploratory cohort of patients who had received prior immune checkpoint therapy or who had carcinosarcomas, 5 of 21 patients who had received prior immunotherapy had a response, while 1 of 9 carcinosarcoma patients did. These results further support the development of anti-angiogenic tyrosine-kinase inhibitors together with immune checkpoint blockade in endometrial cancer.
Cervix cancer An analysis of disease free and disease specific survival in patients with early stage cervical cancer who underwent sentinel lymph node biopsy versus bilateral pelvic lymphadenectomy was reported from two prospective trials (SENTICOL I and II) (#6006) . The study used blue dye and radioactive tracer and was done predominantly via minimally invasive route (study conducted prior to the Laparoscopic Approach to Cervical Cancer (LACC) trial results). The results demonstrated that there were no significant differences between the two groups and that sentinel lymph nodes can be performed in this setting; however, this was a low risk group as only ~12% of the study population required adjuvant therapy, and we anxiously await the evaluation of sentinel lymph node biopsy in patients with high risk cervical cancer. A phase 3 study evaluating the role of adjuvant therapy after radical hysterectomy in cervical cancer (STARS trial) was also presented (#6007) . Huang and colleagues randomized over 1000 patients with intermediate or high risk factors following surgery to radiation therapy, chemoradiation (with cisplatin weekly), or radiation therapy with sequential chemoradiation (2 cycles of cisplatin and paclitaxel before and after radiation). The primary outcome, disease free survival at 3 years, was highest in the sequential chemoradiation arm (90% versus 85% in chemoradiation arm and 82% in radiation arm). This difference was most significant in those with high-risk features and associated with a decreased risk of death. Ongoing trials though the RTOG and GOG evaluating the role of systemic chemotherapy after chemoradiation (±surgery) will help determine if this is truly practice changing, but the data are promising. In an ancillary data analysis of GOG 49, 92, and 141 of early stage cervical cancer patients, Levinson et al. re-evaluated the Sedlis criteria of intermediate risk factors (#6019) . They noted that recurrence in squamous cell carcinoma was associated independently with lymphovascular space invasion, depth of invasion, and tumor size; however, for adenocarcinoma, only tumor size >4 cm was associated with recurrence. This presentation suggests that recurrence rates may be influenced by different factors in squamous cell and adenocarcinoma of the cervix. The novel combination of camrelizumab (PD1 inhibitor) plus apatinib (tyrosine kinase inhibitor targeting VEFGR2) was studied in a phase II open label trial of 45 women with advanced cervical cancer after prior systemic chemotherapy (#6021) . The authors noted two complete responses and 23 partial responses for an objective response rate of 59.6%. Though PDL1 expression resulted in a longer PFS, responses were noted in both populations. The regimen was associated with tolerable toxicities and demonstrates promising anti-tumor activity in a typically poor prognostic group.
Rare tumors The first prospective trial of immunotherapy in gestational trophoblastic neoplasia, TROPHIMMUNE, was presented (#6008) . In this study, 15 patients received avelumab after demonstrating resistance to monotherapy. Fifty percent of patients achieved monotherapy, including 5 patients who would have received polychemotherapy after failing both single agent regimens. The five patients who were resistant to avelumab were able to achieve a complete response after receiving chemotherapy. Toxicities were mild and one patient successfully delivered an infant following treatment. Blanc and colleagues presented a study of high dose chemotherapy and autologous stem cell rescue (HDC-aSCR) in patients with ovarian small cell carcinoma, hypercalcemic type (#6023) . This intense regimen included surgery followed by chemotherapy (cisplatin, doxorubicin, etoposide, and cyclophosphamide) and if a complete response was noted, patients received HDC-aSCR with or without pelvic radiation. They noted a median overall survival of 36.4 months and a 2-year event free survival of 40%. If pelvic radiation were administered, a 57% recurrence free survival was noted. This study provides additional insights into the management of a rare but aggressive histology with improvements in outcome.
Conclusions This year's virtual ASCO meeting provided continued progress and outcomes in areas such as PARPi and, secondary cytoreductive surgery in ovarian cancer, and evaluation of novel agents (e.g. adavosertib, camrelizumab, and mirvetuximab) or approaches (HDC-aSCR) across all gynecologic malignancies. lists selected trials in progress presented at the meeting. These and other ongoing trials are likely to accelerate progress at faster paces than seen before. Even though we couldn't be together for the meeting, the theme of “unite and conquer” resonates strongly.
Ritu Salani has no relevant conflict of interest related to the manuscript. Joyce F. Liu reports advisory board participation for AstraZeneca, Clovis, Genentech, Merck, Regeneron, and Tesaro/GSK, outside the submitted work; and funding to her institution for study conduct as PI on trials from 2X Oncology, Aravive, Arch Oncology, AstraZeneca, Bristol-Myers Squibb, Clovis Oncology, CytomX Therapeutics, GlaxoSmithKline, Regeneron, Surface Oncology, Tesaro, and Vigeo Therapeutics, outside the submitted work. Both authors were involved in the conception, design, review and editing of the manuscript.
|
Novel Phenotypical and Biochemical Findings in Mucolipidosis Type II | a560c847-59af-4be1-8206-31bbed8adb36 | 11941985 | Biochemistry[mh] | Mucolipidosis (ML) II (MIM#252500) is an autosomal recessive lysosomal storage disorder caused by mutations in the GNPTAB gene producing N-acetylglucosamine-1-phosphotransferase (GlcNAc-PTase, EC 2.7.8.17) deficiency. The final consequence is a defect in the mannose 6-phosphate (M6P) targeting signal function, which generates the hypersecretion of lysosomal enzymes out of cells. Their respective substrates gather inside lysosomes in several tissues (fibroblasts, secretory organs, and connective tissue are severely affected); their accumulation along with the altered traffic of lysosomal enzymes produces the clinical signs and symptoms of these diseases . These alterations are cell type and tissue specific, predominantly affecting fibroblasts and other mesenchymal cells; in other tissues, such as the brain, Man-6-P-independent mechanisms have been found to direct lysosomal enzymes to lysosomes and partially replace its function but cannot fully prevent neurologic manifestations . Also, M6P dependency for lysosomal enzyme trafficking could differ between enzymes . ML is a rare disease with an estimated prevalence of 0.22–2.7 per 100,000 live births . The combined birth prevalence is estimated to be between 1/37,000 and 455,000 worldwide, whereas the reported birth prevalence for ML II is estimated to be between 1/123,000 and 2,000,000 in Europe ( www.orpha.net , accessed 3 January 2025). A founder effect has been reported in the Saguenay-Lac-St-Jean region of Quebec where a birth prevalence is estimated at 1/6000 . ML II is considered a multisystemic disease mainly characterized by skeletal abnormalities (dysostosis multiplex), which cause significant growth restriction with short stature and low weight. Other frequent signs and symptoms (>10%) include dysmorphic facial features, developmental delay, and restricted joint range of motion . Gingival hyperplasia is a distinctive feature of this disease. Cardiac involvement with heart valve disease, cardiomyopathy, or left ventricular hypertrophy can also be found, as well as respiratory problems due to mucosal thickening and stiffening of the thoracic cage. Other occasional clinical features include hepatosplenomegaly, abdominal wall defects (diastasis recti, umbilical and inguinal hernia), recurrent respiratory infection, and hypotonia . Even though most lysosomal diseases have a nonspecific neonatal presentation, in most cases, symptoms are already present at birth and the phenotype is already distinctive at this time and can help to establish a diagnostic hypothesis. ML II patients can present early symptoms such as hyperparathyroidism right after birth [ , , ] and radiological evidence may be found even in the fetal period pointing to ML II [ , , ]. However, due to the low prevalence of the disease, data on its clinical presentation, natural history, and biochemical profile are scarce. There is currently no specific treatment for ML II and ML III α/β and the prognosis of the disease is variable but generally very poor, with a median survival of 5 years for the ML II phenotype . Most patients die from pulmonary and cardiac complications. Research on potential therapies is currently underway, with pending results, which may open new horizons for these patients. Identifying these patients remains a significant challenge. In the absence of a biochemical marker, most patients are lately diagnosed by genetic study, probably later than expected in other lysosomal diseases. However, abnormalities in lysosomal enzyme activity in plasma and dried blood spots (DBS) can be detected using Liquid Chromatography-Mass Spectrometry (LC-MS/MS), without a specific profile. Altered glycosaminoglycan (GAGs) levels can be found as well in serum/plasma, DBS, and urine specimens [ , , ]. In this study, enzymatic and proteomic tests are carried out to shed light on poorly described and understood aspects of the disease.
2.1. Clinical and Genetic Characteristics Three patients with a confirmed diagnosis of ML II were enrolled in this study and followed up for an average of 24 months. The demographic, anthropometric, and main clinical characteristics of the patients are summarized in . Patient 1 is a 5-year-old female. Family history included a previous miscarriage due to severe fetal skeletal dysplasia without fetal genetic study. In the third trimester of gestation, she was diagnosed with intrauterine growth restriction. She was born full term, at 37 + 5 weeks of gestational age (GA); at birth, she was admitted to the Neonatal Unit due to low weight ( ). During her admission, cardiac anomalies consisting of patent ductus arteriosus and left ventricular diastolic dysfunction were observed. Moreover, she presented left clubfoot and bilateral hip dysplasia. She was later diagnosed at 5 months of life by clinical whole-exome sequencing (WES), a frameshift variant with premature stop codon [c.738del, p.(Lys246Asnfs*21)] and a deletion including the last two exons of the GNPTAB. (NM_024312.5). This information has been communicated to the Clin Var database (SUB15044432). As for her skeletal dysplasia, imaging tests were performed showing a cranial deformity ( ) with triangular morphology of the anterior portion of the skull and craniosynostosis of the metopic suture, with parietal and temporal bone widening. Anomalies in the thorax with a widening of costal metaphysis were also incidentally found during radiographic procedures ( ). Over the first year of life, she required a left Achilles tenotomy and arthrography of both hips, with adductor tenotomy. Over the years, she has developed mild concentric hypertrophy of the left ventricle, severe at the septal level, with no evidence of left ventricular outflow tract obstruction. She also presents heart valve disease, with moderate mitral insufficiency and thickened aortic valve leaflets leading to mild insufficiency, as well as mild pulmonary insufficiency. Ophthalmologic and otorhinolaryngologic evaluations showed corneal turbidity and adenoid hypertrophy with significant obstruction of the choanae, as well as hypertrophy of the pharyngeal mucosa. On physical examination, she has progressively developed some distinctive features such as infiltrative facies and gingival hyperplasia. She maintains significant growth retardation and weight stagnation ( ). During early childhood, she presented recurrent respiratory infections, so at 4 years of age immunological studies were carried out. Results showed a possible defect in humoral immunity, with evidence of IgM deficiency (IgM < 5 mg/dL). Regarding her neurodevelopment, there is currently a motor impairment and a slight delay in language development. She can sit without support, but never achieved autonomous ambulation or could stand by herself. At a cognitive level, however, she presents normal development, being able to understand simple sentences, performing symbolic play, and building sentences of two or three words, sometimes with low intelligibility. She suffers from minor feeding difficulties and receives treatment with levothyroxine, due to non-autoimmune hypothyroidism, and captopril. The patient is under multidisciplinary follow-up. Patient 2 is a 2-year-old male. There is no relevant prenatal history, he was born at 37 + 2 weeks of gestational age. Since birth, he presented subtle dysmorphic features, gingival hyperplasia, as well as altered phospho-calcium metabolism. In the first initial hours of life, he was admitted to the Neonatal Unit because he presented jaundice and blood tests revealed thrombocytopenia with non isoimmune hyperbilirubinemia. At 13 days of life, he had to be admitted again because hypophosphatemia was observed; further studies showed elevated PTH ( ) and alkaline phosphatase, being diagnosed with hyperparathyroidism. During the first two months of life, brachycephaly and trigonocephaly were observed, along with delayed motor development and hypotonia. Imaging tests ( ) showed severe brachycephaly and trigonocephaly with craniosynostosis; due to progressive worsening of his bone alterations, surgery—consisting of cranial remodeling—was performed at 7 months of life. Other skeletal anomalies characteristic of the pathology were also found ( ). The results of genetic studies were obtained at that time confirming the ML II diagnosis. Clinical WES showed a frameshift variant in homozygosis (c.3503_3504del) in the GNPTAB gene classified as pathogenic, which is the most prevalent variant of this disease worldwide . Since the patient showed evolutionary feeding difficulties with episodes of choking and weight stagnation ( ), it was decided to start enteral nutrition through a nasogastric tube at one year of life. A gastrostomy was later performed; as in the previous surgery, a difficult airway was observed, requiring the use of a laryngeal mask and intubation through fibro bronchoscopy. Although his motor development is delayed, his neurodevelopment at social and cognitive levels keeps progressing. His joint contractures limit his ability to extend his knees and arms, but he can manipulate objects with both hands. Currently, he does not receive any pharmacological treatment. Patient 3 is an 11-month-old male with a severe neonatal onset of the disease. During prenatal ultrasound evaluations, a delayed intrauterine growth was observed; due to fetal distress risk, a cesarean section was performed at 29 + 4 weeks of gestational age after receiving lung maturation with corticosteroids and neuroprotection with magnesium sulfate. He needed immediate admission to the NICU (Neonatal intensive care unit) due to respiratory distress requiring endotracheal intubation. During admission, the patient was extubated in the first 24 h of life but there was a persistent need for oxygen and non-invasive respiratory support, probably related to bronchopulmonary dysplasia without significant response to medical treatment with inhaled corticosteroids and oral spironolactone. A thoracic CT (Computed Tomography) scan, at two months of age, showed significant bilateral pulmonary involvement with parenchymal fibrosis and volume loss. Along with the chronic lung disease, abdominal wall defects, osteopenia, and dysmorphic features were also identified. The passage of meconium was delayed until the eighth day of life. He presented diastasis recti, with bilateral reducible inguinal hernias and right hydrocele. Dysmorphic features with abnormal cranial morphology have been striking since birth, with severe osteopenia and spontaneous distal fracture of the right tibia diagnosed at 5 weeks of life. Due to secondary neonatal hyperparathyroidism ( ), he was treated with vitamin D, monosodium phosphate, calcium gluconate, and later bisphosphonates. He also presented difficulties in sucking and swallowing from the first days of life, requiring a nasogastric tube for feeding. He was diagnosed at 4 months of age by clinical WES; the patient presented two pathogenic variants in heterozygosis (c.2956C>A and c.3503_3504del), affecting the GNPTAB gene. During the first months of life, he needed several hospital admissions due to respiratory failure, requiring continued support with non-invasive mechanical ventilation. At 8 months of life, it was decided to perform a tracheostomy along with a gastrostomy and surgical correction of the abdominal wall defects. He currently presents a stationary clinical situation at the respiratory level, with occasional decompensation, maintaining mechanical ventilation in spontaneous mode. In terms of neurodevelopment, significant generalized hypotonia persists and he is not able to consistently hold his head up. From an immunological point of view, patient 2 has presented mild respiratory infections while patient 3 has been asymptomatic. Immunological studies were performed in both patients confirming the presence of humoral alterations, with a profile consistent with IgM production deficit for patient 2 (IgM 8 mg/dL) and IgG values at the low end of the normal range for patient 3 (IgG 262 mg/dL). 2.2. Biochemical Characteristics: GAG and Enzymatic Activity Profile, Plasma Proteomic Analysis In urine samples for GAG quantification (neonatal samples obtained for routine newborn screening), we observed patient 1 had elevated dermatan sulfate (DS), heparan sulfate (HS) and chondroitin sulfate (CS), in patient 2 all the values were normal, and patient 3 showed elevate DS and HS. However, in the diagnostic samples, all the values obtained were normal in all 3 patients. The results are shown in . The enzymatic analysis showed high values of acid sphingomyelinase (ASMD) EC 3.1.4.12, alpha-iduronidase (IDUA) EC 3.2.1.76, iduronidate 2-sulfatase (IDS) EC 3.1.6.13, alpha-N-acetyl glucosaminidase (NAGLU) EC 3.2.1.50, and beta-glucuronidase (GUSB) EC 3.2.1.31 in the neonatal samples and diagnostic samples of the three patients, with the only exception being the arylsulfatase B (ARSB) EC 3.1.6.12 value of patient 1 in the neonatal sample. The results are shown in . Alpha-galactosidase (GAA) (EC 3.2.1.22) enzyme analysis showed elevated values only in diagnostic samples, with normal results in neonatal samples. The other enzymes (galactocerebrosidase-GALC (EC 3.2.1.46), alpha-glucosidase-GLA (EC 3.2.1.20), beta-glucosidase-GBA (3.2.1.21), galactose 6-sulfatase-GANLS (EC 2.5.1.5), and beta-galactosidase-GLB1 (EC 3.2.1.23) presented normal values, with a few isolated results slightly above or below the normal level. The results are shown in . In the following figure ( ), we put all the enzymes that participate in the degradation routes of the glycosaminoglycans DS and HS. We marked the enzymes not analyzed in the DBS samples of ML whose results have been elevated in all the analyzed enzymes. 2.3. Quantitative Proteomic Analysis by SWATH Samples from ML II patients, both at birth and diagnosis, and samples from healthy neonates were analyzed using SWATH proteomic technique. In neonatal samples of patients affected by ML II, the upregulated proteins are shown in . The elevated proteins of healthy newborn samples when compared to affected neonates are shown in . In , a volcano plot from the SWATH-MS quantitative proteomics analysis and the results from the string tool analysis are shown. In the proteins upregulated in healthy samples compared with neonatal disease samples, we found PPBG and HEXB proteins, which are related to lysosomal activity, and CEPT protein, which is related to cholesterol transport. In diagnostic samples compared with healthy samples, we found 36 dysregulated proteins (see ). The proteins upregulated in ML II patients at the time of diagnosis are shown in ; the downregulated proteins in ML II patients are shown in . Among these dysregulated proteins, we found the following interesting proteins: APOD protein related to cholesterol; G3P, LDHA, and LDHB proteins related to carbon energy generation; and HEXB protein related to lysosomal activity. Further on, when we compared samples from healthy neonates vs. affected patients (neonatal period and at the time of ML II diagnosis), we found 32 dysregulated proteins (see ). The proteins downregulated in ML II patients are shown in ; the upregulated proteins in ML II patients are shown in . In the comparison between healthy neonates and samples from affected patients, we found proteins previously found such as HEXB, G3P, and CATG. When we compared the dysregulated proteins, we found interesting common proteins ( and ), such as HEXB, elevated in healthy individuals and downregulated in patients (both during the neonatal period and at the time of diagnosis).
Three patients with a confirmed diagnosis of ML II were enrolled in this study and followed up for an average of 24 months. The demographic, anthropometric, and main clinical characteristics of the patients are summarized in . Patient 1 is a 5-year-old female. Family history included a previous miscarriage due to severe fetal skeletal dysplasia without fetal genetic study. In the third trimester of gestation, she was diagnosed with intrauterine growth restriction. She was born full term, at 37 + 5 weeks of gestational age (GA); at birth, she was admitted to the Neonatal Unit due to low weight ( ). During her admission, cardiac anomalies consisting of patent ductus arteriosus and left ventricular diastolic dysfunction were observed. Moreover, she presented left clubfoot and bilateral hip dysplasia. She was later diagnosed at 5 months of life by clinical whole-exome sequencing (WES), a frameshift variant with premature stop codon [c.738del, p.(Lys246Asnfs*21)] and a deletion including the last two exons of the GNPTAB. (NM_024312.5). This information has been communicated to the Clin Var database (SUB15044432). As for her skeletal dysplasia, imaging tests were performed showing a cranial deformity ( ) with triangular morphology of the anterior portion of the skull and craniosynostosis of the metopic suture, with parietal and temporal bone widening. Anomalies in the thorax with a widening of costal metaphysis were also incidentally found during radiographic procedures ( ). Over the first year of life, she required a left Achilles tenotomy and arthrography of both hips, with adductor tenotomy. Over the years, she has developed mild concentric hypertrophy of the left ventricle, severe at the septal level, with no evidence of left ventricular outflow tract obstruction. She also presents heart valve disease, with moderate mitral insufficiency and thickened aortic valve leaflets leading to mild insufficiency, as well as mild pulmonary insufficiency. Ophthalmologic and otorhinolaryngologic evaluations showed corneal turbidity and adenoid hypertrophy with significant obstruction of the choanae, as well as hypertrophy of the pharyngeal mucosa. On physical examination, she has progressively developed some distinctive features such as infiltrative facies and gingival hyperplasia. She maintains significant growth retardation and weight stagnation ( ). During early childhood, she presented recurrent respiratory infections, so at 4 years of age immunological studies were carried out. Results showed a possible defect in humoral immunity, with evidence of IgM deficiency (IgM < 5 mg/dL). Regarding her neurodevelopment, there is currently a motor impairment and a slight delay in language development. She can sit without support, but never achieved autonomous ambulation or could stand by herself. At a cognitive level, however, she presents normal development, being able to understand simple sentences, performing symbolic play, and building sentences of two or three words, sometimes with low intelligibility. She suffers from minor feeding difficulties and receives treatment with levothyroxine, due to non-autoimmune hypothyroidism, and captopril. The patient is under multidisciplinary follow-up. Patient 2 is a 2-year-old male. There is no relevant prenatal history, he was born at 37 + 2 weeks of gestational age. Since birth, he presented subtle dysmorphic features, gingival hyperplasia, as well as altered phospho-calcium metabolism. In the first initial hours of life, he was admitted to the Neonatal Unit because he presented jaundice and blood tests revealed thrombocytopenia with non isoimmune hyperbilirubinemia. At 13 days of life, he had to be admitted again because hypophosphatemia was observed; further studies showed elevated PTH ( ) and alkaline phosphatase, being diagnosed with hyperparathyroidism. During the first two months of life, brachycephaly and trigonocephaly were observed, along with delayed motor development and hypotonia. Imaging tests ( ) showed severe brachycephaly and trigonocephaly with craniosynostosis; due to progressive worsening of his bone alterations, surgery—consisting of cranial remodeling—was performed at 7 months of life. Other skeletal anomalies characteristic of the pathology were also found ( ). The results of genetic studies were obtained at that time confirming the ML II diagnosis. Clinical WES showed a frameshift variant in homozygosis (c.3503_3504del) in the GNPTAB gene classified as pathogenic, which is the most prevalent variant of this disease worldwide . Since the patient showed evolutionary feeding difficulties with episodes of choking and weight stagnation ( ), it was decided to start enteral nutrition through a nasogastric tube at one year of life. A gastrostomy was later performed; as in the previous surgery, a difficult airway was observed, requiring the use of a laryngeal mask and intubation through fibro bronchoscopy. Although his motor development is delayed, his neurodevelopment at social and cognitive levels keeps progressing. His joint contractures limit his ability to extend his knees and arms, but he can manipulate objects with both hands. Currently, he does not receive any pharmacological treatment. Patient 3 is an 11-month-old male with a severe neonatal onset of the disease. During prenatal ultrasound evaluations, a delayed intrauterine growth was observed; due to fetal distress risk, a cesarean section was performed at 29 + 4 weeks of gestational age after receiving lung maturation with corticosteroids and neuroprotection with magnesium sulfate. He needed immediate admission to the NICU (Neonatal intensive care unit) due to respiratory distress requiring endotracheal intubation. During admission, the patient was extubated in the first 24 h of life but there was a persistent need for oxygen and non-invasive respiratory support, probably related to bronchopulmonary dysplasia without significant response to medical treatment with inhaled corticosteroids and oral spironolactone. A thoracic CT (Computed Tomography) scan, at two months of age, showed significant bilateral pulmonary involvement with parenchymal fibrosis and volume loss. Along with the chronic lung disease, abdominal wall defects, osteopenia, and dysmorphic features were also identified. The passage of meconium was delayed until the eighth day of life. He presented diastasis recti, with bilateral reducible inguinal hernias and right hydrocele. Dysmorphic features with abnormal cranial morphology have been striking since birth, with severe osteopenia and spontaneous distal fracture of the right tibia diagnosed at 5 weeks of life. Due to secondary neonatal hyperparathyroidism ( ), he was treated with vitamin D, monosodium phosphate, calcium gluconate, and later bisphosphonates. He also presented difficulties in sucking and swallowing from the first days of life, requiring a nasogastric tube for feeding. He was diagnosed at 4 months of age by clinical WES; the patient presented two pathogenic variants in heterozygosis (c.2956C>A and c.3503_3504del), affecting the GNPTAB gene. During the first months of life, he needed several hospital admissions due to respiratory failure, requiring continued support with non-invasive mechanical ventilation. At 8 months of life, it was decided to perform a tracheostomy along with a gastrostomy and surgical correction of the abdominal wall defects. He currently presents a stationary clinical situation at the respiratory level, with occasional decompensation, maintaining mechanical ventilation in spontaneous mode. In terms of neurodevelopment, significant generalized hypotonia persists and he is not able to consistently hold his head up. From an immunological point of view, patient 2 has presented mild respiratory infections while patient 3 has been asymptomatic. Immunological studies were performed in both patients confirming the presence of humoral alterations, with a profile consistent with IgM production deficit for patient 2 (IgM 8 mg/dL) and IgG values at the low end of the normal range for patient 3 (IgG 262 mg/dL).
In urine samples for GAG quantification (neonatal samples obtained for routine newborn screening), we observed patient 1 had elevated dermatan sulfate (DS), heparan sulfate (HS) and chondroitin sulfate (CS), in patient 2 all the values were normal, and patient 3 showed elevate DS and HS. However, in the diagnostic samples, all the values obtained were normal in all 3 patients. The results are shown in . The enzymatic analysis showed high values of acid sphingomyelinase (ASMD) EC 3.1.4.12, alpha-iduronidase (IDUA) EC 3.2.1.76, iduronidate 2-sulfatase (IDS) EC 3.1.6.13, alpha-N-acetyl glucosaminidase (NAGLU) EC 3.2.1.50, and beta-glucuronidase (GUSB) EC 3.2.1.31 in the neonatal samples and diagnostic samples of the three patients, with the only exception being the arylsulfatase B (ARSB) EC 3.1.6.12 value of patient 1 in the neonatal sample. The results are shown in . Alpha-galactosidase (GAA) (EC 3.2.1.22) enzyme analysis showed elevated values only in diagnostic samples, with normal results in neonatal samples. The other enzymes (galactocerebrosidase-GALC (EC 3.2.1.46), alpha-glucosidase-GLA (EC 3.2.1.20), beta-glucosidase-GBA (3.2.1.21), galactose 6-sulfatase-GANLS (EC 2.5.1.5), and beta-galactosidase-GLB1 (EC 3.2.1.23) presented normal values, with a few isolated results slightly above or below the normal level. The results are shown in . In the following figure ( ), we put all the enzymes that participate in the degradation routes of the glycosaminoglycans DS and HS. We marked the enzymes not analyzed in the DBS samples of ML whose results have been elevated in all the analyzed enzymes.
Samples from ML II patients, both at birth and diagnosis, and samples from healthy neonates were analyzed using SWATH proteomic technique. In neonatal samples of patients affected by ML II, the upregulated proteins are shown in . The elevated proteins of healthy newborn samples when compared to affected neonates are shown in . In , a volcano plot from the SWATH-MS quantitative proteomics analysis and the results from the string tool analysis are shown. In the proteins upregulated in healthy samples compared with neonatal disease samples, we found PPBG and HEXB proteins, which are related to lysosomal activity, and CEPT protein, which is related to cholesterol transport. In diagnostic samples compared with healthy samples, we found 36 dysregulated proteins (see ). The proteins upregulated in ML II patients at the time of diagnosis are shown in ; the downregulated proteins in ML II patients are shown in . Among these dysregulated proteins, we found the following interesting proteins: APOD protein related to cholesterol; G3P, LDHA, and LDHB proteins related to carbon energy generation; and HEXB protein related to lysosomal activity. Further on, when we compared samples from healthy neonates vs. affected patients (neonatal period and at the time of ML II diagnosis), we found 32 dysregulated proteins (see ). The proteins downregulated in ML II patients are shown in ; the upregulated proteins in ML II patients are shown in . In the comparison between healthy neonates and samples from affected patients, we found proteins previously found such as HEXB, G3P, and CATG. When we compared the dysregulated proteins, we found interesting common proteins ( and ), such as HEXB, elevated in healthy individuals and downregulated in patients (both during the neonatal period and at the time of diagnosis).
The overall goal of the present study was to investigate the phenotype, metabolomic, enzymatic, and proteomic profiles involved in MLII patients and to explore the clinical utility. From a clinical point of view, patient 3 presented a severe neonatal form of the disease while patients 1 and 2 presented later and more moderate forms. The bone involvement of our patients is similar to that which is previously described in the literature . All three patients showed skeletal malformations detected during the first year of life, and cranial malformation ( ) seems to be a very consistent finding in these patients . Deciding to perform a craniotomy may be questionable since craniosynostosis cannot be considered to be unequivocally caused by bone involvement . Thoracic alterations in the ribs are subtle but may be an incidental finding of interest. In two of the three patients, gingival hyperplasia and hypoparathyroidism were also detected as an early distinctive feature. Although ML II is a rare disease, early biochemical findings together with skeletal alterations (especially intrauterine or spontaneous fractures and cranial malformation) and characteristic facial features may be clinically suggestive of the disease. Respiratory tract involvement is described in ML II ; however, we consider the chronic lung disease shown by patient 3 to be related to his prematurity since this type of pulmonary findings has not been described in any other patient with the disease. It is noteworthy that, as expected, two of the three patients required advanced airway management during surgery due to a complex airway probably related to mucosal thickening and deposit of metabolic substrate . A potentially life-threatening airway distortion has been previously described but more knowledge of the possible causes and their management is needed. Two patients required gastrostomy placement due to their inability to feed. Even so, over time all individuals have presented severe growth problems with failure to thrive. The origin of this growth problem remains to be clarified. From a neurodevelopmental point of view, a severe delay in the motor development of the three patients has been found, related to hypotonia. None of the patients achieved autonomous ambulation. In contrast, cognitive dysfunction was present but it was not as significant compared with motor impairment. In other patients, significant sensory-motor impairment has also been described, in contrast to much less affected social and communication skills . The evolution followed by our patients shows a delay in the acquisition of cognitive items but with continuous progress. The presence or absence of possible neurodegeneration in the disease seems controversial [ , , ]. Further studies are needed to elucidate the effect that the loss of the Man-6-P targeting signal has on the central nervous system. Although recurrent respiratory infections have been reported as part of the disease, the humoral immunity deficit affecting ML II patients is poorly understood. The pathogenesis and clinical implications of these findings remain to be clarified, research in affected mice suggests a dysfunction in M6P transport routes could lead to alterations in B cell functions, while DC and T cell functions remain normal due to M6P-independent targeting pathways . Previously identified pathogenic variants of the disease have been found in patients 2 and 3. Patient 1 presents a new heterozygous variant (c.738del, p.(Lys246Asnfs*21)), which has not been previously described, in combination with a deletion affecting exons 20 and 21 of the gene. When enzymatic studies are carried out, the analysis conditions replicate the ideal conditions of lysosomes due to the acidic pH that the enzymes need to be functional. This has always made diagnosing the disease challenging, as the location of the enzymes within the cells is not well understood (many are found in the cytoplasm, while others are within the lysosome). As a result, genetic studies have been essential for diagnosing this condition . The results obtained from the enzymatic activity of this study do have a special impact on the pathophysiology and diagnosis of ML II. In this case, 12 different enzymes have been used, all of them hydrolases whose site of action is the lysosome. The enzymes found to be significantly high during the analysis (IDUA, IDS, NAGLU, GUSB, and ARSB) are part of the HS and DS degradation pathways ( ). The enzyme ASMD is involved in cholesterol degradation. All these enzymes need a mannose 6-phosphate ligand to reach the lysosome. The data obtained are of great diagnostic value as these alterations are maintained regardless of age and, given their presence since the neonatal period, they constitute a key diagnostic tool for early diagnosis of the disease. The results obtained for the GAA, GALC, GLA, GANLS, and GLB1 enzymes are not conclusive, as is the case with the GAGs. Some of these tests are mentioned in articles as being able to be used as a suspicion of this disease. The same is said about the elevated IDUA and IDS; however, until now they had not all been studied together . Data obtained from proteomic studies, in the comparison of ML II neonates vs. healthy neonates, show elevated levels of several proteins. These include the ODPB protein, which participates in the Krebs cycle, the lysosomal protein CATG cathepsin G, which activates extracellular matrix degradation, the MMP8 metalloprotease, which degrades extracellular matrix , and AACT alpha-1-antichymotrypsin, which inactivates cathepsin G. In addition, we also found the VDAC2 protein, which binds to various lipids such as the sphingolipid ceramide, among others, and cholesterol sterols . In the healthy neonate vs. ML II neonate comparison, the CETP protein stands out. This protein is involved in lipid transfer, including the transfer of cholesterol and triglyceride esters, allowing the movement of cholesterol esters. In addition, it also regulates reverse cholesterol transport; by this process, excess cholesterol is removed from peripheral tissues . The proteins we found that were elevated, together with the enzyme activities that were also elevated, show us a process of activation of matrix degradation to extract sugars from GAGs for energy production . Due to a malfunction of lysosomal trafficking, caused by M6P deficiency, lysosomal hydrolases are not able to reach the lysosome, resulting in a large increase in the production of these enzymes to compensate and to obtain energy from sugars. These processes are common in situations of absence of nutrients in the cell (autophagy situations) and could explain why GAGs may be found normal or minimally elevated. Similarly, the finding of low CEPT in patients with MLII compared to healthy individuals reveals difficulties in cholesterol transport and esterification; data obtained from ASMD activity also support this disturbance in cholesterol utilization. It is also important to mention the alteration found in the LAMP2 protein. This lysosomal transmembrane protein helps to internalize hydrolases . It has been found to be high in neonatal ML II but its relevance disappears in samples from older patients (at diagnosis). This loss of function has been previously described by Takanobu Otomo et al. . PPGB proteins are protective proteins that appear to be essential for both beta-galactosidase and neuraminidase activity. PPGB exerts a protective function necessary for cell stability and could help prevent the generation of gangliosides, molecules that accumulate in MLII, and other lysosomal diseases . The HEXB protein is always higher in the healthy group than in the group affected by ML II. Authors such as Whelan et al. found the activity of this enzyme to be deficient in the cells but elevated in plasma. The low number of patients is the main limitation of this research. Due to the very low prevalence of the disease and its high mortality, it is difficult to identify and include patients in research projects. In addition, the proteomic studies have been performed on DBS samples. Although the results are very relevant, we consider they should be validated on liquid samples.
4.1. Patients Three patients, two male and one female, have been identified in our center presenting GNPTAB variants. Written consent for publication of patient data was obtained from the families of all individuals included. Demographic information and clinical data including fetal, early, and evolutionary clinical characteristics of the patients were collected. Laboratory variables performed include phosphocalcic metabolism study (PTH, calcium, phosphorus, alkaline phosphatase, vitamin D) and immunological profile including quantification of immunoglobulins IgA, IgG (IgG1, IgG2, IgG3, IgG4) and IgM, total number of lymphocytes, CD4 and CD8 lymphocytes, B-CD19 lymphocytes, B-CD20 lymphocytes, memory B lymphocytes, and NK lymphocytes (NK-CD19, NK-CD56, NK-CD57) developed in our local laboratory. The normal range for the biochemical variables according to age: PTH 15–60 pg/mL; Alkaline phosphatase 1 week–6 months < 1076 U/L, 7 months–1 year < 1107. Normal range for immunoglobulins according to age: IgM 3–6 months 5–50 mg/dL, 6–12 months 8–70 mg/dL, 12–24 months 10–100 mg/dL, 2 years–6 years 50–180 mg/dL; IgG 3–6 months 170–560 mg/dL, 6–12 months 200–670 mg/dL, 12–24 months 300–1160 mg/dL, 2 years–6 years 400–1100 mg/dL. 4.2. Methods 4.2.1. Molecular Diagnosis Genomic DNA was obtained from blood samples and variants were identified using WES, performed as part of routine diagnostic protocols. WES was performed using Illumina NextSeq (Illumina, Inc., San Diego, CA, USA). The library preparation method used was the TruSeq DNA Library Prep Kit (Illumina, Inc.), and the capture of regions of interest was performed using IDT xGen Exome ResearchPanel (Integrated DNA Technologies, Inc., Coralville, IA, USA). 4.2.2. Biochemical and Enzymatic Profile To carry out the studies of GAG and enzymatic activity, DBS and dry urine spot (DUS) samples were selected from the 3 clinical cases described above. Neonatal samples for routine newborn screening were selected (collected 24 h after the first lactation); samples collected as part of this study of a possible lysosomal disease (diagnosis) of each patient were also used. GAG levels were quantified in DUS and enzymatic analysis was performed in DBS specimens using LC-MS/MS. The following enzymes were studied in the DBS samples: ASMD, IDUA, IDS, NAGLU, ARSB, GUSB, GALC, alpha-glucosidase, alpha-galactosidase, beta-glucosidase, galactose 6-sulfatase, and beta-galactosidase. Subsequent proteomic analyses were also conducted at our laboratory. 4.2.3. GAG Analysis For the quantification of urinary GAGs, we adopted a method previously described . Briefly, DUS from urine samples and calibrators (DS, HS, CS, and Creatinine) was subjected to methanolysis for 1 h at 65 °C in a 96-well plate. The resulting supernatant was evaporated under a nitrogen stream and subsequently resuspended in a solution containing the internal standard. LC-MS/MS was performed using an AB Sciex API 4500 QTRAP in positive ion mode to quantify the three types of GAGs and Creatinine. presents the results of the GAG analysis in the neonate period and at diagnosis from the three patients. 4.2.4. Enzymatic Analysis Further biochemical characterization was performed through enzymatic analysis. For this purpose, two kits from Revvity (Waltham, MA, USA) were used. The first, the NeoLSD™ MSMS Kit (cat. 3093-0010), allows for the multiplex quantitative measurement of the activity of several enzymes: acid-β-glucocerebrosidase (Gaucher disease), acid-sphingomyelinase (acid sphingomyelinase deficiency), acid α-glucosidase (Pompe disease), β-galactocerebrosidase (Krabbe disease), α-galactosidase A (Fabry disease), and α-L-iduronidase (Mucopolysaccharidosis type I disease—MPS I). The method was adapted to convert flow injection into gradient measurement . Quantification was performed using an AB Sciex API 4500 QTRAP in positive ion mode using a UPLC column from Waters (X Select CSH C18 3.5 µm, 2.1 × 50 mm (Cat. 186005255) held at 55 °C. The flow rate was 0.7 mL/min. Solvent A was 30% acetonitrile with 0.1% formic acid, and solvent B was 65% isopropanol/35% acetonitrile with 0.1% formic acid. The gradient is shown in . The second kit (cat. 4416-0010) allows for the multiplex quantitative measurement of the activity of: iduronate 2-sulfatase (MPS II), α-N-acetylglucosaminidase (MPS IIIB), galactose 6-sulfatase (MPS IV A), β-galactosidase (MPS IV B), N-acetylgalactosamine (MPS VI), β-glucoronidase (MPS VII), and tripeptidyl peptidase (CLN2). Adaptations were also made for gradient measurement . The same equipment, UPLC column, and gradient solvents were used as in the NeoLSD kit. The gradient used for MPS measurement is shown in . 4.2.5. Proteomic Studies Proteomic studies are quantitative studies using SWATH (sequential window acquisition of all theoretical fragment ion spectra), an information-dependent acquisition (IDA) technique was used. A human library containing 2150 proteins has been used, from which 1.021 have been quantified. 4.3. Experimental Design and Statistical Analysis Proteomic Analysis by TripleTOF 6600 using liquid chromatography–tandem mass spectrometry (LC-MS/MS): (see for further information). Protein extraction: Protein from Dried Blood Samples was extracted by incubating the paper in 100 μL of 100 mM ammonium bicarbonate at room temperature for 1 h. The sample was centrifuged for 20 min at 13,000× g , and the supernatant was transferred to a new tube. Then the protein was precipitated by the MeOH/CHCl 3 method, and the protein concentration was measured using an RC DC™ Protein Assay (reducing agent and detergent compatible) (BioRad, Hercules, CA, USA). Protein Digestion: For protein identification, equal amounts of protein from each sample ( n = 3 per group and 4 healthy controls) were loaded onto a 10% SDS-PAGE gel. The resulting condensed protein bands underwent gel digestion using Trypsin and were processed, as previously described in the by our group . Protein Quantification by SWATH-MS Analysis [ , , , ]: To build the MSMS spectral libraries, peptide solutions were analyzed by shotgun data-dependent acquisition (DDA) using micro-liquid chromatography–tandem mass spectrometry (LC-MS/MS), as described in the and previously by our group. The MSMS spectra of the identified peptides were then used to generate the spectral library for SWATH peak extraction using the add-in for PeakView Software (version 2.2, Sciex), MSMS ALL with SWATH Acquisition MicroApp (version 2.0, Sciex). Peptides with a confidence score >99% (obtained from the ProteinPilot database search) were included in the spectral library. For relative quantification by SWATHMS analysis, SWATH-MS acquisition was performed on a Triple TOF 6600 LC–MS–MS system (Sciex) using SWATH mode. The acquisition mode consisted of a 250 ms survey MS scan from 400 to 1250 m / z , followed by an MSMS scan from 100 to 1500 m / z (25 ms acquisition time) of the top 65 precursor ions from the survey scan, for a total cycle time of 2.8 s. The fragmented precursors were then added to a dynamic exclusion list for 15 s. Any singly charged ions were excluded from the MSMS analysis. Targeted data extraction from the SWATH MS runs was performed by PeakView v.2.2 (Sciex, Redwood City, CA, USA) using the SWATH-MS Acquisition MicroApp v.2.0 (Sciex, USA). Data were processed using the spectral library created from DDA. SWATH-MS quantization was attempted for all proteins in the ion library that were identified by ProteinPilot TM 5.0.1 with a false discovery rate (FDR) < 1%. PeakView computed an FDR and a score for each assigned peptide based on the chromatographic and spectra components: only peptides with an FDR < 1%, 10 peptides, and 7 transitions per peptide were used for protein quantization. The integrated peak areas were processed by MarkerView software version 1.3.1 (Sciex, USA) for a data-independent method for relative quantitative analysis. A most likely ratio normalization was performed to control for possible uneven sample loss across the different samples during the sample preparation process . Unsupervised multivariate statistical analysis using PCA was performed to compare data across samples. 4.4. Functional and Pathway Analysis Pathway analysis was performed using Reactome ( https://reactome.org/ , accessed on 20 July 2023), which applies a statistical (hypergeometric distribution) test to determine whether specific pathways are over-represented (enriched) and produces a probability score, which is corrected for FDR using the Benjamini–Hochberg method. The most enriched pathways were represented using Reactome pathway diagrams. Protein interactions were evaluated using STRING ( https://string-db.org/ , accessed on 25 July 2023), applying a minimum required interaction score of PPI = 0.9 (protein–protein interaction) and an FDR < 0.05. Venn diagrams were generated using http://www.interactivenn.net/ , accessed on 2 June 2023) and box plots using GraphPad Prism 9. Statistical analyses were performed using Markerview v 1.3.1 or Scaffold software v 5.2.2. Volcano plots and box plots were generated using GraphPad Prism 9.0.0.
Three patients, two male and one female, have been identified in our center presenting GNPTAB variants. Written consent for publication of patient data was obtained from the families of all individuals included. Demographic information and clinical data including fetal, early, and evolutionary clinical characteristics of the patients were collected. Laboratory variables performed include phosphocalcic metabolism study (PTH, calcium, phosphorus, alkaline phosphatase, vitamin D) and immunological profile including quantification of immunoglobulins IgA, IgG (IgG1, IgG2, IgG3, IgG4) and IgM, total number of lymphocytes, CD4 and CD8 lymphocytes, B-CD19 lymphocytes, B-CD20 lymphocytes, memory B lymphocytes, and NK lymphocytes (NK-CD19, NK-CD56, NK-CD57) developed in our local laboratory. The normal range for the biochemical variables according to age: PTH 15–60 pg/mL; Alkaline phosphatase 1 week–6 months < 1076 U/L, 7 months–1 year < 1107. Normal range for immunoglobulins according to age: IgM 3–6 months 5–50 mg/dL, 6–12 months 8–70 mg/dL, 12–24 months 10–100 mg/dL, 2 years–6 years 50–180 mg/dL; IgG 3–6 months 170–560 mg/dL, 6–12 months 200–670 mg/dL, 12–24 months 300–1160 mg/dL, 2 years–6 years 400–1100 mg/dL.
4.2.1. Molecular Diagnosis Genomic DNA was obtained from blood samples and variants were identified using WES, performed as part of routine diagnostic protocols. WES was performed using Illumina NextSeq (Illumina, Inc., San Diego, CA, USA). The library preparation method used was the TruSeq DNA Library Prep Kit (Illumina, Inc.), and the capture of regions of interest was performed using IDT xGen Exome ResearchPanel (Integrated DNA Technologies, Inc., Coralville, IA, USA). 4.2.2. Biochemical and Enzymatic Profile To carry out the studies of GAG and enzymatic activity, DBS and dry urine spot (DUS) samples were selected from the 3 clinical cases described above. Neonatal samples for routine newborn screening were selected (collected 24 h after the first lactation); samples collected as part of this study of a possible lysosomal disease (diagnosis) of each patient were also used. GAG levels were quantified in DUS and enzymatic analysis was performed in DBS specimens using LC-MS/MS. The following enzymes were studied in the DBS samples: ASMD, IDUA, IDS, NAGLU, ARSB, GUSB, GALC, alpha-glucosidase, alpha-galactosidase, beta-glucosidase, galactose 6-sulfatase, and beta-galactosidase. Subsequent proteomic analyses were also conducted at our laboratory. 4.2.3. GAG Analysis For the quantification of urinary GAGs, we adopted a method previously described . Briefly, DUS from urine samples and calibrators (DS, HS, CS, and Creatinine) was subjected to methanolysis for 1 h at 65 °C in a 96-well plate. The resulting supernatant was evaporated under a nitrogen stream and subsequently resuspended in a solution containing the internal standard. LC-MS/MS was performed using an AB Sciex API 4500 QTRAP in positive ion mode to quantify the three types of GAGs and Creatinine. presents the results of the GAG analysis in the neonate period and at diagnosis from the three patients. 4.2.4. Enzymatic Analysis Further biochemical characterization was performed through enzymatic analysis. For this purpose, two kits from Revvity (Waltham, MA, USA) were used. The first, the NeoLSD™ MSMS Kit (cat. 3093-0010), allows for the multiplex quantitative measurement of the activity of several enzymes: acid-β-glucocerebrosidase (Gaucher disease), acid-sphingomyelinase (acid sphingomyelinase deficiency), acid α-glucosidase (Pompe disease), β-galactocerebrosidase (Krabbe disease), α-galactosidase A (Fabry disease), and α-L-iduronidase (Mucopolysaccharidosis type I disease—MPS I). The method was adapted to convert flow injection into gradient measurement . Quantification was performed using an AB Sciex API 4500 QTRAP in positive ion mode using a UPLC column from Waters (X Select CSH C18 3.5 µm, 2.1 × 50 mm (Cat. 186005255) held at 55 °C. The flow rate was 0.7 mL/min. Solvent A was 30% acetonitrile with 0.1% formic acid, and solvent B was 65% isopropanol/35% acetonitrile with 0.1% formic acid. The gradient is shown in . The second kit (cat. 4416-0010) allows for the multiplex quantitative measurement of the activity of: iduronate 2-sulfatase (MPS II), α-N-acetylglucosaminidase (MPS IIIB), galactose 6-sulfatase (MPS IV A), β-galactosidase (MPS IV B), N-acetylgalactosamine (MPS VI), β-glucoronidase (MPS VII), and tripeptidyl peptidase (CLN2). Adaptations were also made for gradient measurement . The same equipment, UPLC column, and gradient solvents were used as in the NeoLSD kit. The gradient used for MPS measurement is shown in . 4.2.5. Proteomic Studies Proteomic studies are quantitative studies using SWATH (sequential window acquisition of all theoretical fragment ion spectra), an information-dependent acquisition (IDA) technique was used. A human library containing 2150 proteins has been used, from which 1.021 have been quantified.
Genomic DNA was obtained from blood samples and variants were identified using WES, performed as part of routine diagnostic protocols. WES was performed using Illumina NextSeq (Illumina, Inc., San Diego, CA, USA). The library preparation method used was the TruSeq DNA Library Prep Kit (Illumina, Inc.), and the capture of regions of interest was performed using IDT xGen Exome ResearchPanel (Integrated DNA Technologies, Inc., Coralville, IA, USA).
To carry out the studies of GAG and enzymatic activity, DBS and dry urine spot (DUS) samples were selected from the 3 clinical cases described above. Neonatal samples for routine newborn screening were selected (collected 24 h after the first lactation); samples collected as part of this study of a possible lysosomal disease (diagnosis) of each patient were also used. GAG levels were quantified in DUS and enzymatic analysis was performed in DBS specimens using LC-MS/MS. The following enzymes were studied in the DBS samples: ASMD, IDUA, IDS, NAGLU, ARSB, GUSB, GALC, alpha-glucosidase, alpha-galactosidase, beta-glucosidase, galactose 6-sulfatase, and beta-galactosidase. Subsequent proteomic analyses were also conducted at our laboratory.
For the quantification of urinary GAGs, we adopted a method previously described . Briefly, DUS from urine samples and calibrators (DS, HS, CS, and Creatinine) was subjected to methanolysis for 1 h at 65 °C in a 96-well plate. The resulting supernatant was evaporated under a nitrogen stream and subsequently resuspended in a solution containing the internal standard. LC-MS/MS was performed using an AB Sciex API 4500 QTRAP in positive ion mode to quantify the three types of GAGs and Creatinine. presents the results of the GAG analysis in the neonate period and at diagnosis from the three patients.
Further biochemical characterization was performed through enzymatic analysis. For this purpose, two kits from Revvity (Waltham, MA, USA) were used. The first, the NeoLSD™ MSMS Kit (cat. 3093-0010), allows for the multiplex quantitative measurement of the activity of several enzymes: acid-β-glucocerebrosidase (Gaucher disease), acid-sphingomyelinase (acid sphingomyelinase deficiency), acid α-glucosidase (Pompe disease), β-galactocerebrosidase (Krabbe disease), α-galactosidase A (Fabry disease), and α-L-iduronidase (Mucopolysaccharidosis type I disease—MPS I). The method was adapted to convert flow injection into gradient measurement . Quantification was performed using an AB Sciex API 4500 QTRAP in positive ion mode using a UPLC column from Waters (X Select CSH C18 3.5 µm, 2.1 × 50 mm (Cat. 186005255) held at 55 °C. The flow rate was 0.7 mL/min. Solvent A was 30% acetonitrile with 0.1% formic acid, and solvent B was 65% isopropanol/35% acetonitrile with 0.1% formic acid. The gradient is shown in . The second kit (cat. 4416-0010) allows for the multiplex quantitative measurement of the activity of: iduronate 2-sulfatase (MPS II), α-N-acetylglucosaminidase (MPS IIIB), galactose 6-sulfatase (MPS IV A), β-galactosidase (MPS IV B), N-acetylgalactosamine (MPS VI), β-glucoronidase (MPS VII), and tripeptidyl peptidase (CLN2). Adaptations were also made for gradient measurement . The same equipment, UPLC column, and gradient solvents were used as in the NeoLSD kit. The gradient used for MPS measurement is shown in .
Proteomic studies are quantitative studies using SWATH (sequential window acquisition of all theoretical fragment ion spectra), an information-dependent acquisition (IDA) technique was used. A human library containing 2150 proteins has been used, from which 1.021 have been quantified.
Proteomic Analysis by TripleTOF 6600 using liquid chromatography–tandem mass spectrometry (LC-MS/MS): (see for further information). Protein extraction: Protein from Dried Blood Samples was extracted by incubating the paper in 100 μL of 100 mM ammonium bicarbonate at room temperature for 1 h. The sample was centrifuged for 20 min at 13,000× g , and the supernatant was transferred to a new tube. Then the protein was precipitated by the MeOH/CHCl 3 method, and the protein concentration was measured using an RC DC™ Protein Assay (reducing agent and detergent compatible) (BioRad, Hercules, CA, USA). Protein Digestion: For protein identification, equal amounts of protein from each sample ( n = 3 per group and 4 healthy controls) were loaded onto a 10% SDS-PAGE gel. The resulting condensed protein bands underwent gel digestion using Trypsin and were processed, as previously described in the by our group . Protein Quantification by SWATH-MS Analysis [ , , , ]: To build the MSMS spectral libraries, peptide solutions were analyzed by shotgun data-dependent acquisition (DDA) using micro-liquid chromatography–tandem mass spectrometry (LC-MS/MS), as described in the and previously by our group. The MSMS spectra of the identified peptides were then used to generate the spectral library for SWATH peak extraction using the add-in for PeakView Software (version 2.2, Sciex), MSMS ALL with SWATH Acquisition MicroApp (version 2.0, Sciex). Peptides with a confidence score >99% (obtained from the ProteinPilot database search) were included in the spectral library. For relative quantification by SWATHMS analysis, SWATH-MS acquisition was performed on a Triple TOF 6600 LC–MS–MS system (Sciex) using SWATH mode. The acquisition mode consisted of a 250 ms survey MS scan from 400 to 1250 m / z , followed by an MSMS scan from 100 to 1500 m / z (25 ms acquisition time) of the top 65 precursor ions from the survey scan, for a total cycle time of 2.8 s. The fragmented precursors were then added to a dynamic exclusion list for 15 s. Any singly charged ions were excluded from the MSMS analysis. Targeted data extraction from the SWATH MS runs was performed by PeakView v.2.2 (Sciex, Redwood City, CA, USA) using the SWATH-MS Acquisition MicroApp v.2.0 (Sciex, USA). Data were processed using the spectral library created from DDA. SWATH-MS quantization was attempted for all proteins in the ion library that were identified by ProteinPilot TM 5.0.1 with a false discovery rate (FDR) < 1%. PeakView computed an FDR and a score for each assigned peptide based on the chromatographic and spectra components: only peptides with an FDR < 1%, 10 peptides, and 7 transitions per peptide were used for protein quantization. The integrated peak areas were processed by MarkerView software version 1.3.1 (Sciex, USA) for a data-independent method for relative quantitative analysis. A most likely ratio normalization was performed to control for possible uneven sample loss across the different samples during the sample preparation process . Unsupervised multivariate statistical analysis using PCA was performed to compare data across samples.
Pathway analysis was performed using Reactome ( https://reactome.org/ , accessed on 20 July 2023), which applies a statistical (hypergeometric distribution) test to determine whether specific pathways are over-represented (enriched) and produces a probability score, which is corrected for FDR using the Benjamini–Hochberg method. The most enriched pathways were represented using Reactome pathway diagrams. Protein interactions were evaluated using STRING ( https://string-db.org/ , accessed on 25 July 2023), applying a minimum required interaction score of PPI = 0.9 (protein–protein interaction) and an FDR < 0.05. Venn diagrams were generated using http://www.interactivenn.net/ , accessed on 2 June 2023) and box plots using GraphPad Prism 9. Statistical analyses were performed using Markerview v 1.3.1 or Scaffold software v 5.2.2. Volcano plots and box plots were generated using GraphPad Prism 9.0.0.
The diagnostic process of ML is complex due to its multiorgan involvement; the diagnostic methods up to now have not been properly established because they are indirect enzymatic methods. Since procedures such as the measurement of M6P-containing hydrolases depend on which one is selected and studied, results can be inconclusive. In this article, we describe a group of enzymes that should always be elevated in this type of disease, such as ASMD, and HS- and DS-degrading enzymes. Multiplex enzyme studies, which are increasingly used in LC-MS/MS techniques, can facilitate diagnosis by simply measuring ASMD and IDUA. In addition, proteomic data are sufficiently robust to demonstrate these enzymes are elevated due to compensatory mechanisms for energy production in cells. This alternative process to generate energy produces negative side effects on cells that become chronic over time. Certain proteins found in this research may be of relevance for the design of future studies and to shed light on the pathology of the disease and the molecular processes involved.
|
A Case of Epithelioid Angiosarcoma Diagnosed From Gross Examination of a Pulmonary Tumor Utilizing Imprint Cytology and Immunocytochemistry | c4185aed-d259-4d53-ac56-92c92307de39 | 11498060 | Anatomy[mh] | Introduction Angiosarcomas represent a rare category of malignant vascular neoplasms, comprising less than 2%–4% of soft tissue sarcomas, officially recognized as a distinct entity in the 2019 World Health Organization classification of tumors of soft tissue and bone . These tumors can manifest in cutaneous sites, soft tissues, and various organs including bone . The histologic presentation of angiosarcoma is highly variable, featuring diverse growth patterns such as papillary, spindled, and epithelioid morphologies . Particularly, a distinct morphologic subtype of angiosarcoma, referred to as epithelioid angiosarcoma (EAS), is characterized by malignant endothelial cells, primarily exhibiting epithelial characteristic . EAS poses a diagnostic challenge due to its morphologic resemblance to carcinoma . Furthermore, vascular tumors such as angiosarcoma often contain a significant amount of blood, complicating the acquisition of sufficient tumor cells for a definitive diagnosis through small preoperative biopsies . Conclusive diagnosis based on cytological material is similarly challenging . While there are reports in the literature of cytological findings or immunocytochemistry (ICC) of EAS, there are no reports of cytology leading to an earlier diagnosis than traditional pathology. One contributing factor to the delay in rapid cytological diagnosis might be the absence of detailed immunocytochemical staining protocols in the literature. The immunostaining protocol for EAS presented in this study addresses this gap, enabling a diagnosis using solely cytological specimens. In this report, we present a case of EAS diagnosed through cytomorphological features and ICC, employing touch imprint cytology prompted by the gross findings observed in a lung resection sample. Case A 48‐year‐old male was admitted to Kanazawa University Hospital on January 9, 2020 due to an enlarged and painful mass on his left back. He had been experiencing discomfort without pain or visible abnormalities in his left dorsal region for over 10 years. Recently, he noticed a mass and pain in the left dorsal region. A contrast‐enhanced CT scan revealed osteolytic changes on the dorsal aspect of the left 11th rib, along with a substantial mass measuring 73 mm. Additionally, masses were detected in the mediastinal and lung areas (Figure ). Considering the clinical course and imaging findings, potential differentials included pulmonary metastases of tumors in the ribs, multiple metastases of lung cancer, and multiple metastases of cancer of unknown origin. The patient underwent an echo‐guided percutaneous biopsy of the back tumor. Histological examination revealed a few atypical cells with poor connectivity against a necrotic background. Immunostaining results showed positive expression for AE1/AE3 and CK7, with focal positivity for CAM5.2. However, they tested negative for EMA, Desmin, CD34, S100, and TTF‐1. Consequently, based on these findings, the biopsy diagnosis raised suspicion of metastasis from an undifferentiated carcinoma of unknown origin. To investigate the possibility of primary lung cancer, the patient underwent a partial resection of the lower lobe of the right lung. The gross examination of the resected lung revealed a dark reddish‐brown hemorrhagic nodule (Figure ). Considering the gross findings, we contemplated the potential for a vascular neoplasm and proceeded to prepare a stamp‐imprinted cytology specimen of the tumor. The neoplasm was stamped onto glass slides and fixed in either 95% alcohol or air‐dried. The alcohol‐fixed slide was treated with Papanicolaou stain, while the air‐dried slide was treated with Giemsa stain. Cytological analysis revealed the presence of small to medium‐sized clusters and isolated scattered atypical cells on a bloody background. Some of these clusters exhibited rosette‐like structures (Figure ). The atypical cells exhibited an epithelioid or plasmacytoid morphology with fine to coarse vacuolated cytoplasm (Figure ). The nuclei demonstrated pleomorphism, characterized by gyrate nuclei, and prominent nucleoli (Figure ). Additionally, some atypical cells exhibited mitosis (Figure ), and the cytoplasm contained erythrocytes, a phenomenon known as erythrophagocytosis (Figure ). The cytological findings collectively suggested the type of tumor was an epithelioid vascular neoplasm. Moreover, the highly atypical morphology and observed mitotic activity raised suspicion of EAS, with adenocarcinoma considered as a potential differential diagnosis. To distinguish angiosarcoma from adenocarcinoma, immunocytochemical staining with anti‐AE1/AE3 antibody and endothelial markers, including anti‐CD31 antibody, anti‐ERG antibody, and anti‐FLI‐1 antibody, was performed on the cytology specimen (Table ). The immunocytochemical results demonstrated the positivity of atypical cells for CK AE1/AE3, CD31, ERG, and FLI‐1, supporting the diagnosis of EAS (Figure ). In this case, the diagnosis was facilitated through the combination of stamp cytology and ICC, complementing the gross findings from the surgically resected lung. The subsequent histological diagnosis confirmed angiosarcoma based on both histological and immunohistochemical evidence. The histological examination revealed the presence of atypical cells with irregular nuclei and prominent nucleoli in a hemorrhagic background (Figure ). The tumor cells exhibited the formation of thin‐walled lumens, within which hobnail‐like atypical cells were identified, and the lumens were filled with blood (Figure ). Immunostaining further supported the diagnosis by demonstrating the positivity of atypical cells for CD31, ERG, FLI‐1, and AE1/AE3 (Figure ). Consequently, these observations led to the establishment of a definitive diagnosis of EAS. Discussion We encountered a case where a vascular neoplasm was suspected based on the gross findings of the surgical specimen, and the diagnosis of EAS was established through a combination of touch imprint cytology and ICC. Previous studies have reported very few cases of EAS diagnosed solely on cytological specimens, including Geller et al. and Klijanienko et al. . This report highlights the potential for diagnosing EAS using cytological samples, which could be valuable in situations where obtaining tissue samples is not feasible. While the cytological diagnosis of EAS is highly challenging, the ability to suggest its presence is crucial for achieving an accurate diagnosis. EAS is characterized by round to oval cells, polygonal shapes, epithelioid clusters, erythrophagocytosis, and backgrounds with hemorrhage and neutrophils . Erythrophagocytosis is reported to be a less sensitive but highly specific finding . Furthermore, Wakely, Frable, and Kneisl have reported the presence of cytoplasmic vacuoles . In this case, all of these cellular findings were observed. Additionally, Liu and Layfield emphasize hemosiderin deposition in malignant cells as a diagnostic feature of angiosarcoma; however, it was not observed in this case . Preoperative diagnosis of EAS based solely on cytological material is infrequent . However, with careful consideration of clinical and imaging findings, we posit that a definitive diagnosis can be achieved through cytology and by correctly selecting antibodies for angiosarcoma diagnosis via immunostaining (Table ) . The key to diagnosing EAS through immunostaining lies in the careful selection of antibodies. Firstly, given the frequent expression of epithelial markers in EAS , they should be included in the immunostaining panel. In particular, CK AE1/AE3 is positive in many EAS cases . Additionally, CD31, ERG, and Fli‐1 should incorporated as useful endothelial markers for EAS diagnosis . Wu, Li, and Liu reported that CD31 and Fli‐1 exhibit greater sensitivity as markers for endothelial cells compared with CD34 and Factor VIII‐related antigen . Sullivan et al. reported that ERG demonstrates comparable sensitivity to CD31 and is valuable in diagnosing angiosarcoma . Establishing evidence of vascular differentiation in epithelioid tumor cells through positive staining for endothelial markers is particularly crucial for diagnosing EAS. In this case, CK7 positivity was observed in immunohistochemistry (IHC) conducted during the biopsy. While Lin et al. reported CK7 positivity in thyroid EAS, to the best of my knowledge, this is the only report of CK7 positivity in EAS . The significance of this CK7 positivity remains unclear, and accumulation of further cases is necessary to clarify this issue. When performing the definitive diagnosis of EAS using cytological material, the reliability of ICC becomes crucial. Unlike IHC, ICC lacks established staining methods, which can lead to potential false negatives and false positives. To address this, IHC was also performed on tissue specimens from the same case. False negatives and false positives were eliminated by comparing staining characteristics between IHC and ICC, confirming specific staining by observing positivity with the same antibodies and similar cellular localization. Interestingly, ICC yielded clearer positive images compared with IHC (Table ; Figures and ). The differential diagnoses for EAS include poorly differentiated carcinoma, melanoma, mesothelioma, epithelioid hemangioendothelioma (EHE), and other epithelioid sarcomas. Carcinoma typically shows negativity for endothelial markers, while malignant mesothelioma is negative for endothelial markers but positive for mesothelial markers such as calretinin and D2‐40. EHE typically exhibits positivity for CAMTA1 or TFE3 . Melanoma is negative for endothelial markers but positive for melanocytic markers such as Melan‐A, S‐100, and HMB‐45. Our report underscores that with a comprehensive understanding of the cytological characteristics of EAS and the appropriate immunostaining panel, a reliable diagnosis of EAS can be achieved using cytological samples alone. Positive results for both endothelial and epithelial markers, even with a limited quantity of collected cells, significantly contribute to the diagnosis of EAS. The most important aspect of this paper is that it details the method of immunocytochemical staining using cytology specimens for the diagnosis of angiosarcoma. Providing a detailed description of the immunocytochemical staining method is important to enable diagnosis based solely on cytology. While previous reports have described the results of immunocytochemical staining and methods of concurrent immunohistochemical staining, no papers have detailed the methods of immunocytochemical staining specifically for diagnosing angiosarcoma . By using this method, a definitive diagnosis can be made from cytology specimens even in cases where tissue sampling is difficult. In conclusion, this report provides a diagnostic method for EAS using cytology. We prepared touch imprint cytology specimens from lung tumors excised during surgery and demonstrated the immunocytochemical staining method. When EAS is clinically and cytologically suspected, performing immunocytochemical staining using our described method enables a definitive diagnosis of EAS based solely on cytology. Tatsuya mori: conceptualization (equal), investigation (equal), writing – original draft (lead). Keishi Mizuguchi: conceptualization (equal), methodology (lead), writing – review and editing (lead). Chie Shimaguchi: investigation (equal). Kaori Sakano: investigation (equal). Tsubasa Shimoda: investigation (equal). Urara Okawa: investigation (equal). Miyu Okuda: investigation (equal). Mayo Usui: investigation (equal). Hiroko Ikeda: supervision (lead). The study adhered to the tenets of the Declaration of Helsinki. The retrospective case report described in this manuscript includes no patient‐identifying information. Ethical approval was not required. We obtained a written statement of informed consent from the patient for the publication of case details and the use of images. The authors declare no conflicts of interest. |
Academic benchmarks for leaders in Otolaryngology - Head & Neck Surgery: a Canadian perspective | 266dfc34-3c7b-4cae-91ff-af0a5e64131b | 7201551 | Otolaryngology[mh] | In 2015, Eloy and colleagues published a cross-sectional overview of academic leaders in Otolaryngology – Head & Neck Surgery (OHNS). This study examined the profile of academic leaders including department chairpersons (CPs), vice CPs, and program directors (PDs) across 99 OHNS programs in the United States. Their findings centered on research and publication impact, additional training including fellowships and graduate degrees, and gender distribution across leadership positions. The study noted that CPs had more years of experience, higher publication impact, and increased research funding, with low representation of women in leadership positions, accounting for only 3% of CPs and 15% of PDs . A similar overview has not been conducted in the Canadian setting. Examination of the characteristics and productivity of Canadian academic leaders will provide an understanding of the pathways for promotion in a Canadian context. It may also reveal possible issues with diverse representation within academic leadership. Therefore, the aim of the present study is to summarize the demographics, subspecialty training, and academic productivity of contemporary leaders in OHNS training programs across Canada. A secondary objective is to evaluate the impact of academic appointment on research productivity.
Demographic data on departmental CPs and PDs from 13 accredited Canadian OHNS programs were obtained from publicly-available faculty listings after July 1, 2019. Current leaders were identified from the Canadian Society of Otolaryngology – Head & Neck Surgery ( https://www.entcanada.org/cso/canadian-departments-otolaryngology/ ) and Royal College of Physicians and Surgeons of Canada ( http://www.royalcollege.ca/rcsite/documents/arps/otolaryngology-e ) websites and validated using departmental websites. Data included employment institution, location of residency and/or fellowship training, nature of subspecialty training, gender, and years of experience following medical school graduation. Two individuals (one PD and one CP) with general OHNS subspecialty training reported on faculty listings were categorized according to their focus of clinical practice (pediatric otolaryngology and otology, respectively) after consensus agreement among all authors. Research impact and productivity were measured using the h-index and average annual number of publications obtained from Scopus. The h-index is a composite score that incorporates both the number of publications and number of citations of publications for an individual [ – ]. Individuals with multiple listings on Scopus were cross-referenced and non-duplicate publications were summed. In the event of multiple listings, the highest reported h-index was adopted. Data unavailable from public listings were supplemented with direct communication with individual physician leaders. Statistical analysis was used to explore the strength of correlation between training and employment location using a Cramér’s V. Characteristics were compared between chairs and program directors using a t-test for continuous variables and a Chi-Square or Fisher’s exact test for categorical variables. To achieve the secondary objective of evaluating the impact of academic appointment on research productivity, the average number of publications over a standardized measure of time before and after academic appointment was plotted and compared using a t-test. Inclusion criteria for this analysis included appointment prior to January 2018 (allowing at least 1.5 years of post-appointment data) and h-index ≥ 1. Data was aggregated using Microsoft Excel 2018 (Microsoft Corp., Redmond, Washington), and statistical analyses were performed using SPSS version 24.0 (SPSS Inc., Chicago, Illinois). A p -value less than 0.05 was considered statistically significant.
Data was gathered from all 13 accredited OHNS training programs, with information available from 27 academic leaders, summarized in Table . There were 13 CPs and 14 PDs (one training program had 2 PDs). All academic leaders were fellowship trained. Among CPs, head and neck surgical oncology (77%), facial plastic surgery (15%), and laryngology (15%) were the most common areas of fellowship training (Fig. a). Among PDs, head and neck surgical oncology (43%) and pediatric otolaryngology (43%) were the most common fellowship-trained subspecialties (Fig. b). Head and neck surgical oncology (59%) was the most common area of subspecialty training among academic leaders overall. Females represented 3 of 14 (21%) PDs, none of the CPs, and altogether only 11% of the academic leaders overall (Fig. ). There was a significant association between location of residency training and employment, with 56% (15/27) of physicians working where they had trained ( p = 0.001, Fisher’s exact test; φ = 2.63, p = 0.001). Compared to PDs, CPs were more likely to have achieved the academic rank of full professor (77% vs. 21%). The remainder of the CPs were associate professors (23%). Overall, there was a significant difference in the composition of academic ranks ( p = 0.008; Fig. ). Not surprisingly, CPs accrued significantly more years of post-graduate experience (29.7 vs. 21.3 years, p = 0.008) and higher mean h-index scores (14.5 vs. 8.14, p = 0.04; Fig. ). In total, 9 academic leaders (33%) earned a graduate (Master’s) degree, including 5 CPs (38%) and 4 PDs (29%). There was no difference in the proportion of CPs and PDs with graduate degrees ( p = 0.51). Individuals with a graduate degree had a significantly higher h-index (17.7 vs 7.4, p = 0.001) and average number of publications (106 vs. 52, p = 0.02) compared to those without a graduate degree. The average number of publications over a standardized measure of time before/after academic appointment was plotted for 19 of 27 (70%) academic leaders meeting our inclusion criteria. We found that there were no significant differences before and after academic appointment (1.8 vs. 2.0, p = 0.64). However, we noted subjective trends of increasing research productivity leading up to appointment and peaking 3 years after appointment, followed by a notable decline in productivity (Fig. ).
This is the first study to review the characteristics of academic leaders in Canadian OHNS programs. The primary goal of our study was to summarize the demographics, training characteristics, and academic productivity of contemporary leaders in OHNS training programs across Canada, with a secondary objective of evaluating the impact of academic appointment on research productivity. Our study found that all CPs and PDs had at least 1 year of fellowship subspecialty training. This is not surprising as care is often subspecialized within academic centers in Canada. This finding may reflect the trend towards additional fellowship training as a prerequisite for academic practice over the past two decades [ , , ]. The most common area of fellowship training was head and neck surgical oncology among both CPs and PDs. This finding was consistent with the study by Eloy et al., and reflects the historical influence and longevity of head and neck surgical oncology as the one of the earliest OHNS subspecialties to emerge . Pediatric otolaryngology was also heavily represented among PDs, and may reflect the fact that most academic pediatric otolaryngologists are remunerated through an alternate funding plan. In particular, salary-based compensation may afford pediatric subspecialists comparatively greater time and opportunity to pursue administrative and academic leadership roles. The impact of physician renumeration on academic involvement and research productivity was not formally evaluated in this study due to lack of publicly available data and could be explored with future research. CPs had a significantly greater number of publications, a higher h-index , more years of post-graduate experience, and a higher proportion of full professors compared to PDs. Again, this finding was consistent with the previous study of U.S. OHNS programs . This finding is not surprising as CPs typically represent more senior physicians with more clinical and academic experience. In several instances, previous PDs have transitioned to the role of CP over time. Our study found that academic leaders with a Master’s degree had, on average, a significantly higher h-index score and greater number of publications, suggesting an added value to graduate-level training for acquiring additional skills and increasing academic productivity. However, the potential benefits of a Master’s degree must be balanced with the significant time and monetary investment, and therefore remains a highly individual- and institutional-dependent decision. The secondary objective of this study was to examine the impact of academic appointment on research productivity, as measured by average annual publication rates before and after academic appointment. Although there was no significant difference in productivity between periods, we observed a trend of increasing productivity prior to appointment that peaked at 3 years post-appointment. In addition, there was a general decline over time beyond 3 years post-appointment. The observed trends must be interpreted cautiously due to the small sample size. One could speculate that the observed peak at 3 years post-appointment may correspond with an institutional three-year review or the completion of previous work leading up to the academic appointment. The subsequent decrease in productivity may be secondary to the increased administrative responsibility associated with leadership roles. On the other hand, some leaders were successful in maintaining their research productivity by incorporating new research interests in complementary areas such as medical education or curriculum development. The impact of administrative burden on productivity has particular importance to aspiring leaders considering the optimal time to pursue a leadership appointment. Data published by the Canadian Medical Association (2018) has estimated that females represent 23% of practicing OHNS surgeons in Canada . Despite females accounting for a minority of Canadian OHNS surgeons, our study found that even fewer still are represented among academic leaders and, more broadly, in academic otolaryngology. A review of faculty listings across Canadian academic centers, at the time of this writing, revealed that females represent approximately 19% of all academic otolaryngologists across Canada. Furthermore, our present study demonstrated that females represented only 11% of Canadian OHNS leaders overall. This figure is consistent with U. S studies which found that females were represented in 7 to 10% of American OHNS academic leadership roles . Choi et al. surveyed over 10,000 American Academy of Otolaryngology – Head & Neck Surgery members in 2010 and found that women represented only 4% of CPs and 12% of PDs . In contrast to our study, Choi et al. noted that women were, in fact, proportionally represented after accounting for the number of females practicing OHNS. They also noted that women were proportionally represented in journal and special society leadership positions, and that female OHNS surgeons were relatively younger compared to their male counterparts, with over 63% of females under the age of 45 years . This discrepancy can be explained by geographic variation in training, certification, employment opportunities, and patient population. Despite the underrepresentation of females in otolaryngology and academic leadership, there is evidence that the gender proportions are becoming more balanced. The 23% female proportion of Canadian OHNS surgeons shown in 2018 CMA report represents a significant increase from the previous figure of less than 10% in the early 2000’s . Representation is expected to further increase as females currently outnumber males entering medical school (56%), and represent approximately 40% of Canadian OHNS residents and 47% of surgical residents . We must continue to monitor the representation of females in our specialty and ensure equitable representation through ongoing efforts to attract and retain women in leadership positions. Study Limitations The cross-sectional study design provided a snapshot of the current academic leadership profile and did not examine trends over time or describe the characteristics of other otolaryngologists across Canada. Data on h-index and academic productivity were gathered from Scopus similar to the methods of Eloy et al. ; these data are subject to potential variability depending on the type of citation tool used (i.e. web of science, Google scholar, etc.) . Moreover, the study was not designed to consider confounding variables such as factors impacting choice of training, employment location, and interest in research. Furthermore, the study did not examine other aspects of leadership such as participation in teaching, committees, and clinical leadership roles. A more thorough review of leaders’ curriculum vitae would have captured a more robust assessment of their academic productivity. Future qualitative studies with a survey design and/or semi-structured interviews could also be used to capture more granular aspects of academic leadership.
The cross-sectional study design provided a snapshot of the current academic leadership profile and did not examine trends over time or describe the characteristics of other otolaryngologists across Canada. Data on h-index and academic productivity were gathered from Scopus similar to the methods of Eloy et al. ; these data are subject to potential variability depending on the type of citation tool used (i.e. web of science, Google scholar, etc.) . Moreover, the study was not designed to consider confounding variables such as factors impacting choice of training, employment location, and interest in research. Furthermore, the study did not examine other aspects of leadership such as participation in teaching, committees, and clinical leadership roles. A more thorough review of leaders’ curriculum vitae would have captured a more robust assessment of their academic productivity. Future qualitative studies with a survey design and/or semi-structured interviews could also be used to capture more granular aspects of academic leadership.
This study provides a cross-sectional overview of academic leaders in Canadian OHNS programs, and demonstrated the following key findings: 1) fellowship training was universal; 2) head and neck surgical oncology is the most common fellowship training subspecialty; 3) leaders were more likely to be employed at the institution where they trained; 4) a graduate (Master’s) degree may be associated with increased research productivity; 5) there is a potential risk of decreased academic productivity after appointment to a leadership position; and 6) women are currently underrepresented in academic leadership roles.
|
Parenting Styles, Food Parenting Practices, Family Meals, and Weight Status of African American Families | 038b641d-0639-489e-a9dd-8117d4ac9aab | 9864142 | Family Medicine[mh] | Obesity is a major public health issue and a growing concern universally. The obesity prevalence is around 21% among older adolescents aged 12–19 years, which is higher than that of younger children and adolescents . The National Center for Health Statistics reported that the obesity prevalence is higher among African American youth in comparison to White youth . Moreover, overweight and obese children and adolescents are more likely to be overweight and develop a chronic disease in adulthood . Environmental, behavioral, and personal factors exert an influence on eating habits and behaviors as an element involved in developing or preventing obesity in children and adolescents . Lifestyle choices, psychological factors, family factors, and socioeconomic factors are the most remarkable etiologies for childhood obesity . Parenting (or caregiver) styles (PSs), food parenting practices (FPPs), and frequency of family meals are major factors among different environmental factors that impact children and adolescents as the first and most influential community that they join. PSs refer to the engagement and responsiveness level of parents in different situations with their child. FPPs are postulated to impact children’s eating behaviors . FPPs were also identified as a predictor of children and adolescents’ health outcomes in adulthood . A high frequency of family meals provides several benefits for families, which include improving weight status and promoting healthy eating habits . An absence of family meals is associated with unhealthy eating patterns and poor diet quality . Furthermore, there is a negative association between the frequency of family meals and obesity development . However, the question of how eating meals together as a family is related to other aspects of the family environment, such as different PSs and FPPs, and whether this relationship is associated with obesity status among African American minority groups is unresolved. African American parents were shown to use an authoritarian PS, which is characterized by high restriction and monitoring of children’s food consumption . Many African American children and adolescents do not meet the recommended dietary intake of fruits, vegetables, and whole grains due to low socioeconomic status, which can lead to a higher risk of obesity . Family system theory (FST) was used as the theoretical framework for this study. This theory emphasizes the importance of the family as a system to understand and explain individual behaviors in the context of family interactions . FST suggests that any change in family structure or the role of family members can have an impact on the behavior of the entire family over time . Previous studies revealed that a warm and supportive PS correlates with the number of desirable healthy behaviors practiced. This can impact adolescent weight status and dietary patterns . In contrast, restrictive FPPs, such as pressure to eat, restrictions on youth’s access to foods, and parental concerns about adolescents’ weight status, are associated with poorer diet quality . There are very few studies addressing the impact of the family environment on the obesity status of both parents and adolescents, especially among minorities. The importance of this work is to study the effect of three influential factors of family environment, including PSs, FPPs, and family meal frequency together on obesity status among African American families. This study becomes even more important considering that obesity is a major problem among adolescents of minority groups, especially African Americans. The goal of this study is to help elucidate which family environmental factors have a positive impact on controlling the weight status of African American families and communities, and to determine which PSs and/or FPPs may lead to a higher family meal frequency. The results indicate higher family meal frequency with positive correlation with healthier weight status among African American adolescents, and authoritative PS and monitoring, reasoning, and modeling FPPs with higher frequency of family meals.
2.1. Research Design, Participants, and Procedure The protocol of the current study was approved by the institutional review board (IRB) of the University of the District of Columbia. A total of 211 African American parent–adolescent dyads participated in this cross-sectional study. The dyads were recruited by Qualtrics from November to December 2021 to complete the survey. The inclusion criteria included the following: parents or caregivers willing to participate in the study with their 10–17-year-old adolescents; access to the Internet; being comfortable reading and writing in English; being responsible for providing food for the adolescent. All participants signed a parental consent or adolescent assent form before participating as a prerequisite for the survey. 2.2. Parents’ Survey The parents completed a 20–25 min online survey. The survey used questions from the 85-item Comprehensive General Parenting Questionnaire (CGPQ), which facilitate research exploring how parenting impacts a child’s weight-related behaviors and items used by a Monroe-Lord et al. (2021) study on African American families . The demographic characteristics of both the adolescents and the parents, household food security, household acculturation, and participation in federal food assistance programs were evaluated. The parents self-reported anthropometric measurements including height and weight for themselves and their children. Body mass index (BMI) for parents and BMI percentile for adolescents were used in this study. BMI was calculated with participants’ weight divided by the square of height used for parents. As BMI increases with age during childhood and adolescence, and it is different between males and female, BMI-for-age percentile based on CDC growth charts were used for obesity status for adolescents. BMI was categorized into three groups, including normal weight if BMI was between 18.5 and 24.9, overweight if BMI was 25.0–29.9, and obesity if BMI was 30.0 and above. BMI percentile also was categorized into three groups, including normal weight if the BMI percentile was equal to or greater than the 5th percentile and less than 85th percentile, overweight if BMI percentile was at or the 85th percentile but less than the 95th percentile, and obesity if BMI percentile was at or above the 95th percentile for specific age, gender, and height . The survey also included the following question: “During the past 7 days, how many times did all, or most, of your family in your house eat a meal together?” The answer included six options, from “Never” to “More than 7 times.” For this study, frequency of family meals was categorized into three groups after combining the answer choices, namely, two times or less, three to six times, and 7seven times or more . 2.3. Statistical Analysis Two exploratory factor analyses were run to identify the PSs and FPPs. Once the factors were identified, average factor scores for each parent were calculated. Spearman’s rank correlation (when weight status was considered as a continuous variable) and the Wilcoxon rank sum test (when weight status was considered as a categorical variable) were used to test the relationship between BMI percentile and BMI for both the adolescents and the parents, respectively, and PSs and FPPs. The Wilcoxon rank sum test was used to examine the relationships between family meal frequency and weight status, as well as PSs and FPPs. Spearman’s correlation was used to test the relationship between adolescent BMI percentile and parental BMI. SAS 9.4 (SAS Institute, Cary, NC, USA) was used for statistical analysis in this study. The results are considered significant at p < 0.05.
The protocol of the current study was approved by the institutional review board (IRB) of the University of the District of Columbia. A total of 211 African American parent–adolescent dyads participated in this cross-sectional study. The dyads were recruited by Qualtrics from November to December 2021 to complete the survey. The inclusion criteria included the following: parents or caregivers willing to participate in the study with their 10–17-year-old adolescents; access to the Internet; being comfortable reading and writing in English; being responsible for providing food for the adolescent. All participants signed a parental consent or adolescent assent form before participating as a prerequisite for the survey.
The parents completed a 20–25 min online survey. The survey used questions from the 85-item Comprehensive General Parenting Questionnaire (CGPQ), which facilitate research exploring how parenting impacts a child’s weight-related behaviors and items used by a Monroe-Lord et al. (2021) study on African American families . The demographic characteristics of both the adolescents and the parents, household food security, household acculturation, and participation in federal food assistance programs were evaluated. The parents self-reported anthropometric measurements including height and weight for themselves and their children. Body mass index (BMI) for parents and BMI percentile for adolescents were used in this study. BMI was calculated with participants’ weight divided by the square of height used for parents. As BMI increases with age during childhood and adolescence, and it is different between males and female, BMI-for-age percentile based on CDC growth charts were used for obesity status for adolescents. BMI was categorized into three groups, including normal weight if BMI was between 18.5 and 24.9, overweight if BMI was 25.0–29.9, and obesity if BMI was 30.0 and above. BMI percentile also was categorized into three groups, including normal weight if the BMI percentile was equal to or greater than the 5th percentile and less than 85th percentile, overweight if BMI percentile was at or the 85th percentile but less than the 95th percentile, and obesity if BMI percentile was at or above the 95th percentile for specific age, gender, and height . The survey also included the following question: “During the past 7 days, how many times did all, or most, of your family in your house eat a meal together?” The answer included six options, from “Never” to “More than 7 times.” For this study, frequency of family meals was categorized into three groups after combining the answer choices, namely, two times or less, three to six times, and 7seven times or more .
Two exploratory factor analyses were run to identify the PSs and FPPs. Once the factors were identified, average factor scores for each parent were calculated. Spearman’s rank correlation (when weight status was considered as a continuous variable) and the Wilcoxon rank sum test (when weight status was considered as a categorical variable) were used to test the relationship between BMI percentile and BMI for both the adolescents and the parents, respectively, and PSs and FPPs. The Wilcoxon rank sum test was used to examine the relationships between family meal frequency and weight status, as well as PSs and FPPs. Spearman’s correlation was used to test the relationship between adolescent BMI percentile and parental BMI. SAS 9.4 (SAS Institute, Cary, NC, USA) was used for statistical analysis in this study. The results are considered significant at p < 0.05.
3.1. Demographic Analysis The details of the sample characteristics are presented in . The adolescent sample was composed of 41% male and 59% female individuals with a mean age of 14.28 years. The mean BMI percentile was 71.35. The obesity rate of the adolescents was 19.6% (the national estimate for African American youth is 22%). Approximately 82% of the caregiver participants in this study were the parents of the adolescents (we use the term parents for them). Most of the parents were female (70%), and 57% of the parents were overweight or obese. Approximately 56% of the parents had a college education or above. Furthermore, approximately 52% of the adolescents lived in single-parent households, and more than half of the families had family meals three to six times per week. 3.2. Parenting Styles and Food Parenting Practices Two exploratory factor analyses were run for sets of 35 (for PSs) and 33 (for FPPs) items. The first factor analysis for the identification of PSs produced four factors, which were named authoritative, authoritarian, setting rules/expectations, and neglecting. One item was excluded because it did not load with any of the other four factors. The items for each PS are listed in . A second factor analysis was run for the identification of FPPs, at which point five items were excluded from the final factors. The analysis produced four factors, which were named monitoring, reasoning, copying, and role modeling. Monitoring is defined as parents keeping track of what and how much their children eat. Reasoning or teaching is defined as parents reasoning with the child about the benefits of healthy food and teaching them healthy eating habits. Copying is defined as when parents intentionally or unintentionally encourage the child to copy their eating behaviors. Role modeling is defined as parents exhibiting healthy eating behaviors to encourage similar behaviors in their children. The items for each FPP factor are listed in . The factor loadings and the details of the factor analyses for each PS and FPP are shown in . The internal reliability of each factor was good (Cronbach’s alpha > 0.8) or acceptable (Cronbach’s alpha > 0.7) for all PSs and FPPs. The parents received a score on all eight factors. The highest median scores were for the authoritative and setting rules PSs. Setting rules, authoritative, neglecting, and authoritarian were the PSs applied by African American parents most prevalently and, respectively, while role modeling, copying, reasoning, and monitoring were used most prevalently and, respectively, as FPPs. 3.3. Relationship of Different Demographic Data with Weight Status of Both Adolescents and Parents The relationship between the weight status of both parents and adolescents and different demographic variables were examined and reported in . Based on the results, both adolescent and parent sex were meaningfully related to BMI percentile of adolescent ( p = 0.012 and p = 0.0485, respectively). Fewer male adolescents (40.6%) were in the normal weight group compared with female adolescents (63.5%), and adolescents whose main caregiver was female were in the better weight status compared with those whose caregiver was male. Male parents significantly had higher BMI compared to female parents ( p = 0.0081). In addition, there was a significant trend toward lower BMI among the younger parents ( p = 0.0286). Interestingly, we could not find any relationship between socioeconomic factors, including parental education and household income, and obesity status of both parent and adolescents of African American. 3.4. Relationship of Parent Weight Status with Adolesscent Weight Status The relationship between parent BMI and adolescent BMI percentiles was examined. The relationship between parent BMI and adolescent BMI percentiles was examined. A highly significant correlation between parent BMI and adolescent BMI percentile was found (r = 0.42, p < 0.0001). 3.5. Relationship of PSs, FPPs, and Family Meal Frequency with Adolescents’ Weight Status BMI percentile was considered a categorical variable (normal weight, overweight, and obese) to evaluate whether the PSs and FPPs are correlated with the obesity status of the African American adolescents. No meaningful relationship was found between the categorized adolescents’ BMI percentiles and PSs and FPPs . In addition, no correlation was found between parent’s weight status and PSs, FPPs, and family meal. Family meal frequency was associated with the adolescents’ BMI percentile ( p = 0.03). The median BMI percentile score was 87.06, which indicates overweight, for those adolescents with two or fewer family meals, while it was 62.45, which indicates a normal weight, for those adolescents with more than seven family meals per week. Although no significant correlation was found between parental BMI and family meal frequency ( p = 0.33), there was a positive trend, with a decrease in parental BMI when having more family meals . 3.6. Relationship of Family Meal Frequency with PSs and FPPs Among different studied PSs, only the authoritative PS was positively related to family meal frequency ( p = 0.0004). The authoritative score was one score higher in those families with seven or more family meals compared to those with two or fewer family meals. However, among the four different FPPs, three of them—monitoring, reasoning, and modeling—were correlated with the frequency of family meals ( p = 0.0002, p = 0.0017, and p = 0.0008, respectively) .
The details of the sample characteristics are presented in . The adolescent sample was composed of 41% male and 59% female individuals with a mean age of 14.28 years. The mean BMI percentile was 71.35. The obesity rate of the adolescents was 19.6% (the national estimate for African American youth is 22%). Approximately 82% of the caregiver participants in this study were the parents of the adolescents (we use the term parents for them). Most of the parents were female (70%), and 57% of the parents were overweight or obese. Approximately 56% of the parents had a college education or above. Furthermore, approximately 52% of the adolescents lived in single-parent households, and more than half of the families had family meals three to six times per week.
Two exploratory factor analyses were run for sets of 35 (for PSs) and 33 (for FPPs) items. The first factor analysis for the identification of PSs produced four factors, which were named authoritative, authoritarian, setting rules/expectations, and neglecting. One item was excluded because it did not load with any of the other four factors. The items for each PS are listed in . A second factor analysis was run for the identification of FPPs, at which point five items were excluded from the final factors. The analysis produced four factors, which were named monitoring, reasoning, copying, and role modeling. Monitoring is defined as parents keeping track of what and how much their children eat. Reasoning or teaching is defined as parents reasoning with the child about the benefits of healthy food and teaching them healthy eating habits. Copying is defined as when parents intentionally or unintentionally encourage the child to copy their eating behaviors. Role modeling is defined as parents exhibiting healthy eating behaviors to encourage similar behaviors in their children. The items for each FPP factor are listed in . The factor loadings and the details of the factor analyses for each PS and FPP are shown in . The internal reliability of each factor was good (Cronbach’s alpha > 0.8) or acceptable (Cronbach’s alpha > 0.7) for all PSs and FPPs. The parents received a score on all eight factors. The highest median scores were for the authoritative and setting rules PSs. Setting rules, authoritative, neglecting, and authoritarian were the PSs applied by African American parents most prevalently and, respectively, while role modeling, copying, reasoning, and monitoring were used most prevalently and, respectively, as FPPs.
The relationship between the weight status of both parents and adolescents and different demographic variables were examined and reported in . Based on the results, both adolescent and parent sex were meaningfully related to BMI percentile of adolescent ( p = 0.012 and p = 0.0485, respectively). Fewer male adolescents (40.6%) were in the normal weight group compared with female adolescents (63.5%), and adolescents whose main caregiver was female were in the better weight status compared with those whose caregiver was male. Male parents significantly had higher BMI compared to female parents ( p = 0.0081). In addition, there was a significant trend toward lower BMI among the younger parents ( p = 0.0286). Interestingly, we could not find any relationship between socioeconomic factors, including parental education and household income, and obesity status of both parent and adolescents of African American.
The relationship between parent BMI and adolescent BMI percentiles was examined. The relationship between parent BMI and adolescent BMI percentiles was examined. A highly significant correlation between parent BMI and adolescent BMI percentile was found (r = 0.42, p < 0.0001).
BMI percentile was considered a categorical variable (normal weight, overweight, and obese) to evaluate whether the PSs and FPPs are correlated with the obesity status of the African American adolescents. No meaningful relationship was found between the categorized adolescents’ BMI percentiles and PSs and FPPs . In addition, no correlation was found between parent’s weight status and PSs, FPPs, and family meal. Family meal frequency was associated with the adolescents’ BMI percentile ( p = 0.03). The median BMI percentile score was 87.06, which indicates overweight, for those adolescents with two or fewer family meals, while it was 62.45, which indicates a normal weight, for those adolescents with more than seven family meals per week. Although no significant correlation was found between parental BMI and family meal frequency ( p = 0.33), there was a positive trend, with a decrease in parental BMI when having more family meals .
Among different studied PSs, only the authoritative PS was positively related to family meal frequency ( p = 0.0004). The authoritative score was one score higher in those families with seven or more family meals compared to those with two or fewer family meals. However, among the four different FPPs, three of them—monitoring, reasoning, and modeling—were correlated with the frequency of family meals ( p = 0.0002, p = 0.0017, and p = 0.0008, respectively) .
In this study, the relationship between different parental influences (i.e., PSs, FPPs, and family meals) and African American families’ obesity status was evaluated. Notably, the existing literature examining the influence of PSs and FPPs on obesity status among both parents and adolescents in a minority population, specifically African American, is sparse. The findings of this study reveal that African American families establish set rules and expectations more than other PSs, and authoritarian was the least prevalent PS. A previous study revealed that PS characterized by rigidity, restriction, and high control, which are classed as authoritarian styles, is more prevalent among African Americans . This type of PS evokes a sense of safety and nurturance among adolescents . Setting a large number of rules and expectations is a form of behavioral control by parents, which can be perceived as an authoritarian style by adolescents. Thus, it can play a negative role in health behaviors among adolescents, instead of resulting in improvements in their health status. It is important to have a supportive and alternative plan for adolescents while establishing rules and expectations. Moreover, role modeling was the dominant FPP among the African American adolescents in this study, while monitoring was the least prevalent. This finding is consistent with a previous study with a small sample size that claimed a higher score for role modeling compared to other FPPs among African American families . This study could not find any relationship between the BMI percentiles/BMI of the adolescents/parents and PSs and FPPs. This finding is consistent with two recent studies and one older study that reported no specific correlation between FPPs and being overweight or obesity in children and adolescents . In addition, two other studies confirmed that maternal weight status is an independent factor of FPPs . However, a previous study showed that greater parental responsiveness, which is characteristic of an authoritative PS, is significantly correlated with a lower BMI percentile . These different findings regarding the correlation between authoritative PSs and obesity status could result from the different questionnaire design used in the two studies. It is important to note that although we could not find any statistically meaningful correlation, the association between authoritative PSs and the obesity status of the adolescents was negative. It is important to consider the impact of PSs or FPPs on obesity status in adulthood. Its impact can be highlighted in the future of adolescents. Family meal frequency was associated with the weight status of the adolescents. The adolescents who had more family meals per week had a lower BMI percentile. Although no significant correlation was found between parental BMI and family meal frequency, a positive impact on the parents’ obesity status was observed, as also those adolescents with three or more family meals per week were in the normal weight status group, compared to those who had two or fewer family meals who were overweight. Previous findings also showed that family meal frequency can control the development of obesity among children and adolescents . In addition, family meals during adolescence not only maintain adolescents’ normal weight status, but also help to protect them from the development of becoming overweight and obese in young adulthood . This is due to learning how to choose healthier and more nutrient-dense foods during family meals, which impacts their dietary habits and aids in the prevention of the consumption of unhealthy snacks and foods. Authoritative was the only PS positively associated with the frequency of family meals. Previously, a study with a small sample size of overweight and obese African American adolescents demonstrated that an authoritative style contributes to improving the family meal frequency . Three out of four FPPs (i.e., monitoring, reasoning, and modeling) significantly impacted the frequency of family meals. Paying attention to FPPs can help to create positive changes in establishing healthier behaviors, such as family meals in comparison to PSs. FPPs may not only promote healthier diets among adolescents, but may also help to promote better psychological health for family members . The strength of this study is the consideration of African American individuals as a minority group who are one of the most vulnerable populations to obesity. Future studies can use other methods, such as bio-impedance, waist circumference, and dual X-ray absorptiometry to examine the correlation between obesity status and parental influences. The majority of the previous studies on PSs, FPPs, family meals, and adolescents’ weight status focused on one of the parent or caregiver variables, especially maternal influences, in minority groups such as African American families. Future studies can focus on the father’s styles and practices in terms of the weight status and dietary habits of adolescents. In addition, intervention studies can also help us to understand which and how different parental influences (i.e., PSs, FPPs, and family meal frequency) can be most useful in maintaining a normal weight status while considering the cultural values of minority groups.
This study focused on different parental influences, including PSs, FPPs, and family meal frequency, and their relationships with the weight status of African American families. We also examined how each PS and FPP can impact the family meal frequency. The results indicate that family meal frequency plays a more important role in ensuring a healthy weight status among African American adolescents in comparison to PSs and FPPs. An authoritative PS was the only style correlated with a higher family meal frequency, while monitoring, reasoning, and modeling practices were correlated with a higher frequency of family meals.
|
Cyclosporine A Delivery Platform for Veterinary Ophthalmology—A New Concept for Advanced Ophthalmology | 8e339c6f-133d-4b68-b3ab-f4e54b4f19ff | 9599649 | Ophthalmology[mh] | Cyclosporine A (CsA) is one of the most important transplantation drugs that was discovered and isolated by Jean Borel and co-workers in 1970, from the fungus Tolypocladium inflatum . In 1976, their results demonstrated that cyclosporine has immunosuppressive characteristics , which were crucial in transplantology and immunopharmacology. Nowadays, it is widely used as human and veterinary medicine in transplantation procedures, and some immune-mediated inflammatory diseases (IMIDs) . Nevertheless, it has several side effects. For instance, a recent study demonstrated that up to 50% of patients have CsA-associated neurotoxicity in both intravenous and oral administrations which makes the CsA mechanism of action remain ambiguous . This review aims to summarize the effectiveness and safety of CsA delivery platforms in veterinary ophthalmology. However, it is important to highlight numerous studies and great progress in human ophthalmology such as the intracameral drug-delivery system for high-risk penetrating keratoplasty , modified intraocular lens to inhibit post-cataract surgery uveitis and preventing posterior capsular opacification , silica-thermogel nanohybrids sustainably releasing drugs after subconjunctival injection , or a glaucoma drainage device containing CsA and poly(lactic-co-glycolic acid) (PLGA) to prevent postoperative fibrosis .
Cyclosporine A is a natural cyclic hydrophobic peptide with eleven amino acid residues (cyclo[MeBmt 1 -Abu 2 -Sar 3 -MeLeu 4 -Val 5 -MeLeu 6 -Ala 7 -d-Ala 8 -MeLeu 9 -MeLeu 10 -MeVal 11 ], seven peptide bonds (N-methylated), four intra-molecular hydrogen bonds responsible for the cyclic structure, and a molecular weight of 1202.6 g/mol . The CsA is highly soluble in organic solvents such as methanol, ethanol, acetone, ether and chloroform, yet with a different degree of solubility in each one of those solvents, according to Gonzalez et al. and Czogalla et al. . For instance, the CsA has the lowest solubility in water with 0.04 mg/g. However, the highest solubility occurred in chloroform, acetonitrile, dimethyl sulfoxide, methanol, ethyl acetate, isopropyl alcohol, ethanol, polyethylene glycol acetate, isopropyl alcohol, ethanol, polyethylene glycol, propylene glycol, N,N-dimethylacetamide, glycofurol 75, N-methylpyrrolidone, sesame oil, labrafil, labrafac, oleic acid, Tween-20, and Solutol HS with >100 mg/g . In addition, there are some parameters that require further consideration. For instance, the temperature and the pH. The temperature-dependent parameter works in an inversely proportional way; thus, the CsA solubility in water at temperatures ranging between 5 °C to 37 °C showed the highest solubility at 5 °C (101.5 µg/mL) and the lowest at 37 °C (7.3 µg/mL). Therefore, the behavior mechanism associated with D-Ala amino acid residual position number 8 because of its hydration water is lost at high temperatures. Moreover, the selected pH values that represent the pH range of the stomach and the small intestine showed no significant effect on the CsA solubility at pH 1.2 and 6.6 . Poor membrane permeation after topical application, especially with oral administration, related to low water solubility linked with high lipophilic characteristics (log P = 2.92 at pH 7.4), and significant rigidity of the cyclic structure of the CsA results in a limited absorption of the peptide across the gastrointestinal membrane. For that reason, Cyclosporine A (CsA) was classified as a Class IV under the biopharmaceutics classification system . Therefore, the proper selection of drug administration is essential for successful therapy.
Cyclosporine A (CsA) affects the proliferation of helper subset T lymphocytes and cytokine production. The mechanism is linked to two major CsA pathways on the calcineurin/NAFT pathway, as well as JNK and p38 signaling pathways . Cyclosporine A blocks the T cells’ infiltration and subsequently the expression of the inflammatory cytokines such as IL-2 and IL-4 via the calcineurin/NAFT pathway by cyclosporine-cyclophilins interaction in the cytoplasm of T-cells, which causes an increase in Ca 2+ in the cell. A high concentration of Ca 2+ combined with an inducing T cell receptor (TCR) activates calmodulin and binds to protein serine/threonine phosphatases known as calcineurin (calmodulin-dependent protein phosphatase). The calcineurin has catalytic (CnA) and regulatory (CnB) subunits. However, the calcineurin catalytic (CnA) is the dominant in T-cells. The calmodulin- calcineurin A interaction causes the inhibitor domain active site in the CnA to be released and inhibits the phosphatase activity. Therefore, the combination mechanism of the cyclosporine-cyclophilin binding to calcineurin A dephosphorylates the nuclear factor of activated T cell (NFAT) family members (NFAT1, NFAT2, and NFAT4), and prevents the translocation of NAFT family members into the nucleus and transcription of lymphokines genes . The JNK, p38 and ERK signaling pathways are subgroups of the mitogen-activated protein kinase (MAPK) superfamily. However, JNK and p38 signaling pathways showed a greater selectivity of effect to cyclosporine A (CsA) compared to the ERK pathway. Both pathways are synergistically activated during stress responses, such as inflammation and apoptosis, as well as when T cells are stimulated by TCR and CD28 costimulatory receptors. It has been demonstrated that any mutation that blocks JNK and p38 signaling pathways revokes the NF-AT cis-element transcription activation which has binding sites for NFAT family members and Activator protein 1 (AP-1), and both are involved in IL-2 expression . Moreover, the inhibition of both JNK and p38 signaling pathways is related to the upstream level of MAPKK-K activation for example, the MEKK1/MAPKK-Ks participates in the JNK and p38 signaling pathways through MKK7 and MKK6 . Other potential indirect actions of CsA suggested that the JNK and p38 signaling pathways inhibition could be related to over-expression of Vav1/Vav2/Dbl and GEF for Rac1 or Cdc42. Moreover, the HPK1 (Rac1-independent) contributes to JNK activation in an indirect manner. However, the CsA mechanism of action on JNK and p38 signaling pathways remains obscure .
Systemic administration of cyclosporine is commonly used in transplantation procedures but also showed effectiveness in topical treatment, especially during the inflammatory process of the eye . In ophthalmic diseases, direct ocular administration is preferable because of the systemic administration-related side effects such as nephrotoxicity, digestive tract disorders and hypertension . There are two topical formulations of CsA registered in human medicine for the treatment of keratoconjunctivitis sicca (KCS): Restasis ® in the USA and Ikervis ® in Europe . In veterinary ophthalmology, Optimmune ® ointment containing 0.2% cyclosporine is registered for analogous purposes and additionally for chronic superficial keratitis (CSK) . However, there are many examples of immune-mediated ocular disorders in which cyclosporine is beneficial : graft versus host disease (GVHD) , recurrent anterior uveitis , vernal keratoconjunctivitis in humans, and equine recurrent uveitis and immune-mediated keratitis in horses. Because of the physiochemical properties of the CsA, new formulations and devices are investigated . To ensure high penetration capability, long-term effects, and constant drug delivery or to minimize problems with administering eye drops by animal owners, CsA-incorporated implants were developed.
Pearson et al. formulated one of the first CsA delivery devices , based on a sustained-release ganciclovir intravitreal implant . The device containing 5 mg of CsA was implanted intravitreally in eighteen New Zealand albino rabbits and an analogous implant containing 6 mg of CsA was used in three cynomolgus monkeys. The study aimed to determine the toxicity of an intravitreal device that provides long-term delivery of CsA. However, the results showed no evidence of toxicity in the cynomolgus monkeys, but in the rabbits lens opacification in the vicinity of the implant was observed as well as a decrease of the b-wave amplitude in the ERG . In addition, a similar study was established by Enyedi et al. , who investigated the intraocular device containing a combination of dexamethasone (2 mg) and CsA (100 ug) in New Zealand albino rabbits. The 2.5 mm diameter drug pellet of the implant was coated with polyvinyl alcohol (PVA) and ethylene vinyl acetate (EVA) . The highest levels of CsA were detected in the lens compared to low levels in the sclera, cornea, iris and aqueous . Moreover, Jaffe et al. have determined the effectiveness of the intravitreal CsA-sustained delivery device, that proposed by Pearson et al., in the treatment of experimental uveitis in rabbits. The inflammation in the treated eye was considerably less than in the control eye. The therapeutic level of CsA in the vitreous was detected 6 months after implantation . 5.1. Keratoconjunctivitis Sicca (KCS) Keratoconjunctivitis sicca is an inflammatory disease that affects the gland of the third eyelid and lacrimal glands, causing a decrease of tear film, quantitative or qualitative disorder, that could be the result of a congenital, metabolic, drug-induced, neurogenic or immune-mediated defect . The most common signs of KCS are mucopurulent ocular discharge, conjunctivitis and keratitis which can lead to corneal ulcers; some patients develop blepharitis and, in chronic cases, corneal pigmentation and scarring occur . An indispensable element of ophthalmic examination is the Schirmer test (STT-1) which in normal dogs shows tear production around 15 to 25 mm/min; STT-1 in the course of KCS ranges from 9–14 mm/min as mild, >4 to 8 mm/min as moderate and <4 mm/min as a severe stage . The histopathology and serologic results suggested that most of the cases may be immune-mediated (more than 30% of dogs with KCS); the confirmation of which is a positive reaction to immunosuppressive drugs . Kim et al. formulated three similar silicone-based matrix CsA implants (with 20–30% wt / wt ), which were used in several subsequent studies. For instance, the study aimed to provide drug delivery devices that were effective in treating lacrimal gland GVHD (graft versus host disease after transplantation of allogeneic stem cells) and assessed the rate of CsA release (in vitro), implant toxicity, pharmacokinetics and pharmacodynamics in normal rabbits, dogs and dogs with KCS. The results after six months revealed the safety of the delivery device; no ocular toxicity and abnormalities in blood examination were observed. The implant provided therapeutic levels of CsA in the lacrimal gland, conjunctiva and cornea; dogs (with clinical signs of KCS and Schirmer test below 5 mm/min) after implantation did not need further local treatment and the Schirmer test results were above 10 mm/min during the study period . Acton et al. reported a case of keratoconjunctivitis sicca in a red wolf ( Canis rufus ) . After a positive response to topical 2% cyclosporine (initial Schirmer test result at 0 mm/min and after two weeks of combination therapy with triple antibiotic with dexamethasone, the tear production levels were 15 mm/min in the left eye and 16 mm/min in the right eye) they performed implantation of episcleral sustained-release CsA devices (10% matrix CsA-silicone). After two weeks of implantation the tear production remained at the physiological level above 13 mm/min twelve months after surgery . Penetrating keratoplasty (PKP) is a common allograft performed in humans. However, there is one major drawback of this procedure—its high rejection rate (65%). To overcome this obstacle Lee et al. have proposed the use of episcleral CsA implant described previously: CsA powder mixed with silicone; wt / wt 30% . Two implants with different total release were used to determine short (implant B—7.7 mg CsA per implant) and long-term (implant A—12 mg CsA per implant) pharmacokinetics in rabbits and dogs, therefore, the cumulative release observed over the 400-days was approximately 3.8 mg (implant A) and 2.3 mg (implant B). This study showed effective penetration into the cornea and no signs of ocular toxicity. Moreover, CsA concentrations in the cornea were approximately 0.1 µg/mg three hours after implantation and ensured the suppression of T-cell and vascular endothelial cells for over a year. Pharmacokinetics evaluation of CsA in the rabbit model was detected in buccal lymph nodes at 1 h, which suggests that lymphatic vessels in conjunctiva support the rapid dissolution of the drug to the cornea and surrounding tissue . Numerous studies have shown the effectiveness and safety of a cyclosporin episcleral implant with a silicone matrix, an implant of 1.9 cm × 2 mm × 1 mm, containing 12 mg of CsA and ensuring its release at an average level of 17 µg/day for at least 6 months is particularly useful in anterior segment disease . Choi et al. proposed hydrogel contact lenses (CLs) loaded with CsA and determined its efficiency in the rabbit model of dry eye . Previous studies using drug-soaked lenses showed low efficiency in sustained release of the drugs . Therefore, they used a supercritical fluid (SCF) technique to modify and control the degree and the rate of releasing CsA. The in vivo study showed that adequate concentration of CsA was maintained for over 48 h in the cornea, conjunctiva, and crystalline lens. In comparison with control groups, the CsA-CL group exhibited higher density of the goblet cell, tear volume, lower staining score, and reduction of the inflammatory process through immunomodulatory effects. Several clinical trials presented new methods for extended-release drug delivery. For Sight Vision, owned by Allergan, proposed a peri-conjunctival ring currently used for delivery bimatoprost in glaucoma patients . Work is well under way to deliver a CsA ring based on the same technology. In cooperation with NC State University, they conducted clinical trials on dogs with KCS, in which a conjunctival ring releasing CsA was well tolerated and as long as it rested on the conjunctiva under the upper and lower eyelids, the results were satisfying. The therapeutic effect lasted about a month with 75% retention in the eye (unpublished data, ESVO Webinar, 21 February 2021). Ocular Therapeutix™ is working on a group of drug-eluting intracanalicular drug inserts. A study funded by Ocular Therapeutix™ and conducted by Vanslette et al. evaluated pharmacokinetics of Cyclosporine Intracanalicular Insert (OTX-CSI) in Beagle dogs with surgically induced Dry Eye. Intracanalicular devices combines two treatments of dry eye disease: sustained release delivery of cyclosporine and punctal occlusion which aids tear conservation. OTX-SCI contains 0.36 mg CsA in fully biodegradable polyethylene glycol hydrogel. It was designed to provide effective therapy for 12 weeks. The study showed successfully released CsA and its higher concentration in tear fluid in dogs with dry eye was probably due to less dilution on the ocular surface. OTX-CSI was well tolerated and assured immunomodulatory levels in tear fluid . Although OTX-CSI is a promising device for the treatment of immunological diseases in human and veterinary ophthalmology, Ocular Therapeutix™ published results of a Phase 2 clinical trial in which OTX-SCI did not meet the primary endpoint of increased tear production at 12 weeks; therefore, more research is required. 5.2. Chronic Superficial Keratitis (CSK) Chronic superficial keratitis, also known as Pannus, is another immune-mediated disease that affects dogs, with chronic corneal lesions characteristic, mostly in the lateral quadrant: vascularization, progressive pigmentation, and sometimes white opacity . The etiology is still not fully understood but CD4+ T lymphocyte infiltration from cornea stroma suggests an immunological background . The German shepherd dog is a predisposed breed, but it can also occur in the Australian shepherd, collie, border collie, golden retriever, Akita, vizsla, and others . Several studies have found that excessive ultraviolet exposure increases the risk of CSK . Topical immunomodulators such as steroids and calcineurin inhibitors are the standard therapy . Dogs that are responsive to topical CsA, and for whom therapy must be continuous, might be good candidates for an episcleral cyclosporine implant proposed for treating keratoconjunctivitis sicca. However, further studies are required . 5.3. Immune-Mediated Keratitis (IMMK) Immune-mediated keratitis is nonulcerative, primary keratitis (NUK) that occurs in horses . Although the etiopathology has not been thoroughly investigated, the absence of microorganisms and significant improvement after implementation of immunosuppressive therapy suggest an immunological background. The common symptoms of IMMK are nonulcerative corneal opacity, corneal edema and neovascularization, cellular infiltration, and no features of uveitis. Horses with IMMK experience no or mild discomfort . IMMK is classified into four types based on its location in the cornea: epithelial, superficial stromal, middle stromal, and endothelial, with the superficial stroma being the most frequent site of occurrence . CsA topical application is most effective in the case of epithelial and superficial IMMK, but efficiency decreases along with the posterior layers of the cornea . Thus, another route of CsA distribution to all layers of the cornea is being investigated. Gilger et al. proposed the use of a silicone matrix CsA episcleral implant in nineteen horses with different types of IMMK . More than two devices, described previously , were implanted per eye in the dorso temporal episcleral space. The study demonstrated good tolerance of the implants with no significant deviations between the number of devices implanted. Superficial and endothelial immune-mediated keratitis were considered controlled in all treated eyes, although in three cases of endothelial IMMK topical bromfenac was also administered. The worst response was observed in the case of midstromal IMMK. Implants were unable to control inflammation. In vivo, the therapeutic effect of CsA in the case of superficial IMMK was determined for 12–18 months . 5.4. Equine Recurrent Uveitis (ERU) Equine recurrent uveitis (moon blindness, periodic ophthalmia, iridocyclitis) is a condition in which immune-mediated active episodes of panuveitis reoccur every few weeks to months . ERU is still one of the most common causes of horse blindness. Around 30% of horses who presented for examination due to ERU symptoms were unilaterally or bilaterally blind . Horses experience spontaneous relapses similar to humans . Initial causes of recurrent uveitis are not always known, but genetic predisposition or microbes such as Leptospira sp. might be involved . According to recent research, the retinal expression of neuraminidase 1 (NEU1) plays an important role in ERU. Furthermore, horses with recurrent uveitis had higher levels of NEU1 in Müller glial cells in the retina. Therefore, NEU1 might be a new marker of activated Müller glial cells in uveitis . Clinical signs associated with ERU can include anterior segment: blepharospasm, increased lacrimation, photophobia, miosis, edema and vascularization, aqueous flare, cellular infiltration, hypopyon and hyphema, low IOP; posterior segment: vitreous, chorioretinitis, and retinal degeneration . Gilger et al. used an intravitreal cyclosporine delivery device, previously described by Pearson et al., 1996 , Enyedi et al., 1996 , and Jaffe et al., 1998 , in horses with experimental uveitis . The study found that the CsA intravitreal implant reduced the severity and duration of symptoms (but the inflammatory suppression was incomplete), cellular infiltrate was less intense compared to the control eye (PVA/EVA devices without CsA), and the CsA-delivery device was well tolerated. Moreover, the concentration of cyclosporine in the vitreous humor was below therapeutic levels. Nevertheless, tissue levels were not measured . Additionally, the long-term study shows that intravitreal sustained-release CsA delivery devices are safe for at least 12 months . A similar implant was evaluated in horses with ERU that occurred naturally. Devices releasing 4 µg of CsA per day (in a previous study 2 µq/d) were implanted in the eyes of sixteen horses with unilateral uveitis and history of disease recurrence. Follow-up was performed between 6 and 24 months after implantation. After surgery, less than 20% of horses developed uveitis, but as reported by owners, the symptoms were less severe and responded better to anti-inflammatory medication. Complications were noted in four patients, including vision loss due to cataracts or complete retinal detachment as well as glaucoma . A different study by Gilger et al. evaluated episcleral and deep scleral bioerodible cyclosporine implants . Intravitreal delivery devices showed some good results but also revealed complications after implantation, such as cataracts caused by lens injury, endophthalmitis or increased risk of retinal detachment . Thus, the use of implants that do not require entry into the eye has been proposed . Gilger et al. conducted an in vitro study of transscleral diffusion of CsA from a biodegradable matrix-reservoir CsA implant, formulated by Robinson et al. from the National Eye Institute, that suggested the release duration of CsA around 38 months and poor penetration through the sclera. This study aimed to determine the pharmacokinetics and safety of episcleral as well as deep scleral lamellar CsA devices in horses. However, episcleral implantation of the device did not reduce the frequency of relapses due to limited penetration through the sclera. Moreover, the CsA concentration in retina-choroid and vitreous was below the required minimum to treat inflammation . In addition, the deep sclera CsA device was well tolerated, and no toxicity was observed. Therapeutic drug concentration was observed in vitreous and sclera, choroid-retina and optic nerve tissue, although there was no detection of CsA in the aqueous humor, cornea, and samples of peripheral blood. Follow-ups were performed on average after 14 months and a reduction in flare-ups was noted. Blindness occurred in 15% of the eyes as a result of glaucoma, uncontrolled uveitis, cataract, fungal keratitis, and retinal detachment. At the end of the study period, 68/80 of the eyes had vision after surgery . A long-term study on 133 horses (151 eyes) confirmed the promising results from the previous survey but also noted complications such as glaucoma, persistent uveitis, cataracts, and retinal detachment .
Keratoconjunctivitis sicca is an inflammatory disease that affects the gland of the third eyelid and lacrimal glands, causing a decrease of tear film, quantitative or qualitative disorder, that could be the result of a congenital, metabolic, drug-induced, neurogenic or immune-mediated defect . The most common signs of KCS are mucopurulent ocular discharge, conjunctivitis and keratitis which can lead to corneal ulcers; some patients develop blepharitis and, in chronic cases, corneal pigmentation and scarring occur . An indispensable element of ophthalmic examination is the Schirmer test (STT-1) which in normal dogs shows tear production around 15 to 25 mm/min; STT-1 in the course of KCS ranges from 9–14 mm/min as mild, >4 to 8 mm/min as moderate and <4 mm/min as a severe stage . The histopathology and serologic results suggested that most of the cases may be immune-mediated (more than 30% of dogs with KCS); the confirmation of which is a positive reaction to immunosuppressive drugs . Kim et al. formulated three similar silicone-based matrix CsA implants (with 20–30% wt / wt ), which were used in several subsequent studies. For instance, the study aimed to provide drug delivery devices that were effective in treating lacrimal gland GVHD (graft versus host disease after transplantation of allogeneic stem cells) and assessed the rate of CsA release (in vitro), implant toxicity, pharmacokinetics and pharmacodynamics in normal rabbits, dogs and dogs with KCS. The results after six months revealed the safety of the delivery device; no ocular toxicity and abnormalities in blood examination were observed. The implant provided therapeutic levels of CsA in the lacrimal gland, conjunctiva and cornea; dogs (with clinical signs of KCS and Schirmer test below 5 mm/min) after implantation did not need further local treatment and the Schirmer test results were above 10 mm/min during the study period . Acton et al. reported a case of keratoconjunctivitis sicca in a red wolf ( Canis rufus ) . After a positive response to topical 2% cyclosporine (initial Schirmer test result at 0 mm/min and after two weeks of combination therapy with triple antibiotic with dexamethasone, the tear production levels were 15 mm/min in the left eye and 16 mm/min in the right eye) they performed implantation of episcleral sustained-release CsA devices (10% matrix CsA-silicone). After two weeks of implantation the tear production remained at the physiological level above 13 mm/min twelve months after surgery . Penetrating keratoplasty (PKP) is a common allograft performed in humans. However, there is one major drawback of this procedure—its high rejection rate (65%). To overcome this obstacle Lee et al. have proposed the use of episcleral CsA implant described previously: CsA powder mixed with silicone; wt / wt 30% . Two implants with different total release were used to determine short (implant B—7.7 mg CsA per implant) and long-term (implant A—12 mg CsA per implant) pharmacokinetics in rabbits and dogs, therefore, the cumulative release observed over the 400-days was approximately 3.8 mg (implant A) and 2.3 mg (implant B). This study showed effective penetration into the cornea and no signs of ocular toxicity. Moreover, CsA concentrations in the cornea were approximately 0.1 µg/mg three hours after implantation and ensured the suppression of T-cell and vascular endothelial cells for over a year. Pharmacokinetics evaluation of CsA in the rabbit model was detected in buccal lymph nodes at 1 h, which suggests that lymphatic vessels in conjunctiva support the rapid dissolution of the drug to the cornea and surrounding tissue . Numerous studies have shown the effectiveness and safety of a cyclosporin episcleral implant with a silicone matrix, an implant of 1.9 cm × 2 mm × 1 mm, containing 12 mg of CsA and ensuring its release at an average level of 17 µg/day for at least 6 months is particularly useful in anterior segment disease . Choi et al. proposed hydrogel contact lenses (CLs) loaded with CsA and determined its efficiency in the rabbit model of dry eye . Previous studies using drug-soaked lenses showed low efficiency in sustained release of the drugs . Therefore, they used a supercritical fluid (SCF) technique to modify and control the degree and the rate of releasing CsA. The in vivo study showed that adequate concentration of CsA was maintained for over 48 h in the cornea, conjunctiva, and crystalline lens. In comparison with control groups, the CsA-CL group exhibited higher density of the goblet cell, tear volume, lower staining score, and reduction of the inflammatory process through immunomodulatory effects. Several clinical trials presented new methods for extended-release drug delivery. For Sight Vision, owned by Allergan, proposed a peri-conjunctival ring currently used for delivery bimatoprost in glaucoma patients . Work is well under way to deliver a CsA ring based on the same technology. In cooperation with NC State University, they conducted clinical trials on dogs with KCS, in which a conjunctival ring releasing CsA was well tolerated and as long as it rested on the conjunctiva under the upper and lower eyelids, the results were satisfying. The therapeutic effect lasted about a month with 75% retention in the eye (unpublished data, ESVO Webinar, 21 February 2021). Ocular Therapeutix™ is working on a group of drug-eluting intracanalicular drug inserts. A study funded by Ocular Therapeutix™ and conducted by Vanslette et al. evaluated pharmacokinetics of Cyclosporine Intracanalicular Insert (OTX-CSI) in Beagle dogs with surgically induced Dry Eye. Intracanalicular devices combines two treatments of dry eye disease: sustained release delivery of cyclosporine and punctal occlusion which aids tear conservation. OTX-SCI contains 0.36 mg CsA in fully biodegradable polyethylene glycol hydrogel. It was designed to provide effective therapy for 12 weeks. The study showed successfully released CsA and its higher concentration in tear fluid in dogs with dry eye was probably due to less dilution on the ocular surface. OTX-CSI was well tolerated and assured immunomodulatory levels in tear fluid . Although OTX-CSI is a promising device for the treatment of immunological diseases in human and veterinary ophthalmology, Ocular Therapeutix™ published results of a Phase 2 clinical trial in which OTX-SCI did not meet the primary endpoint of increased tear production at 12 weeks; therefore, more research is required.
Chronic superficial keratitis, also known as Pannus, is another immune-mediated disease that affects dogs, with chronic corneal lesions characteristic, mostly in the lateral quadrant: vascularization, progressive pigmentation, and sometimes white opacity . The etiology is still not fully understood but CD4+ T lymphocyte infiltration from cornea stroma suggests an immunological background . The German shepherd dog is a predisposed breed, but it can also occur in the Australian shepherd, collie, border collie, golden retriever, Akita, vizsla, and others . Several studies have found that excessive ultraviolet exposure increases the risk of CSK . Topical immunomodulators such as steroids and calcineurin inhibitors are the standard therapy . Dogs that are responsive to topical CsA, and for whom therapy must be continuous, might be good candidates for an episcleral cyclosporine implant proposed for treating keratoconjunctivitis sicca. However, further studies are required .
Immune-mediated keratitis is nonulcerative, primary keratitis (NUK) that occurs in horses . Although the etiopathology has not been thoroughly investigated, the absence of microorganisms and significant improvement after implementation of immunosuppressive therapy suggest an immunological background. The common symptoms of IMMK are nonulcerative corneal opacity, corneal edema and neovascularization, cellular infiltration, and no features of uveitis. Horses with IMMK experience no or mild discomfort . IMMK is classified into four types based on its location in the cornea: epithelial, superficial stromal, middle stromal, and endothelial, with the superficial stroma being the most frequent site of occurrence . CsA topical application is most effective in the case of epithelial and superficial IMMK, but efficiency decreases along with the posterior layers of the cornea . Thus, another route of CsA distribution to all layers of the cornea is being investigated. Gilger et al. proposed the use of a silicone matrix CsA episcleral implant in nineteen horses with different types of IMMK . More than two devices, described previously , were implanted per eye in the dorso temporal episcleral space. The study demonstrated good tolerance of the implants with no significant deviations between the number of devices implanted. Superficial and endothelial immune-mediated keratitis were considered controlled in all treated eyes, although in three cases of endothelial IMMK topical bromfenac was also administered. The worst response was observed in the case of midstromal IMMK. Implants were unable to control inflammation. In vivo, the therapeutic effect of CsA in the case of superficial IMMK was determined for 12–18 months .
Equine recurrent uveitis (moon blindness, periodic ophthalmia, iridocyclitis) is a condition in which immune-mediated active episodes of panuveitis reoccur every few weeks to months . ERU is still one of the most common causes of horse blindness. Around 30% of horses who presented for examination due to ERU symptoms were unilaterally or bilaterally blind . Horses experience spontaneous relapses similar to humans . Initial causes of recurrent uveitis are not always known, but genetic predisposition or microbes such as Leptospira sp. might be involved . According to recent research, the retinal expression of neuraminidase 1 (NEU1) plays an important role in ERU. Furthermore, horses with recurrent uveitis had higher levels of NEU1 in Müller glial cells in the retina. Therefore, NEU1 might be a new marker of activated Müller glial cells in uveitis . Clinical signs associated with ERU can include anterior segment: blepharospasm, increased lacrimation, photophobia, miosis, edema and vascularization, aqueous flare, cellular infiltration, hypopyon and hyphema, low IOP; posterior segment: vitreous, chorioretinitis, and retinal degeneration . Gilger et al. used an intravitreal cyclosporine delivery device, previously described by Pearson et al., 1996 , Enyedi et al., 1996 , and Jaffe et al., 1998 , in horses with experimental uveitis . The study found that the CsA intravitreal implant reduced the severity and duration of symptoms (but the inflammatory suppression was incomplete), cellular infiltrate was less intense compared to the control eye (PVA/EVA devices without CsA), and the CsA-delivery device was well tolerated. Moreover, the concentration of cyclosporine in the vitreous humor was below therapeutic levels. Nevertheless, tissue levels were not measured . Additionally, the long-term study shows that intravitreal sustained-release CsA delivery devices are safe for at least 12 months . A similar implant was evaluated in horses with ERU that occurred naturally. Devices releasing 4 µg of CsA per day (in a previous study 2 µq/d) were implanted in the eyes of sixteen horses with unilateral uveitis and history of disease recurrence. Follow-up was performed between 6 and 24 months after implantation. After surgery, less than 20% of horses developed uveitis, but as reported by owners, the symptoms were less severe and responded better to anti-inflammatory medication. Complications were noted in four patients, including vision loss due to cataracts or complete retinal detachment as well as glaucoma . A different study by Gilger et al. evaluated episcleral and deep scleral bioerodible cyclosporine implants . Intravitreal delivery devices showed some good results but also revealed complications after implantation, such as cataracts caused by lens injury, endophthalmitis or increased risk of retinal detachment . Thus, the use of implants that do not require entry into the eye has been proposed . Gilger et al. conducted an in vitro study of transscleral diffusion of CsA from a biodegradable matrix-reservoir CsA implant, formulated by Robinson et al. from the National Eye Institute, that suggested the release duration of CsA around 38 months and poor penetration through the sclera. This study aimed to determine the pharmacokinetics and safety of episcleral as well as deep scleral lamellar CsA devices in horses. However, episcleral implantation of the device did not reduce the frequency of relapses due to limited penetration through the sclera. Moreover, the CsA concentration in retina-choroid and vitreous was below the required minimum to treat inflammation . In addition, the deep sclera CsA device was well tolerated, and no toxicity was observed. Therapeutic drug concentration was observed in vitreous and sclera, choroid-retina and optic nerve tissue, although there was no detection of CsA in the aqueous humor, cornea, and samples of peripheral blood. Follow-ups were performed on average after 14 months and a reduction in flare-ups was noted. Blindness occurred in 15% of the eyes as a result of glaucoma, uncontrolled uveitis, cataract, fungal keratitis, and retinal detachment. At the end of the study period, 68/80 of the eyes had vision after surgery . A long-term study on 133 horses (151 eyes) confirmed the promising results from the previous survey but also noted complications such as glaucoma, persistent uveitis, cataracts, and retinal detachment .
Cyclosporine is a calcineurin inhibitor with immunomodulatory and immunosuppressive properties. Lipophilic, high molecular weight, poor solubility in water, and numerous side effects require an urgent need to develop new formulations and devices to deliver this therapeutic agent. The implants described so far were well tolerated and provided therapeutic levels in the target tissues. In cases of KCS, they ensured increased tear production, while controlling inflammation in other diseases. Positive results in animals promise better treatment prospects for not only graft versus host disease in humans, but also keratoconjunctivitis sicca and recurrent uveitis. However, longer and more thorough examinations may be necessary to obtain the most effective and sustainable delivery devices, ensuring a long-lasting and constant supply of the drug and high penetration, which are the critical factors for the therapy’s effectiveness and the elimination of the problem of administering drugs to uncooperative animals. The implants proposed so far require anesthetized patients and an invasive surgical procedure. To reach posterior segments of the eye, deep sclera devices are needed which come with the risk of side effects such as cataracts and high intraocular pressure. Accordingly, the future requires biodegradable implants with a long duration of action and a minimally invasive implantation procedure. Furthermore, episcleral injection-based hydrogel carriers seem to be a promising solution; therefore, there is an urgent need for further research on the form of cyclosporine administration and cyclosporine carriers.
|
Development of a scoring system to predict endovascular crossing of femoropopliteal artery chronic total occlusions: the Endo VAscular CROSsing Score for Chronic Total Occlusions (EVACROSS-CTO) | 0b79d084-148f-458e-85ac-e7c26783e90d | 11919073 | Cardiovascular System[mh] | Chronic total occlusions (CTOs) of infrainguinal arteries are frequently observed in patients with peripheral arterial disease (PAD). They are often associated with high rates of treatment failure and complications during endovascular treatment (EVT). Chronic total occlusions have posed significant challenges due to various factors affecting successful lesion crossing. Factors like calcification burden, lesion length, collateral circulation, and cap morphology are considered some of the most critical determinants, and extensive research has aimed to establish their correlation with recanalization success. Several advanced techniques have been developed to enhance technical success, including STAR, CART, and SAFARI, complemented by cutting-edge equipment, such as re-entry catheters and CTO-dedicated guidewires. Recently, there has been a growing interest in the development of scoring systems aimed at predicting the likelihood of successful crossing in cases of CTOs within coronary and peripheral arteries. Notably, 2 distinct predictive tools, specifically designed for below-the-knee (BTK) occlusions, have been proposed. These tools provide clinicians with objective criteria aiding them in selecting the optimal crossing approach, based on the angiographic evaluation. , However, it is worth noting that a comparable predictive tool targeting femoropopliteal CTOs, utilizing pre-procedural data such as CT angiography or MR angiography, has yet to be proposed. The potential impact of such a tool on clinical decision-making is considerable, as it could contribute to the determination of whether EVT or surgical bypass is the more suitable option. Furthermore, it could assist in the selection of specific equipment and the application of appropriate endovascular techniques, enhancing the precision and success of interventions in this specific anatomical context, thereby improving patient outcomes and resource allocation. The aim of this manuscript was to develop and test a predictive score that can be used in everyday clinical practice to predict successful endovascular crossing of peripheral artery CTOs.
Patient recruitment In this retrospective observational descriptive single-centre study, all consecutive patients meeting specific inclusion criteria were enrolled: (1) symptomatic atherosclerotic PAD patients undergoing EVT and (2) CT angiography depicting a CTO in superficial femoral artery (SFA) and/or popliteal artery (POPA). Exclusion criteria were (1) patients without pre-operative CTA, (2) lesions extending below the trifurcation, and (3) prior endovascular intervention in the same lesion. A total of 139 patients with CTOs were retrospectively recruited, but ultimately, 84 patients (68 male-16 female) met the inclusion criteria and were incorporated into this study. The remaining patients were excluded from further analysis due to not meeting the specified exclusion criteria. The patient data were retrieved from the hospital’s electronic database spanning the period between 2016 and 2022, encompassing details regarding the procedures performed and pre-procedural imaging assessments. The dataset was split 70%:30% in cohort A used to construct the score and cohort B to test the score. The initial 59 consecutive patients (cohort A) were used to build the score, while the subsequent 25 consecutive patients (cohort B) were used to assess the predictive capacity of the score. The study has received approval from the Ethical Review Board of our University Hospital and all patients have signed an informed consent to undergo the procedure. Imaging and procedural definitions All procedures were performed by 2 experienced endovascular specialists with over 30 years and 15 years of expertise in peripheral artery endovascular interventions, respectively. Technical success was determined by successfully crossing the CTO with a guidewire, either via the true lumen or sub-intimally and it was evaluated by using the procedural records and images. Pre-procedural CTA images were obtained using Revolution HD (General Electric Medical Systems, Waukesha, WI, USA) CT scanner and reviewed using commercial software (VesselIQ, GE Medical Systems). The CT images were compared with the digital subtraction angiography data. Pre-procedural images were reviewed to identify and describe lesion characteristics, such as a tapered proximal or distal cap, which was characterized by a ‘V’ or ‘U’ shape. Flush occlusion was defined as a cap displaying a flat shape, with or without a sloppy morphology, or a reverse ‘U’ shape. Proximal or distal side branches were identified as branches whose orifice were located at the level of the proximal or distal cap, respectively, in which the guidewire could easily slip through. Bridging collaterals were identified as slender vessels located alongside the occlusion, serving as connectors between the non-occluded proximal and distal ends of the segment. Occlusion length was assessed from the proximal to the distal cap of the lesion. The calcification burden of the occluded vessel was assessed by segmenting the whole lesion in 3D Slicer (slicer.org) and segmentation of calcifications by means of thresholding. The total volume of the CTO (Vol total ) and the volume of the calcified part (Vol calcification ) were calculated and the calcification percentage was calculated as Percentage calcification = Vol calcification Vol total × 100 . The target lesions were primarily managed via an ipsilateral antegrade approach. In instances where this approach was unsuccessful, an alternative strategy involving a retrograde pedal approach was considered and implemented. Assessment of outcome predictors and score development Predictors determining technical success were assessed using logistic regression. Predictors included age >50, the presence of a flush occlusion, the presence of proximal or distal side branches, calcification percentage of the lesion >10%, the presence of a tapered proximal cap, the presence of bridging collaterals, and the length of the occlusion. Univariable logistic regression was used to assess the significance of all the available predictors for the success of the treatment in cohort A ( n = 59). A multivariable logistic regression model using significant predictors ( P < .05) in univariable analysis was built with the success or failure of the procedure as an outcome. A predictive score was built as previously described. For example, factors with P < 0.1 identified from the multivariable logistic regression model (cohort A) were used as score variables. Scores were assigned to each predictor based on the odds ratio of the predictor in multivariable analysis to build a final scoring system. The predictive capacity of the score was assessed on an independent set of patients (cohort B [ n = 25]) collected over a different time frame (∼3 years) to avoid biases introduced by the random selection of test cases. An optimal score threshold was selected based on the receiver operating characteristics (ROC) curve analysis. Statistical analysis Continuous variables were expressed as mean±SD, whereas categorical variables were expressed as frequencies. For logistic regression analysis, odds ratios with the respective 95% CIs were calculated for all included variables, and goodness of fit was assessed with the use of the Hosmer-Lemeshow test. Receiver operating characteristics curves were used to assess the performance of the score in predicting the success of endovascular crossing. The optimal threshold for the score was defined as the ROC curve point where both sensitivity and specificity are maximized. Accuracy, sensitivity, specificity, and area under the curve (AUC) were reported at the optimal score threshold. Statistical analysis was performed with the use of SPSS v. 29 and significance was defined with P ≤ .05.
In this retrospective observational descriptive single-centre study, all consecutive patients meeting specific inclusion criteria were enrolled: (1) symptomatic atherosclerotic PAD patients undergoing EVT and (2) CT angiography depicting a CTO in superficial femoral artery (SFA) and/or popliteal artery (POPA). Exclusion criteria were (1) patients without pre-operative CTA, (2) lesions extending below the trifurcation, and (3) prior endovascular intervention in the same lesion. A total of 139 patients with CTOs were retrospectively recruited, but ultimately, 84 patients (68 male-16 female) met the inclusion criteria and were incorporated into this study. The remaining patients were excluded from further analysis due to not meeting the specified exclusion criteria. The patient data were retrieved from the hospital’s electronic database spanning the period between 2016 and 2022, encompassing details regarding the procedures performed and pre-procedural imaging assessments. The dataset was split 70%:30% in cohort A used to construct the score and cohort B to test the score. The initial 59 consecutive patients (cohort A) were used to build the score, while the subsequent 25 consecutive patients (cohort B) were used to assess the predictive capacity of the score. The study has received approval from the Ethical Review Board of our University Hospital and all patients have signed an informed consent to undergo the procedure.
All procedures were performed by 2 experienced endovascular specialists with over 30 years and 15 years of expertise in peripheral artery endovascular interventions, respectively. Technical success was determined by successfully crossing the CTO with a guidewire, either via the true lumen or sub-intimally and it was evaluated by using the procedural records and images. Pre-procedural CTA images were obtained using Revolution HD (General Electric Medical Systems, Waukesha, WI, USA) CT scanner and reviewed using commercial software (VesselIQ, GE Medical Systems). The CT images were compared with the digital subtraction angiography data. Pre-procedural images were reviewed to identify and describe lesion characteristics, such as a tapered proximal or distal cap, which was characterized by a ‘V’ or ‘U’ shape. Flush occlusion was defined as a cap displaying a flat shape, with or without a sloppy morphology, or a reverse ‘U’ shape. Proximal or distal side branches were identified as branches whose orifice were located at the level of the proximal or distal cap, respectively, in which the guidewire could easily slip through. Bridging collaterals were identified as slender vessels located alongside the occlusion, serving as connectors between the non-occluded proximal and distal ends of the segment. Occlusion length was assessed from the proximal to the distal cap of the lesion. The calcification burden of the occluded vessel was assessed by segmenting the whole lesion in 3D Slicer (slicer.org) and segmentation of calcifications by means of thresholding. The total volume of the CTO (Vol total ) and the volume of the calcified part (Vol calcification ) were calculated and the calcification percentage was calculated as Percentage calcification = Vol calcification Vol total × 100 . The target lesions were primarily managed via an ipsilateral antegrade approach. In instances where this approach was unsuccessful, an alternative strategy involving a retrograde pedal approach was considered and implemented.
Predictors determining technical success were assessed using logistic regression. Predictors included age >50, the presence of a flush occlusion, the presence of proximal or distal side branches, calcification percentage of the lesion >10%, the presence of a tapered proximal cap, the presence of bridging collaterals, and the length of the occlusion. Univariable logistic regression was used to assess the significance of all the available predictors for the success of the treatment in cohort A ( n = 59). A multivariable logistic regression model using significant predictors ( P < .05) in univariable analysis was built with the success or failure of the procedure as an outcome. A predictive score was built as previously described. For example, factors with P < 0.1 identified from the multivariable logistic regression model (cohort A) were used as score variables. Scores were assigned to each predictor based on the odds ratio of the predictor in multivariable analysis to build a final scoring system. The predictive capacity of the score was assessed on an independent set of patients (cohort B [ n = 25]) collected over a different time frame (∼3 years) to avoid biases introduced by the random selection of test cases. An optimal score threshold was selected based on the receiver operating characteristics (ROC) curve analysis.
Continuous variables were expressed as mean±SD, whereas categorical variables were expressed as frequencies. For logistic regression analysis, odds ratios with the respective 95% CIs were calculated for all included variables, and goodness of fit was assessed with the use of the Hosmer-Lemeshow test. Receiver operating characteristics curves were used to assess the performance of the score in predicting the success of endovascular crossing. The optimal threshold for the score was defined as the ROC curve point where both sensitivity and specificity are maximized. Accuracy, sensitivity, specificity, and area under the curve (AUC) were reported at the optimal score threshold. Statistical analysis was performed with the use of SPSS v. 29 and significance was defined with P ≤ .05.
Patient characteristics Patient occlusions were found in 40 left and 44 right lower limbs with a mean patient age of 67.69 ± 10.9 years. The majority of occlusions ( n = 71) were found at the level of the SFA, 22 of which extended distally to the POPA, and one of them presented as flush occlusion at the origin of the SFA. Thirteen occlusions involved only the POPA. Fifty-six out of 84 occlusions had a tapered proximal cap, 57 had a proximal and 27 had a distal side branch, whereas 37 occlusions had bridging collaterals. Analysis of calcification showed that in 54/84 (64.3%) more than 10% of the lesion volume was calcified. In 26 out of 84 patients (30.95%), the procedure failed due to inability to cross the lesion with the guidewire. Score construction Univariate analysis indicated that all examined variables were significant apart from the length of the occlusion and the presence of bridging collaterals. Multivariate logistic regression including all variables identified in univariate analysis, indicated a series of factors with P < .1 eligible for inclusion in the score. These included age > 50 ( P = .036, odds ratio (OR) 53.7 with 95% CI, 1.284-2245), calcification percentage of the occlusion >10% ( P = .011, OR 16.63 with 95% CI, 1.899-145.654), the presence of a flush occlusion ( P = 0.02, OR 15.564 with 95% CI, 1.531-158.204), the presence of a distal side branch ( P = 0.018, OR 9.879 with 95% CI, 1.493-65.377), and the presence of a proximal side branch ( P = 0.064, OR 23.369 with 95% CI, 0.831-657.218). Relative to their OR, the factors were assigned the following EndoVAscular CROSsing Score for Chronic Total Occlusion (EVACROSS-CTO) scores: +10 points for age > 50 years, +3 points for a flush occlusion, +4 points for a proximal, and +2 points for a distal side branch, as well as +3 points for calcification of the CTO of >10%. The maximum achievable score is 22 points ( and ). Score assessment Patients of cohort B were scored based on the EVACROSS-CTO score and an ROC curve was created based on the scores and the respective outcome. EVACROSS-CTO was able to predict the success of endovascular crossing with an AUC-ROC of 79.8% (95% CI, 58.5%-100%). A score >16 yielded a sensitivity of 75% with a specificity of 70.6% in cohort B ( and ). The procedure to calculate the score prior to the procedure is outlined in .
Patient occlusions were found in 40 left and 44 right lower limbs with a mean patient age of 67.69 ± 10.9 years. The majority of occlusions ( n = 71) were found at the level of the SFA, 22 of which extended distally to the POPA, and one of them presented as flush occlusion at the origin of the SFA. Thirteen occlusions involved only the POPA. Fifty-six out of 84 occlusions had a tapered proximal cap, 57 had a proximal and 27 had a distal side branch, whereas 37 occlusions had bridging collaterals. Analysis of calcification showed that in 54/84 (64.3%) more than 10% of the lesion volume was calcified. In 26 out of 84 patients (30.95%), the procedure failed due to inability to cross the lesion with the guidewire.
Univariate analysis indicated that all examined variables were significant apart from the length of the occlusion and the presence of bridging collaterals. Multivariate logistic regression including all variables identified in univariate analysis, indicated a series of factors with P < .1 eligible for inclusion in the score. These included age > 50 ( P = .036, odds ratio (OR) 53.7 with 95% CI, 1.284-2245), calcification percentage of the occlusion >10% ( P = .011, OR 16.63 with 95% CI, 1.899-145.654), the presence of a flush occlusion ( P = 0.02, OR 15.564 with 95% CI, 1.531-158.204), the presence of a distal side branch ( P = 0.018, OR 9.879 with 95% CI, 1.493-65.377), and the presence of a proximal side branch ( P = 0.064, OR 23.369 with 95% CI, 0.831-657.218). Relative to their OR, the factors were assigned the following EndoVAscular CROSsing Score for Chronic Total Occlusion (EVACROSS-CTO) scores: +10 points for age > 50 years, +3 points for a flush occlusion, +4 points for a proximal, and +2 points for a distal side branch, as well as +3 points for calcification of the CTO of >10%. The maximum achievable score is 22 points ( and ).
Patients of cohort B were scored based on the EVACROSS-CTO score and an ROC curve was created based on the scores and the respective outcome. EVACROSS-CTO was able to predict the success of endovascular crossing with an AUC-ROC of 79.8% (95% CI, 58.5%-100%). A score >16 yielded a sensitivity of 75% with a specificity of 70.6% in cohort B ( and ). The procedure to calculate the score prior to the procedure is outlined in .
Herein, various determinants affecting the success of recanalization procedures for peripheral artery CTOs, particularly concerning lesions in the SFA and POPA, have been identified. These determinants were observed in pre-procedural CT angiography and correlated with the successful or unsuccessful guidewire crossing of the CTOs. Subsequently, these factors were utilized in developing the EVACROSS-CTO score, designed to predict the likelihood of a successful EVT outcome of a CTO in one or a combination of the aforementioned arterial segments. After conducting logistic regression, 5 key variables were identified: age over 50, flush occlusion, proximal side branch, distal side branch, and calcification burden over 10%. Treatment of CTO in coronary and peripheral arteries is technically challenging with a high failure rate ranging between 20% and 40%. , Technical success in the current study is in line with the literature. The demand for improved patient selection and treatment planning has led to the development of new scoring algorithms. , , Recent studies , , employed a variety of demographic, clinical, and anatomical variables to predict the success of peripheral artery CTO interventions. Common variables identified included patient age, diabetes, and history of previous revascularization. Anatomical factors such as the length and severity of occlusion, degree of calcification, morphological classifications, and specific lesion locations (eg, proximal superficial femoral artery and distal POPA) were critical in their predictive models. These studies demonstrated that their scoring systems provided high predictive accuracy, with an AUC ranging between 0.82 and 0.87. This emphasizes the effectiveness of a multifactorial approach in developing predictive scores for peripheral artery CTO interventions. Notably, the abovementioned scores have been predominantly directed towards addressing lesions in BTK and coronary arteries. The current study is dedicated to establishing a CTO scoring system, specifically tailored for femoropopliteal lesions, known for their frequent incidence with significant length and high calcification burden. In contrast to prior research endeavours, scoring in this study relied on pre-procedural CTA scans, an imaging method previously validated for treatment planning and prognostication. , The selection of determinants was primarily guided by previous studies, , , , supplemented by observations made during interventions, such as the propensity of the guidewire to traverse into a side branch, particularly in conjunction with a flush occlusion. The study yielded a predictive accuracy of 79.8%, which is a good performance given the multiple factors affecting the score. Notably, the inclusion of a substantial number of patients with lengthy occlusions did not show a significant impact on the success of the interventions, consistent with findings from prior research. In contrast, studies focusing on BTK CTO lesions have suggested that lesions exceeding 200 mm may influence procedural success. , This discrepancy underscores the potential variations in CTO characteristics based on their specific vascular locations. To ensure the EVACROSS-CTO score is of practical clinical use, its primary utility could potentially focus on prioritizing patients for angioplasty. By incorporating variables such as patient age, occlusion characteristics, and calcification burden, the score effectively identifies those most likely to benefit from the procedure. This prioritization not only enhances patient selection by minimizing unnecessary interventions for those with lower probabilities of success but also optimizes resource allocation within the healthcare system. Additionally, the physician would be equipped to preemptively identify scenarios where access might encounter a higher failure rate and could prepare with available devices to improve efficacy. Furthermore, this scoring system might additionally assist operators in determining the level of effort required to achieve technical success. Finally, patients could receive precise information regarding the complexity of their CTO and the probability of a favourable treatment outcome. Despite the valuable insights, there are notable limitations. Such scores can be complex, relying on comprehensive and subjective assessments, which may impact reproducibility and consistency across different clinical settings. The predictive accuracy of these models can vary based on the patient population and CTO characteristics, making them less universally applicable. The need for thorough validation in diverse populations remains a challenge to ensure the broad utility and reliability of these scoring tools. The study has several potential limitations that need consideration. Firstly, it is an observational, single-centre, retrospective study, which might limit the generalizability of the findings to broader populations. However, conducting this scoring system in various centres could verify its accuracy and applicability across different settings. Secondly, despite including a considerable number of patients with CTOs, the final sample size might still be relatively limited. Moreover, although the EVACROSS-CTO score demonstrated predictive capacity in cohort B, its ability to independently guide clinical decision-making might be limited. Further validation in larger and more diverse cohorts is essential to assess its reliability and applicability across different patient populations and clinical settings. Additionally, the score was specifically developed using data from patients with atherosclerotic PAD. Its relevance to other conditions, such as Berger’s disease or post-embolic and thrombotic occlusions, remains unclear, as these conditions were not represented in the study population. Given the distinct pathophysiological characteristics of these diseases, the score’s applicability to such cases may be limited. Future versions of the EVACROSS-CTO score could be refined to include factors that predict potential procedural complications, such as vessel perforation or distal embolization. This would not only increase its clinical value but also help provide a more holistic approach to decision-making. Additionally, validating the score in broader patient populations, including those with conditions like thrombotic occlusions, could significantly enhance its applicability and relevance across diverse clinical scenarios.
The development and validation of the EVACROSS-CTO score represents a significant step forward in predicting the success of EVT for CTOs in femoropopliteal arteries. This predictive tool incorporates variables derived from pre-procedural CT angiography and demonstrates promising predictive capacity in determining procedural outcomes. Further validation of the applicability of this score across diverse populations and clinical settings is needed in order to confirm its reliability and broader utility. Despite these considerations, the use of this scoring system holds considerable potential in aiding clinicians to make informed decisions and optimize patient care strategies in managing femoropopliteal CTOs.
|
Improving the Obstetrics and Gynecology Learning Environment Through Faculty Development | d366780d-fc52-409d-ba97-094e2fc98a48 | 9061934 | Gynaecology[mh] | By the end of this activity, learners will be able to: 1. Analyze their role in creating a positive learning environment. 2. Identify mistreatment and aspects of a suboptimal learning environment by critiquing cases. 3. Reflect on their own practice in creating an optimal learning environment. 4. Create their own strategies to combat mistreatment in the learning environment. 5. Commit to routinely incorporating at least one strategy into their teaching repertoire. The clinical learning environment is shaped by multiple factors, including the formal curriculum; interactions with faculty, residents, staff, and peers; and other aspects of the hidden curriculum, which is defined as the “set of influences that function at the level of organizational structure and culture.” The learning environment impacts students’ ability to learn, the depth and breadth of their clinical experiences, student wellness, and academic achievement and satisfaction. Experiences of mistreatment among medical students , reflect the health of the learning environment. Unfortunately, mistreatment is common, with up to 83% of students reporting at least one experience of mistreatment by residents, faculty, or staff. These experiences range from microaggressions (with 61% of medical students in a recent national survey reporting experiencing weekly episodes of microaggressions) to neglect, , inadequate student supervision, and perceptions of disrespect. In the OB/GYN clerkship in particular, students consistently report mistreatment, with up to one out of four clerkship students reporting mistreatment in a longitudinal study of over 800 students at one institution. A smaller study with 18 students similarly found that 25% reported experiencing mistreatment on their OB/GYN rotation. Baecher-Lind and colleagues hypothesized that in stressful clinical environments with high acuity, such as labor and delivery, communication breakdown contributes to feelings of perceived disrespect and neglect and missed educational opportunities. In particular, labor and delivery encompass a variety of patients in different care settings, from a triage area serving as the obstetric emergency room to an inpatient unit with increasingly complex patients to an operating room handling emergent deliveries. Teaching across this broad clinical spectrum carries with it a unique set of challenges due to the different skills needed in each arena. The challenges of procedural teaching, such as the complexity of patient cases and lack of continuity with preceptors, may lead to multiple episodes of neglect or disrespect, which degrade interpersonal interactions and contribute to an adverse learning environment. At our institution, reports from end-of-course evaluations over several years demonstrated episodes of learner mistreatment. Nearly 25% of students witnessed the use of unprofessional or derogatory language, did not receive constructive feedback, and noted a lack of empathy from OB/GYNs. Seventeen percent of students experienced a lack of respect on their OB/GYN clerkship. When compared to other clerkships at the institution, the OB/GYN clerkship ratings were consistently lower, and our hospital's site ratings were lower than other clerkship sites for OB/GYN within the academic institution. When we examined resident data based on an ACGME survey, faculty engagement was identified as an area of growth. These findings prompted the OB/GYN medical education committee (graduate medical education, undergraduate medical education, and department leadership) to commit to change. The first step we undertook was to promote faculty involvement in changing the learning environment through a faculty development session, which was inspired by the 2018 Macy Foundation conference detailing faculty development interventions to improve the learning environment. To address concerns about the learning environment in our OB/GYN clerkship, we created a faculty development workshop to increase awareness, promote discussion, and problem-solve about this topic. Our workshop is grounded in the principles of adult learning theory that encourage learners to take responsibility for their learning and to have an active role in the learning experience. We utilize the concept of communities of practice for our workshop. A community of practice is structured around a group of people who care about the same issues and interact with each other regularly to learn from each other and address issues together. Our workshop targets faculty in the OB/GYN department, many of whom work with each other on a regular basis and all of whom are impacted by these concerns around the learning environment. We capitalize on people's acknowledgment that they all face the same problem to encourage them to problem-solve together. A community of practice is thus created in this workshop by bringing people together explicitly for the purpose of learning with and from each other. By having colleagues work through problems they identify, the workshop engages participants so that they feel invested in the process of finding solutions that will work for their group. Other workshops , have described how to train participants to identify mistreatment or have asked for identification of barriers to creating a positive learning environment. Our workshop expands on this prior work by having participants identify mistreatment and develop strategies to combat it. Thus, this workshop adds to the literature by describing a strategy that encourages participant involvement in creating solutions, along with a commitment to follow through on those solutions. We base this strategy on data illustrating that adult learners who are allowed to choose their own actions demonstrate higher rates of positive change. While the workshop and curriculum are focused on OB/GYN faculty, other areas of medicine, particularly family medicine and surgery, can easily adapt the curriculum to their own field. This workshop was offered as part of our department's quarterly faculty development grand rounds series, which was open to all faculty, residents, students, and staff in the department. Attendance at grand rounds was mandatory for all faculty and residents. Due to the COVID-19 pandemic, this workshop was offered virtually in 2021. There were no prerequisites to attend the workshop. Although both residents and faculty attended grand rounds, only faculty participants were invited to complete an electronic preworkshop survey by email. Members of the department's medical education committee reviewed end-of-clerkship evaluations from the medical students from the prior 6 months. Developed by the medical school, this evaluation was based on the AAMC Medical School Graduation Questionnaire. The medical education committee focused on the domains in which the clerkship was most deficient, based on feedback from the medical students. These domains included constructive feedback, respectful interactions with students, using professional language, conflict resolution, and showing empathy. The committee worked with a research scientist with expertise in qualitative research to create a 12-item preworkshop survey to determine faculty teaching skills and confidence in engaging with and involving medical students and residents in clinical care. The survey also queried demographics, including age, gender identity, race and ethnicity, and number of years at the institution. Faculty were asked to complete the survey via an email sent 3 days prior to the event and sent again 1 day prior to initial nonresponders. The introductory portion of the 60-minute workshop consisted of a PowerPoint presentation and occurred in a virtual large-group setting. The first half of the presentation consisted of a summary of the quarterly student clerkship learning environment survey, as well as mistreatment reports at the institutional and departmental levels. The second half of the presentation focused on effective strategies to achieve positive change in the learning environment. We discussed the effect of physician stress and burnout on effective teaching. We discussed the importance of culture change and our ability to do so using clinical teaching strategies that help to shape our learning environment. These clinical teaching strategies were adapted from Chuang and colleagues, who developed the report on behalf of the Association of Professors of Gynecology and Obstetrics Undergraduate Medical Education Committee. The strategies included creating a climate of humanism, recognizing and discussing seminal events, role modeling, actively engaging learners, being relevant and practical, and synthesizing multiple strategies. This introduction took approximately 15 minutes. After the introduction, participants were divided into six discussion groups of approximately 10 people each, facilitated by an experienced educator within the department. These breakout groups were conducted via Zoom, and an administrator randomly assigned participants to one of the breakout groups at the conclusion of the introduction. All the facilitators had a formal role within graduate or undergraduate medical education and served on the medical education committee for the department. All the faculty facilitators had at least 5 years of clinical and teaching experience. Each of the facilitators was assigned to cover one of four cases . These cases had been created by the medical education committee, members of which were also group facilitators for the workshop. The cases covered a range of behaviors by faculty, staff, and residents ranging from minor (passive neglect) to blatant (belittling, overt humiliation) and were loosely modeled on events that had occurred within the institution. The cases were based on the medical student learning environment data indicating areas for improvement. Each case focused on two of the clinical teaching strategies from Chuang and colleagues and highlighted aspects of the hidden curriculum that were utilized as prompts for further discussion. After initial development of the cases, six of the medical education committee members revised them to ensure that each case would meet the teaching objectives and be aligned with two of the clinical teaching strategies. The committee members matched the teaching strategies by identifying the most significant aspects of mistreatment in a case and connecting these to positive behaviors that could have been implemented instead. The cases were reviewed by the entire committee prior to finalization. The medical education committee also created a facilitator guide with questions to guide the discussion . Each group chose a member to be the reporter. The groups read the assigned case, and the facilitator then led discussion through a series of questions focused on the related key domains. Facilitators encouraged participants to develop their own strategies to address specific aspects of the learning environment they found challenging. Discussion focused specifically on eliciting examples that promoted a positive learning environment. Breakout groups were allocated 25–30 minutes for reviewing and discussing the cases. After that time, the administrator closed the breakout rooms, and all groups then reconvened for a report-out, with a group debriefing and discussion. A total of 15 minutes was allocated for this portion of the workshop. As participants reported out, one facilitator noted specific strategies for each aspect of the learning environment in a table that was later shared via email with participants at the conclusion of the workshop. Each participant was asked to commit to using one strategy described by their fellow participants in the coming months. Immediately after the workshop concluded, the faculty participants received a postworkshop survey sent via email . A reminder to complete the survey was sent a week later. The postworkshop survey included all the questions in the preworkshop survey and added three new ones: whether the participant's interactions with medical students and residents would change based on the workshop, the relevance of the workshop, and whether the participant had learned new skills in the workshop. A research assistant collated the open-ended responses and analyzed the quantitative data with simple descriptive statistics. After the workshop, the medical education committee also reviewed data from the learning environment questionnaire, which the medical school distributed to the next set of students completing the OB/GYN clerkship, to evaluate any changes in clerkship ratings. Data were presented either as median with interquartile range or as proportion. Categorical data were compared using the chi-square or Fisher's exact test, whereas continuous data were compared using the Wilcoxon rank sum test. McNemar's test was used to calculate statistical differences between paired proportions. We considered p values less than .05 statistically significant. Data were analyzed with SAS 9.4 (SAS Institute). Institutional review board approval was granted as an exempt application. Sixty faculty members attended the workshop, representing the range of specialties in OB/GYN; 57% completed the preworkshop survey, and 33% completed the postworkshop survey . The median age of the participants was 50, and the median number of years at the institution was 6. Three-quarters of the participants were female, and 50% were White. Eighteen residents participated in the workshop but did not receive the survey; medical students were not present during the workshop. Prior to the session, while the majority (94%) of faculty who responded to the survey believed that they contributed to the learning environment, fewer (76%) felt they had the skills needed to be an effective teacher, and only 65% reported consistently trying to use effective teaching strategies. The vast majority (97%) aimed to create an inclusive environment, but only 68% of faculty reported they routinely involved medical students in the clinical care of patients. Only 62% felt prepared to engage medical students when they were the assigned faculty preceptor, although 76% felt prepared to engage residents as the assigned faculty preceptor. After the session, the majority of participants (75%) reported that they would consistently try to use effective teaching strategies . The majority of participants (85%) reported that after the workshop, they were more aware of issues around the learning environment. Most participants (85%) felt that their interactions with medical students would change in a positive way because of the workshop and the discussions it generated. Similarly, 80% of participants felt that their interactions with residents would change in a positive way because of the workshop, and they committed to creating an inclusive learning environment after the workshop . However, none of these changes were statistically significant. One area in which there were significant differences was faculty's self-rated ability to engage learners of different levels. When comparing faculty's self-perceived ability to engage residents versus students, the difference was statistically significant ( p = .014). After the workshop, we found that faculty more frequently felt well prepared to engage both medical students (75%) and residents (85%), although they still were more likely to engage residents ; this difference was statistically significant ( p = .046). There was also a statistically significant difference when we compared faculty's involvement of medical students versus residents specifically in clinical care. Faculty involved medical students in clinical care only 60% of the time, compared to 90% of the time for residents ( p = .01; ). We collected and collated the strategies that the smaller discussion groups shared with the entire workshop during report-out and arranged them by themes. At the end of the session, participants were invited to commit to using one or more of these strategies in the future. The strategies that participants committed to ranged from simple (asking the student's name) to more complex (e.g., reminding themselves to be patient and encouraging increased communication). These strategies are summarized in . Review of the free-text responses to the postsession evaluation demonstrated that multiple respondents appreciated the session's interactivity. In particular, respondents mentioned that videoconferencing promoted small-group discussion. Nearly all the respondents (90%) rated the session as being relevant to their needs, and 70% reported learning a new pedagogical skill during the workshop . One participant wrote, “We need to be reminded that in these incredibly stressful times, we must make space for effective teaching.” The next learning environment questionnaire administered by the medical school after the workshop demonstrated significant improvement, with 100% of students reporting that faculty used professional language, demonstrated empathy, and were respectful towards students. The majority of students (86%) reported receiving constructive feedback. The hidden curriculum of a learning environment contributes substantially to students’ development of professional behavior, medical knowledge, and clinical skills, and a positive learning environment has been demonstrated to support the acquisition of skills. This workshop provides a framework to discuss the hidden curriculum and its impact on the learning environment, as well as strategies to improve the learning environment through case-based discussion among educators and faculty. We relied on the principles of adult learning theory to empower learners to take responsibility for their learning and explicitly created a community of practice to allow them to discuss shared experiences and devise solutions applicable to the group. We provided participants a framework with which to address aspects of the hidden curriculum, but the participants themselves arrived at suggestions to improve the learning environment in six key domains. This workshop was successful and relevant in part because the participants were from one department and knew each other well, thereby providing a safe environment among trusted colleagues and enabling active engagement and honest discussion. The scenarios were modeled on actual experiences and demonstrated aspects of the hidden curriculum (including microaggressions, neglect, and humiliation) that impact the learning environment. We would recommend that in the smaller breakout sessions, one person be designated at the very beginning to report out to the large group in order to facilitate their taking notes or preparing to share the group's discussion. Having breakout groups of approximately 10 participants was helpful because each group was small enough to encourage participation but large enough that, if a few people did not participate, the discussion was not negatively impacted. One hour was sufficient time for the workshop. Given the sensitive nature of the discussion, facilitators for the workshop should be experienced medical educators. Multiple participants commented on the interactivity of the workshop; particularly when conferences are virtual, an interactive component is key. Sharing educational strategies demonstrated a commitment to medical education by participants that other faculty and residents could emulate. While other curricula have described workshops discussing the learning environment, ours specifically provided participants with a framework to critically analyze the learning environment and then develop their own solutions using that framework. Asking participants to commit to using at least one strategy confirmed their dedication to addressing the learning environment and reinforced the need for change. Data from reviews of motivational interviewing, in which participants reach their own decisions about how to implement behavioral change, have demonstrated increased rates of positive change and increased self-efficacy. Extrapolating these data to our workshop, we anticipate that the commitments to change generated by the participants themselves will have some ongoing impact. Although faculty felt more prepared to engage residents than students and included residents more often in clinical work both before and after the workshop, overall there was an upward trend for feeling better prepared when engaging medical students. Interestingly, while prior to the workshop 68% of faculty involved medical students in clinical care, immediately after the workshop 60% of faculty involved medical students in clinical care. We believe that these numbers do not represent a true decline in faculty's involvement of medical students but rather may be due to the fact that the number of faculty who completed the postworkshop survey was small. We acknowledge that faculty may find it easier to engage and teach residents who are in the program for 4 years, rather than students who have rotations lasting just a few weeks, with time fragmented between different services on the rotation. This lack of continuity with students may make it more challenging for faculty to feel invested in individual students and to expend their time teaching. In addition, some faculty may prefer or find it easier to teach learners who are already committed to the specialty rather than students who may still be exploring career interests. Within a procedural specialty, faculty may also be struggling to balance learners at different levels and may fear that focusing on one learner could detract from the learning opportunities of the others. The difference between faculty preparedness for teaching and inclusion in clinical care between students and residents remained statistically significant even after the workshop; these findings merit further research. We hope to include a future session with a focus on teaching to one's highest abilities with multiple levels of learners. Limitations of this work included a small number of participants from a single department at one academic medical center. Our cases are directly relevant to OB/GYN, but other specialties could adapt the cases to reflect clinical examples from their field. Another limitation was a small response rate for the postworkshop survey; response rates could have been improved if time was allocated at the end of the session to complete the postworkshop survey, which took approximately 5 minutes. In addition, pre- and postsurveys were not matched to directly compare respondents’ answers. If nonresponders are significantly less likely to have improved then responders, we may be overestimating the effect of our workshop. Finally, we have limited long-term data regarding changes in the learning environment, thereby making it difficult to determine the long-term impact of the workshop, although we plan to continue tracking data pertaining to the learning environment. Since the workshop, initial updated data about the learning environment from our institution demonstrated an internal upward trend in 80% of the domains examined by the institution, as well as higher ratings in 100% of the domains when compared with the other sites at which medical students rotate. We will continue to track data for 12–18 months after the workshop to assess for long-term improvement. In summary, this interactive, virtual workshop enabled a vital, guided discussion around the hidden curriculum and the learning environment and empowered participants to create their own strategies for how to improve the learning environment. The workshop encouraged participants to commit to one teaching strategy to address deficits in the learning environment, which served to strengthen their commitment to medical education. Limited data collected from our institution demonstrated improvement in the learning environment and an upward trend for our hospital in particular when compared to the other sites. The learning environment is intertwined with students’ academic success and personal well-being. Engaging the faculty who help shape that learning environment is crucial if there is to be impactful, effective change that will help maximize our learners’ success. As we train the next generation of physicians, we have an obligation to our patients and to society to ensure that the milieu in which they learn how to care for others is supportive, encouraging, and fulfilling. Preworkshop Survey.docx PowerPoint for the Learning Environment.pptx Cases for the Learning Environment.docx Facilitator Guide.docx Postworkshop Survey.docx All appendices are peer reviewed as integral parts of the Original Publication. |
Intestinal schistosomiasis in Uganda at high altitude (>1400 m): malacological and epidemiological surveys on Mount Elgon and in Fort Portal crater lakes reveal extra preventive chemotherapy needs | 4b4df7fe-4c9f-4202-804e-a716d1d7ee37 | 5292801 | Preventive Medicine[mh] | Please see Additional file for translations of the abstract into the six official working languages of the United Nations. Schistosomiasis is of considerable public health importance in sub-Saharan Africa . Across the continent there are several national control programmes (NCPs) operating at various levels of scale-up, as guided by expectations set out within the WHO 2012–2020 Roadmap for neglected tropical diseases . In Uganda, since 2003 there has been an active NCP primarily concerned with the delivery of praziquantel to school-aged children . Whilst each form of schistosomiasis can be found in Uganda, intestinal schistosomiasis, caused by Schistosoma mansoni (and transmitted by Biomphalaria spp.), is most common and widespread . Just under 20 million people in 73 districts are estimated to be at-risk of infection . The disease-endemic zone is most obvious within shoreline habitats of the Great Lakes and River Nile as well as in its vicinities nearby . Outside of these however, large-scale ecological and epidemiological predictions have broadly set aside such areas as being unlikely to sustain natural transmission, for example, too cold (altitude > 1400 m) or too arid (annual rainfall < 90 cm) . Alongside general surveillance and monitoring activities within the Ugandan NCP, an associated programme of operational and cross-country collaborative research has been undertaken . This has explored the dynamics of environmental transmission in East Africa, especially in areas poorly sampled previously . Schistosome-snail ecology has been addressed at either macro- and micro-epidemiological levels; many heterogeneities in biotic and (or) non-biotic factors have come to light . In addition, more sensitive diagnostics tools have been introduced which go beyond routine parasitological sampling which has in turn increased abilities to better detect and differentiate schistosome infection(s) either in people or in snails . Furthermore, with advances in geographical information systems and more affordable digital cartography , new opportunities arise to review and refine earlier eco-epidemiological predictions. This includes targeted epidemiological surveys to ‘ground-truth’ assertions now better armed with new diagnostics. This paper is an attempt to determine the risk of intestinal schistosomiasis transmission at higher altitudes in Uganda using data from two prospective malacological and epidemiological surveys conducted on Mount Elgon and in Fort Portal crater lakes. The surveys also make reference to Lake Albert and Rwenzori Mountains, assessing if natural transmission of intestinal schistosomiasis occurs at altitudes near to or exceeding 1400 m. Malacological surveys and laboratory investigations At each freshwater habitat surveyed, global position system (GPS) coordinates, altitude and location photographs were taken with an Oregon 650 receiver (Garmin, Olathe, Kansas, USA). Water temperature (°C), pH and conductivity (μS) were recorded with a HI-98129 Pocket EC/TDS and pH Tester (Hanna Instruments Ltd, Leighton Buzzard, Bedfordshire, UK). Two snail collectors searched for Biomphalaria spp. by hand and with metal scoops, for over 20 min at each site. All collected snails were counted then transferred into plastic cups containing mineral water, exposed to light for two hours, then checked for shedding cercariae under the dissecting microscope. Thirty seven sites (1139 m–3937 m) were surveyed in June 2011 in the Mount Elgon area, to include areas located close to sampled schools (see below). Within the Fort Portal crater lakes, 23 sites (1123 m–1567 m) were surveyed in June 2012, including a selection of those previously visited by Rubaihayo et al . in 2006 , with sampling taking place in areas of the lakes known to be used by local communities for fishing and domestic activities. Eight further higher altitude sites were surveyed by foot in the Rwenzori Mountains (1620 m–4050 m), plus an additional two by car on the southern shore of Lake Albert (616 m and 624 m) as low altitude reference. Biomphalaria spp. collected from Mount Elgon was placed in absolute ethanol, transferred to the UK and genomic DNA was extracted according to standard protocols. A total of 118 snails (33 from < 1400 m and 85 from > 1400 m) were then screened by polymerase chain reaction (PCR) for schistosome DNA following protocols described by Kane et al . using schistosome–specific primers, RAKqIGSF (5’ AAA GTC GGA AAA ATG AAA 3’) and RAKqIGSR (5’ TAT GAA TGA AAT CGG TTA 3’) for a sub-region of the nuclear ribosomal intergenic spacer . Amplicons were separated by 1.5% agarose gel electrophoresis and stained with ethidium bromide. A snail was judged to be infected with S. mansoni if a 300 base pairs fragment that could also be digested with the restriction enzyme Acc 1 was observed . Epidemiological surveys and diagnostic assays The GPS coordinates and elevation of each sampled school were recorded. Within the Mount Elgon area, six universal primary education (UPE) schools were included: 3 at higher altitudes (1856 m–2072 m) and three at lower altitudes (1150 m–1268 m). Schools were sampled randomly within two elevation strata (<1500 m and >1500 m). Within each school, 50 children were randomly selected (25 boys, 25 girls) from classes Primary 5 – Primary 7 (age range of 10–13 years). Within the Fort Portal Crater Lakes area, 14 UPE schools (1165 m–1526 m) were selected, six of which had been surveyed previously by Rubaihayo et al . in 2006 . The additional schools were selected purposively such that they covered a range of altitudes and were close to a crater lake. An additional school was sampled close to the southern shores of Lake Albert (621 m), an area where transmission is known to be high, as a comparison. Within each school, 30 children were examined. No school was sampled in the Rwenzori Mountains as the survey took place within the Rwenzori National Park. Specimen collection and diagnostic testing took place over a two day period with methods described previously . Urine, stool and finger-prick blood samples were taken on the first day, with an additional stool sample obtained on the second day. Faecal microscopy involved inspection of duplicate thick Kato-Katz smears from each faecal sample with egg counting at ×100 magnification. A urine-CCA dipstick test (Rapid Medical Diagnostics, Pretoria, South Africa) was used to detect intestinal schistosomiasis and 3 μl of harvested sera was used for detection of antibodies against soluble egg antigen (SEA-ELISA) with a commercially available kit (IVD Inc.; Carlsbad, USA). All sampled children were treated on site with praziquantel (40 mg/kg, Cipla, Mumbai, India) and albendazole (400 mg, GSK, Hertford, Hertfordshire, UK) irrespective of infection status. Spatial and statistical analyses The relationship between altitude and the survey results was explored through the production of maps and scatter plots using the geographical information software QGIS (version 2.18.1) and the statistical software R (version 3.3.1) . In addition to mapping the geographical coordinates of the survey sites, elevation data at a 90 m resolution was obtained from NASA’s Shuttle Radar Topography Mission (SRTM) and displayed using elevation bands of less than 1000 m, 1000–1200 m, 1200–1400 m, 1400–1600 m, 1600–1800 m, 1800–2000 m and greater than 2000 m for the whole of Uganda. With regards to the malacological survey data, in addition to plotting elevation against number of Biomphalaria spp. caught at each site, plots of water temperature (°C), pH and conductivity against elevation were produced in order to explore further this relationship. Plots of school prevalence using all three diagnostic methods were produced and a relationship between the pooled prevalence results using each of the three diagnostic methods and elevation was assessed by fitting a logistic regression model to the data. The goodness of fit of these models was assessed by calculating the deviance statistic . In order to explore the currently available information on high altitude schistosomiasis transmission excluding the surveys presented in this paper, georeferenced schistosomiasis prevalence data obtained primarily using Kato-Katz were downloaded from the Global Atlas of Helminth Infections (GAHI) website . This database is a collation of prevalence data from multiple sources including the published literature and the national NTD control programme . The elevation of each georeferenced point was extracted from the 90 m resolution SRTM data, and summaries of surveys, including the number undertaken and the reported prevalence ranges by elevation bands, were reported. The altitude of each georeferenced point was extracted from the 90 m resolution SRTM data, and summaries of surveys, including the number undertaken and the reported prevalence ranges by elevation bands, were reported. Finally, using 100 m gridded population data for 2015 obtained from WorldPop ( http://www.worldpop.org.uk/ ) the number of people living within elevation bands of less than 1400 m, 1400–2000 m and greater than 2000 m was calculated in order to obtain a crude estimate of the number of people at-risk. At each freshwater habitat surveyed, global position system (GPS) coordinates, altitude and location photographs were taken with an Oregon 650 receiver (Garmin, Olathe, Kansas, USA). Water temperature (°C), pH and conductivity (μS) were recorded with a HI-98129 Pocket EC/TDS and pH Tester (Hanna Instruments Ltd, Leighton Buzzard, Bedfordshire, UK). Two snail collectors searched for Biomphalaria spp. by hand and with metal scoops, for over 20 min at each site. All collected snails were counted then transferred into plastic cups containing mineral water, exposed to light for two hours, then checked for shedding cercariae under the dissecting microscope. Thirty seven sites (1139 m–3937 m) were surveyed in June 2011 in the Mount Elgon area, to include areas located close to sampled schools (see below). Within the Fort Portal crater lakes, 23 sites (1123 m–1567 m) were surveyed in June 2012, including a selection of those previously visited by Rubaihayo et al . in 2006 , with sampling taking place in areas of the lakes known to be used by local communities for fishing and domestic activities. Eight further higher altitude sites were surveyed by foot in the Rwenzori Mountains (1620 m–4050 m), plus an additional two by car on the southern shore of Lake Albert (616 m and 624 m) as low altitude reference. Biomphalaria spp. collected from Mount Elgon was placed in absolute ethanol, transferred to the UK and genomic DNA was extracted according to standard protocols. A total of 118 snails (33 from < 1400 m and 85 from > 1400 m) were then screened by polymerase chain reaction (PCR) for schistosome DNA following protocols described by Kane et al . using schistosome–specific primers, RAKqIGSF (5’ AAA GTC GGA AAA ATG AAA 3’) and RAKqIGSR (5’ TAT GAA TGA AAT CGG TTA 3’) for a sub-region of the nuclear ribosomal intergenic spacer . Amplicons were separated by 1.5% agarose gel electrophoresis and stained with ethidium bromide. A snail was judged to be infected with S. mansoni if a 300 base pairs fragment that could also be digested with the restriction enzyme Acc 1 was observed . The GPS coordinates and elevation of each sampled school were recorded. Within the Mount Elgon area, six universal primary education (UPE) schools were included: 3 at higher altitudes (1856 m–2072 m) and three at lower altitudes (1150 m–1268 m). Schools were sampled randomly within two elevation strata (<1500 m and >1500 m). Within each school, 50 children were randomly selected (25 boys, 25 girls) from classes Primary 5 – Primary 7 (age range of 10–13 years). Within the Fort Portal Crater Lakes area, 14 UPE schools (1165 m–1526 m) were selected, six of which had been surveyed previously by Rubaihayo et al . in 2006 . The additional schools were selected purposively such that they covered a range of altitudes and were close to a crater lake. An additional school was sampled close to the southern shores of Lake Albert (621 m), an area where transmission is known to be high, as a comparison. Within each school, 30 children were examined. No school was sampled in the Rwenzori Mountains as the survey took place within the Rwenzori National Park. Specimen collection and diagnostic testing took place over a two day period with methods described previously . Urine, stool and finger-prick blood samples were taken on the first day, with an additional stool sample obtained on the second day. Faecal microscopy involved inspection of duplicate thick Kato-Katz smears from each faecal sample with egg counting at ×100 magnification. A urine-CCA dipstick test (Rapid Medical Diagnostics, Pretoria, South Africa) was used to detect intestinal schistosomiasis and 3 μl of harvested sera was used for detection of antibodies against soluble egg antigen (SEA-ELISA) with a commercially available kit (IVD Inc.; Carlsbad, USA). All sampled children were treated on site with praziquantel (40 mg/kg, Cipla, Mumbai, India) and albendazole (400 mg, GSK, Hertford, Hertfordshire, UK) irrespective of infection status. The relationship between altitude and the survey results was explored through the production of maps and scatter plots using the geographical information software QGIS (version 2.18.1) and the statistical software R (version 3.3.1) . In addition to mapping the geographical coordinates of the survey sites, elevation data at a 90 m resolution was obtained from NASA’s Shuttle Radar Topography Mission (SRTM) and displayed using elevation bands of less than 1000 m, 1000–1200 m, 1200–1400 m, 1400–1600 m, 1600–1800 m, 1800–2000 m and greater than 2000 m for the whole of Uganda. With regards to the malacological survey data, in addition to plotting elevation against number of Biomphalaria spp. caught at each site, plots of water temperature (°C), pH and conductivity against elevation were produced in order to explore further this relationship. Plots of school prevalence using all three diagnostic methods were produced and a relationship between the pooled prevalence results using each of the three diagnostic methods and elevation was assessed by fitting a logistic regression model to the data. The goodness of fit of these models was assessed by calculating the deviance statistic . In order to explore the currently available information on high altitude schistosomiasis transmission excluding the surveys presented in this paper, georeferenced schistosomiasis prevalence data obtained primarily using Kato-Katz were downloaded from the Global Atlas of Helminth Infections (GAHI) website . This database is a collation of prevalence data from multiple sources including the published literature and the national NTD control programme . The elevation of each georeferenced point was extracted from the 90 m resolution SRTM data, and summaries of surveys, including the number undertaken and the reported prevalence ranges by elevation bands, were reported. The altitude of each georeferenced point was extracted from the 90 m resolution SRTM data, and summaries of surveys, including the number undertaken and the reported prevalence ranges by elevation bands, were reported. Finally, using 100 m gridded population data for 2015 obtained from WorldPop ( http://www.worldpop.org.uk/ ) the number of people living within elevation bands of less than 1400 m, 1400–2000 m and greater than 2000 m was calculated in order to obtain a crude estimate of the number of people at-risk. Malacological findings Figure presents an overview of the locations of the snail sampling sites across Uganda, with all data in Additional file : Table S1. Within the Mount Elgon area, Biomphalaria spp. were found at 30% (11/37) of sampled sites at altitudes as high as 1951 m whereas in the Fort Portal crater lakes were found 92% (21/23) of sampled sites, at a maximum altitude of 1567 m. No Biomphalaria spp. were found in the Rwenzori Mountains. Biomphalaria spp. were found in one of the two low altitude sites on Lake Albert. Scatter plots of number of Biomphalaria spp. against altitude (Fig. ) indicated that no Biomphalaria spp. were found above 2000 m. There was no discernible pattern in the number of snails and altitude below this value. Plots of water temperature, pH and conductivity against altitude, indicate that whilst there appears to be a negative linear relationship between temperature and altitude, see Fig. , the water in Mount Elgon has a lower pH (median = 8.3) than that of Fort Portal crater lakes area (median = 11.3) and the Rwenzori Mountains (median = 9.5). There is no clear trend between pH and altitude or Biomphalaria spp.. A negative association of conductivity with increasing altitude was observed, with the lowest values (<100 μS) being found at sites at very high altitudes (>3000 m). A total of 264 Biomphalaria spp. collected from Mount Elgon were observed for shedding cercariae under natural conditions on at least one occasion. While non-human cercariae were seen, no snail shed schistosome cercariae. Upon PCR analysis conducted in the UK, 16 snails from a total of 33 screened from below 1400 m and six snails from a total of 85 screened from above 1400 m were judged infected with S. mansoni with a prevalence of 48.5 and 7.1%, respectively. In total, 235 Biomphalaria spp. collected from Fort Portal crater lakes were observed for shedding cercariae, with three snails from Nyinambuga and Lyantonde, two independent crater lakes, patently shedding schistosome cercariae. Epidemiological findings Table presents the overall prevalence of S. mansoni within schools sampled from Mount Elgon, Fort Portal crater lakes and Lake Albert as measured by Kato-Katz, CCA and SEA-ELISA. Prevalence by SEA-ELISA testing in the Mount Elgon and Fort Portal crater lakes areas are shown in Fig. , with Fig. depicting school-level prevalence against altitude for each of the three diagnostic methods. Prevalence as obtained with each diagnostic decreases with increasing altitude (SEA-ELISA: Lake Albert = 100%, Fort Portal crater lakes = 60.4%, Mount Elgon = 42.3%). Individual school data can be found in Additional file : Table S2. According to SEA-ELISA testing there is strong evidence of infection at higher altitudes, such that a prevalence of 26.7% (40/150) was detected in the three most elevated schools (1856–2072 m), falling to 10.0% (15/150) and 1.3% (2/150) using CCA and Kato-Katz, respectively. Of note, is that these schools are within reasonable proximity (approximately 10 km) to a location where numerous Biomphalaria spp. were found, Fig. . The logistic regression model fitted to the prevalence data obtained using each of the three diagnostic methods, with elevation as the only risk factor, are presented in Table . The resulting odds ratios were 0.9977 (95% CI : 0.9947–0.9966), 0.9974 (95% CI : 0.9967–0.9980) and 0.9957 (95% CI : 0.9947–0.9966) for SEA-ELISA, CCA and Kato-Katz respectively. Thus, at the 1400 m threshold the fitted models predict prevalence values of 59.0% (95% CI : 55.2%–62.7%), 29.8% (95% CI : 26.4–33.4%), 7.8% (95% CI : 5.9–10.2%) for ELISA, CCA and Kato-Katz, respectively. It should be noted however that these models had a poor goodness of fit ( P < 0.05). Of the 671 geolocated survey sites in the GAHI database (excluding the seven surveys whose coordinates were outside of the country boundary), 93.7% (629/671) were below an altitude of 1400 m. Of the remaining 42 surveys, 17 were within the 1400–1600 m range, five were within 1600 m–1800 m and 14 were within 1800 m–2000 m. The majority of surveys (94.3%, 633/671) used single Kato-Katz faecal smears to estimate prevalence of S. mansoni with egg-patent prevalence ranging from 0.0 to 57.0% (1400 m–1600 m), 0.0–1.7% (1600–1800 m) and 0.0%–3.4% (1800–2000 m). No cases were found at any sites at an altitude greater than 2000 m. Estimating the at-risk population Using the gridded population estimates from WorldPop (100 m) and SRTM elevation (90 m), of the 39.2 million people living in Uganda, 82.8% (32.5 million) of people live at altitudes below 1400 m. Of the remainder, 15.4% (6 M) live within an altitude range of 1400 m–2000 m and the remaining 1.8% (0.7 million) live above 2000 m. Figure presents an overview of the locations of the snail sampling sites across Uganda, with all data in Additional file : Table S1. Within the Mount Elgon area, Biomphalaria spp. were found at 30% (11/37) of sampled sites at altitudes as high as 1951 m whereas in the Fort Portal crater lakes were found 92% (21/23) of sampled sites, at a maximum altitude of 1567 m. No Biomphalaria spp. were found in the Rwenzori Mountains. Biomphalaria spp. were found in one of the two low altitude sites on Lake Albert. Scatter plots of number of Biomphalaria spp. against altitude (Fig. ) indicated that no Biomphalaria spp. were found above 2000 m. There was no discernible pattern in the number of snails and altitude below this value. Plots of water temperature, pH and conductivity against altitude, indicate that whilst there appears to be a negative linear relationship between temperature and altitude, see Fig. , the water in Mount Elgon has a lower pH (median = 8.3) than that of Fort Portal crater lakes area (median = 11.3) and the Rwenzori Mountains (median = 9.5). There is no clear trend between pH and altitude or Biomphalaria spp.. A negative association of conductivity with increasing altitude was observed, with the lowest values (<100 μS) being found at sites at very high altitudes (>3000 m). A total of 264 Biomphalaria spp. collected from Mount Elgon were observed for shedding cercariae under natural conditions on at least one occasion. While non-human cercariae were seen, no snail shed schistosome cercariae. Upon PCR analysis conducted in the UK, 16 snails from a total of 33 screened from below 1400 m and six snails from a total of 85 screened from above 1400 m were judged infected with S. mansoni with a prevalence of 48.5 and 7.1%, respectively. In total, 235 Biomphalaria spp. collected from Fort Portal crater lakes were observed for shedding cercariae, with three snails from Nyinambuga and Lyantonde, two independent crater lakes, patently shedding schistosome cercariae. Table presents the overall prevalence of S. mansoni within schools sampled from Mount Elgon, Fort Portal crater lakes and Lake Albert as measured by Kato-Katz, CCA and SEA-ELISA. Prevalence by SEA-ELISA testing in the Mount Elgon and Fort Portal crater lakes areas are shown in Fig. , with Fig. depicting school-level prevalence against altitude for each of the three diagnostic methods. Prevalence as obtained with each diagnostic decreases with increasing altitude (SEA-ELISA: Lake Albert = 100%, Fort Portal crater lakes = 60.4%, Mount Elgon = 42.3%). Individual school data can be found in Additional file : Table S2. According to SEA-ELISA testing there is strong evidence of infection at higher altitudes, such that a prevalence of 26.7% (40/150) was detected in the three most elevated schools (1856–2072 m), falling to 10.0% (15/150) and 1.3% (2/150) using CCA and Kato-Katz, respectively. Of note, is that these schools are within reasonable proximity (approximately 10 km) to a location where numerous Biomphalaria spp. were found, Fig. . The logistic regression model fitted to the prevalence data obtained using each of the three diagnostic methods, with elevation as the only risk factor, are presented in Table . The resulting odds ratios were 0.9977 (95% CI : 0.9947–0.9966), 0.9974 (95% CI : 0.9967–0.9980) and 0.9957 (95% CI : 0.9947–0.9966) for SEA-ELISA, CCA and Kato-Katz respectively. Thus, at the 1400 m threshold the fitted models predict prevalence values of 59.0% (95% CI : 55.2%–62.7%), 29.8% (95% CI : 26.4–33.4%), 7.8% (95% CI : 5.9–10.2%) for ELISA, CCA and Kato-Katz, respectively. It should be noted however that these models had a poor goodness of fit ( P < 0.05). Of the 671 geolocated survey sites in the GAHI database (excluding the seven surveys whose coordinates were outside of the country boundary), 93.7% (629/671) were below an altitude of 1400 m. Of the remaining 42 surveys, 17 were within the 1400–1600 m range, five were within 1600 m–1800 m and 14 were within 1800 m–2000 m. The majority of surveys (94.3%, 633/671) used single Kato-Katz faecal smears to estimate prevalence of S. mansoni with egg-patent prevalence ranging from 0.0 to 57.0% (1400 m–1600 m), 0.0–1.7% (1600–1800 m) and 0.0%–3.4% (1800–2000 m). No cases were found at any sites at an altitude greater than 2000 m. Using the gridded population estimates from WorldPop (100 m) and SRTM elevation (90 m), of the 39.2 million people living in Uganda, 82.8% (32.5 million) of people live at altitudes below 1400 m. Of the remainder, 15.4% (6 M) live within an altitude range of 1400 m–2000 m and the remaining 1.8% (0.7 million) live above 2000 m. The analysis presented in this paper demonstrates the potential for intestinal schistosomiasis transmission at altitudes greater than 1400 m. This was clearly demonstrated both by the presence of disease and by the presence of intermediate snail hosts in Mount Elgon and Fort Portal crater lakes, respectively. The prevalence of disease at these higher altitudes should not be overlooked, and as such should be considered for inclusion in the national control programme. For natural transmission of intestinal schistosomiasis to occur, several critical aspects within the lifecycle of S. mansoni need to be fulfilled . Foremost perhaps, is the presence of permissive populations of Biomphalaria spp., alongside sufficient environmental opportunity for larval stages of the schistosome to encounter and successfully develop within both definitive and intermediate hosts. As might be expected, there is an optimal range of temperatures operating across these processes both in time and space which may facilitate or, where thermal boundaries are exceeded by being too cold or too hot, stall natural transmission . Hence there is ample reason and often sufficient ecological evidence to attempt to predict where transmission is or is not possible. There are, however, well-known conceptual problems broadly grouped as issues of geographical scale . In terms of public health, the implication of this can be profound for it may under-estimate or over-estimate the preventive chemotherapy needs of people living within areas predicted or proven to be at-risk . Where resources permit, it is therefore sensible to have iterative cycles that progressively refine any prediction(s) such that subsequent validations are undertaken to ‘ground-truth’ any newly envisaged scenario. The first formal attempt to describe the distribution of Biomphalaria spp. in Uganda was by Georg Mandahl-Barth in his 1954 monograph entitled “ The freshwater mollusks of Uganda and adjacent territories ” . His attention focused most on Biomphalaria spp. from the great lakes and in lowland areas but mention was made of the crater lake fauna south of Fort Portal where he considered Biomphalaria adowensis adowensis (Bourguignat, 1879) to occur. Realising that his knowledge of the distribution and ecology of many species was scant, he provided only a broad overview but observed that Biomphalaria rṻppellii (Dunker, 1848) was found in highland areas in the South-West of the country within Lakes Mutanda (1800 m) and Bunyonyi (1962 m) . Since then many of the older species names of Biomphalaria have been synonymised as the genus has been downwardly revised to represent some 12 species in total . Most importantly all currently named species have some natural or experimental compatibility with S. mansoni hence the presence of Biomphalaria spp. alone is sufficient to raise suspicion of local transmission potential . Clearly from the information provided here, Biomphalaria spp. has been proven to be found at altitudes very close to 2000 m on Mount Elgon and with reference to Lake Bunyonyi was first reported in the 1930s and is present today. Even with the introduction of molecular DNA typing methods for species delineation, it remains difficult to describe and record precisely the distribution of each species of Biomphalaria in Uganda . A good example is the status of populations in Lake Victoria although there have been successful attempts to map, describe and predict general distributions throughout the lake . In terms of assessing transmission, recourse to a combination of traditional methods to inspect snails for evidence of infection, by cercarial shedding for example, alongside new DNA assays to detect schistosomes in snails is a powerful way to assess transmission in nature . With the observation of schistosome cercariae, it is apparent that active transmission was caught in action and was occurring during the survey of the Fort Portal crater lakes. It is perhaps unsurprising that the previous survey of Rubaihayo et al . found egg-patent prevalence of 27.8% around the crater lakes within the altitudinal range of 1487 m–1682 m . There are also several recent reports of travellers contracting intestinal schistosomiasis locally from these crater lakes . While shedding cercariae were not observed in the Mount Elgon survey, 7.1% of examined snails did have evidence of schistosome DNA which at the very least, demonstrates either recent encounters with schistosome miracidia or that snails were incubating sporocysts within the pre-patent period before cercariogenesis . With hindsight our malacological surveys should have inspected a sub-sample of snails by crushing to visualise the presence of sporocyts and bolster this with recourse to real-time PCR approaches which can better quantify levels of schistosome DNA than traditional PCR methods using gel electrophoresis. With better quantification of DNA it should be possible to set a detection threshold which if exceeded, differentiates miracidial-contamination from cercarial-emergence events. With the introduction of more sensitive diagnostics for intestinal schistosomiasis in people which clearly evidence the limitations of Kato-Katz, it is not surprising that a greater number of infections were encountered . Figs. and demonstrate that while egg-patent prevalence was less than 5.0%, using SEA-ELISA or CCA the prevalence would fall somewhere between 5.0 and 37.5% in UPE schools above 1400 m. It is therefore safe to assume that rather than schistosomiasis being absent, it only appears absent due to the use of insensitive diagnostic tests. Table shows that while there is a negative association with altitude there is a significant amount of infection between 1400 and 1600 m, with a likely absolute boundary exceeding 2000 m, although it should be noted however that these logistic regression models had a poor goodness of fit ( P < 0.05) possibly due to the lack of additional important risk factors in the model or the small sample size. More data are therefore needed to explore this finding further. As an example, a prior survey for intestinal schistosomiasis at Hamukaaka (also known as Amasiko) village was undertaken on the shoreline of Lake Buynonyi (~1960 m) in 2006 and tested both pre-school children ( n = 26) and their mothers ( n = 29) by finger prick SEA-ELISA, urine-CCA dipsticks and stool concentration methods. Although Biomphalaria spp. could be found locally, there was no evidence of schistosome infection in children or in adults . For those people living above 1400 m who are egg-negative by Kato-Katz but are positive by SEA-ELISA or urine-CCA test should have access to praziquantel treatment. Moreover, the significance of ‘asymptomatic’ schistosomiasis has been debated, concluding that there is a tangible benefit to treating those who have any evidence of infection . Using available data on population density it is safe to assume that some 6.7 million people live within inhabited areas above 1400 m and that a sizeable fraction, perhaps a quarter to a half, will likely have intestinal schistosomiasis. However, these people have not yet been reached with treatment. Rather than set aside these areas for control, there is now an imperative for expanded access to preventive chemotherapy to be undertaken. This is especially true if control ultimately aims to eliminate morbidity and transmission, otherwise intestinal schistosomiasis could continue on within such high altitude refugia. These conjoint parasitological and malacological surveys undertaken on Mount Elgon and in the Fort Portal Crater Lakes clearly show that natural transmission of S. mansoni occurs at altitudes above 1400 m, with a putative upper boundary of 2000 m, and reveals additional preventive chemotherapy needs. In future, collected snails should also be examined for the presence of developing sporocysts by microscopy as well as implement real-time PCR techniques to better quantify levels of schistosome DNA snails with pre-patent infection(s). In these highlands, there is an appreciable burden of intestinal schistosomiasis in school children attending UPE schools. Using spatial epidemiological predictions, this now places some extra six million Ugandans at-risk of disease and calls for a much needed expansion of preventive chemotherapy by the Ugandan NCP in these highland areas. This will not only provide more equitable distribution of praziquantel treatment but also prevent schistosomes from making transmission refugia at higher altitude. |
Assessment of efficacy and safety by CalliSpheres versus HepaSpheres for drug-eluting bead transarterial chemoembolization in unresectable large hepatocellular carcinoma patients | 22872788-6ab2-4d30-95bd-833f4bb3b20f | 8245102 | Pharmacology[mh] | Nowadays, liver cancer is one of the world’s most frequent and lethal cancer with an estimated incidence of 9.3 and a mortality of 8.5 per 100,000 person-years, and hepatocellular carcinoma (HCC) is the predominant type of liver cancers (Bray et al., ). HCC treatment is predominantly based on the Barcelona Clinic Liver Cancer (BCLC) staging system, and patients in the intermediate-late stage are the most common in the clinical setting; however, the treatment options for these patients are quite limited (Pinero et al., 2018). In recent decades, the introduction of transarterial chemoembolization (TACE) has vastly improved the HCC patients’ prognosis, especially for intermediate-late stage patients, among which, despite that conventional TACE is the most common in the clinical setting, drug-eluting bead TACE (DEB-TACE) is recently becoming more and more popular thanks to its benefit in treatment responses and tolerance profile (Li et al., ; Xiang et al., ; Wang et al., ). Nonetheless, whether the category of microspheres used in DEB-TACE has impact on efficacy and safety in HCC patients is largely unknown. HepaSpheres and CalliSpheres are two competitive microsphere products in the treatment using DEB-TACE among HCC patients. HepaSpheres has been applied in the clinical practice for almost 30 years, it is a vinyl alcohol-sodium acrylate microsphere featured by good quality in chemotherapeutic absorbing/releasing abilities and its acceptable flexibility in the vessels (Jordan et al., ; Zurstrassen et al., ). As for CalliSpheres, it is the first microsphere product completely produced in China and utilized for DEB-TACE in recent years, and there are many clinical studies and trials elucidating that it presents with satisfying efficacy and safety in DEB-TACE treatments for HCC patients (Ren et al., ; Zhang et al., ; Duan et al., ; Liang et al., ). Moreover, the DEB-TACE using microsphere products also presents with considerable efficacy and tolerance in some advanced HCC patients, such as the patients with large HCC tumor (Song et al., ). However, to our best knowledge, the comparison of efficacy or tolerance between HepaSpheres and CalliSpheres in large HCC patients has not been done. Therefore, the aim of the present study was to compare the efficacy and safety of HepaSpheres and CalliSpheres in unresectable large HCC patients.
Study population In this multicenter retrospective study, we analyzed 127 unresectable large HCC patients who received DEB-TACE treatment with CalliSpheres or HepaSpheres microspheres in Affiliated Zhongshan Hospital of Dalian University, Beijing Tsinghua Changgung Hospital, and Linyi Cancer Hospital, from July 2016 to July 2018. Inclusion criteria were as follows: (a) diagnosed as primary HCC; (b) BCLC stage B or C; (c) maximum diameter of a single tumor ≥5 cm; (d) Child-Pugh stage A or B; (e) age ≤80 years; (f) Eastern Cooperative Oncology Group performance status (ECOG PS) score of 0–2 points; (g) treated by DEB-TACE with CalliSpheres or HepaSpheres microspheres. Exclusion criteria included (a) previously underwent radiofrequency ablation or radio-chemotherapy before TACE; (b) complicated with other malignancies; (c) data missing including clinical data, imaging evaluations, or follow-up records. Among 127 patients, 67 patients who received DEB-TACE with CalliSpheres were categorized as CalliSpheres group, and 60 patients who received DEB-TACE with HepaSpheres were categorized as HepaSpheres group. The study complied with the Declaration of Helsinki, and the Ethics Committee of the Affiliated Zhongshan Hospital of Dalian University (principal research center) approved it. All patients provided the written informed consent forms. Clinical data collection Medical records of patients were reviewed to collect the basic clinical features for study analysis, which mainly included age, gender, history of liver disease, ECOG PS score, model for end-stage liver disease (MELD) score, Child-Pugh stage, tumor size, numbers of tumors, vascular invasion status, extrahepatic metastasis status, BCLC stage, and preoperative alpha-fetoprotein (AFP). DEB-TACE treatment Before operation, drug-loading was performed. For CalliSpheres group, 50 mg pirarubicin solution was mixed with 1.0 g CalliSpheres (diameter: 300–500 μm, Suzhou Hengrui Galisheng Biomedical Technology Co., Ltd., Suzhou, China), then the contrast agent was added in a ratio of 1:1, and the mixture was placed for 30 minutes for further use. For HepaSpheres group, ‘4-fold method’ was used for drug-loading, as follows: 50 mg pirarubicin was dissolved in 20 mL of 0.9% NaCl, then 10 mL of the solution was injected into a bottle of HepaSpheres (1.0 g, diameter: 50–100 μm, Biosphere Medical, Inc., South Jordan, UT), followed by shaking up every three minutes for 10 minutes. The remaining 10 mL solution of pirarubicin was extracted into a syringe, next, the suspension in the bottle was also extracted into the syringe, followed by shaking up every five minutes for 15 minutes. After that, the mixture in the syringe was transferred to operating table, followed by full sedimentation. Subsequently, the liquid supernatant was pushed out, and the remaining drug-loaded HepaSphere was diluted with 20 mL contrast agent for further use. After preparation of drug-eluting microspheres, DEB-TACE was carried out. In brief, arterial angiography was performed to identify all arteries that supply blood to the tumor, then the microcatheter was introduced into the target vessel by superselective catheterization. Following that, the prepared drug-eluting microspheres and contrast agent were injected into the target vessel slowly and alternately until the complete disappearance of tumor staining. When necessary, gelatin sponge particles were used for supplementary embolization. Criteria for ending embolization were stoppage of blood flow in the tumor blood vessel and complete disappearance of tumor staining under imaging. Efficacy, safety, and survival evaluation Treatment response was assessed by enhanced computed tomography or magnetic resonance imaging (CT/MRI) in terms of the modified RECIST (mRECIST) assessment for HCC (Lencioni & Llovet, ). Evaluation data of treatment response at 1 month, 3 months, and 6 months after treatment were collected for study analysis. Meanwhile, patients’ Karnofsky performance status (KPS) scores assessed at 1 month, 3 months, and 6 months after treatment were also collected. For safety assessment, the adverse events and the main liver function indexes (before treatment, seven days after treatment, and 30 days after treatment) were collected and analyzed as well. As for survival evaluation, according to the surveillance and follow-up records, time to progression (TTP) and overall survival (OS) were summarized and analyzed in the study. Statistical analysis We performed statistical analyses using SPSS (Social Package for Social Sciences) 20.0 software (IBM, Chicago, IL). Descriptive statistical analysis of the clinicopathological data was performed using mean values, standard deviations, numbers, and percentages. The χ 2 test or Fisher’s exact test was used to analyze the unordered categorical data, and the Wilcoxon rank sum test was used to analyze the ordered categorical data. Student’s t test was used to determine the quantitative data. The Kaplan–Meier method was used to plot the survival curves. Log-rank (Mantel-Cox) test was applied for determining survival analysis. p Values <.05 were considered statistically significant.
In this multicenter retrospective study, we analyzed 127 unresectable large HCC patients who received DEB-TACE treatment with CalliSpheres or HepaSpheres microspheres in Affiliated Zhongshan Hospital of Dalian University, Beijing Tsinghua Changgung Hospital, and Linyi Cancer Hospital, from July 2016 to July 2018. Inclusion criteria were as follows: (a) diagnosed as primary HCC; (b) BCLC stage B or C; (c) maximum diameter of a single tumor ≥5 cm; (d) Child-Pugh stage A or B; (e) age ≤80 years; (f) Eastern Cooperative Oncology Group performance status (ECOG PS) score of 0–2 points; (g) treated by DEB-TACE with CalliSpheres or HepaSpheres microspheres. Exclusion criteria included (a) previously underwent radiofrequency ablation or radio-chemotherapy before TACE; (b) complicated with other malignancies; (c) data missing including clinical data, imaging evaluations, or follow-up records. Among 127 patients, 67 patients who received DEB-TACE with CalliSpheres were categorized as CalliSpheres group, and 60 patients who received DEB-TACE with HepaSpheres were categorized as HepaSpheres group. The study complied with the Declaration of Helsinki, and the Ethics Committee of the Affiliated Zhongshan Hospital of Dalian University (principal research center) approved it. All patients provided the written informed consent forms.
Medical records of patients were reviewed to collect the basic clinical features for study analysis, which mainly included age, gender, history of liver disease, ECOG PS score, model for end-stage liver disease (MELD) score, Child-Pugh stage, tumor size, numbers of tumors, vascular invasion status, extrahepatic metastasis status, BCLC stage, and preoperative alpha-fetoprotein (AFP).
Before operation, drug-loading was performed. For CalliSpheres group, 50 mg pirarubicin solution was mixed with 1.0 g CalliSpheres (diameter: 300–500 μm, Suzhou Hengrui Galisheng Biomedical Technology Co., Ltd., Suzhou, China), then the contrast agent was added in a ratio of 1:1, and the mixture was placed for 30 minutes for further use. For HepaSpheres group, ‘4-fold method’ was used for drug-loading, as follows: 50 mg pirarubicin was dissolved in 20 mL of 0.9% NaCl, then 10 mL of the solution was injected into a bottle of HepaSpheres (1.0 g, diameter: 50–100 μm, Biosphere Medical, Inc., South Jordan, UT), followed by shaking up every three minutes for 10 minutes. The remaining 10 mL solution of pirarubicin was extracted into a syringe, next, the suspension in the bottle was also extracted into the syringe, followed by shaking up every five minutes for 15 minutes. After that, the mixture in the syringe was transferred to operating table, followed by full sedimentation. Subsequently, the liquid supernatant was pushed out, and the remaining drug-loaded HepaSphere was diluted with 20 mL contrast agent for further use. After preparation of drug-eluting microspheres, DEB-TACE was carried out. In brief, arterial angiography was performed to identify all arteries that supply blood to the tumor, then the microcatheter was introduced into the target vessel by superselective catheterization. Following that, the prepared drug-eluting microspheres and contrast agent were injected into the target vessel slowly and alternately until the complete disappearance of tumor staining. When necessary, gelatin sponge particles were used for supplementary embolization. Criteria for ending embolization were stoppage of blood flow in the tumor blood vessel and complete disappearance of tumor staining under imaging.
Treatment response was assessed by enhanced computed tomography or magnetic resonance imaging (CT/MRI) in terms of the modified RECIST (mRECIST) assessment for HCC (Lencioni & Llovet, ). Evaluation data of treatment response at 1 month, 3 months, and 6 months after treatment were collected for study analysis. Meanwhile, patients’ Karnofsky performance status (KPS) scores assessed at 1 month, 3 months, and 6 months after treatment were also collected. For safety assessment, the adverse events and the main liver function indexes (before treatment, seven days after treatment, and 30 days after treatment) were collected and analyzed as well. As for survival evaluation, according to the surveillance and follow-up records, time to progression (TTP) and overall survival (OS) were summarized and analyzed in the study.
We performed statistical analyses using SPSS (Social Package for Social Sciences) 20.0 software (IBM, Chicago, IL). Descriptive statistical analysis of the clinicopathological data was performed using mean values, standard deviations, numbers, and percentages. The χ 2 test or Fisher’s exact test was used to analyze the unordered categorical data, and the Wilcoxon rank sum test was used to analyze the ordered categorical data. Student’s t test was used to determine the quantitative data. The Kaplan–Meier method was used to plot the survival curves. Log-rank (Mantel-Cox) test was applied for determining survival analysis. p Values <.05 were considered statistically significant.
Clinical characteristics of HCC patients No difference was found between the HepaSpheres group and CalliSpheres group regarding all the characteristics of HCC patients . The mean age was 65.5 ± 8.7 years in HepaSpheres group and was 64.3 ± 9.0 years in CalliSpheres group ( p = .431). The number of males was 42 (70.0%) in the HepaSpheres group and was 53 (79.1%) in the CalliSpheres group ( p = .238). In addition, the numbers of patients with ECOG score of 0, 1, and 2 were 37 (61.7%), 20 (33.3%), and 3 (5.0%) in HepaSpheres group, then were 42 (62.7%), 21 (31.3%), as well as 4 (6.0%) in CalliSpheres group ( p = .946). There were 34 (56.7%) and 26 (43.3%) patients in HepaSpheres group, and 40 (59.7%) and 27 (40.3%) patients in CalliSpheres group who had Child-Pugh stage A and B, respectively ( p = .729). Besides, the mean value of tumor size was 7.9 ± 2.5 cm in the HepaSpheres group and was 7.5 ± 2.3 cm in the CalliSpheres group ( p = .350). The numbers of patients in BCLC B stage and C stage were 27 (45.0%) and 33 (55.0%) in the HepaSpheres group, and were 32 (47.8%) and 35 (52.2%) in the CalliSpheres group ( p = .755). The information of other characteristics is displayed in . Comparison of treatment response between the two groups The general treatment response was more favorable in the CalliSpheres group compared to HepaSpheres group, presenting as that the CR and PR were higher while SD was lower in CalliSpheres group compared to HepaSpheres group at 1 month after treatment, while there was no PD in CalliSpheres group ( p = .026) . However, the general treatment response at 3 months ( p = .863) and 6 months ( p = .853) were of no difference between the two groups. In terms of ORR and DCR, the ORR ( p = .030) and DCR ( p = .030) were both increased in CalliSpheres group compared to HepaSpheres group at 1 month after treatment . However, the ORR ( p = .817) and DCR ( p = .867) at 3 months after treatment were similar between the two groups ; in addition, the ORR ( p = .153) as well as DCR ( p = .796) at 6 months were also of no difference between the two groups . Comparison of performance status, liver function, and adverse events between the two groups Moreover, the performance status by KPS was also compared between the CalliSpheres group and HepaSpheres group, which disclosed that the KPS score was similar between the two groups at 1 month ( p = .251), 3 months ( p = .695), and 6 months ( p = .432) after treatment . The liver function index levels including ALT, AST, ALB, and TBIL at seven days and 30 days after treatment were similar between the CalliSpheres group and HepaSpheres group (all p > .05) . In regard to the adverse events, which consisted of nausea/vomiting, pain, fever, myelosuppression, biloma, and abscess, were of no difference between the two groups (all p > .05) . As for the adverse events in grade I–II, they were similar between the two groups as well (all p > .05), and so did the adverse events in grade III–IV (all p > .05). Comparison of survival profile between the two groups The TTP was similar between the CalliSpheres group and HepaSpheres group, and the median value of TTP was 6.3 months (95%CI: 5.9–6.6 months) in the CalliSpheres group, and was 6.0 months (95%CI: 5.6–6.4 months) in the HepaSpheres group ( p = .082) . As for OS, it was also of no difference between the two groups, and the median value of OS was 23.0 months (95%CI: 20.1–25.9 months) in the CalliSpheres group, and was 22.0 months (95%CI: 20.2–23.8 months) in the HepaSpheres group ( p = .571) . In addition, the numbers of patients at risk regarding TTP and OS were also shown at the bottom of , respectively.
No difference was found between the HepaSpheres group and CalliSpheres group regarding all the characteristics of HCC patients . The mean age was 65.5 ± 8.7 years in HepaSpheres group and was 64.3 ± 9.0 years in CalliSpheres group ( p = .431). The number of males was 42 (70.0%) in the HepaSpheres group and was 53 (79.1%) in the CalliSpheres group ( p = .238). In addition, the numbers of patients with ECOG score of 0, 1, and 2 were 37 (61.7%), 20 (33.3%), and 3 (5.0%) in HepaSpheres group, then were 42 (62.7%), 21 (31.3%), as well as 4 (6.0%) in CalliSpheres group ( p = .946). There were 34 (56.7%) and 26 (43.3%) patients in HepaSpheres group, and 40 (59.7%) and 27 (40.3%) patients in CalliSpheres group who had Child-Pugh stage A and B, respectively ( p = .729). Besides, the mean value of tumor size was 7.9 ± 2.5 cm in the HepaSpheres group and was 7.5 ± 2.3 cm in the CalliSpheres group ( p = .350). The numbers of patients in BCLC B stage and C stage were 27 (45.0%) and 33 (55.0%) in the HepaSpheres group, and were 32 (47.8%) and 35 (52.2%) in the CalliSpheres group ( p = .755). The information of other characteristics is displayed in .
The general treatment response was more favorable in the CalliSpheres group compared to HepaSpheres group, presenting as that the CR and PR were higher while SD was lower in CalliSpheres group compared to HepaSpheres group at 1 month after treatment, while there was no PD in CalliSpheres group ( p = .026) . However, the general treatment response at 3 months ( p = .863) and 6 months ( p = .853) were of no difference between the two groups. In terms of ORR and DCR, the ORR ( p = .030) and DCR ( p = .030) were both increased in CalliSpheres group compared to HepaSpheres group at 1 month after treatment . However, the ORR ( p = .817) and DCR ( p = .867) at 3 months after treatment were similar between the two groups ; in addition, the ORR ( p = .153) as well as DCR ( p = .796) at 6 months were also of no difference between the two groups .
Moreover, the performance status by KPS was also compared between the CalliSpheres group and HepaSpheres group, which disclosed that the KPS score was similar between the two groups at 1 month ( p = .251), 3 months ( p = .695), and 6 months ( p = .432) after treatment . The liver function index levels including ALT, AST, ALB, and TBIL at seven days and 30 days after treatment were similar between the CalliSpheres group and HepaSpheres group (all p > .05) . In regard to the adverse events, which consisted of nausea/vomiting, pain, fever, myelosuppression, biloma, and abscess, were of no difference between the two groups (all p > .05) . As for the adverse events in grade I–II, they were similar between the two groups as well (all p > .05), and so did the adverse events in grade III–IV (all p > .05).
The TTP was similar between the CalliSpheres group and HepaSpheres group, and the median value of TTP was 6.3 months (95%CI: 5.9–6.6 months) in the CalliSpheres group, and was 6.0 months (95%CI: 5.6–6.4 months) in the HepaSpheres group ( p = .082) . As for OS, it was also of no difference between the two groups, and the median value of OS was 23.0 months (95%CI: 20.1–25.9 months) in the CalliSpheres group, and was 22.0 months (95%CI: 20.2–23.8 months) in the HepaSpheres group ( p = .571) . In addition, the numbers of patients at risk regarding TTP and OS were also shown at the bottom of , respectively.
HCC is an aggressive and thus hard to treat solid caner, often presenting with advanced disease at the time of diagnosis that blocks many potential curative therapies (Grandhi et al., ; Chedid et al., ). DEB-TACE is effective and safe for HCC patients, although conventional TACE is still the mainstay of TACE treatments, the utilization of DEB-TACE presents with a growing trend in cancer patients. As for other embolization therapies, there is evidence elucidating that radioembolization and bland embolization are comparable regarding efficacy when compared to the chemoembolization in liver cancer patients; however, more efforts are needed to compare the efficacy and safety among different TACE treatments in liver cancer patients (Facciorusso et al., , ). However, most authors emphasize the superiority or equal value of DEB-TACE compared to conventional TACE in treating HCC, very few authors investigate the impact of the microsphere category on efficacy and safety. For instance, there are many studies reporting the efficacy of DEB-TACE using HepaSpheres in HCC patients. A previous cohort study with 30 HCC patients treated by DEB-TACE using 50–100 μm HepaSpheres reveals that, at 1 month post treatment, the ORR is 63.3% and the DCR is 86.7% (Sun et al., ). Another prospective cohort study illustrates that in 18 HCC patients treated with DEB-TACE, the application of 50–100 μm HepaSpheres achieves an ORR of 53.3%, and the BCLC stage is correlated with treatment response rate (Zurstrassen et al., ). In regard to the CalliSpheres, a cohort study illuminates that in 50 middle stage HCC patients treated with DEB-TACE using CalliSpheres, CR, PR, SD, and PD are 35.4%, 29.4%, 17.6%, and 17.6%, respectively, in patients treated with 100–300 μm beads, and are 33.1%, 23.1%, 20.8%, as well as 23.0% respectively in patients treated with 300–500 μm beads (Wang et al., ). Another study consisting of a cohort of 90 HCC patients receiving DEB-TACE treatment using CalliSpheres reports that, the DCR at 3 months, 6 months, 1 year, and 2 years post treatment are 93.33%, 88.89%, 36.67%, and 12.22%, respectively; in addition, the 2-year survival rate is 45.56% (Cao et al., ). In terms of the efficacy of DEB-TACE for patients with large HCC tumor, the studies are very rare, only few studies could be referred to. For instance, in a study of 81 elderly patients with advanced HCC with the largest tumor size of 5–10 cm, DEB-TACE using 300–500 μm CalliSpheres achieves a DCR of 75.8%, 42.4%, and 12.1% at 1, 3, and 6 months post treatment, respectively (Yang et al., ). Nevertheless, no effort has been done to explore the effect of different microsphere products on the efficacy and safety of DEB-TACE for treating patients with unresectable large HCC. In the present study, we found that the general treatment response of unresectable large HCC patients at 1 month after treatment was more favorable by CalliSpheres compared to HepaSpheres, but the treatment response at 3 months or 6 months after treatment was not. Additionally, the KPS scores at 1 month, 3 months, and 6 months after treatment were of no difference between the two groups. In terms of survival profile, there was no difference regarding TTP or OS between the two groups, either. For the increased treatment response rates by CalliSpheres compared to HepaSpheres at 1 month after treatment, we presumed that it could be caused by the following reasons. One of the possible reasons was that, the drug elution speed was lower when the microsphere diameter was greater, which indicated that the CalliSpheres (300–500 μm) used in our study was better in the slow-release effect of the chemotherapeutics compared to HepaSpheres (50–100 μm), and this might contribute to the more favorable efficacy by CalliSpheres at 1 month after treatment (Han et al., ; Chen et al., , ). Another reason could be that, HepaSpheres is reported to present with fractures during the release of the drug, which could weaken the embolization effect in the vessel, thus caused a worse tumor necrosis effect (Jordan et al., ). Regarding the safety profile, a previous study uncovers that in 15 patients (colorectal cancer liver metastasis and intrahepatic cholangiocarcinoma) treated with DEB-TACE using HepaSpheres, there are no severe adverse events during and post the treatment, with the most common adverse event of embolization syndrome (Poggi et al., ). Besides, another study reports that in unresectable HCC patients treated with DEB-TACE using HepaSpheres, the incidence of embolization syndrome, treatment related mortality and treatment related morbidity are 89%, 1.9%, and 9.4%, respectively (Kucukay et al., ). In regard to the safety profile of CalliSpheres, the rates of embolization syndrome, transient liver injury, liver abscess, ascites, myelosuppression, and granulocyte reduction are 62.5%, 46%, 4.1%, 13%, 4.1%, and 8.3%, respectively (Wu et al., ). In addition, a prospective cohort study elucidates that the incidences of liver function injury, pain, nausea, vomiting, and fever are 43.9%, 40.9%, 33.3%, 19.7%, and 56.1%, respectively, in 66 HCC patients treated with CalliSpheres DEB-TACE (Zhang et al., ). As for the safety of DEB-TACE in patients with large HCC, one study reveals that using 300–500 μm CalliSpheres in DEB-TACE to treat elderly patients with large HCC, no severe complication exists and the most common adverse event is embolization syndrome (Yang et al., ). In this study, we evaluated the liver function index levels at seven days and 30 days and adverse events after treatment, they were of no difference between the HepaSpheres group and CalliSpheres group. These indicated that CalliSpheres was equally tolerable compared to HepaSpheres in DEB-TACE for treatment of patients with large HCC. However, in the clinical setting, CalliSpheres is much more economical compared to HepaSpheres, which might indicate that patient could obtain comparable efficacy and safety but cost less if choosing CalliSpheres for DEB-TACE treatment (Kadam & Chuan, ; Zhou et al., ). Another issue that should be discussed here was the limitation of our study. First, this was a retrospective observational study, which could cause some bias, such as the information bias. Second, the sample of 127 HCC patients was relatively small and this may interfere with our statistical power to some extent. Third, merely one size of each microsphere product (HepaSpheres and CalliSpheres) was evaluated in our study, thus, the impact of other microspheres sizes on efficacy and safety in unresectable large HCC patients, such as the 100–300 μm CalliSpheres, was still obscure. In conclusion, CalliSpheres seems to be superior in short-term efficacy and equal in long-term efficacy as well as safety compared to HepaSpheres for DEB-TACE treatment in patients with unresectable large HCC.
|
Validation of the PEN-FAST Score in a Pediatric Population | 1808dc8d-5760-4dd7-a1c3-87c728f099b5 | 9486451 | Pediatrics[mh] | The pediatric prevalence of self-reported drug allergies is 10%, which carries significant health and economic implications. Following direct oral penicillin challenge, 94.6% of such labels are removed. , However, despite published algorithms, there are no validated pediatric decision rules to guide clinician management. The aim of this cohort study was to examine the previously validated PEN-FAST adult score in children. Using a Canadian prospective pediatric cohort from 3 centers, we examined the PEN-FAST score in 2028 children with 2031 penicillin allergy labels (eMethods in the ). Data were collected from August 8, 2011, to March 3, 2021. This cohort study was approved by the McGill University Ethics Committee and the Research Ethics Board at the University of Manitoba, and written informed consent was collected. Sample characteristics are presented as median (IQR) and frequency . The PEN-FAST score and area under the curve (AUC) were calculated. Logistic regression with components of the score was performed. Sensitivity analysis with different time categories and removal of severe cutaneous adverse reaction (SCAR) was performed, and subgroup analysis for immediate and delayed reactions and various age groups were performed. All analyses were performed in Stata version 16.1 (StataCorp). The median (IQR) age for the 2028 children in the cohort was 4.3 (2.1-8.0) years, with mostly male participants (1091 [53.7%]). Most reported reactions occurred in the past 5 years or at an unknown time (1661 [81.8%]), with amoxicillin suspected in 2022 reactions (99.6%) . Anaphylaxis and angioedema were reported in 229 cases (11.3%). Treatment (or unknown) was administered for 1231 cases (60.6%). The AUC for the PEN-FAST score was calculated at 0.528, showing poor discrimination ability. Using the published adult PEN-FAST cutoff of 3 or greater, the AUC was 0.510 (95% CI, 0.47-0.56), and sensitivity and specificity were 57.0% (95% CI, 47.1%-66.5%) and 45.7% (95% CI, 43.5%-48.0%), respectively. The negative predictive value was 95.0% (95% CI, 93.4%-96.3%), considered poor in the context of a low prevalence positive challenge (5%). Furthermore, none of the individual variables were associated with a positive test. Changing the coding for timing (<1 year) or removing the angioedema reported symptom did not improve the performance of the PEN-FAST tool in this pediatric population . A subgroup analysis for the positive skin testing or challenges based on immediate vs delayed reaction or the time of the reported allergy showed similar results . When the tool was used in children 13 years or older, the AUC was 0.622, indicating that despite variable adjustment, the tool is not useful . In this Canadian pediatric prospective multicenter cohort, the PEN-FAST tool did not help identify low-risk penicillin allergies. This previously validated tool in an adult population was not useful for risk stratification in children younger than 12 years. In teenagers (≥13 years), the predictive ability of the tool increased (higher AUC, specificity, and NPV but lower sensitivity), indicating that the tool could have some value in this population following further study. The extrapolation of the results is limited by the small number of teenagers included. The PEN-FAST performed similarly for immediate and delayed reactions, and the timing of the reported reaction was not associated with the outcome of the allergy investigation. In this context, it is possible that the low validity of the PEN-FAST tool in children is explained by the increased prevalence of viral-induced reactions compared with true drug hypersensitivity. This analysis adds to the evidence that true drug allergies are rare among children and that they are often incorrectly labeled during a viral infection. Furthermore, the criteria included in the PEN-FAST score might not provide adequate information considering different index reactions in the pediatric population compared with the adult population. This study highlights that children are not little adults and clinical decision rules need to be derived and validated in the target population. These findings suggest that the PEN-FAST drug allergy clinical decision rule should not be adapted to a pediatric population younger than 12 years at this time. New validated point-of-care clinical tools are required to identify low-risk penicillin allergies in a pediatric population, and validation of PEN-FAST in adolescents requires further examination in extended international cohorts. |
Barriers and facilitators experienced by patients, carers and healthcare professionals when managing symptoms in infants, children and young people at end-of-life: a mixed methods systematic review protocol | c9710b6a-5804-40bc-bb2c-3de0eee90918 | 6661662 | Pediatrics[mh] | Approximately 40 000 infants, children and young people (ICYP) are living with a life-threatening or life-limiting condition in England. These include congenital anomalies, cancer and neurological, haematological, respiratory, genitourinary, perinatal, metabolic, circulatory and gastrointestinal conditions. There were nearly 3000 child deaths due to medical conditions in England in 2017, of which over 2350 were due to a known life-limiting condition or neonatal death. ICYP’s palliative care needs often differ from those of adults, and the diversity of conditions in this population means that practitioners must manage a wide range of complex symptoms. A particular challenge is managing continuous ‘background’ pain as well as bouts of severe, sudden-onset ‘breakthrough pain’, both of which are common in ICYP with a terminal illness and are known to be underassessed and undertreated. Family carers play a vital role in supporting ICYP with a terminal illness, allowing patients to be cared for and die at home where possible. However, there is little research on carers’ experiences of administering medicines for symptom relief to ICYP receiving palliative care. Managing symptoms such as pain is potentially difficult for carers of children at home. They may lack the necessary skills and confidence required to balance symptom relief and side effects while fear of errors can lead to insufficient or inappropriate doses of analgesics. Families will move ICYP away from their preferred place of care if symptoms, including pain, are not managed effectively. Community nurses and doctors may also lack the skills and experience required to support carers. A systematic review found that GPs experience anxiety regarding their competency to deliver appropriate palliative care while healthcare support workers providing end-of-life care in the community require training in palliative care to cope with emotionally demanding situations. The recent National Institute for Health and Care Excellence (NICE) guideline is based on evidence from 20 systematic reviews investigating different aspects of planning and management of end-of-life care for ICYP with life-limiting conditions. These include reviews on what information is perceived as helpful and what social and practical support is effective for ICYP and their caregivers. The findings indicate that timely, honest and consistent information that meets individuals’ needs (eg, developmentally appropriate for patients) is beneficial, including information about access to services, community and medical resources. One study also found that parents wanted information on how to use equipment that a child/young person required. However, symptom management was not identified as a major theme in these reviews. Four other reviews looked at the effectiveness of pharmacological and non-pharmacological interventions for pain management, agitation, respiratory distress and seizures. Only the pain management review found any studies that met the inclusion criteria and all of these involved pharmacological interventions only. Although these reviews provide essential guidance in managing end-of-life care for ICYP, to our knowledge, no systematic review has examined the barriers and facilitators to symptom management in ICYP at end-of-life for healthcare professionals, caregivers and patients. NICE highlights pain management in palliative care as a research priority and a greater understanding of this could inform the design of evidence-based interventions to support more effective medicine management. Objectives The main objective of this systematic review is to identify and synthesise the existing literature that explores the barriers and facilitators experienced by children and young people themselves and their carers and healthcare professionals when managing symptoms in ICYP at end-of-life.
The main objective of this systematic review is to identify and synthesise the existing literature that explores the barriers and facilitators experienced by children and young people themselves and their carers and healthcare professionals when managing symptoms in ICYP at end-of-life.
This protocol follows the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) guidelines (see online —PRISMA-P checklist) and is registered (ID CRD42019124797) on PROSPERO, an international register of systematic reviews. Any changes to the protocol will be recorded on PROSPERO. 10.1136/bmjopen-2019-030566.supp1 Supplementary data The reporting of the systematic review will be informed by the Centre for Reviews and Dissemination and the Cochrane Qualitative Research Methods Group guidelines and will follow the Enhancing Transparency in Reporting the Synthesis of Qualitative Research (ENTREQ) and the PRISMA statements for reporting systematic reviews (see online ). In the case of sections applicable to qualitative systematic reviews that are included in PRISMA, but are not covered by ENTREQ, these will also be reported. 10.1136/bmjopen-2019-030566.supp2 Supplementary data Eligibility criteria The criteria outlined below will be used for study selection. Following the recommendations of the Cochrane Qualitative and Implementation Methods Group Guidance, we have used STARLITE (Sampling Strategy, Type of study, Approaches, Range of years, Limits, Inclusions and exclusions, Terms used, Electronic sources ) to report our search methods. Sampling strategy This review will consider all studies carried out worldwide that involve carers, healthcare professionals or patients’ views on symptom management in ICYP up to the age of 24 years at end-of-life care. A cut-off age of 24 years will be used since this corresponds to adolescent growth and current understandings of this stage in life. Type of study The review will consider qualitative, quantitative and mixed-method studies including questionnaires, surveys, interviews, focus groups, case studies and observations. Trials, cohort and intervention studies that assess barriers and facilitators to symptom management will all be considered. Approaches In addition to searching electronic databases, the search strategy will include hand searching of reference lists of identified eligible studies. Finally, active researchers in the field who have contributed to this literature will be contacted. Range of years Studies published from the inception of each database will be included. Limits Articles written in any language other than English will not be searched due to a lack of funding for adequate translation, along with masters theses, conference abstracts and reviews. Inclusions Studies reporting barriers and facilitators experienced by carers, healthcare professionals and the patients themselves, when managing symptoms in ICYP with terminal illnesses receiving palliative care and/or at end-of-life will be included. All definitions of ‘end-of-life’ will be included since there are a wide variety of definitions and there is a paucity of research in ICYP symptom management in this area. Data on carers, healthcare professionals and patients’ views, attitudes, opinions, perceptions, beliefs or feelings will be included. Exclusions Studies that focus only on the effectiveness of pharmacological treatments for symptom management will be excluded. Searches Electronic sources The Cochrane Library and PROSPERO will be searched initially to check for any existing systematic reviews on this topic. As recommended by the Cochrane Qualitative and Implementation Methods Group, CINAHL (Cumulative Index of Nursing and Allied Health Literature) via Ebsco and Ovid MEDLINE will be searched, as well as PsycINFO via Ebsco and the Web of Science Core Collection. To identify any additional unpublished work, the ProQuest Dissertations & Theses Database, Evidence Search and OpenGrey will also be searched. The search strategy will include hand searching of reference lists of eligible studies for additional records. All searches will be run during February 2019. Search terms used A search strategy was developed based on the ‘Managing Pain’ search strategy used in the NICE guideline ’End-of-life Care for Infants, Children and Young people with Life-limiting Conditions: Planning and Management (NG61)’. The strategy incorporated search terms in four blocks: (1) ‘Patient Population’; (2) ‘Caregivers and Patients’; (3) ‘End-of-life’; and (4) ‘Pain and Symptoms.’ Additional searches used in the Palliative Care Search Filter were also incorporated into Block 3 for each database. Combinations of keywords, text words, Medical Subject Headings (MeSH) and other terms relevant to the four blocks were selected for each database to optimise the search sensitivity and specificity. The search strategy was piloted and adapted for each database. A professional healthcare research librarian assisted in the development of the strategy. Please see online for the full search strategy for each database. 10.1136/bmjopen-2019-030566.supp3 Supplementary data Data management All records and data will be saved to Endnote X8. This software will be used to identify potential duplicates. The researchers will check this and remove all confirmed duplicated articles. Selection process Articles will be screened by title and abstract by one reviewer (KG) with a second reviewer (SH) assessing 10% of the articles, randomly selected. At this stage, articles will be judged as either (1) ‘not relevant’ or (2) ‘potentially relevant’. Both reviewers will read and screen all remaining potentially relevant articles. The reviewers will independently apply the criteria at all stages of the selection process. Intercoder agreement will be evaluated using Cohen’s kappa coefficient. A minimum kappa value of 0.75 will be taken to represent high agreement. The full text of all remaining potentially relevant articles will then be obtained. If the relevance of a study cannot be ascertained from the abstract, then the full article will be obtained. The full articles will be read by two reviewers independently (KG and SH) to make the final decision about whether they should be chosen for inclusion in the review. A third reviewer (CL) will resolve any uncertainties. Additional information will be sought from authors if necessary at the stage of full-text assessment. Data collection process and items The following information will be extracted into a piloted data collection form for all included studies: study aims; patient population (infant/child/adolescent); participant population (patient/caregiver/healthcare professional); inclusion and exclusion criteria; sample size; recruitment; design; intervention and comparator group (where applicable); date and duration of data collection; setting; country; data collection; analysis methods; data describing the participants’ views/experiences of barriers and facilitators to symptom management. For qualitative data, the authors’ interpretations (presented through themes and categories) will represent these data. KG will extract this information and SH will check it, with any disagreements resolved through discussion with CL. Quality assessment (including risk of bias) A quality appraisal of included studies will be conducted independently by two reviewers (KG and SH). Disagreements will be resolved by discussion between KG and SH, with CL if required. Three checklists will be used depending on each study’s design. These were chosen since they are all validated and have been used in published systematic reviews within healthcare research. For each study type, Cohen’s kappa coefficient will be used to measure inter-rater agreement between the two reviewers. A minimum kappa value of 0.75 will be taken to represent high agreement with disagreements resolved via discussion with CL. As recommended by the Cochrane Qualitative and Implementation Methods Group, we will not calculate total quality scores across domains since domains of quality are not equal. Instead, KG, SH and CL will determine how each study’s methodological limitations affect confidence in the findings via discussion. We will not exclude studies based on poor quality but will record and highlight methodological issues. Qualitative studies will be quality appraised using the Critical Appraisal Skills Programme for Qualitative Studies (CASP). CASP assesses clarity of research aims, research design, recruitment methods, data collection, relationships between participants and researchers, ethical issues, analyses, description of findings and valuableness of the research. It is comprised of nine closed questions (eg, ‘Was there a clear statement of the aims of the research?’ Yes/Can’t tell/No) and one open-ended question (‘How valuable is the research?’). For each question, there is the option to add comments to explain the reasoning for each rating. Currently, CASP is the most frequently used qualitative research synthesis tool in the Cochrane Library and WHO guideline research and has been used in similar systematic reviews assessing barriers and facilitators within healthcare research . However, because the CASP tool does not address aspects of the research validity and can favour papers that are less insightful as long as they comply with ‘expectations of research practice’, in addition, the evaluative criteria of credibility, transferability, dependability and confirmability will be applied. Included studies will be assessed as to whether they apply the techniques suggested for ensuring study quality according to Guba and Lincoln’s criteria, that is, prolonged engagement, persistent observation, peer review, triangulation, negative case analysis, referential adequacy and member checking to ensure credibility; thick description for transferability; inquiry audit for dependability; confirmability audit, audit trail, triangulation and reflexivity to ensure confirmability. Studies will be rated as ‘high quality’ if they meet at least three of the four criteria, ‘medium quality’ if they meet two of the criteria and ‘low quality’ if they meet one or none. The Quality Assessment Tool for Quantitative Studies (QATQS) will be used to assess all clinical studies with or without randomisation and control groups, including quasi-experimental and before-and-after studies. The QATQS is comprised of 22 closed questions and an overall rating of strong, moderate or weak in eight sections: selection bias; study design; confounders; blinding; data collection; withdrawals and dropouts; intervention integrity; analysis. It has been shown to be a valid tool for assessing quality, comparing studies and addressing threats to validity of findings. The Mixed Methods Appraisal Tool (V.11) will be used to assess the quality of any mixed methods studies. This tool consists of five closed questions assessing the research question, research design, integration of qualitative and quantitative methods, integration of qualitative and quantitative data and consideration of methodological limitations in mixed methods studies. As reported by the Cochrane Qualitative and Implementation Methods Group, this tool has been used widely in systematic reviews and has the advantage of being able to assess interdependent qualitative and quantitative elements of mixed-methods research. Outcomes and prioritisation The main outcomes sought are carers’, healthcare professionals’ and patients’ (CYP) views on the barriers and facilitators to effective symptom management in ICYP at end-of-life. Data synthesis Although it is unlikely that the majority of included studies will be quantitative, if this is the case, then random-effects meta-analysis will be conducted to synthesise group means and standard deviation from individual studies using Comprehensive Meta-Analysis version 3. For meta-analysis to be conducted, data must be available from two or more eligible studies reporting similar barriers or facilitators. The studies must report the number of participants reporting that barrier/facilitator and the total number of valid participant responses for that survey item. A random-effects model will be used for all analyses since; unlike a fixed-effects model, this can be used when statistical heterogeneity (I 2 ) is present in the results of the included studies. Where evidence of statistically significant heterogeneity is present, sensitivity analyses will be conducted where possible to verify the robustness of the study conclusions, assessing the impact of methodological quality, study design, sample size and the potential effects of missing data. We will use funnel plots to detect potential reporting biases and small-study effects where data are available from 10 or more studies. If the included studies are all qualitative or a combination of quantitative and qualitative, there are several approaches that could be taken for data synthesis. Some of the most commonly used methods to synthesise qualitative health research include thematic analysis, grounded theory and meta-ethnography. However, there is no consensus on the best approach, which will depend on the type and number of included studies and the form and nature of the research question. As such, we will make a final decision on the most appropriate method after selecting and quality assessing the included articles, as recommended by the Cochrane Qualitative and Implementation Methods Group. We will first analyse and synthesise data related to the experience of patients, care providers and healthcare professionals separately before deciding whether it is appropriate to aggregate data between these groups. These data will likely include themes, concepts and categories of information. If data are relatively ‘thin’, then we will consider using thematic synthesis to undertake line-by-line coding and development of descriptive and analytic themes. If the included articles include sufficient ‘thick’ data (eg, details about the context and background of the studies and participants ), we will consider a more interpretative approach such as meta-ethnography. This method goes beyond aggregating data to generate new interpretations of the findings. As recommended by the Cochrane Qualitative and Implementation Methods Group, the GRADE‐CERQual (Confidence in the Evidence from Reviews of Qualitative research ) will be used to summarise our confidence in synthesised qualitative findings (eg, in the themes that we identify). The CERQual is made up of four key components, that is, methodological limitations of included studies, coherence of the review finding, adequacy of the data contributing to a review finding and relevance of the included studies to the review question. After assessing each of the four components, overall confidence will be graded as high, moderate, low or very low. The barriers and facilitators to symptom management will be divided into overarching themes for each group (patients, healthcare professionals, carers), and presented in a matrix along with our CERQual assessment of confidence in the evidence of each theme and an explanation of this assessment. The GRADE guidelines will be used to appraise the quality of any quantitative findings. The GRADE guidelines include four elements for which quantitative findings will be rated against: risk of bias (‘Study limitations’), inconsistency, indirectness, imprecision, publication bias. Patient and public involvement Consultation with young people, parents and healthcare professionals has been used to determine their perception of the barriers and facilitators they experience when managing symptoms in ICYP at end-of-life. It is based on their perspectives that this systematic review was deemed to be timely and crucial to conduct to inform further research work. Moreover, patient and public involvement (PPI) is represented in the authorship (MJ) of this manuscript. MJ is the PPI representative at the UK National Institute for Health Research (NIHR) Pain and Palliative Care Clinical Studies Group-Children. MJ is a parent of four children, one of whom died from T-cell acute lymphoblastic leukaemia at the age of 12. She has supported many families of children with cancer including those receiving palliative and end-of-life care.
The criteria outlined below will be used for study selection. Following the recommendations of the Cochrane Qualitative and Implementation Methods Group Guidance, we have used STARLITE (Sampling Strategy, Type of study, Approaches, Range of years, Limits, Inclusions and exclusions, Terms used, Electronic sources ) to report our search methods. Sampling strategy This review will consider all studies carried out worldwide that involve carers, healthcare professionals or patients’ views on symptom management in ICYP up to the age of 24 years at end-of-life care. A cut-off age of 24 years will be used since this corresponds to adolescent growth and current understandings of this stage in life. Type of study The review will consider qualitative, quantitative and mixed-method studies including questionnaires, surveys, interviews, focus groups, case studies and observations. Trials, cohort and intervention studies that assess barriers and facilitators to symptom management will all be considered. Approaches In addition to searching electronic databases, the search strategy will include hand searching of reference lists of identified eligible studies. Finally, active researchers in the field who have contributed to this literature will be contacted. Range of years Studies published from the inception of each database will be included. Limits Articles written in any language other than English will not be searched due to a lack of funding for adequate translation, along with masters theses, conference abstracts and reviews. Inclusions Studies reporting barriers and facilitators experienced by carers, healthcare professionals and the patients themselves, when managing symptoms in ICYP with terminal illnesses receiving palliative care and/or at end-of-life will be included. All definitions of ‘end-of-life’ will be included since there are a wide variety of definitions and there is a paucity of research in ICYP symptom management in this area. Data on carers, healthcare professionals and patients’ views, attitudes, opinions, perceptions, beliefs or feelings will be included. Exclusions Studies that focus only on the effectiveness of pharmacological treatments for symptom management will be excluded.
This review will consider all studies carried out worldwide that involve carers, healthcare professionals or patients’ views on symptom management in ICYP up to the age of 24 years at end-of-life care. A cut-off age of 24 years will be used since this corresponds to adolescent growth and current understandings of this stage in life.
The review will consider qualitative, quantitative and mixed-method studies including questionnaires, surveys, interviews, focus groups, case studies and observations. Trials, cohort and intervention studies that assess barriers and facilitators to symptom management will all be considered.
In addition to searching electronic databases, the search strategy will include hand searching of reference lists of identified eligible studies. Finally, active researchers in the field who have contributed to this literature will be contacted.
Studies published from the inception of each database will be included.
Articles written in any language other than English will not be searched due to a lack of funding for adequate translation, along with masters theses, conference abstracts and reviews.
Studies reporting barriers and facilitators experienced by carers, healthcare professionals and the patients themselves, when managing symptoms in ICYP with terminal illnesses receiving palliative care and/or at end-of-life will be included. All definitions of ‘end-of-life’ will be included since there are a wide variety of definitions and there is a paucity of research in ICYP symptom management in this area. Data on carers, healthcare professionals and patients’ views, attitudes, opinions, perceptions, beliefs or feelings will be included.
Studies that focus only on the effectiveness of pharmacological treatments for symptom management will be excluded.
Electronic sources The Cochrane Library and PROSPERO will be searched initially to check for any existing systematic reviews on this topic. As recommended by the Cochrane Qualitative and Implementation Methods Group, CINAHL (Cumulative Index of Nursing and Allied Health Literature) via Ebsco and Ovid MEDLINE will be searched, as well as PsycINFO via Ebsco and the Web of Science Core Collection. To identify any additional unpublished work, the ProQuest Dissertations & Theses Database, Evidence Search and OpenGrey will also be searched. The search strategy will include hand searching of reference lists of eligible studies for additional records. All searches will be run during February 2019. Search terms used A search strategy was developed based on the ‘Managing Pain’ search strategy used in the NICE guideline ’End-of-life Care for Infants, Children and Young people with Life-limiting Conditions: Planning and Management (NG61)’. The strategy incorporated search terms in four blocks: (1) ‘Patient Population’; (2) ‘Caregivers and Patients’; (3) ‘End-of-life’; and (4) ‘Pain and Symptoms.’ Additional searches used in the Palliative Care Search Filter were also incorporated into Block 3 for each database. Combinations of keywords, text words, Medical Subject Headings (MeSH) and other terms relevant to the four blocks were selected for each database to optimise the search sensitivity and specificity. The search strategy was piloted and adapted for each database. A professional healthcare research librarian assisted in the development of the strategy. Please see online for the full search strategy for each database. 10.1136/bmjopen-2019-030566.supp3 Supplementary data
The Cochrane Library and PROSPERO will be searched initially to check for any existing systematic reviews on this topic. As recommended by the Cochrane Qualitative and Implementation Methods Group, CINAHL (Cumulative Index of Nursing and Allied Health Literature) via Ebsco and Ovid MEDLINE will be searched, as well as PsycINFO via Ebsco and the Web of Science Core Collection. To identify any additional unpublished work, the ProQuest Dissertations & Theses Database, Evidence Search and OpenGrey will also be searched. The search strategy will include hand searching of reference lists of eligible studies for additional records. All searches will be run during February 2019.
A search strategy was developed based on the ‘Managing Pain’ search strategy used in the NICE guideline ’End-of-life Care for Infants, Children and Young people with Life-limiting Conditions: Planning and Management (NG61)’. The strategy incorporated search terms in four blocks: (1) ‘Patient Population’; (2) ‘Caregivers and Patients’; (3) ‘End-of-life’; and (4) ‘Pain and Symptoms.’ Additional searches used in the Palliative Care Search Filter were also incorporated into Block 3 for each database. Combinations of keywords, text words, Medical Subject Headings (MeSH) and other terms relevant to the four blocks were selected for each database to optimise the search sensitivity and specificity. The search strategy was piloted and adapted for each database. A professional healthcare research librarian assisted in the development of the strategy. Please see online for the full search strategy for each database. 10.1136/bmjopen-2019-030566.supp3 Supplementary data
All records and data will be saved to Endnote X8. This software will be used to identify potential duplicates. The researchers will check this and remove all confirmed duplicated articles.
Articles will be screened by title and abstract by one reviewer (KG) with a second reviewer (SH) assessing 10% of the articles, randomly selected. At this stage, articles will be judged as either (1) ‘not relevant’ or (2) ‘potentially relevant’. Both reviewers will read and screen all remaining potentially relevant articles. The reviewers will independently apply the criteria at all stages of the selection process. Intercoder agreement will be evaluated using Cohen’s kappa coefficient. A minimum kappa value of 0.75 will be taken to represent high agreement. The full text of all remaining potentially relevant articles will then be obtained. If the relevance of a study cannot be ascertained from the abstract, then the full article will be obtained. The full articles will be read by two reviewers independently (KG and SH) to make the final decision about whether they should be chosen for inclusion in the review. A third reviewer (CL) will resolve any uncertainties. Additional information will be sought from authors if necessary at the stage of full-text assessment.
The following information will be extracted into a piloted data collection form for all included studies: study aims; patient population (infant/child/adolescent); participant population (patient/caregiver/healthcare professional); inclusion and exclusion criteria; sample size; recruitment; design; intervention and comparator group (where applicable); date and duration of data collection; setting; country; data collection; analysis methods; data describing the participants’ views/experiences of barriers and facilitators to symptom management. For qualitative data, the authors’ interpretations (presented through themes and categories) will represent these data. KG will extract this information and SH will check it, with any disagreements resolved through discussion with CL.
A quality appraisal of included studies will be conducted independently by two reviewers (KG and SH). Disagreements will be resolved by discussion between KG and SH, with CL if required. Three checklists will be used depending on each study’s design. These were chosen since they are all validated and have been used in published systematic reviews within healthcare research. For each study type, Cohen’s kappa coefficient will be used to measure inter-rater agreement between the two reviewers. A minimum kappa value of 0.75 will be taken to represent high agreement with disagreements resolved via discussion with CL. As recommended by the Cochrane Qualitative and Implementation Methods Group, we will not calculate total quality scores across domains since domains of quality are not equal. Instead, KG, SH and CL will determine how each study’s methodological limitations affect confidence in the findings via discussion. We will not exclude studies based on poor quality but will record and highlight methodological issues. Qualitative studies will be quality appraised using the Critical Appraisal Skills Programme for Qualitative Studies (CASP). CASP assesses clarity of research aims, research design, recruitment methods, data collection, relationships between participants and researchers, ethical issues, analyses, description of findings and valuableness of the research. It is comprised of nine closed questions (eg, ‘Was there a clear statement of the aims of the research?’ Yes/Can’t tell/No) and one open-ended question (‘How valuable is the research?’). For each question, there is the option to add comments to explain the reasoning for each rating. Currently, CASP is the most frequently used qualitative research synthesis tool in the Cochrane Library and WHO guideline research and has been used in similar systematic reviews assessing barriers and facilitators within healthcare research . However, because the CASP tool does not address aspects of the research validity and can favour papers that are less insightful as long as they comply with ‘expectations of research practice’, in addition, the evaluative criteria of credibility, transferability, dependability and confirmability will be applied. Included studies will be assessed as to whether they apply the techniques suggested for ensuring study quality according to Guba and Lincoln’s criteria, that is, prolonged engagement, persistent observation, peer review, triangulation, negative case analysis, referential adequacy and member checking to ensure credibility; thick description for transferability; inquiry audit for dependability; confirmability audit, audit trail, triangulation and reflexivity to ensure confirmability. Studies will be rated as ‘high quality’ if they meet at least three of the four criteria, ‘medium quality’ if they meet two of the criteria and ‘low quality’ if they meet one or none. The Quality Assessment Tool for Quantitative Studies (QATQS) will be used to assess all clinical studies with or without randomisation and control groups, including quasi-experimental and before-and-after studies. The QATQS is comprised of 22 closed questions and an overall rating of strong, moderate or weak in eight sections: selection bias; study design; confounders; blinding; data collection; withdrawals and dropouts; intervention integrity; analysis. It has been shown to be a valid tool for assessing quality, comparing studies and addressing threats to validity of findings. The Mixed Methods Appraisal Tool (V.11) will be used to assess the quality of any mixed methods studies. This tool consists of five closed questions assessing the research question, research design, integration of qualitative and quantitative methods, integration of qualitative and quantitative data and consideration of methodological limitations in mixed methods studies. As reported by the Cochrane Qualitative and Implementation Methods Group, this tool has been used widely in systematic reviews and has the advantage of being able to assess interdependent qualitative and quantitative elements of mixed-methods research.
The main outcomes sought are carers’, healthcare professionals’ and patients’ (CYP) views on the barriers and facilitators to effective symptom management in ICYP at end-of-life.
Although it is unlikely that the majority of included studies will be quantitative, if this is the case, then random-effects meta-analysis will be conducted to synthesise group means and standard deviation from individual studies using Comprehensive Meta-Analysis version 3. For meta-analysis to be conducted, data must be available from two or more eligible studies reporting similar barriers or facilitators. The studies must report the number of participants reporting that barrier/facilitator and the total number of valid participant responses for that survey item. A random-effects model will be used for all analyses since; unlike a fixed-effects model, this can be used when statistical heterogeneity (I 2 ) is present in the results of the included studies. Where evidence of statistically significant heterogeneity is present, sensitivity analyses will be conducted where possible to verify the robustness of the study conclusions, assessing the impact of methodological quality, study design, sample size and the potential effects of missing data. We will use funnel plots to detect potential reporting biases and small-study effects where data are available from 10 or more studies. If the included studies are all qualitative or a combination of quantitative and qualitative, there are several approaches that could be taken for data synthesis. Some of the most commonly used methods to synthesise qualitative health research include thematic analysis, grounded theory and meta-ethnography. However, there is no consensus on the best approach, which will depend on the type and number of included studies and the form and nature of the research question. As such, we will make a final decision on the most appropriate method after selecting and quality assessing the included articles, as recommended by the Cochrane Qualitative and Implementation Methods Group. We will first analyse and synthesise data related to the experience of patients, care providers and healthcare professionals separately before deciding whether it is appropriate to aggregate data between these groups. These data will likely include themes, concepts and categories of information. If data are relatively ‘thin’, then we will consider using thematic synthesis to undertake line-by-line coding and development of descriptive and analytic themes. If the included articles include sufficient ‘thick’ data (eg, details about the context and background of the studies and participants ), we will consider a more interpretative approach such as meta-ethnography. This method goes beyond aggregating data to generate new interpretations of the findings. As recommended by the Cochrane Qualitative and Implementation Methods Group, the GRADE‐CERQual (Confidence in the Evidence from Reviews of Qualitative research ) will be used to summarise our confidence in synthesised qualitative findings (eg, in the themes that we identify). The CERQual is made up of four key components, that is, methodological limitations of included studies, coherence of the review finding, adequacy of the data contributing to a review finding and relevance of the included studies to the review question. After assessing each of the four components, overall confidence will be graded as high, moderate, low or very low. The barriers and facilitators to symptom management will be divided into overarching themes for each group (patients, healthcare professionals, carers), and presented in a matrix along with our CERQual assessment of confidence in the evidence of each theme and an explanation of this assessment. The GRADE guidelines will be used to appraise the quality of any quantitative findings. The GRADE guidelines include four elements for which quantitative findings will be rated against: risk of bias (‘Study limitations’), inconsistency, indirectness, imprecision, publication bias.
Consultation with young people, parents and healthcare professionals has been used to determine their perception of the barriers and facilitators they experience when managing symptoms in ICYP at end-of-life. It is based on their perspectives that this systematic review was deemed to be timely and crucial to conduct to inform further research work. Moreover, patient and public involvement (PPI) is represented in the authorship (MJ) of this manuscript. MJ is the PPI representative at the UK National Institute for Health Research (NIHR) Pain and Palliative Care Clinical Studies Group-Children. MJ is a parent of four children, one of whom died from T-cell acute lymphoblastic leukaemia at the age of 12. She has supported many families of children with cancer including those receiving palliative and end-of-life care.
This systematic review will be the first to synthesise and report barriers and facilitators experienced by patients, carers and healthcare professionals when managing symptoms in ICYP at end-of-life. The dearth and heterogeneity of the included studies, which may use qualitative, quantitative or mixed-methods approaches, could limit the overall data synthesis we are able to conduct. As we expect there to be a lack of suitable studies, they will not be excluded on the basis of quality, which may limit the confidence in our findings. The review findings will be used to inform our ongoing work to develop a structured educational tool to support carers and healthcare professionals to administer pain and symptom relief to ICYP at the end-of-life.
As this is a systematic review of published literature, ethical approval will not be sought. We will publish the protocol and our findings in peer-reviewed journals aimed at paediatric palliative care clinicians and researchers as well as health commissioners. We will present our work at the growing numbers of national and international meetings focused on paediatric palliative care and pain.
Reviewer comments Author's manuscript
|
Patterns of follow‐up care in adult blood cancer survivors—Prospective evaluation of health‐related outcomes, resource use, and quality of life | f2f394fc-b03e-4237-951b-f40dc6a47a1c | 10979186 | Internal Medicine[mh] | INTRODUCTION Due to improvements in diagnosis and treatment, the number of long‐term cancer survivors is continuously growing. , Because of the risk of relapse and late effects, follow‐up care is an important component of long‐term support. How best to provide this support, is a matter of debate. There are different models of follow‐up care delivery. , , Follow‐up can be provided by follow‐up clinics, medical specialists, general practitioners, and specialized nurses, alone or in combination. The most common combined model is the parallel type where follow‐up care is provided by oncologists and other health care by general practitioners. , In the sequential model, often realized in pediatric oncology, cancer survivors are formally transferred from oncologists to primary care providers. , Another, more complex delivery type is the shared‐care model where oncologists and general practitioners have complementary roles in follow‐up care. , , For blood cancer, the fourth most common cancer, , little is known about follow‐up care. Blood cancer differs from solid tumors by its disseminated nature, an unusually large number of subtypes, , and a leading role of drug‐ and cell‐based therapies. , Surgery, the mainstay of treatment for curable solid tumor stages, has no role in the therapeutic armamentarium of blood cancer. Whether early recognition of relapse is associated with improved prognosis, remains uncertain for most subtypes. Given the favorable prognosis of many types of blood cancer, there is a wealth of information about long‐term treatment side effects, , secondary diseases, , , , and quality of life. , , , How and by whom follow‐up care is delivered, however, remains largely unexplored. Expert opinion‐based recommendations from national and international guidelines remain vague in this respect. , , , Studies comparing different follow‐up models have not yet been performed for blood cancer, and comprehensive information about blood cancer follow‐up practice patterns in Germany is not available. To gather information about the follow‐up care received by blood cancer survivors from the University Hospital of Essen, the oldest and one of the largest comprehensive cancer centers in Germany, and to compare patterns of care, we performed a questionnaire‐based observational study consisting of two parts. The retrospective part of the “ A ftercare in B lood C ancer Survivors” (ABC) study aimed at identifying follow‐up institutions. It was based on information provided by 1551 blood cancer survivors with a median follow‐up time of almost 10 years. The most important result of this part of the study was the identification of three distinct groups of care providers: academic oncologists working at the university hospital, community oncologists working outside the university hospital, and non‐oncological internists or general practitioners. In the German health care system, academic and community oncologists receive the same training. Depending on the time period when the qualification was obtained, oncologists in training spend 6 (before 2006) or 5 years (after 2006) in various fields of internal medicine with an additional 2 years in hematology and medical oncology. Major differences between academic and community oncologists include specific diagnostic and therapeutic procedures and access to specialized inpatient care and the expertise of other disciplines that are more readily available at academic institutions. Although the training of nonspecialized internists and general practitioners differs (6 or 5 years in internal medicine vs. 3 [before 2006] or 5 years [after 2006] in various fields of medicine including internal medicine), they are both entitled to provide primary care. For the purpose of this study, non‐oncological internists and general practitioners were therefore combined in one group (subsequently referred to as primary care physicians). The prospective part of the ABC study, which is the subject of this publication, was based on information provided by the follow‐up physicians of the patients surveyed in the retrospective part. Its main goals were to characterize the three provider types and their patients, specify their information sources, explore their expectations of follow‐up care, and compare their performance with respect to health‐related outcomes (e.g., relapse detection, secondary disease prevention and detection), resource use (e.g., laboratory tests, imaging), and their patients' quality of life. METHODS 2.1 Eligibility The patient eligibility criteria have been previously described. In brief, patients ≥18 years who had been diagnosed with and/or treated for a hematological malignancy at the University Hospital of Essen were eligible, if the interval between study inclusion and date of diagnosis (for untreated patients) or end of last treatment (for primary disease or relapse) was ≥3 years. In patients receiving continuous oral medication or low dose maintenance therapy after intensive induction, eligibility started 3 years after treatment initiation or end of induction, respectively. The 3‐year starting point of the study was chosen because relapse tends to occur early in hematological malignancies, mainly within the first 2 or 3 years. , , Because of its poor prognosis most patients with early relapse are not candidates for long‐term follow‐up care. Conditions included monoclonal gammopathy of undetermined significance (MGUS), multiple myeloma (MM), indolent non‐Hodgkin lymphoma including chronic lymphocytic leukemia (iNHL/CLL), myelodysplastic syndrome (MDS), myeloproliferative neoplasms including chronic myeloid leukemia (MPN/CML), aggressive non‐Hodgkin or Hodgkin's lymphoma (aNHL/HL), and acute myeloid or acute lymphoblastic leukemia (AML/ALL). Irrespective of the underlying disease, patients with a history of allogeneic transplantation were allocated to a separate group (AlloTx), because health issues arising ≥3 years after transplantation are more likely to be related to the procedure than to the disease. Patients were categorized according to the above‐mentioned disease groups and the three follow‐up institutions identified in the retrospective part of the study. 2.2 Study design The observational study was approved by the ethics committee of the University of Duisburg‐Essen (February 17, 2014; no. 14‐5692‐BO). All participating patients and physicians gave written informed consent. In the retrospective part, the patients were asked to name the physicians providing follow‐up care and specify the date of their next scheduled follow‐up visit. Their quality of life was assessed using the EORTC QLQ C‐30 questionnaire (restricted to the broad domains “global health”, “functioning” [physical, role, cognitive, emotional, and social combined] and “symptoms”) and the Hospital Anxiety and Depression Scale (HADS) questionnaires. The quality‐of‐life assessment in the prospective part was independent of the assessment in the retrospective part. The physician questionnaire was developed in three steps by the authors. The first draft was designed by JB, modified by UD, and then discussed among all authors. As a result, the questionnaire was split into three individual documents fulfilling different goals. In the final step, the questionnaires were completed by independent physicians of the Department of Hematology which led to minor changes in wording. The goals of the three documents were to (1) compare the characteristics of participating and nonparticipating physicians (questionnaire 1 [participation form] to be completed by both participating and nonparticipating physicians), (2) compare the attitudes of participating physicians from the three types of follow‐up institutions towards follow‐up care for a particular patient (questionnaire 2 [general aspects], only to be completed by participating physicians), and (3) compare the measures taken during follow‐up visits (questionnaire 3 [visit‐specific], to be completed by participating physicians, if their patient's visit fell within the 18‐month study period). If a patient had more than one follow‐up visit during the study period, questionnaire 3 was used repeatedly. The original questionnaires are provided in the Supporting Information. In brief, the questionnaires addressed the following questions: The 2‐page participation form included questions concerning the physicians' age, the year that they obtained their medical license, their medical specialty, current professional position, and follow‐up care guidelines used. The 3‐page general aspects questionnaire covered the exact blood‐cancer diagnosis of a particular patient as known by the follow‐up physician, comorbidities, frequency of follow‐up visits, and the physician‐perceived importance of follow‐up care for this patient. The 5‐page visit‐specific questionnaire ensured the prospective documentation of follow‐up visits. Events to be documented included disease detection (relapse, second primary malignancy, any other new disease possibly related to the hematological malignancy or its treatment), counseling for disease prevention (cancer and cardiovascular screening programs, vaccinations), other consultation topics, physical examination, laboratory investigations, imaging, and organ function tests. Laboratory tests were divided into basic parameters (blood count including differential; basic plasma coagulation tests; serum lactate dehydrogenase, electrolytes, kidney and liver function parameters, total protein; urinalysis) and extensive investigations (lymphocyte subpopulations; serum protein electrophoresis, immunoglobulins, hormones, vitamins, tumor markers, iron metabolism; molecular analyses). The physicians were also asked to specify the date of the next visit which was documented in the same way if it fell within the 18‐month study period. The physicians were contacted by mail and asked to participate in the study and complete the questionnaires. If they did not respond within 4–6 weeks, they were contacted by mail again, and if they also failed to respond to the second letter, they were contacted by phone. Each questionnaire was rewarded with 15 €. Monetary incentives have previously been shown to increase physician response rates in medical surveys. 2.3 Statistical analysis Frequencies are presented as numbers and compared using the chi test. Unless otherwise stated, percentages refer to the total number of patients, that is, they are not corrected for missing data. Continuous data are presented as median, first and third quartile (interquartile range, IQR), compared using the Kruskal‐Wallis test, and graphically displayed as box‐whisker plots, diamonds representing means. All analyses were exploratory, assuming statistical significance at p ≤ 0.05. The quality‐of‐life scales were normalized to attain maximum power. Details of the procedure have been described before. The statistical analyses were performed using SAS (SAS version 9.4, SAS Institute Inc., Cary, NC, USA). Eligibility The patient eligibility criteria have been previously described. In brief, patients ≥18 years who had been diagnosed with and/or treated for a hematological malignancy at the University Hospital of Essen were eligible, if the interval between study inclusion and date of diagnosis (for untreated patients) or end of last treatment (for primary disease or relapse) was ≥3 years. In patients receiving continuous oral medication or low dose maintenance therapy after intensive induction, eligibility started 3 years after treatment initiation or end of induction, respectively. The 3‐year starting point of the study was chosen because relapse tends to occur early in hematological malignancies, mainly within the first 2 or 3 years. , , Because of its poor prognosis most patients with early relapse are not candidates for long‐term follow‐up care. Conditions included monoclonal gammopathy of undetermined significance (MGUS), multiple myeloma (MM), indolent non‐Hodgkin lymphoma including chronic lymphocytic leukemia (iNHL/CLL), myelodysplastic syndrome (MDS), myeloproliferative neoplasms including chronic myeloid leukemia (MPN/CML), aggressive non‐Hodgkin or Hodgkin's lymphoma (aNHL/HL), and acute myeloid or acute lymphoblastic leukemia (AML/ALL). Irrespective of the underlying disease, patients with a history of allogeneic transplantation were allocated to a separate group (AlloTx), because health issues arising ≥3 years after transplantation are more likely to be related to the procedure than to the disease. Patients were categorized according to the above‐mentioned disease groups and the three follow‐up institutions identified in the retrospective part of the study. Study design The observational study was approved by the ethics committee of the University of Duisburg‐Essen (February 17, 2014; no. 14‐5692‐BO). All participating patients and physicians gave written informed consent. In the retrospective part, the patients were asked to name the physicians providing follow‐up care and specify the date of their next scheduled follow‐up visit. Their quality of life was assessed using the EORTC QLQ C‐30 questionnaire (restricted to the broad domains “global health”, “functioning” [physical, role, cognitive, emotional, and social combined] and “symptoms”) and the Hospital Anxiety and Depression Scale (HADS) questionnaires. The quality‐of‐life assessment in the prospective part was independent of the assessment in the retrospective part. The physician questionnaire was developed in three steps by the authors. The first draft was designed by JB, modified by UD, and then discussed among all authors. As a result, the questionnaire was split into three individual documents fulfilling different goals. In the final step, the questionnaires were completed by independent physicians of the Department of Hematology which led to minor changes in wording. The goals of the three documents were to (1) compare the characteristics of participating and nonparticipating physicians (questionnaire 1 [participation form] to be completed by both participating and nonparticipating physicians), (2) compare the attitudes of participating physicians from the three types of follow‐up institutions towards follow‐up care for a particular patient (questionnaire 2 [general aspects], only to be completed by participating physicians), and (3) compare the measures taken during follow‐up visits (questionnaire 3 [visit‐specific], to be completed by participating physicians, if their patient's visit fell within the 18‐month study period). If a patient had more than one follow‐up visit during the study period, questionnaire 3 was used repeatedly. The original questionnaires are provided in the Supporting Information. In brief, the questionnaires addressed the following questions: The 2‐page participation form included questions concerning the physicians' age, the year that they obtained their medical license, their medical specialty, current professional position, and follow‐up care guidelines used. The 3‐page general aspects questionnaire covered the exact blood‐cancer diagnosis of a particular patient as known by the follow‐up physician, comorbidities, frequency of follow‐up visits, and the physician‐perceived importance of follow‐up care for this patient. The 5‐page visit‐specific questionnaire ensured the prospective documentation of follow‐up visits. Events to be documented included disease detection (relapse, second primary malignancy, any other new disease possibly related to the hematological malignancy or its treatment), counseling for disease prevention (cancer and cardiovascular screening programs, vaccinations), other consultation topics, physical examination, laboratory investigations, imaging, and organ function tests. Laboratory tests were divided into basic parameters (blood count including differential; basic plasma coagulation tests; serum lactate dehydrogenase, electrolytes, kidney and liver function parameters, total protein; urinalysis) and extensive investigations (lymphocyte subpopulations; serum protein electrophoresis, immunoglobulins, hormones, vitamins, tumor markers, iron metabolism; molecular analyses). The physicians were also asked to specify the date of the next visit which was documented in the same way if it fell within the 18‐month study period. The physicians were contacted by mail and asked to participate in the study and complete the questionnaires. If they did not respond within 4–6 weeks, they were contacted by mail again, and if they also failed to respond to the second letter, they were contacted by phone. Each questionnaire was rewarded with 15 €. Monetary incentives have previously been shown to increase physician response rates in medical surveys. Statistical analysis Frequencies are presented as numbers and compared using the chi test. Unless otherwise stated, percentages refer to the total number of patients, that is, they are not corrected for missing data. Continuous data are presented as median, first and third quartile (interquartile range, IQR), compared using the Kruskal‐Wallis test, and graphically displayed as box‐whisker plots, diamonds representing means. All analyses were exploratory, assuming statistical significance at p ≤ 0.05. The quality‐of‐life scales were normalized to attain maximum power. Details of the procedure have been described before. The statistical analyses were performed using SAS (SAS version 9.4, SAS Institute Inc., Cary, NC, USA). RESULTS 3.1 Patients Of 2386 patients meeting the inclusion criteria, 1551 (65.0%) consented to participate in the study (Table ). The characteristics of participating and nonparticipating patients have been reported elsewhere. 3.2 Characteristics of follow‐up physicians (questionnaire 1) The patients named a total of 1070 physicians involved in follow‐up care. Seventy‐two patients named no follow‐up physician (abstention from follow‐up care), 729 named one, 706 named two, and 44 named three physicians. Of these, 223 were hematologists and medical oncologists (which is a single medical specialty in Germany; subsequently referred to as “oncologists”), 366 were internists (without specialization or specialized in fields other than hematology‐oncology), 386 were general practitioners, and 95 were specialists in other disciplines. Eighty‐eight physicians worked at the university hospital, 61 at other hospitals, 498 in individual private practice, and 423 in group private practice (Table ). Four hundred and seventy‐eight physicians consented to participate (44.7%). Reasons for nonparticipation included retirement, disregard of three invitations to participate, and refusal (Figure ). Physicians consenting or refusing to participate were similar with regard to sex (male, 72.6 vs. 73.4%), age (median 61 [range, 30–74] vs. 56 [42–68] years), and work experience (20 [1–48] vs. 19 [3–36] years). Consent was highest among oncologists (119/223, 53.4%), intermediate among general practitioners (181/386, 46.9%) and nononcological internists (162/366, 44.3%), and lowest among specialists in other disciplines (16/95, 16.8%; p < 0.0001). Physicians in group private practice (244/423, 57.7%) were more likely to participate than physicians in other work environments (individual private practice, 171/498, 34.3%; university hospital, 40/88, 45.5%; other hospital, 23/61, 37.7%; p < 0.0001) (Table ). Participation of physicians outside the university hospital was correlated with the number of follow‐up patients that they cared for (median, 1; range, 1–11), increasing from 31.6% for physicians caring for a single patient to 100% for those caring for ≥6 patients. 3.3 Follow‐up guidelines used by physicians (questionnaire 1) Oncologists primarily relied on the concise “Onkopedia” guidelines of the German Society of Hematology and Medical Oncology (74.0%), closely followed by the more comprehensive guidelines of the Association of the Scientific Medical Societies in Germany (61.3%). Some oncologists (12.6%) also used international guidelines (Table S ). Non‐oncological internists (66.7%) and general practitioners (65.2%) primarily relied on information received from physicians previously caring for their patients. Reliance on knowledge acquired during postgraduate medical training was more common among oncologists (58.0%) and internists (51.9%) than among general practitioners (24.9%). The differences between the medical disciplines were statistically highly significant (Table S ; p < 0.0001 for all comparisons). 3.4 Physician‐perceived importance of follow‐up care (questionnaire 2) A major result of the ABC study's retrospective part was the identification of three follow‐up care provider groups : academic oncologists from the university hospital (40 of whom participated in the prospective part); community oncologists (84 participants) working in private practice (66 participants) or outpatient clinics (18 participants); and general practitioners (180 participants), non‐oncological internists (158 participants), and other physicians without specialization in oncology (16 participants) in private practice. Because of overlapping roles in the German health care system, the last‐named disciplines were combined in a single group (collectively referred to as “primary care physicians”). The participating physicians returned 1387 questionnaires addressing general aspects of follow‐up care for 1129 of 1479 participating follow‐up patients (76.3%; academic oncologists, 857 questionnaires; community oncologists, 162; primary care physicians, 368). Eight hundred and seventy one patients were covered by a single questionnaire and 258 were covered by two separate questionnaires from different physicians. All groups of physicians agreed that relapse detection was the most important goal of follow‐up care, followed by second primary malignancies, cardiovascular diseases, and infection. Polyneuropathy, psychosocial and fertility issues received lower scores (Table ). As a rule, primary care physicians attributed higher importance to follow‐up care than oncologists did. This was true for all domains investigated (Table ; p < 0.0001 for most comparisons) and confirmed when the analysis was restricted to patients who were covered by both an oncologist and a primary care physician (Figure S ; p < 0.0001 for all comparisons). The physicians were also asked to rate the importance of preventive care. Counseling for cancer screening, cardiovascular risk factors, and vaccination was considered more important by primary care physicians than by oncologists (Table ; p < 0.0001 for all comparisons). 3.5 Prospective documentation of follow‐up visits (questionnaire 3) Based on the patients' ranking, each patient was allocated to a single institution predominantly responsible for follow‐up care (Table ). The major follow‐up institutions returned 1679 visit‐specific questionnaires, covering 715 of 1479 follow‐up patients (48.3%) from all disease groups except MDS (Table ). Coverage was highest at the university hospital (574 of 1045 follow‐up patients, 54.9%), followed by community oncologists (90/231, 39.0%) and primary care physicians (51/203, 25.1%). The disease spectrum differed at the three institutions ( p < 0.0001). The largest groups in relation to total group size were AlloTx and iNHL/CLL for academic oncologists, MGUS, MM, and MPN/CML for community oncologists, and AML/ALL and aNHL/HL for primary care physicians (Table ). The average number of documented visits was 2 (range, 1–6). Items listed in the questionnaire were regarded as addressed, if they were documented at least once during the study period. 3.5.1 Relapse, second primary malignancy, and other diseases During the 18‐month study period, the physicians reported 58 blood cancer relapses (8.1% of 715 patients), 25 second primary malignancies (3.5%), 22 other noninfectious new diseases (3.1%), 37 acute infections (5.2%), and 47 cancer‐ or cancer‐therapy‐related chronic diseases (6.6%). Except for acute infections, there were no significant differences in the disease detection rates among the three types of follow‐up institutions (Table ). The diseases reported by the three institutions in the seven blood‐cancer disease groups are detailed in Table S . Relapse or progression was most frequent in MM (5 of 18 patients, 27.8%), iNHL/CLL (31/123, 25.2%), and MGUS (1/4 [progression to lymphoplasmacytic lymphoma], 25.0%) and less frequent in AML/ALL (2/29, 6.9%), MPN/CML (3/55, 5.4%), aNHL/HL (6/145, 4.1%), and AlloTx (10/341, 2.9%; p < 0.0001). While academic oncologists and community oncologists reported relapses in each of the blood‐cancer disease groups documented, primary care physicians recorded a relapse only in iNHL/CLL (Table S ). Second primary malignancies were most frequent in iNHL/CLL (7/123, 5.7%) and AlloTx (15/341, 4.4%), rare in MPN/CML (1/55, 1.8%) and aNHL/HL (2/145, 1.4%), and not observed in MGUS, MM, and AML/ALL ( p = 0.3386). They included one squamous and 13 basal cell carcinomas of the skin (iNHL/CLL, 5; MPN/CML, 1; AlloTx, 8), one anal, one esophageal and two oral squamous cell carcinomas (all AlloTx), two pancreatic cancers, and one case each of follicular lymphoma, diffuse large B‐cell lymphoma, breast cancer, liver cancer, and cancer of unknown primary. Academic oncologists observed second primary malignancies in four blood‐cancer disease groups (iNHL/CLL, MPN/CML, aNHL/HL, AlloTx), primary care physicians in three (iNHL/CLL, aNHL/HL, AlloTx), and community oncologists in one (AlloTx) (Table S ). Other noninfectious new diseases were most frequently recorded in MPN/CML (3/55, 5.5%), followed by AlloTx (13/341, 3.8%), iNHL/CLL (3/123, 2.4%), and aNHL/HL (3/145, 2.1%; no cases in MGUS, MM, and AML/ALL; p = 0.6846). Chronic blood cancer‐ or treatment‐related diseases predominated in Allo‐Tx (39/341, 11.4%), followed by iNHL/CLL (5/123, 4.1%), MPN/CML (1/55, 1.8%), and aNHL/HL (2/145, 1.4%; no cases in MGUS, MM, and AML/ALL; p = 0.0002). Acute or chronic complications restricted to AlloTx included graft‐versus‐host disease ( n = 32), femoral head necrosis ( n = 4), depression ( n = 2), and kidney failure ( n = 1). Cutaneous ulcers ( n = 2) were exclusively reported in MPN/CML. Diseases occurring in several blood cancer groups included cardiovascular complications ( n = 10), polyneuropathy ( n = 6), phlebothrombosis ( n = 4), osteoporosis ( n = 3), fatigue ( n = 3), and cataract ( n = 2). Although all three follow‐up institutions cared for a substantial number of AlloTx patients, chronic transplantation‐related sequelae, for example, graft‐versus‐host disease, were only reported by academic oncologists and community oncologists (Table S ). Acute infections were most often recorded in iNHL/CLL (10/123, 8.1%) and AlloTx (23/341, 6.7%), followed by MM (1/18, 5.6%), MPN/CML (2/55; 3.6%), and aNHL/HL (1/145, 0.7%; no cases in MGUS and AML/ALL; p = 0.0638) (Table S ). Acute infections were significantly more often reported by primary care physicians (11.8%) and academic oncologists (5.4%) than by community oncologists (0.0%; p = 0.0087) (Table ). 3.5.2 Counseling for disease prevention Counseling for disease prevention was more extensively done by primary care physicians than by academic oncologists or community oncologists (Table ). The differences were statistically significant for cancer screening (54.9% vs. 49.3% vs. 26.7%; p = 0.0002), cardiovascular risk factors (64.7% vs. 42.0% vs. 31.1%; p = 0.0005), and vaccination (68.6% vs. 44.9% vs. 15.6%; p < 0.0001). 3.5.3 Consultation topics and physical examination Primary care physicians significantly more often addressed questions related to psychosocial issues and sexuality than academic oncologists or community oncologists did (psychosocial issues, 31.4% vs. 16.7% vs. 20.0%, p = 0.0304; sexuality, 15.7% vs. 3.3% vs. 6.7%, p = 0.0002). The converse was true for infections (25.5% vs. 55.9% vs. 42.2%; p < 0.0001). Physical examination was more often performed by academic oncologists (84.1%) and primary care physicians (80.4%) than by community oncologists (62.2%; p < 0.0001) (Table ). 3.5.4 Laboratory investigations Laboratory tests were more frequently ordered by academic oncologists than by community oncologists or primary care physicians (basic tests, 89.4% vs. 68.9% vs. 78.4%, p < 0.0001; extensive tests, 63.8% vs. 38.9% vs. 52.9%, p < 0.0001) (Table ). Extensive tests were more often performed in AlloTx (269/341, 78.9%), MM (14/18, 77.8%), and iNHL/CLL (70/123, 56.9%) than in MPN/CML (18/55, 32.7%), aNHL/HL (43/145, 29.7%), AML/ALL (8/29, 27.6%), or MGUS (1/4, 25.0%; p < 0.0001). Their frequency remained constant over time (year 4–5, 107 of 182 patients, 58.8%; years 6–10, 195/330, 58.8%; year >10, 127/203; 62.6%; p = 0.6772). 3.5.5 Imaging and organ function tests Imaging was more often ordered by primary care physicians than by academic oncologists or community oncologists (90.2 vs. 66.7 vs. 52.4 investigations per 100 patients; p < 0.0001). The same was true for electrocardiography and echocardiography (Table ). Other organ function tests were rarely performed (data not shown). 3.6 Quality of life of patients visiting different follow‐up institutions The quality‐of‐life assessment included 1348 patients on follow‐up care and 59 patients not undergoing follow‐up care. While quality of life was significantly better in patients forgoing follow‐up care than in patients utilizing it, there were no statistically significant differences between the three follow‐up institutions (Figure ). Patients Of 2386 patients meeting the inclusion criteria, 1551 (65.0%) consented to participate in the study (Table ). The characteristics of participating and nonparticipating patients have been reported elsewhere. Characteristics of follow‐up physicians (questionnaire 1) The patients named a total of 1070 physicians involved in follow‐up care. Seventy‐two patients named no follow‐up physician (abstention from follow‐up care), 729 named one, 706 named two, and 44 named three physicians. Of these, 223 were hematologists and medical oncologists (which is a single medical specialty in Germany; subsequently referred to as “oncologists”), 366 were internists (without specialization or specialized in fields other than hematology‐oncology), 386 were general practitioners, and 95 were specialists in other disciplines. Eighty‐eight physicians worked at the university hospital, 61 at other hospitals, 498 in individual private practice, and 423 in group private practice (Table ). Four hundred and seventy‐eight physicians consented to participate (44.7%). Reasons for nonparticipation included retirement, disregard of three invitations to participate, and refusal (Figure ). Physicians consenting or refusing to participate were similar with regard to sex (male, 72.6 vs. 73.4%), age (median 61 [range, 30–74] vs. 56 [42–68] years), and work experience (20 [1–48] vs. 19 [3–36] years). Consent was highest among oncologists (119/223, 53.4%), intermediate among general practitioners (181/386, 46.9%) and nononcological internists (162/366, 44.3%), and lowest among specialists in other disciplines (16/95, 16.8%; p < 0.0001). Physicians in group private practice (244/423, 57.7%) were more likely to participate than physicians in other work environments (individual private practice, 171/498, 34.3%; university hospital, 40/88, 45.5%; other hospital, 23/61, 37.7%; p < 0.0001) (Table ). Participation of physicians outside the university hospital was correlated with the number of follow‐up patients that they cared for (median, 1; range, 1–11), increasing from 31.6% for physicians caring for a single patient to 100% for those caring for ≥6 patients. Follow‐up guidelines used by physicians (questionnaire 1) Oncologists primarily relied on the concise “Onkopedia” guidelines of the German Society of Hematology and Medical Oncology (74.0%), closely followed by the more comprehensive guidelines of the Association of the Scientific Medical Societies in Germany (61.3%). Some oncologists (12.6%) also used international guidelines (Table S ). Non‐oncological internists (66.7%) and general practitioners (65.2%) primarily relied on information received from physicians previously caring for their patients. Reliance on knowledge acquired during postgraduate medical training was more common among oncologists (58.0%) and internists (51.9%) than among general practitioners (24.9%). The differences between the medical disciplines were statistically highly significant (Table S ; p < 0.0001 for all comparisons). Physician‐perceived importance of follow‐up care (questionnaire 2) A major result of the ABC study's retrospective part was the identification of three follow‐up care provider groups : academic oncologists from the university hospital (40 of whom participated in the prospective part); community oncologists (84 participants) working in private practice (66 participants) or outpatient clinics (18 participants); and general practitioners (180 participants), non‐oncological internists (158 participants), and other physicians without specialization in oncology (16 participants) in private practice. Because of overlapping roles in the German health care system, the last‐named disciplines were combined in a single group (collectively referred to as “primary care physicians”). The participating physicians returned 1387 questionnaires addressing general aspects of follow‐up care for 1129 of 1479 participating follow‐up patients (76.3%; academic oncologists, 857 questionnaires; community oncologists, 162; primary care physicians, 368). Eight hundred and seventy one patients were covered by a single questionnaire and 258 were covered by two separate questionnaires from different physicians. All groups of physicians agreed that relapse detection was the most important goal of follow‐up care, followed by second primary malignancies, cardiovascular diseases, and infection. Polyneuropathy, psychosocial and fertility issues received lower scores (Table ). As a rule, primary care physicians attributed higher importance to follow‐up care than oncologists did. This was true for all domains investigated (Table ; p < 0.0001 for most comparisons) and confirmed when the analysis was restricted to patients who were covered by both an oncologist and a primary care physician (Figure S ; p < 0.0001 for all comparisons). The physicians were also asked to rate the importance of preventive care. Counseling for cancer screening, cardiovascular risk factors, and vaccination was considered more important by primary care physicians than by oncologists (Table ; p < 0.0001 for all comparisons). Prospective documentation of follow‐up visits (questionnaire 3) Based on the patients' ranking, each patient was allocated to a single institution predominantly responsible for follow‐up care (Table ). The major follow‐up institutions returned 1679 visit‐specific questionnaires, covering 715 of 1479 follow‐up patients (48.3%) from all disease groups except MDS (Table ). Coverage was highest at the university hospital (574 of 1045 follow‐up patients, 54.9%), followed by community oncologists (90/231, 39.0%) and primary care physicians (51/203, 25.1%). The disease spectrum differed at the three institutions ( p < 0.0001). The largest groups in relation to total group size were AlloTx and iNHL/CLL for academic oncologists, MGUS, MM, and MPN/CML for community oncologists, and AML/ALL and aNHL/HL for primary care physicians (Table ). The average number of documented visits was 2 (range, 1–6). Items listed in the questionnaire were regarded as addressed, if they were documented at least once during the study period. 3.5.1 Relapse, second primary malignancy, and other diseases During the 18‐month study period, the physicians reported 58 blood cancer relapses (8.1% of 715 patients), 25 second primary malignancies (3.5%), 22 other noninfectious new diseases (3.1%), 37 acute infections (5.2%), and 47 cancer‐ or cancer‐therapy‐related chronic diseases (6.6%). Except for acute infections, there were no significant differences in the disease detection rates among the three types of follow‐up institutions (Table ). The diseases reported by the three institutions in the seven blood‐cancer disease groups are detailed in Table S . Relapse or progression was most frequent in MM (5 of 18 patients, 27.8%), iNHL/CLL (31/123, 25.2%), and MGUS (1/4 [progression to lymphoplasmacytic lymphoma], 25.0%) and less frequent in AML/ALL (2/29, 6.9%), MPN/CML (3/55, 5.4%), aNHL/HL (6/145, 4.1%), and AlloTx (10/341, 2.9%; p < 0.0001). While academic oncologists and community oncologists reported relapses in each of the blood‐cancer disease groups documented, primary care physicians recorded a relapse only in iNHL/CLL (Table S ). Second primary malignancies were most frequent in iNHL/CLL (7/123, 5.7%) and AlloTx (15/341, 4.4%), rare in MPN/CML (1/55, 1.8%) and aNHL/HL (2/145, 1.4%), and not observed in MGUS, MM, and AML/ALL ( p = 0.3386). They included one squamous and 13 basal cell carcinomas of the skin (iNHL/CLL, 5; MPN/CML, 1; AlloTx, 8), one anal, one esophageal and two oral squamous cell carcinomas (all AlloTx), two pancreatic cancers, and one case each of follicular lymphoma, diffuse large B‐cell lymphoma, breast cancer, liver cancer, and cancer of unknown primary. Academic oncologists observed second primary malignancies in four blood‐cancer disease groups (iNHL/CLL, MPN/CML, aNHL/HL, AlloTx), primary care physicians in three (iNHL/CLL, aNHL/HL, AlloTx), and community oncologists in one (AlloTx) (Table S ). Other noninfectious new diseases were most frequently recorded in MPN/CML (3/55, 5.5%), followed by AlloTx (13/341, 3.8%), iNHL/CLL (3/123, 2.4%), and aNHL/HL (3/145, 2.1%; no cases in MGUS, MM, and AML/ALL; p = 0.6846). Chronic blood cancer‐ or treatment‐related diseases predominated in Allo‐Tx (39/341, 11.4%), followed by iNHL/CLL (5/123, 4.1%), MPN/CML (1/55, 1.8%), and aNHL/HL (2/145, 1.4%; no cases in MGUS, MM, and AML/ALL; p = 0.0002). Acute or chronic complications restricted to AlloTx included graft‐versus‐host disease ( n = 32), femoral head necrosis ( n = 4), depression ( n = 2), and kidney failure ( n = 1). Cutaneous ulcers ( n = 2) were exclusively reported in MPN/CML. Diseases occurring in several blood cancer groups included cardiovascular complications ( n = 10), polyneuropathy ( n = 6), phlebothrombosis ( n = 4), osteoporosis ( n = 3), fatigue ( n = 3), and cataract ( n = 2). Although all three follow‐up institutions cared for a substantial number of AlloTx patients, chronic transplantation‐related sequelae, for example, graft‐versus‐host disease, were only reported by academic oncologists and community oncologists (Table S ). Acute infections were most often recorded in iNHL/CLL (10/123, 8.1%) and AlloTx (23/341, 6.7%), followed by MM (1/18, 5.6%), MPN/CML (2/55; 3.6%), and aNHL/HL (1/145, 0.7%; no cases in MGUS and AML/ALL; p = 0.0638) (Table S ). Acute infections were significantly more often reported by primary care physicians (11.8%) and academic oncologists (5.4%) than by community oncologists (0.0%; p = 0.0087) (Table ). 3.5.2 Counseling for disease prevention Counseling for disease prevention was more extensively done by primary care physicians than by academic oncologists or community oncologists (Table ). The differences were statistically significant for cancer screening (54.9% vs. 49.3% vs. 26.7%; p = 0.0002), cardiovascular risk factors (64.7% vs. 42.0% vs. 31.1%; p = 0.0005), and vaccination (68.6% vs. 44.9% vs. 15.6%; p < 0.0001). 3.5.3 Consultation topics and physical examination Primary care physicians significantly more often addressed questions related to psychosocial issues and sexuality than academic oncologists or community oncologists did (psychosocial issues, 31.4% vs. 16.7% vs. 20.0%, p = 0.0304; sexuality, 15.7% vs. 3.3% vs. 6.7%, p = 0.0002). The converse was true for infections (25.5% vs. 55.9% vs. 42.2%; p < 0.0001). Physical examination was more often performed by academic oncologists (84.1%) and primary care physicians (80.4%) than by community oncologists (62.2%; p < 0.0001) (Table ). 3.5.4 Laboratory investigations Laboratory tests were more frequently ordered by academic oncologists than by community oncologists or primary care physicians (basic tests, 89.4% vs. 68.9% vs. 78.4%, p < 0.0001; extensive tests, 63.8% vs. 38.9% vs. 52.9%, p < 0.0001) (Table ). Extensive tests were more often performed in AlloTx (269/341, 78.9%), MM (14/18, 77.8%), and iNHL/CLL (70/123, 56.9%) than in MPN/CML (18/55, 32.7%), aNHL/HL (43/145, 29.7%), AML/ALL (8/29, 27.6%), or MGUS (1/4, 25.0%; p < 0.0001). Their frequency remained constant over time (year 4–5, 107 of 182 patients, 58.8%; years 6–10, 195/330, 58.8%; year >10, 127/203; 62.6%; p = 0.6772). 3.5.5 Imaging and organ function tests Imaging was more often ordered by primary care physicians than by academic oncologists or community oncologists (90.2 vs. 66.7 vs. 52.4 investigations per 100 patients; p < 0.0001). The same was true for electrocardiography and echocardiography (Table ). Other organ function tests were rarely performed (data not shown). Relapse, second primary malignancy, and other diseases During the 18‐month study period, the physicians reported 58 blood cancer relapses (8.1% of 715 patients), 25 second primary malignancies (3.5%), 22 other noninfectious new diseases (3.1%), 37 acute infections (5.2%), and 47 cancer‐ or cancer‐therapy‐related chronic diseases (6.6%). Except for acute infections, there were no significant differences in the disease detection rates among the three types of follow‐up institutions (Table ). The diseases reported by the three institutions in the seven blood‐cancer disease groups are detailed in Table S . Relapse or progression was most frequent in MM (5 of 18 patients, 27.8%), iNHL/CLL (31/123, 25.2%), and MGUS (1/4 [progression to lymphoplasmacytic lymphoma], 25.0%) and less frequent in AML/ALL (2/29, 6.9%), MPN/CML (3/55, 5.4%), aNHL/HL (6/145, 4.1%), and AlloTx (10/341, 2.9%; p < 0.0001). While academic oncologists and community oncologists reported relapses in each of the blood‐cancer disease groups documented, primary care physicians recorded a relapse only in iNHL/CLL (Table S ). Second primary malignancies were most frequent in iNHL/CLL (7/123, 5.7%) and AlloTx (15/341, 4.4%), rare in MPN/CML (1/55, 1.8%) and aNHL/HL (2/145, 1.4%), and not observed in MGUS, MM, and AML/ALL ( p = 0.3386). They included one squamous and 13 basal cell carcinomas of the skin (iNHL/CLL, 5; MPN/CML, 1; AlloTx, 8), one anal, one esophageal and two oral squamous cell carcinomas (all AlloTx), two pancreatic cancers, and one case each of follicular lymphoma, diffuse large B‐cell lymphoma, breast cancer, liver cancer, and cancer of unknown primary. Academic oncologists observed second primary malignancies in four blood‐cancer disease groups (iNHL/CLL, MPN/CML, aNHL/HL, AlloTx), primary care physicians in three (iNHL/CLL, aNHL/HL, AlloTx), and community oncologists in one (AlloTx) (Table S ). Other noninfectious new diseases were most frequently recorded in MPN/CML (3/55, 5.5%), followed by AlloTx (13/341, 3.8%), iNHL/CLL (3/123, 2.4%), and aNHL/HL (3/145, 2.1%; no cases in MGUS, MM, and AML/ALL; p = 0.6846). Chronic blood cancer‐ or treatment‐related diseases predominated in Allo‐Tx (39/341, 11.4%), followed by iNHL/CLL (5/123, 4.1%), MPN/CML (1/55, 1.8%), and aNHL/HL (2/145, 1.4%; no cases in MGUS, MM, and AML/ALL; p = 0.0002). Acute or chronic complications restricted to AlloTx included graft‐versus‐host disease ( n = 32), femoral head necrosis ( n = 4), depression ( n = 2), and kidney failure ( n = 1). Cutaneous ulcers ( n = 2) were exclusively reported in MPN/CML. Diseases occurring in several blood cancer groups included cardiovascular complications ( n = 10), polyneuropathy ( n = 6), phlebothrombosis ( n = 4), osteoporosis ( n = 3), fatigue ( n = 3), and cataract ( n = 2). Although all three follow‐up institutions cared for a substantial number of AlloTx patients, chronic transplantation‐related sequelae, for example, graft‐versus‐host disease, were only reported by academic oncologists and community oncologists (Table S ). Acute infections were most often recorded in iNHL/CLL (10/123, 8.1%) and AlloTx (23/341, 6.7%), followed by MM (1/18, 5.6%), MPN/CML (2/55; 3.6%), and aNHL/HL (1/145, 0.7%; no cases in MGUS and AML/ALL; p = 0.0638) (Table S ). Acute infections were significantly more often reported by primary care physicians (11.8%) and academic oncologists (5.4%) than by community oncologists (0.0%; p = 0.0087) (Table ). Counseling for disease prevention Counseling for disease prevention was more extensively done by primary care physicians than by academic oncologists or community oncologists (Table ). The differences were statistically significant for cancer screening (54.9% vs. 49.3% vs. 26.7%; p = 0.0002), cardiovascular risk factors (64.7% vs. 42.0% vs. 31.1%; p = 0.0005), and vaccination (68.6% vs. 44.9% vs. 15.6%; p < 0.0001). Consultation topics and physical examination Primary care physicians significantly more often addressed questions related to psychosocial issues and sexuality than academic oncologists or community oncologists did (psychosocial issues, 31.4% vs. 16.7% vs. 20.0%, p = 0.0304; sexuality, 15.7% vs. 3.3% vs. 6.7%, p = 0.0002). The converse was true for infections (25.5% vs. 55.9% vs. 42.2%; p < 0.0001). Physical examination was more often performed by academic oncologists (84.1%) and primary care physicians (80.4%) than by community oncologists (62.2%; p < 0.0001) (Table ). Laboratory investigations Laboratory tests were more frequently ordered by academic oncologists than by community oncologists or primary care physicians (basic tests, 89.4% vs. 68.9% vs. 78.4%, p < 0.0001; extensive tests, 63.8% vs. 38.9% vs. 52.9%, p < 0.0001) (Table ). Extensive tests were more often performed in AlloTx (269/341, 78.9%), MM (14/18, 77.8%), and iNHL/CLL (70/123, 56.9%) than in MPN/CML (18/55, 32.7%), aNHL/HL (43/145, 29.7%), AML/ALL (8/29, 27.6%), or MGUS (1/4, 25.0%; p < 0.0001). Their frequency remained constant over time (year 4–5, 107 of 182 patients, 58.8%; years 6–10, 195/330, 58.8%; year >10, 127/203; 62.6%; p = 0.6772). Imaging and organ function tests Imaging was more often ordered by primary care physicians than by academic oncologists or community oncologists (90.2 vs. 66.7 vs. 52.4 investigations per 100 patients; p < 0.0001). The same was true for electrocardiography and echocardiography (Table ). Other organ function tests were rarely performed (data not shown). Quality of life of patients visiting different follow‐up institutions The quality‐of‐life assessment included 1348 patients on follow‐up care and 59 patients not undergoing follow‐up care. While quality of life was significantly better in patients forgoing follow‐up care than in patients utilizing it, there were no statistically significant differences between the three follow‐up institutions (Figure ). DISCUSSION The major results of the ABC study's prospective part are the following: First, less than half of follow‐up physicians consented to participate in the study. Second, most blood cancer survivors with a consenting follow‐up physician received care at the university hospital, with only a minority being cared for by community oncologists or primary care physicians. Third, the disease spectrum differed among follow‐up institutions. Fourth, although physicians of different disciplines used different follow‐up guidelines, they agreed on the goals of follow‐up care. Finally, relapse and secondary disease detection rates and the patients' reports about their quality of life were similar at all follow‐up institutions, but there were significant differences in other domains. Compared to other questionnaire‐based survivorship studies and despite a monetary incentive, the participation rate among follow‐up physicians appeared low (45% vs. 62%–76% in previous studies). , , This may have been related to the fact that the patients were asked to name one or more follow‐up physicians. In general, only the first‐named physician had a vital role in follow‐up care. Therefore, patient coverage by at least one physician was much higher (76%) than overall physician participation (45%). Unfortunately, considerable attrition occurred between the general aspects questionnaire (covering 76% of participating patients) and the visit‐specific questionnaire (covering only 48%), in particular in patients treated outside the university hospital. This may have been related to the size of the visit‐specific questionnaire. Survey length is negatively correlated with response rate. Overall, 80% of patients with fully‐documented follow‐up visits were treated by academic oncologists, 13% by community oncologists, and 7% by primary care physicians. The disease spectrum differed at the three types of follow‐up institutions. Patients at high risk of relapse (iNHL/CLL) or treatment‐related long‐term problems (AlloTx) were predominantly followed up at the university hospital (80–90% of all iNHL/CLL or AlloTx patients). By contrast, patients in stable condition with diseases requiring continuous monitoring with or without oral maintenance therapy (MGUS, MM, MPN/CML) were frequently seen by community oncologists (30–100% of such patients). Survivors seen by primary care physicians most often had a history of a curable disease and were in stable long‐term remission (10–30% of aNHL/HL and AML/ALL patients). The few iNHL/CLL or AlloTx patients followed up by primary care physicians appeared to be a particularly favorable subgroup, because chronic treatment‐related disease states were not reported and relapses were rare. The major information sources used for follow‐up care were national guidelines , for oncologists, and recommendations by other physicians for primary care providers. Ideally, such recommendations would be included in an individualized survivorship care plan accessible for all involved physicians. , , The implementation of survivorship care plans, however, has been difficult because of time constraints, responsibility and reimbursement issues, and a paucity of data demonstrating its positive impact on patient outcome. In Germany, communication between hospitals and private practices is mainly based on discharge letters that may or may not contain recommendations for follow‐up care. Both in Germany and other countries, , primary care physicians have complained about insufficient information from cancer hospitals. This may have contributed to the low number of primary care providers participating in the ABC study. Despite the use of different information sources, relapse detection, second primary malignancies, and cardiovascular diseases were consistently viewed as the most important goals of follow‐up care. With regard to these endpoints, the performance of the three institutions appeared similar. The resources used, however, differed. While laboratory investigations, crucial and recommended for all types of blood cancer, , were most extensively performed at the university hospital, imaging was most often ordered by primary care physicians. The role of imaging in blood cancer follow‐up care is limited, in particular in survivors with a history of leukemia and lymphoma , that were disproportionately often seen by primary care physicians. National and international guidelines explicitly discourage the use of routine surveillance scans in lymphoma. , , Primary care providers have previously been reported to have an incommensurate tendency to order imaging tests for cancer survivors. Explanations included defensive medicine, reimbursement incentives, and uncertainty about guidelines. Follow‐up care includes management of psychosocial consequences, promotion of a healthy life style, and disease prevention. , , , , These needs were better addressed by primary care physicians than by oncologists. Differences were particularly pronounced in counseling for disease prevention. Our findings are consistent with a report from the USA where cancer survivors expected their follow‐up physicians to provide preventive health care irrespective of medical qualification. While most general practitioners agreed to provide preventive care, only a minority of oncologists considered screening for other cancers their responsibility. Although the patients' needs appeared to be more comprehensively addressed by primary care physicians than by oncologists, one cannot conclude from our data that follow‐up care should be shifted to the former. First, the proportion of patients seen by primary care physicians was small, decreasing from 13.7% in the retrospective part of the study to 7.1% in the prospective part. Second, the disease spectrum differed in different follow‐up institutions. While survivors seen at the university hospital tended to be at high risk for relapse or late adverse events, most survivors seen by primary care physicians had a history of a curable disease and were in stable remission. Third, primary care providers participating in the ABC study were likely a selection of physicians with exceptional skills and motivations. In many countries, primary care physicians feel insufficiently trained, informed and equipped to provide comprehensive follow‐up care. , , , This is in line with our observation that, despite a multitude of primary care providers named as follow‐up physicians by their patients, only a minority played a dominant role. Those who did tended to rely on information received from others to provide adequate care. A recent report from Canada demonstrates that primary care providers can play a more important role in follow‐up care, if they receive adequate training. Whether this also applies to hematological diseases with a high risk of relapse and late complications, remains to be shown. Apart from the physician selection bias mentioned above, our study has several other limitations. First, the number of follow‐up visits documented by physicians outside the university hospital was lower than expected. Second, the disease spectrum was wide, some disease groups were small, the duration of follow‐up varied, and the features of survivors seen by different groups of care providers differed, making comparisons difficult. Third, the original study population was restricted to patients from a single comprehensive cancer center. While the results are likely to be representative for many medical institutions in Germany, they may not apply to countries with other health care systems. Finally, the observational nature of our study did not allow us to verify the data provided. Strengths of our study are its large size and the prospective capture of events during a reasonably long follow‐up period. Conclusions from the ABC study must be drawn with caution, because differences in the spectrum of hematological and secondary diseases among follow‐up institutions precluded an unbiased comparison. With this caveat, detection of relapse and secondary diseases was similar in all three follow‐up institutions. Psychosocial issues and preventive health care appeared to be better addressed by primary care physicians than by oncologists, while the converse was true for a judicious use of medical resources. For patients with curable diseases in stable remission, transfer of follow‐up care from oncologists to primary care providers seems feasible, provided the latter receive adequate information about the required procedures. Hildegard Lax: Data curation (supporting); formal analysis (supporting); methodology (supporting); visualization (supporting); writing – review and editing (supporting). Julia Baum: Data curation (lead); investigation (supporting); methodology (supporting); project administration (supporting); supervision (supporting); writing – review and editing (supporting). Nils Lehmann: Conceptualization (supporting); data curation (supporting); formal analysis (supporting); methodology (supporting); writing – review and editing (supporting). Anja Merkel‐Jens: Data curation (supporting); resources (supporting); writing – review and editing (supporting). Dietrich Beelen: Project administration (supporting); supervision (supporting); writing – review and editing (supporting). Karl‐Heinz Jöckel: Conceptualization (supporting); formal analysis (supporting); funding acquisition (supporting); methodology (supporting); project administration (supporting); supervision (supporting); writing – review and editing (supporting). Ulrich Dührsen: Conceptualization (lead); formal analysis (lead); funding acquisition (lead); investigation (lead); methodology (lead); project administration (lead); resources (lead); supervision (lead); visualization (lead); writing – original draft (lead); writing – review and editing (lead). The study was supported by the Federal Ministry of Education and Research of Germany (Bundesministerium für Bildung und Forschung, grant no. 01GY1341). The authors report no potential conflicts of interest relevant to this article. The study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of the University of Duisburg‐Essen (February 17, 2014; no. 14‐5692‐BO). Data S1. |
Broad application prospects of bone turnover markers in pediatrics | c005258c-b21b-412b-8009-e5205d579ce3 | 9459349 | Pediatrics[mh] | INTRODUCTION Osteoporosis is a prevalent condition in the middle‐aged and elderly population, due to insufficient bone mass during childhood and adolescence. The level of peak bone mass acquired in adolescence and the rate of bone loss in adulthood are two important factors of osteoporosis. Peak bone mass attained in adolescence determine the likelihood of pathological fractures and osteoporosis in adulthood. People who do not attain adequate bone mass during childhood and adolescence are at risk for osteoporosis, even if they do not have accelerated bone loss in adulthood. Thus, even if osteoporosis does not occur in childhood, the achievement of bone mass during adolescent development can influence skeletal development in adulthood. Dual‐energy X‐ray absorptiometry (DXA) and quantitative computed tomography (CT) using densitometry techniques are recognized as gold standards for assessing bone mineral contents and density for children. Normally, DXA and QCT had low radiation exposure, with the advantages of high precision and fast detection. However, there are some problems when using DXA and QCT to diagnosis for children, such as fussiness, irritability, and crying during their examination, resulting in poor reproducibility, and inability to reflect the real‐time biology of the bone at that time point. , Therefore, BMD measurements have limited the use of skeletal health assessment in childhood growth disorders for a long time. As they are minimally invasive and can be dynamically monitored, bone metabolic marker assays have been widely used in the field of childhood growth and development‐related diseases, and their levels not only reflect the skeletal metabolic health of children but indirectly reflect the growth and development of children. Thus, BTMs provide a basis for early identification, diagnosis, and treatment monitoring of childhood growth and development diseases. In this article, the characteristics of BTMs are discussed, including their classification, application in monitoring bone growth in children and adolescents. This review also discusses the relevant important studies and application of bone metabolism biomarkers in energy metabolism, the endocrine environment, osteoporosis, and childhood diseases.
BONE TURNOVER MARKERS 2.1 Specific BTMs provided by clinical laboratories Bone turnover markers ( BTMs) are biochemical or cellular compounds produced during the continuum of bone resorption or formation. BTMs can be divided into two categories: bone formation and bone resorption markers (Figure and Table ). The former represents the activity of osteoblasts and the state of bone formation, while the latter mainly reflects the activity of osteoclasts and the level of bone resorption. Procollagen type I N‐terminal propeptide (PINP), procollagen type I C‐terminal propeptide (PICP), and osteocalcin (OC) are three frequently used markers of bone formation. Pyridinoline (PYD), deoxypyridinoline (DPD), N‐terminal cross‐linked telopeptide (NTX), and C‐terminal cross‐linked telopeptide (CTX) are bone resorption markers. In addition, some important cytokines implicated in the regulation of bone turnover by controlling the activity of osteoblasts or osteoclasts, such as osteoprotegerin (OPG) and receptor activator of nuclear factor κB ligand (RANKL), and these can be considered regulators of bone turnover rather than classical biochemical markers of bone turnover. As BTMs are mostly excreted by the kidneys, their levels are frequently detected in urine. Due to the difficulty of collecting 24‐hour urine samples and the need to adjust for creatinine, blood testing is a preferable sample collection technique over urine sampling. Laboratory testing of BTMs is minimally invasive and relatively affordable and, if used and interpreted properly, it can be a useful tool for assessing metabolic bone disease, treatment outcomes, and patient compliance. , CTX and PINP are the commonly used bone turnover markers in clinical research and are recommended by the International Osteoporosis Foundation (IOF) and the International Federation of clinical laboratory medicine (IFCC). With the development of fully automated platforms, the analytical variability of bone markers has been greatly improved. The European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) has shown the biological variability (CV%) of some BTMs, such as ALP, PINP, CTX, and osteocalcin, and with small interlaboratory variation. In this review, we describe several of the most widely used BTMs in clinical pediatrics. 2.2 Procollagen type I N‐terminal propeptide (PINP) Type I collagen is a procollagen secreted by osteoblasts, which creates a triple helix consisting of PINP and PICP. Since these propeptides are cleaved in their extracellular area and released into the circulation as metabolites, PINP and PICP concentrations can represent the level of bone formation. Although the clearance of PINP may be less susceptible to hormonal changes than PICP, PINP is a more sensitive marker compared with PICP. Several studies have noted that PINP remains stable even after repeated freezing and thawing, and its levels are not affected by circadian rhythms, so it is not necessary to consider the timing of sampling. 2.3 Osteocalcin Osteocalcin is the most common noncollagenous protein in bone, and it is secreted primarily by mature osteoblasts, although it is also secreted during bone resorption. Thus, serum osteocalcin is an important indicator of bone turnover, indicating bone formation and bone resorption. , Because serum osteocalcin concentrations have a distinct circadian rhythm and are highest in the morning, blood samples must be taken in the morning to ensure the accuracy and comparability of the test results. In addition, several studies have shown that seasonal changes and diet did not affect PINP or osteocalcin levels. , , , 2.4 Bone alkaline phosphatase (BALP) Human ALP is classified as tissue non‐specific ALP (TNSALP), intestinal type, placental type, and placental‐like type. TNSALP activity is represented by the products of the ALPL gene, which indicates bone anabolic activity. ALPL gene mutation leads to abnormal skeletal mineralization. BALP is a common biochemical marker of bone formation and a specific marker for osteogenesis, as well as one type of TNSALP, along with liver‐ and kidney‐type ALP. The expression of bone ALP occurs early in the development of osteoblasts from mesenchymal progenitors and is vital in the degradation of pyrophosphate, a natural inhibitor of mineralization. It is hard to identify bone‐type from liver‐type ALP using current immunoassays, as they share the same amino acid sequences. Clinically, the application and interpretation of bone ALP as a BTM should be unaffected by liver diseases because bone ALP is normally cleared from the serum by liver. Children have higher bone ALP activity than adults due to higher bone formation rates. Determination of bone ALP activity also seems to be helpful for the diagnosis of hypophosphatasia and hyperphosphatasia. 2.5 C‐ and N‐Terminal Telopeptides of Type I Collagen (CTX and NTX) During bone degradation, osteocalcin breaks down the bone matrix and releases CTX and NTX. Although both CTX and NTX can be measured from urine samples, CTX has gained prominence because it can also be determined from blood tests on some automated platforms. Additionally, it is the preferred biomarker for detecting bone resorption activity. The designed method determined the particular amino acid sequence of telopeptide type I collagen is known as cross lap, and β‐aspartic acid was called β‐CTX. The International Osteoporosis Foundation recommends CTX as an appropriate bone marker for investigating bone resorption in clinical and research settings. Since circadian rhythms and food consumption have effects on circulation β‐CTX, it should be collected during the period of fasting in the morning.
Specific BTMs provided by clinical laboratories Bone turnover markers ( BTMs) are biochemical or cellular compounds produced during the continuum of bone resorption or formation. BTMs can be divided into two categories: bone formation and bone resorption markers (Figure and Table ). The former represents the activity of osteoblasts and the state of bone formation, while the latter mainly reflects the activity of osteoclasts and the level of bone resorption. Procollagen type I N‐terminal propeptide (PINP), procollagen type I C‐terminal propeptide (PICP), and osteocalcin (OC) are three frequently used markers of bone formation. Pyridinoline (PYD), deoxypyridinoline (DPD), N‐terminal cross‐linked telopeptide (NTX), and C‐terminal cross‐linked telopeptide (CTX) are bone resorption markers. In addition, some important cytokines implicated in the regulation of bone turnover by controlling the activity of osteoblasts or osteoclasts, such as osteoprotegerin (OPG) and receptor activator of nuclear factor κB ligand (RANKL), and these can be considered regulators of bone turnover rather than classical biochemical markers of bone turnover. As BTMs are mostly excreted by the kidneys, their levels are frequently detected in urine. Due to the difficulty of collecting 24‐hour urine samples and the need to adjust for creatinine, blood testing is a preferable sample collection technique over urine sampling. Laboratory testing of BTMs is minimally invasive and relatively affordable and, if used and interpreted properly, it can be a useful tool for assessing metabolic bone disease, treatment outcomes, and patient compliance. , CTX and PINP are the commonly used bone turnover markers in clinical research and are recommended by the International Osteoporosis Foundation (IOF) and the International Federation of clinical laboratory medicine (IFCC). With the development of fully automated platforms, the analytical variability of bone markers has been greatly improved. The European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) has shown the biological variability (CV%) of some BTMs, such as ALP, PINP, CTX, and osteocalcin, and with small interlaboratory variation. In this review, we describe several of the most widely used BTMs in clinical pediatrics.
Procollagen type I N‐terminal propeptide (PINP) Type I collagen is a procollagen secreted by osteoblasts, which creates a triple helix consisting of PINP and PICP. Since these propeptides are cleaved in their extracellular area and released into the circulation as metabolites, PINP and PICP concentrations can represent the level of bone formation. Although the clearance of PINP may be less susceptible to hormonal changes than PICP, PINP is a more sensitive marker compared with PICP. Several studies have noted that PINP remains stable even after repeated freezing and thawing, and its levels are not affected by circadian rhythms, so it is not necessary to consider the timing of sampling.
Osteocalcin Osteocalcin is the most common noncollagenous protein in bone, and it is secreted primarily by mature osteoblasts, although it is also secreted during bone resorption. Thus, serum osteocalcin is an important indicator of bone turnover, indicating bone formation and bone resorption. , Because serum osteocalcin concentrations have a distinct circadian rhythm and are highest in the morning, blood samples must be taken in the morning to ensure the accuracy and comparability of the test results. In addition, several studies have shown that seasonal changes and diet did not affect PINP or osteocalcin levels. , , ,
Bone alkaline phosphatase (BALP) Human ALP is classified as tissue non‐specific ALP (TNSALP), intestinal type, placental type, and placental‐like type. TNSALP activity is represented by the products of the ALPL gene, which indicates bone anabolic activity. ALPL gene mutation leads to abnormal skeletal mineralization. BALP is a common biochemical marker of bone formation and a specific marker for osteogenesis, as well as one type of TNSALP, along with liver‐ and kidney‐type ALP. The expression of bone ALP occurs early in the development of osteoblasts from mesenchymal progenitors and is vital in the degradation of pyrophosphate, a natural inhibitor of mineralization. It is hard to identify bone‐type from liver‐type ALP using current immunoassays, as they share the same amino acid sequences. Clinically, the application and interpretation of bone ALP as a BTM should be unaffected by liver diseases because bone ALP is normally cleared from the serum by liver. Children have higher bone ALP activity than adults due to higher bone formation rates. Determination of bone ALP activity also seems to be helpful for the diagnosis of hypophosphatasia and hyperphosphatasia.
C‐ and N‐Terminal Telopeptides of Type I Collagen (CTX and NTX) During bone degradation, osteocalcin breaks down the bone matrix and releases CTX and NTX. Although both CTX and NTX can be measured from urine samples, CTX has gained prominence because it can also be determined from blood tests on some automated platforms. Additionally, it is the preferred biomarker for detecting bone resorption activity. The designed method determined the particular amino acid sequence of telopeptide type I collagen is known as cross lap, and β‐aspartic acid was called β‐CTX. The International Osteoporosis Foundation recommends CTX as an appropriate bone marker for investigating bone resorption in clinical and research settings. Since circadian rhythms and food consumption have effects on circulation β‐CTX, it should be collected during the period of fasting in the morning.
APPLICATION OF BTMs Many clinical and osseous manifestations of metabolic bone disease are more prevalent in children. However, too few studies on the use of these BTMs in children have potential use implications. Here, we summarize the role of BTMs in clinical pediatric care. 3.1 Application of BTMs in growth and development Children and adolescents are important periods of skeletal growth and exhibit high rates of bone growth and rapid bone turnover. BTMs reach the first peak within 1 year after birth, with little differences between boys and girls. Then, BTMs begin to show a downward trend and reach the second peak in early adolescence at the age of eight, with gender differences. The peak of BTMs occurring in girls is earlier than that in boys, while the magnitude of the peak was lower than in boys. This could be related to secondary sexual characteristics and hormone levels. The decline in BTMs levels occurred earlier in girls than in boys and was more significant than in boys during late adolescence, which can explain the differences in bone peak and bone mineral content (PBC) between boys and girls during puberty. A comparison of the reference interval ranges of BTMs in adults and children showed that the levels of BTMs were higher in childhood than those in adults and did not approach adult levels until late adolescence. , , , , , Many factors can influence childhood growth and development, including genetics, nutrition, endocrine status, medication usage, and tumors. These disorders directly or indirectly affect PBC in childhood. Several studies have evaluated the relationship between BTMs and skeletal growth, , , , and BTMs may be a strong predictor of skeletal status in childhood and adolescence, as well as a good predictor of future skeletal growth. Serum BTM levels in CDGP boys were found to be comparable to those of healthy children. , Meanwhile, PINP, OC, and CTX values are lower in preschool and school age, decline during adolescence, and decrease rapidly after puberty, similar to the growth characteristics of children. , , , , , Gascoin et al. found that idiopathic short stature (ISS) children had lower PINP concentrations than normal children of the same age, and that height growth correlated with PINP concentrations during the first year of GH treatment. In addition, a substantial positive correlation among BAP, OC, and insulin‐like growth factor 1 (IGF‐1) in children with ISS. BAP and OC reflect skeletal growth dynamics and skeletal growth outcomes and can be used as monitoring indicators to assess the current growth status of children with ISS and to monitor treatment effects. 3.2 Application of BTMs in rickets Rickets is a childhood disorder associated with mineralization and ossification defects, the most common of which is vitamin D deficiency. Despite significant improvements in early screening and quality of life, epidemiological surveys have revealed that the prevalence of nutritional rickets in children in rural areas remains as high as 10%. , Due to the dramatic increase in serum ALP in children with this disease, studies have confirmed the use of total serum ALP or BALP as an early screening indicator for differential nutritional rickets due to the good correlation between total serum ALP and BALP in childhood, with normal levels suggesting a low likelihood of rickets. , In addition, Chatterjee et al. , discovered that ALP is strongly expressed in nutritional rickets and is a more reliable marker than osteocalcin, PICP, and NTX and that its expression level may be utilized clinically to predict disease severity and prognosis. However, P1NP, β‐CTX, PTH, and 25(OH)D3 have hardly been studied in the field of nutritional rickets in children, and their application value needs to be further discovered and evaluated. 3.3 Primary osteoporosis Children's primary osteoporosis is a genetic disease caused by mutations, and osteogenesis imperfect is the most common disease caused by 17 identified genetic defects. The main manifestation is increased bone fragility. , Moreover, Abdulmoein showed reduced CTX and osteocalcin levels in children with primary osteoporosis 3 months after the treatment with zoledronic acid, suggesting that bone metabolism can be inhibited via bone resorption and short‐term side effects after early treatment. However, due to their limited predictive value for the diagnosis of fractures, BTMs are not currently used in the diagnosis of osteoporosis. It would be possible to monitor the rate of bone loss throughout drug therapy if baseline BTM levels were compared with follow‐up values. 3.4 Secondary osteoporosis Secondary osteoporosis can be caused by multiple factors, such as primary disease and associated therapy. Harada and Rodan firstly found that osteoblasts and osteoclasts release active compounds are important to the physiological activity of other organs. Previous studies have indicated that bone is not only the structural scaffold of the human body, but also can be important endocrine and hormone target organ. , In addition, many young patients develop secondary osteoporosis due to chronic diseases and the medications used. 3.4.1 Application of BTMs in diabetes According to the International Diabetes Federation, Type 1 diabetes (T1D) is most common in children and adolescents, with more than 130,000 people under the age of 20 diagnosed each year. Bone has been identified as an endocrine organ that regulates glucose and energy metabolism. According to research conducted by Schwartz, enhanced glycemic management reduced the incidence of fractures associated with osteoporosis, suggesting that OC is a highly sensitive biomarker of the bone conversion process, which is reduced in patients with poor glycemia. This may be related to the fact that OC regulates energy metabolism. Thus, osteocalcin may reflect early alterations in bone metabolism in diabetic patients and may serve as an indicator of bone turnover. 3.4.2 Application of BTMs for obesity Childhood obesity is a chronic nutritional disease caused by excessive body fat accumulation. According to the World Health Organization (WHO), the number of overweight or obese children under the age of five reached 60 million worldwide in 2020, and the number of obese children aged 0–7 years in China reached 5.31 million, with the trend continuing to rise. Although studies have shown that the incidence of osteoporosis and fracture risk is significantly higher in children with obesity compared with healthy children, it is worth noting that DXA found no differences in osteoporosis between the two, implying that monitoring bone metabolic status is critical for the bone health of children with obesity. Obese children had significantly lower levels of calcium, phosphorus, ALP, 25(OH)D3, P1NP, and OC than healthy children of the same age, as well as lower mean height than normal children of the same age. , Obesity in children not only affects normal growth and development but also leads to metabolic diseases such as diabetes. A meta‐analysis confirmed that OC was reduced in almost all children with type 1 diabetes, and OC was negatively correlated with glycosylated hemoglobin. , The OC may be related to the regulation of glucose metabolism and bone metabolism, as it can reduce the incidence of osteoporosis‐related fractures by improving blood glucose. Although there are no follow‐up studies on the risk of osteoporosis in obese children and adolescents in early adulthood, numerous studies have confirmed the importance of BTMs for assessing bone health levels in children with obesity and metabolic diseases, as well as aiding in the diagnosis and treatment monitoring. 3.4.3 Application of BTMs in hematological diseases and bone tumors hematologic diseases Hematologic disorders (acute lymphocyte leukemia (ALL), thalassemia, and idiopathic thrombocytopenic purpura (ITP)) and bone‐tumor (multiple myeloma and osteosarcoma) have a direct impact on the skeleton of pediatric patients. , , , , Bone has long been recognized as the most common target organ for malignant metastases, and skeletal cell metastases can lead to osteolysis. DXA measurements of bone mass would be underestimated due to the lack of growth hormone or a delay in pubertal development. Young adults suffer a considerable loss in BMD and insufficient density of bone minerals, which may be related to cancer itself, its therapy, or complications such as endocrine problems (decreased gonadal function, growth hormone shortage, etc.), all of which contribute to reduced bone mass gain. Some ALL children have significant osteoporosis at the time of diagnosis, and most of them will acquire this condition during the therapy process. Osteoporosis might occur during the diagnosis in some therapy cases of ALL children, and the majority will acquire this condition during the therapy process. Within 24 months of chemotherapy maintenance, 64% of children had a reduction in bone mineral content (BMC), and children showed alterations in bone conversion with osteocalcin and CTX. , In addition, young adult survivors of ALL had a decrease in BMD because of the possibility of the direct impacts of chemotherapy, steroids, or both on the skeleton during childhood and thus on the increase in bone mass. In the long run, Delvin's view is that 251 individuals who had been cured of leukemia did not show abnormalities in bone turnover markers. For thalassemia major patients, the decrease in osteocalcin may be due to osteoblastosis caused by iron overload. Nevertheless, no significant difference was found in serum alkaline phosphatase levels. In another research group, normal 25(OH)D concentrations may maintain normal calcium homeostasis in patients with thalassemia, suggesting that a normal vitamin D level is important in the pathogenesis of thalassemia bone disease. Furthermore, in patients with chronic ITP, OC, and type I collagen C‐terminal propeptide (PICP) concentrations were lower, urine DPD output was higher, and bone mineral density (BMD) was significantly lower in both the spine and hip Z‐scores. Moreover, BALP is statistically significantly increased with osteosarcoma in patients, while N‐MID osteocalcin and CTX are not significantly different in adolescent and adult groups. Nonetheless, high BALP has the diagnosis value of adult osteosarcoma but is not recommended for the differential diagnosis of adolescent patients, as BALP is affected by age, pubertal stage, and growth rate. Both N‐MID osteocalcin and CTX have limited use in the differential diagnosis of primary bone tumors. However, since there have been so few reports of solid tumors in children, it has to be studied further. 3.4.4 Application of BTMs in Juvenile Rheumatoid Arthritis (JRA) Juvenile rheumatoid arthritis (JRA) is typically connected with osteoporosis, particularly in children's long bones. In addition to impaired skeletal health, children with rheumatic disorders have other risk factors such as delayed growth, malnutrition, decreased weight‐bearing exercise, inflammation, and glucocorticoid medication. Children with RA almost generally fail to produce adequate bone mineralization, as indicated by bone formation failure and lower ICTP. , Inflammation is a crucial factor in the development of osteoporosis, and the majority of research on the OPG/RANKL/RANK axis in rheumatic disorders has been conducted on adults with rheumatoid arthritis (RA). Several studies have found that the degree of osteoporosis in children with chronic arthritis can be affected by the intensity of inflammation, even in the absence of corticosteroid treatment. Meanwhile, high RANKL concentrations are seen in acute RA. Osteoblasts and osteocytes secrete RANKL, which is also produced by B lymphocytes. It facilitates bone resorption and plays an important role in osteoclastogenesis. There has been a reduction in bone formation markers including osteocalcin and bone ALP.
Application of BTMs in growth and development Children and adolescents are important periods of skeletal growth and exhibit high rates of bone growth and rapid bone turnover. BTMs reach the first peak within 1 year after birth, with little differences between boys and girls. Then, BTMs begin to show a downward trend and reach the second peak in early adolescence at the age of eight, with gender differences. The peak of BTMs occurring in girls is earlier than that in boys, while the magnitude of the peak was lower than in boys. This could be related to secondary sexual characteristics and hormone levels. The decline in BTMs levels occurred earlier in girls than in boys and was more significant than in boys during late adolescence, which can explain the differences in bone peak and bone mineral content (PBC) between boys and girls during puberty. A comparison of the reference interval ranges of BTMs in adults and children showed that the levels of BTMs were higher in childhood than those in adults and did not approach adult levels until late adolescence. , , , , , Many factors can influence childhood growth and development, including genetics, nutrition, endocrine status, medication usage, and tumors. These disorders directly or indirectly affect PBC in childhood. Several studies have evaluated the relationship between BTMs and skeletal growth, , , , and BTMs may be a strong predictor of skeletal status in childhood and adolescence, as well as a good predictor of future skeletal growth. Serum BTM levels in CDGP boys were found to be comparable to those of healthy children. , Meanwhile, PINP, OC, and CTX values are lower in preschool and school age, decline during adolescence, and decrease rapidly after puberty, similar to the growth characteristics of children. , , , , , Gascoin et al. found that idiopathic short stature (ISS) children had lower PINP concentrations than normal children of the same age, and that height growth correlated with PINP concentrations during the first year of GH treatment. In addition, a substantial positive correlation among BAP, OC, and insulin‐like growth factor 1 (IGF‐1) in children with ISS. BAP and OC reflect skeletal growth dynamics and skeletal growth outcomes and can be used as monitoring indicators to assess the current growth status of children with ISS and to monitor treatment effects.
Application of BTMs in rickets Rickets is a childhood disorder associated with mineralization and ossification defects, the most common of which is vitamin D deficiency. Despite significant improvements in early screening and quality of life, epidemiological surveys have revealed that the prevalence of nutritional rickets in children in rural areas remains as high as 10%. , Due to the dramatic increase in serum ALP in children with this disease, studies have confirmed the use of total serum ALP or BALP as an early screening indicator for differential nutritional rickets due to the good correlation between total serum ALP and BALP in childhood, with normal levels suggesting a low likelihood of rickets. , In addition, Chatterjee et al. , discovered that ALP is strongly expressed in nutritional rickets and is a more reliable marker than osteocalcin, PICP, and NTX and that its expression level may be utilized clinically to predict disease severity and prognosis. However, P1NP, β‐CTX, PTH, and 25(OH)D3 have hardly been studied in the field of nutritional rickets in children, and their application value needs to be further discovered and evaluated.
Primary osteoporosis Children's primary osteoporosis is a genetic disease caused by mutations, and osteogenesis imperfect is the most common disease caused by 17 identified genetic defects. The main manifestation is increased bone fragility. , Moreover, Abdulmoein showed reduced CTX and osteocalcin levels in children with primary osteoporosis 3 months after the treatment with zoledronic acid, suggesting that bone metabolism can be inhibited via bone resorption and short‐term side effects after early treatment. However, due to their limited predictive value for the diagnosis of fractures, BTMs are not currently used in the diagnosis of osteoporosis. It would be possible to monitor the rate of bone loss throughout drug therapy if baseline BTM levels were compared with follow‐up values.
Secondary osteoporosis Secondary osteoporosis can be caused by multiple factors, such as primary disease and associated therapy. Harada and Rodan firstly found that osteoblasts and osteoclasts release active compounds are important to the physiological activity of other organs. Previous studies have indicated that bone is not only the structural scaffold of the human body, but also can be important endocrine and hormone target organ. , In addition, many young patients develop secondary osteoporosis due to chronic diseases and the medications used. 3.4.1 Application of BTMs in diabetes According to the International Diabetes Federation, Type 1 diabetes (T1D) is most common in children and adolescents, with more than 130,000 people under the age of 20 diagnosed each year. Bone has been identified as an endocrine organ that regulates glucose and energy metabolism. According to research conducted by Schwartz, enhanced glycemic management reduced the incidence of fractures associated with osteoporosis, suggesting that OC is a highly sensitive biomarker of the bone conversion process, which is reduced in patients with poor glycemia. This may be related to the fact that OC regulates energy metabolism. Thus, osteocalcin may reflect early alterations in bone metabolism in diabetic patients and may serve as an indicator of bone turnover. 3.4.2 Application of BTMs for obesity Childhood obesity is a chronic nutritional disease caused by excessive body fat accumulation. According to the World Health Organization (WHO), the number of overweight or obese children under the age of five reached 60 million worldwide in 2020, and the number of obese children aged 0–7 years in China reached 5.31 million, with the trend continuing to rise. Although studies have shown that the incidence of osteoporosis and fracture risk is significantly higher in children with obesity compared with healthy children, it is worth noting that DXA found no differences in osteoporosis between the two, implying that monitoring bone metabolic status is critical for the bone health of children with obesity. Obese children had significantly lower levels of calcium, phosphorus, ALP, 25(OH)D3, P1NP, and OC than healthy children of the same age, as well as lower mean height than normal children of the same age. , Obesity in children not only affects normal growth and development but also leads to metabolic diseases such as diabetes. A meta‐analysis confirmed that OC was reduced in almost all children with type 1 diabetes, and OC was negatively correlated with glycosylated hemoglobin. , The OC may be related to the regulation of glucose metabolism and bone metabolism, as it can reduce the incidence of osteoporosis‐related fractures by improving blood glucose. Although there are no follow‐up studies on the risk of osteoporosis in obese children and adolescents in early adulthood, numerous studies have confirmed the importance of BTMs for assessing bone health levels in children with obesity and metabolic diseases, as well as aiding in the diagnosis and treatment monitoring. 3.4.3 Application of BTMs in hematological diseases and bone tumors hematologic diseases Hematologic disorders (acute lymphocyte leukemia (ALL), thalassemia, and idiopathic thrombocytopenic purpura (ITP)) and bone‐tumor (multiple myeloma and osteosarcoma) have a direct impact on the skeleton of pediatric patients. , , , , Bone has long been recognized as the most common target organ for malignant metastases, and skeletal cell metastases can lead to osteolysis. DXA measurements of bone mass would be underestimated due to the lack of growth hormone or a delay in pubertal development. Young adults suffer a considerable loss in BMD and insufficient density of bone minerals, which may be related to cancer itself, its therapy, or complications such as endocrine problems (decreased gonadal function, growth hormone shortage, etc.), all of which contribute to reduced bone mass gain. Some ALL children have significant osteoporosis at the time of diagnosis, and most of them will acquire this condition during the therapy process. Osteoporosis might occur during the diagnosis in some therapy cases of ALL children, and the majority will acquire this condition during the therapy process. Within 24 months of chemotherapy maintenance, 64% of children had a reduction in bone mineral content (BMC), and children showed alterations in bone conversion with osteocalcin and CTX. , In addition, young adult survivors of ALL had a decrease in BMD because of the possibility of the direct impacts of chemotherapy, steroids, or both on the skeleton during childhood and thus on the increase in bone mass. In the long run, Delvin's view is that 251 individuals who had been cured of leukemia did not show abnormalities in bone turnover markers. For thalassemia major patients, the decrease in osteocalcin may be due to osteoblastosis caused by iron overload. Nevertheless, no significant difference was found in serum alkaline phosphatase levels. In another research group, normal 25(OH)D concentrations may maintain normal calcium homeostasis in patients with thalassemia, suggesting that a normal vitamin D level is important in the pathogenesis of thalassemia bone disease. Furthermore, in patients with chronic ITP, OC, and type I collagen C‐terminal propeptide (PICP) concentrations were lower, urine DPD output was higher, and bone mineral density (BMD) was significantly lower in both the spine and hip Z‐scores. Moreover, BALP is statistically significantly increased with osteosarcoma in patients, while N‐MID osteocalcin and CTX are not significantly different in adolescent and adult groups. Nonetheless, high BALP has the diagnosis value of adult osteosarcoma but is not recommended for the differential diagnosis of adolescent patients, as BALP is affected by age, pubertal stage, and growth rate. Both N‐MID osteocalcin and CTX have limited use in the differential diagnosis of primary bone tumors. However, since there have been so few reports of solid tumors in children, it has to be studied further. 3.4.4 Application of BTMs in Juvenile Rheumatoid Arthritis (JRA) Juvenile rheumatoid arthritis (JRA) is typically connected with osteoporosis, particularly in children's long bones. In addition to impaired skeletal health, children with rheumatic disorders have other risk factors such as delayed growth, malnutrition, decreased weight‐bearing exercise, inflammation, and glucocorticoid medication. Children with RA almost generally fail to produce adequate bone mineralization, as indicated by bone formation failure and lower ICTP. , Inflammation is a crucial factor in the development of osteoporosis, and the majority of research on the OPG/RANKL/RANK axis in rheumatic disorders has been conducted on adults with rheumatoid arthritis (RA). Several studies have found that the degree of osteoporosis in children with chronic arthritis can be affected by the intensity of inflammation, even in the absence of corticosteroid treatment. Meanwhile, high RANKL concentrations are seen in acute RA. Osteoblasts and osteocytes secrete RANKL, which is also produced by B lymphocytes. It facilitates bone resorption and plays an important role in osteoclastogenesis. There has been a reduction in bone formation markers including osteocalcin and bone ALP.
Application of BTMs in diabetes According to the International Diabetes Federation, Type 1 diabetes (T1D) is most common in children and adolescents, with more than 130,000 people under the age of 20 diagnosed each year. Bone has been identified as an endocrine organ that regulates glucose and energy metabolism. According to research conducted by Schwartz, enhanced glycemic management reduced the incidence of fractures associated with osteoporosis, suggesting that OC is a highly sensitive biomarker of the bone conversion process, which is reduced in patients with poor glycemia. This may be related to the fact that OC regulates energy metabolism. Thus, osteocalcin may reflect early alterations in bone metabolism in diabetic patients and may serve as an indicator of bone turnover.
Application of BTMs for obesity Childhood obesity is a chronic nutritional disease caused by excessive body fat accumulation. According to the World Health Organization (WHO), the number of overweight or obese children under the age of five reached 60 million worldwide in 2020, and the number of obese children aged 0–7 years in China reached 5.31 million, with the trend continuing to rise. Although studies have shown that the incidence of osteoporosis and fracture risk is significantly higher in children with obesity compared with healthy children, it is worth noting that DXA found no differences in osteoporosis between the two, implying that monitoring bone metabolic status is critical for the bone health of children with obesity. Obese children had significantly lower levels of calcium, phosphorus, ALP, 25(OH)D3, P1NP, and OC than healthy children of the same age, as well as lower mean height than normal children of the same age. , Obesity in children not only affects normal growth and development but also leads to metabolic diseases such as diabetes. A meta‐analysis confirmed that OC was reduced in almost all children with type 1 diabetes, and OC was negatively correlated with glycosylated hemoglobin. , The OC may be related to the regulation of glucose metabolism and bone metabolism, as it can reduce the incidence of osteoporosis‐related fractures by improving blood glucose. Although there are no follow‐up studies on the risk of osteoporosis in obese children and adolescents in early adulthood, numerous studies have confirmed the importance of BTMs for assessing bone health levels in children with obesity and metabolic diseases, as well as aiding in the diagnosis and treatment monitoring.
Application of BTMs in hematological diseases and bone tumors hematologic diseases Hematologic disorders (acute lymphocyte leukemia (ALL), thalassemia, and idiopathic thrombocytopenic purpura (ITP)) and bone‐tumor (multiple myeloma and osteosarcoma) have a direct impact on the skeleton of pediatric patients. , , , , Bone has long been recognized as the most common target organ for malignant metastases, and skeletal cell metastases can lead to osteolysis. DXA measurements of bone mass would be underestimated due to the lack of growth hormone or a delay in pubertal development. Young adults suffer a considerable loss in BMD and insufficient density of bone minerals, which may be related to cancer itself, its therapy, or complications such as endocrine problems (decreased gonadal function, growth hormone shortage, etc.), all of which contribute to reduced bone mass gain. Some ALL children have significant osteoporosis at the time of diagnosis, and most of them will acquire this condition during the therapy process. Osteoporosis might occur during the diagnosis in some therapy cases of ALL children, and the majority will acquire this condition during the therapy process. Within 24 months of chemotherapy maintenance, 64% of children had a reduction in bone mineral content (BMC), and children showed alterations in bone conversion with osteocalcin and CTX. , In addition, young adult survivors of ALL had a decrease in BMD because of the possibility of the direct impacts of chemotherapy, steroids, or both on the skeleton during childhood and thus on the increase in bone mass. In the long run, Delvin's view is that 251 individuals who had been cured of leukemia did not show abnormalities in bone turnover markers. For thalassemia major patients, the decrease in osteocalcin may be due to osteoblastosis caused by iron overload. Nevertheless, no significant difference was found in serum alkaline phosphatase levels. In another research group, normal 25(OH)D concentrations may maintain normal calcium homeostasis in patients with thalassemia, suggesting that a normal vitamin D level is important in the pathogenesis of thalassemia bone disease. Furthermore, in patients with chronic ITP, OC, and type I collagen C‐terminal propeptide (PICP) concentrations were lower, urine DPD output was higher, and bone mineral density (BMD) was significantly lower in both the spine and hip Z‐scores. Moreover, BALP is statistically significantly increased with osteosarcoma in patients, while N‐MID osteocalcin and CTX are not significantly different in adolescent and adult groups. Nonetheless, high BALP has the diagnosis value of adult osteosarcoma but is not recommended for the differential diagnosis of adolescent patients, as BALP is affected by age, pubertal stage, and growth rate. Both N‐MID osteocalcin and CTX have limited use in the differential diagnosis of primary bone tumors. However, since there have been so few reports of solid tumors in children, it has to be studied further.
Application of BTMs in Juvenile Rheumatoid Arthritis (JRA) Juvenile rheumatoid arthritis (JRA) is typically connected with osteoporosis, particularly in children's long bones. In addition to impaired skeletal health, children with rheumatic disorders have other risk factors such as delayed growth, malnutrition, decreased weight‐bearing exercise, inflammation, and glucocorticoid medication. Children with RA almost generally fail to produce adequate bone mineralization, as indicated by bone formation failure and lower ICTP. , Inflammation is a crucial factor in the development of osteoporosis, and the majority of research on the OPG/RANKL/RANK axis in rheumatic disorders has been conducted on adults with rheumatoid arthritis (RA). Several studies have found that the degree of osteoporosis in children with chronic arthritis can be affected by the intensity of inflammation, even in the absence of corticosteroid treatment. Meanwhile, high RANKL concentrations are seen in acute RA. Osteoblasts and osteocytes secrete RANKL, which is also produced by B lymphocytes. It facilitates bone resorption and plays an important role in osteoclastogenesis. There has been a reduction in bone formation markers including osteocalcin and bone ALP.
PROSPECTS BTMs are important indicators to evaluate bone health status, and their assays are minimally invasive and reproducible. They are currently utilized in clinical practice to diagnose osteoporosis, predict fracture risk, evaluate treatment effects, determine tumor bone metastases, etc. The role of the bone as an endocrine organ is also being recognized by medical researchers due to the close connection between bone metabolism and glucose metabolism. Furthermore, BTMs played important role in the research field of childhood growth and development‐related diseases. Moreover, BTMs are being used for early screening, diagnosis, and monitoring of the efficacy of growth and development‐related diseases for children. However, the relationship between BTMs and DXA, the gold standard for clinical measurement of bone mineral density in children, is still limited to correlational studies, and the clinical use of their combined testing remains to be investigated further. Therefore, we need to further study to make sure a reliable combination of them in clinical application. We look forward to conducting more analysis on the detection and application of BTMs in children and adolescents, this will enable us to have a more comprehensive application in the growth and prevention of bone metabolic disorders. Therefore, pediatrician should be pay more attention on the use of BTMs in pediatrics.
This work was supported by the Sichuan Province Science and Technology Support Program (Grant No. 2020YFS0107) and the Chengdu Science and Technology Support Program (Grant No. 2021‐YF05‐01500‐SN).
The authors declare that they have no competing interests.
|
Comment on: ‘Is undergraduate ophthalmology teaching in the United Kingdom still fit for purpose?’ | 7ad30b5e-b2c4-468d-8206-b838c139ac51 | 8852967 | Ophthalmology[mh] | Thank you to the authors of the article ‘Is undergraduate ophthalmology teaching in the United Kingdom still fit for purpose?’ for an interesting perspective on how to improve the undergraduate student experience within ophthalmology . The authors conclude that the use of technology and blended learning within undergraduate ophthalmology teaching will improve the knowledge and engagement of students. I believe it is also worth noting that the blended approach to teaching, rather than previous didactic formats, may also increase awareness and interest in ophthalmology as a career. It is known that future career choices are influenced by the length of experience within a specialty at medical school . As an Undergraduate Clinical Teaching Fellow, I can appreciate the immense impact COVID-19 has had on medical education. Medical schools during the pandemic suspended clinical placement and face-to-face teaching . When students did return to hospitals, emphasis was on catching up with core medical and surgical placements. Although clinical placements have now recommenced, opportunities are still limited. Some Trusts are unable to run full ophthalmology theatre lists and may be running clinics under increased time pressure. Consequently, medical students have had very little educational exposure to ophthalmology over the past 18 months. There are some medical students that may never even get to see the most commonly performed surgical procedure in the United Kingdom; a cataract operation . Blended learning techniques help address the likely long-term impact of COVID-19 on ophthalmological exposure and subsequent interest in the specialty. It is likely to increase interest in those considering it as a career choice, but also those within allied specialties that frequently treat ophthalmological patients; for example, general practitioners who may then consider extended ophthalmology roles. By adopting technology within blended learning, it is likely to attract more technologically inclined applicants. As technology continues to play an increasing role in healthcare provision, this passion within a workforce is likely to become hugely beneficial in advancing the specialty. To conclude, adopting a blended learning approach in ophthalmology undergraduate education is likely to improve medical students’ interest in the specialty and subsequent career choice.
|
Definition and Predictors of Early Recurrence in Neoadjuvantly Treated Esophageal and Gastroesophageal Adenocarcinoma: a Dual-Center Retrospective Cohort Study | 77a4d308-28eb-41a9-96d7-7c4526f766dd | 11811458 | Surgical Procedures, Operative[mh] | Study Design This study presents a multicenter cohort analysis of surgically treated patients with esophageal or gastroesophageal adenocarcinoma after neoadjuvant treatment. Data were extracted from two prospectively managed databases. The study was performed in compliance with the Strengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines. The study was approved by the Ethics Committee of Heidelberg University (S-649-2012) and complied with the 1964 Helsinki Declaration and its later amendments. All included patients filed informed consent for data use upon treatment. Participants Patients with EAC who underwent surgical resection with curative intent between 2012 and 2021 at Heidelberg University Hospital, Germany and the Clarunis University Digestive Health Care Center in Basel, Switzerland were included. Patients with clinically staged UICC stage I disease were excluded. Therefore, only patients with an indication for neoadjuvant therapy according to European Society for Medical Oncology (ESMO) and National Comprehensive Cancer Network (NCCN) guidelines were included. , Further exclusion criteria encompassed pathologically proven synchronous metastatic disease, macroscopic positive resection margin (R2), neoadjuvant treatment with regimens other than FLOT (5-fluorouracil, leucovorin, oxaliplatin, and docetaxel) or platinum-based radiochemotherapy (CROSS), 90-day mortality without recurrence, and inadequate follow-up data. Inadequate follow-up data was defined as loss to follow-up without event (recurrence or death) within 12 months and incomplete medical records for the primary outcomes, e.g., owing to follow-up at another institution. Treatment Detailed treatment algorithms are described elsewhere. In brief, treatment decisions including mode of neoadjuvant therapy and type of surgery were made in an interdisciplinary tumor board. The decision on the treatment plan was made independently of this study. Neoadjuvant (radio)chemotherapy was administered with either four cycles of FLOT or platinum-based radiochemotherapy according to the CROSS protocol. Staging was performed by endoscopy and cross-sectional imaging [computed tomography (CT), magnetic resonance imaging (MRI), and/or positron emission tomography (PET)-CT] before and after neoadjuvant treatment. Surgery was performed by abdominothoracic esophagectomy with Ivor-Lewis reconstruction (ILE) for cancers proximal to the gastroesophageal junction (GEJ) or transhiatal esophagectomy with total gastrectomy and Roux-Y reconstruction (THG) in GEJ Siewert type 3 cancers based on upper endoscopy results. The decision regarding ILE or THG in GEJ 2 cancers was based on randomization under study conditions or surgeon preference. Signet ring cell positive cancers with known extension to the stomach were preferably treated with THG. A gastric conduit was routinely used for reconstruction in ILE, with a colon conduit as an alternative when reconstruction with the former method was not possible. Follow-Up and Data Collection Patients were followed up with regular clinical visits, serologic, radiographic, and endoscopic diagnostics every 3 months within the first 2 years after surgery and every 6 months thereafter until 60 months postoperatively. Site of recurrence was extracted from results of CT or MRI imaging, endoscopy with or without ultrasound, and biopsies as well as cytologic analyses of peritoneal fluid, where applicable. Further follow-up data were obtained through electrical records of clinical visits, telephone interviews with patients or their primary care provider, and death certificates. Outcomes The American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) 8th edition was used for tumor–node–metastasis (TNM) staging. Treatment response was assessed according to the Becker or Mandard classification. , Becker 1a and Mandard 1 (complete regression) as well as Becker 1b (< 10% vital tumor cells) and Mandard 2 (rare residual cancer cells scattered through the fibrosis) were defined as major histopathologic treatment response. Overall survival (OS) was defined as time from diagnosis to death. When time of diagnosis was not available, the starting date of neoadjuvant treatment was used. Time to progression (TTP) was defined as time from surgery to cancer recurrence, and survival after recurrence (SAR) was defined as time of diagnosis of recurrence to time of death. Sites of recurrence were defined as local, carcinomatosis, or systemic recurrence. If multiple of the aforementioned sites were involved, then the recurrence site was stratified as multiple. Systemic recurrence encompassed all distant recurrence locations including liver, lung, nonregional lymph nodes, bone, brain, adrenal gland, muscles, skin, or multiple of these sites. Carcinosis recurrence encompassed peritoneal and pleural carcinosis as well as Krukenberg tumors. Local recurrence was defined as intra- or extraluminal locoregional recurrence. Statistical Analysis Statistical analysis was performed using R statistical software (version 4.2.0). The Survival and Survminer packages were used for Kaplan–Meier and logistic regression analyses. Ggplot2 and Forester were used for data visualization. A two-sided p -value of < 0.05 was considered statistically significant. Differences in the distribution of categorical data were compared using the chi-squared test, and continuous data were compared using the Wilcox rank-sum test. The Kruskal–Wallis test was used for comparisons of multiple groups with nonparametric data. Missing data are mentioned in the tables but removed for group comparisons. The optimal cutoff to differentiate early from late recurrence was calculated by using the most significant difference in SAR as described by Hothorn et al. This method uses a repetitive approach where, at each instance, a specific time to recurrence is used to split the patient population into those who have a value below and above the tested cutoff value. For each cutoff, the association with the outcome (in this case, SAR) is examined. The optimal cutoff point corresponds to the value of the variable that results in the best discrimination of SAR. Another method used in medical literature is manually searching for the lowest p -value between the two respective groups. Uni- and multivariable logistic regression analyses with backward elimination were used to identify variables associated with early recurrence. Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were calculated for each pretreatment-assessed variable and post-surgery available variable separately. Survival comparisons are visualized using Kaplan–Meier curves. Subgroup analyses were performed for different treatment regimens. This study presents a multicenter cohort analysis of surgically treated patients with esophageal or gastroesophageal adenocarcinoma after neoadjuvant treatment. Data were extracted from two prospectively managed databases. The study was performed in compliance with the Strengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines. The study was approved by the Ethics Committee of Heidelberg University (S-649-2012) and complied with the 1964 Helsinki Declaration and its later amendments. All included patients filed informed consent for data use upon treatment. Patients with EAC who underwent surgical resection with curative intent between 2012 and 2021 at Heidelberg University Hospital, Germany and the Clarunis University Digestive Health Care Center in Basel, Switzerland were included. Patients with clinically staged UICC stage I disease were excluded. Therefore, only patients with an indication for neoadjuvant therapy according to European Society for Medical Oncology (ESMO) and National Comprehensive Cancer Network (NCCN) guidelines were included. , Further exclusion criteria encompassed pathologically proven synchronous metastatic disease, macroscopic positive resection margin (R2), neoadjuvant treatment with regimens other than FLOT (5-fluorouracil, leucovorin, oxaliplatin, and docetaxel) or platinum-based radiochemotherapy (CROSS), 90-day mortality without recurrence, and inadequate follow-up data. Inadequate follow-up data was defined as loss to follow-up without event (recurrence or death) within 12 months and incomplete medical records for the primary outcomes, e.g., owing to follow-up at another institution. Detailed treatment algorithms are described elsewhere. In brief, treatment decisions including mode of neoadjuvant therapy and type of surgery were made in an interdisciplinary tumor board. The decision on the treatment plan was made independently of this study. Neoadjuvant (radio)chemotherapy was administered with either four cycles of FLOT or platinum-based radiochemotherapy according to the CROSS protocol. Staging was performed by endoscopy and cross-sectional imaging [computed tomography (CT), magnetic resonance imaging (MRI), and/or positron emission tomography (PET)-CT] before and after neoadjuvant treatment. Surgery was performed by abdominothoracic esophagectomy with Ivor-Lewis reconstruction (ILE) for cancers proximal to the gastroesophageal junction (GEJ) or transhiatal esophagectomy with total gastrectomy and Roux-Y reconstruction (THG) in GEJ Siewert type 3 cancers based on upper endoscopy results. The decision regarding ILE or THG in GEJ 2 cancers was based on randomization under study conditions or surgeon preference. Signet ring cell positive cancers with known extension to the stomach were preferably treated with THG. A gastric conduit was routinely used for reconstruction in ILE, with a colon conduit as an alternative when reconstruction with the former method was not possible. Patients were followed up with regular clinical visits, serologic, radiographic, and endoscopic diagnostics every 3 months within the first 2 years after surgery and every 6 months thereafter until 60 months postoperatively. Site of recurrence was extracted from results of CT or MRI imaging, endoscopy with or without ultrasound, and biopsies as well as cytologic analyses of peritoneal fluid, where applicable. Further follow-up data were obtained through electrical records of clinical visits, telephone interviews with patients or their primary care provider, and death certificates. The American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) 8th edition was used for tumor–node–metastasis (TNM) staging. Treatment response was assessed according to the Becker or Mandard classification. , Becker 1a and Mandard 1 (complete regression) as well as Becker 1b (< 10% vital tumor cells) and Mandard 2 (rare residual cancer cells scattered through the fibrosis) were defined as major histopathologic treatment response. Overall survival (OS) was defined as time from diagnosis to death. When time of diagnosis was not available, the starting date of neoadjuvant treatment was used. Time to progression (TTP) was defined as time from surgery to cancer recurrence, and survival after recurrence (SAR) was defined as time of diagnosis of recurrence to time of death. Sites of recurrence were defined as local, carcinomatosis, or systemic recurrence. If multiple of the aforementioned sites were involved, then the recurrence site was stratified as multiple. Systemic recurrence encompassed all distant recurrence locations including liver, lung, nonregional lymph nodes, bone, brain, adrenal gland, muscles, skin, or multiple of these sites. Carcinosis recurrence encompassed peritoneal and pleural carcinosis as well as Krukenberg tumors. Local recurrence was defined as intra- or extraluminal locoregional recurrence. Statistical analysis was performed using R statistical software (version 4.2.0). The Survival and Survminer packages were used for Kaplan–Meier and logistic regression analyses. Ggplot2 and Forester were used for data visualization. A two-sided p -value of < 0.05 was considered statistically significant. Differences in the distribution of categorical data were compared using the chi-squared test, and continuous data were compared using the Wilcox rank-sum test. The Kruskal–Wallis test was used for comparisons of multiple groups with nonparametric data. Missing data are mentioned in the tables but removed for group comparisons. The optimal cutoff to differentiate early from late recurrence was calculated by using the most significant difference in SAR as described by Hothorn et al. This method uses a repetitive approach where, at each instance, a specific time to recurrence is used to split the patient population into those who have a value below and above the tested cutoff value. For each cutoff, the association with the outcome (in this case, SAR) is examined. The optimal cutoff point corresponds to the value of the variable that results in the best discrimination of SAR. Another method used in medical literature is manually searching for the lowest p -value between the two respective groups. Uni- and multivariable logistic regression analyses with backward elimination were used to identify variables associated with early recurrence. Hazard ratios (HRs) and 95% confidence intervals (95% CIs) were calculated for each pretreatment-assessed variable and post-surgery available variable separately. Survival comparisons are visualized using Kaplan–Meier curves. Subgroup analyses were performed for different treatment regimens. Patient Cohort A total of 968 patients received surgical treatment for esophageal and gastroesophageal cancer of any type at the Heidelberg University Hospital or the Clarunis University Digestive Health Care Center, Basel between 2012 and 2021. Of those, 573 were treated for adenocarcinoma with UICC stage II or III. After serial exclusion due to not having received neoadjuvant treatment ( N = 78), treatment regimen other than FLOT or CROSS ( N = 47), 90-day mortality ( N = 20), or insufficient data or loss to follow-up ( N = 96), a total of 334 patients were included in the final analysis. Median follow-up was 38.4 months for surviving patients. Eighty-four percent were treated with FLOT, and 16% with CROSS. Other clinicopathological characteristics are presented in Table . Recurrence The median TTP of the whole cohort was 44.1 months (95% CI: 38.8 months to not reached). Of 334 included patients, 160 (48%) were diagnosed with recurrence after a median of 8.8 months (95% CI 7.0–9.9 months) after surgery. Median SAR was 10.5 months (95%CI 8.0–12.7 months). The remaining 52% had no recurrence after median follow-up of 37.4 months. Demographic and clinicopathological characteristics for patients with and without recurrence are summarized in Table . Notably, no preoperatively available variable was associated with recurrence. However, significant associations with recurrence were found for postoperative pathological pT-Stage ( p < 0.001), pN-Stage ( p < 0.001), lymph node ratio ( p < 0.001), and response to neoadjuvant treatment ( p < 0.001). Even after complete pathological response of the primary tumor, recurrence was observed in 18% of patients (Supplementary Fig. 1). Recurrence was diagnosed as local recurrence in 9.2%, carcinomatosis in 9.2%, and systemic in 60.5%. The other 21.1% had multiple recurrence sites involved. Patients with recurrence had significantly shorter OS as compared with patients without recurrence (median OS: 27.7 months versus not reached, p < 0.001; Fig. a). Consequently, the estimated 5-year survival for patients with recurrence was lower at 22%, versus 77% for patients without recurrence ( p < 0.001). Early recurrence was diagnosed in 35% after conventional CT and 41% who received additional PET-CT for presurgical staging of disease ( p = 0.416). Definitions for Early and Late Recurrence In this study, the optimal length to distinguish early from late recurrence was 17.9 months (Fig. ). Patients with early recurrence ( N = 121) had shorter median OS compared with patients with late recurrence ( N = 39) (19.9 versus 71.2 months, p < 0.001; Fig. b) and, by definition, a shorter median SAR with 9.1 months in the early and 17.8 months in the late recurrence cohort ( p = 0.039; Fig. c). Compared with late recurrence, early recurrence was associated with preoperative clinically node positive disease (95% versus 82%, p = 0.017) and a higher lymph node ratio (mean 0.21 versus 0.09, p = 0.010; Table ). There were no significant differences for patterns of recurrence in early versus late recurrence, with 8.8% versus 10% local recurrences, 9.7% versus 7.7% peritoneal recurrences, 58% versus 67% systemic recurrences, and 23% versus 15% with multiple-site recurrence ( p = 0.726). There was also no significant difference in the rate of treatment received for recurrence in the early versus late cohorts (93.5% versus 86.3%, p = 0.275). Predictors of Early Recurrence The early recurrence cohort was compared with the patients who had either late or no recurrence. The results of univariate logistic regression analysis are presented in Table and were calculated for presurgically and postoperative available variables separately in the multivariable analysis. Notably, in pretreatment available variables, only clinically positive nodal status at time of diagnosis was a predictor of early recurrence (OR 3.02, 95% CI:1.22–7.50, p = 0.017). The excellent positive predictive value of 95% is, however, opposed by the poor negative predictive value of 18%. Univariate analysis of postoperatively available variables yielded significant associations of early recurrence to advanced pathological tumor extension (pT3–4: OR 3.51, 95% CI: 2.08–5.91, p < 0.001), nodal positive disease (reference pN0, OR 3.19, 95% CI: 1.97–5.17, p < 0.001), insufficient response to neoadjuvant treatment (reference Mandard TRG 1–2, OR 3.90, 95% CI: 2.23–6.80, p < 0.001). On multivariable analysis, advanced T-stage (pT3–4, adjusted OR 2.52, 95% CI: 1.32–4.91, p = 0.006), nodal positive disease (pN1–pN3, adjusted OR 2.06, 95% CI: 1.17–3.66, p = 0.013), minor histopathological response (> 10% residual tumor, adjusted OR 2.25, 95% CI: 1.18–4.40, p = 0.015), and no adjuvant therapy (adjusted OR 1.74, 95% CI: 1.01–3.04, p = 0.048) predicted early recurrence. Subanalyses for Type of Chemotherapy and Type of Surgery Patients treated with FLOT were significantly younger ( p = 0.007), had more distal tumor location ( p < 0.001), and were therefore more often treated with gastrectomy with transhiatal distal esophagectomy ( p < 0.001, Supplementary Table 1). Furthermore, major complications (Clavien–Dindo ≥ 3) were observed more often after treatment with FLOT compared with CROSS (38% versus 23%, p = 0.044). Adjuvant treatment was more often given to patients after FLOT (76% versus 19%, p < 0.001), with 175 patients (92%) receiving additional cycles of FLOT and 15 (8%) radiochemotherapy. Despite no significant differences in clinically staged T and N stages, histopathological analyses yielded more advanced T stage ( p = 0.032) and more nodal positive disease ( p = 0.018) in the FLOT cohort. The R1 resection rate was similar in the two cohorts (R1: 7.8% versus 7.7%, p = 0.999). However, the treatment regimen was not associated with the occurrence of early recurrence (37% versus 33%, p = 0.564) or recurrence overall (49% versus 42%, p = 0.379). Local recurrence was observed in 8.7% of the FLOT group versus 2.0% of the CROSS group ( p = 0.095), while peritoneal recurrence was observed in 11.1% and 19.6% ( p = 0.091), respectively. There was no difference in the occurrence of systemic recurrence between the two regimens (37.6% versus 37.3%, p = 0.959). Comparisons for gastrectomy with transhiatal distal esophagectomy versus transthoracic esophagectomy with Ivor-Lewis reconstruction yielded significant differences in tumor location ( p < 0.001), type of neoadjuvant treatment ( p < 0.001), and pathological nodal positive disease ( p = 0.039, Supplementary Table 2). The resection margin was positive in 13% after THG compared with 6.1% after ILE ( p = 0.054). However, again, there was no difference in the occurrence of early recurrence (38% versus 36%, p = 0.772) or recurrence overall (47% versus 48%, p = 0.774). A total of 968 patients received surgical treatment for esophageal and gastroesophageal cancer of any type at the Heidelberg University Hospital or the Clarunis University Digestive Health Care Center, Basel between 2012 and 2021. Of those, 573 were treated for adenocarcinoma with UICC stage II or III. After serial exclusion due to not having received neoadjuvant treatment ( N = 78), treatment regimen other than FLOT or CROSS ( N = 47), 90-day mortality ( N = 20), or insufficient data or loss to follow-up ( N = 96), a total of 334 patients were included in the final analysis. Median follow-up was 38.4 months for surviving patients. Eighty-four percent were treated with FLOT, and 16% with CROSS. Other clinicopathological characteristics are presented in Table . The median TTP of the whole cohort was 44.1 months (95% CI: 38.8 months to not reached). Of 334 included patients, 160 (48%) were diagnosed with recurrence after a median of 8.8 months (95% CI 7.0–9.9 months) after surgery. Median SAR was 10.5 months (95%CI 8.0–12.7 months). The remaining 52% had no recurrence after median follow-up of 37.4 months. Demographic and clinicopathological characteristics for patients with and without recurrence are summarized in Table . Notably, no preoperatively available variable was associated with recurrence. However, significant associations with recurrence were found for postoperative pathological pT-Stage ( p < 0.001), pN-Stage ( p < 0.001), lymph node ratio ( p < 0.001), and response to neoadjuvant treatment ( p < 0.001). Even after complete pathological response of the primary tumor, recurrence was observed in 18% of patients (Supplementary Fig. 1). Recurrence was diagnosed as local recurrence in 9.2%, carcinomatosis in 9.2%, and systemic in 60.5%. The other 21.1% had multiple recurrence sites involved. Patients with recurrence had significantly shorter OS as compared with patients without recurrence (median OS: 27.7 months versus not reached, p < 0.001; Fig. a). Consequently, the estimated 5-year survival for patients with recurrence was lower at 22%, versus 77% for patients without recurrence ( p < 0.001). Early recurrence was diagnosed in 35% after conventional CT and 41% who received additional PET-CT for presurgical staging of disease ( p = 0.416). In this study, the optimal length to distinguish early from late recurrence was 17.9 months (Fig. ). Patients with early recurrence ( N = 121) had shorter median OS compared with patients with late recurrence ( N = 39) (19.9 versus 71.2 months, p < 0.001; Fig. b) and, by definition, a shorter median SAR with 9.1 months in the early and 17.8 months in the late recurrence cohort ( p = 0.039; Fig. c). Compared with late recurrence, early recurrence was associated with preoperative clinically node positive disease (95% versus 82%, p = 0.017) and a higher lymph node ratio (mean 0.21 versus 0.09, p = 0.010; Table ). There were no significant differences for patterns of recurrence in early versus late recurrence, with 8.8% versus 10% local recurrences, 9.7% versus 7.7% peritoneal recurrences, 58% versus 67% systemic recurrences, and 23% versus 15% with multiple-site recurrence ( p = 0.726). There was also no significant difference in the rate of treatment received for recurrence in the early versus late cohorts (93.5% versus 86.3%, p = 0.275). The early recurrence cohort was compared with the patients who had either late or no recurrence. The results of univariate logistic regression analysis are presented in Table and were calculated for presurgically and postoperative available variables separately in the multivariable analysis. Notably, in pretreatment available variables, only clinically positive nodal status at time of diagnosis was a predictor of early recurrence (OR 3.02, 95% CI:1.22–7.50, p = 0.017). The excellent positive predictive value of 95% is, however, opposed by the poor negative predictive value of 18%. Univariate analysis of postoperatively available variables yielded significant associations of early recurrence to advanced pathological tumor extension (pT3–4: OR 3.51, 95% CI: 2.08–5.91, p < 0.001), nodal positive disease (reference pN0, OR 3.19, 95% CI: 1.97–5.17, p < 0.001), insufficient response to neoadjuvant treatment (reference Mandard TRG 1–2, OR 3.90, 95% CI: 2.23–6.80, p < 0.001). On multivariable analysis, advanced T-stage (pT3–4, adjusted OR 2.52, 95% CI: 1.32–4.91, p = 0.006), nodal positive disease (pN1–pN3, adjusted OR 2.06, 95% CI: 1.17–3.66, p = 0.013), minor histopathological response (> 10% residual tumor, adjusted OR 2.25, 95% CI: 1.18–4.40, p = 0.015), and no adjuvant therapy (adjusted OR 1.74, 95% CI: 1.01–3.04, p = 0.048) predicted early recurrence. Patients treated with FLOT were significantly younger ( p = 0.007), had more distal tumor location ( p < 0.001), and were therefore more often treated with gastrectomy with transhiatal distal esophagectomy ( p < 0.001, Supplementary Table 1). Furthermore, major complications (Clavien–Dindo ≥ 3) were observed more often after treatment with FLOT compared with CROSS (38% versus 23%, p = 0.044). Adjuvant treatment was more often given to patients after FLOT (76% versus 19%, p < 0.001), with 175 patients (92%) receiving additional cycles of FLOT and 15 (8%) radiochemotherapy. Despite no significant differences in clinically staged T and N stages, histopathological analyses yielded more advanced T stage ( p = 0.032) and more nodal positive disease ( p = 0.018) in the FLOT cohort. The R1 resection rate was similar in the two cohorts (R1: 7.8% versus 7.7%, p = 0.999). However, the treatment regimen was not associated with the occurrence of early recurrence (37% versus 33%, p = 0.564) or recurrence overall (49% versus 42%, p = 0.379). Local recurrence was observed in 8.7% of the FLOT group versus 2.0% of the CROSS group ( p = 0.095), while peritoneal recurrence was observed in 11.1% and 19.6% ( p = 0.091), respectively. There was no difference in the occurrence of systemic recurrence between the two regimens (37.6% versus 37.3%, p = 0.959). Comparisons for gastrectomy with transhiatal distal esophagectomy versus transthoracic esophagectomy with Ivor-Lewis reconstruction yielded significant differences in tumor location ( p < 0.001), type of neoadjuvant treatment ( p < 0.001), and pathological nodal positive disease ( p = 0.039, Supplementary Table 2). The resection margin was positive in 13% after THG compared with 6.1% after ILE ( p = 0.054). However, again, there was no difference in the occurrence of early recurrence (38% versus 36%, p = 0.772) or recurrence overall (47% versus 48%, p = 0.774). In this analysis, an evidence-based cutoff for separating early from late recurrence was found at 1.5 years after surgical resection. Patients with early recurrence had significantly shorter survival after recurrence and consequently much shorter overall survival compared with patients after late recurrence. Independent risk factors for early recurrence encompassed pathological advanced T stage, nodal positive disease, minor histopathological response, and no adjuvant therapy. However, except for nodal positivity at time of diagnosis, no variables that are known prior to surgical treatment could predict early recurrence. Except from providing adjuvant therapy, treatment modalities such as type of neoadjuvant treatment or surgical approach have no influence on early recurrence and recurrence overall. In the literature, varying cutoffs, such as 6 or 12 months, are used for early recurrence in esophageal cancer. , To our knowledge, this is the first evidence-based statistical cutoff to be defined for this cancer entity. This optimal time from surgery to recurrence to separate early and late recurrence was defined by determining the best discrimination in SAR using log-rank testing, similar to the minimum p value approach. While the value of the cutoff derived from the unique dataset used may be influenced by patient numbers and treatment modalities and center-specific follow-up strategies, this multicenter study is representative for Western high-volume centers. Interestingly, the more aggressive the gastrointestinal tumor entity, the shorter the time to recurrence. With 18 months, the cutoff for esophageal adenocarcinoma is higher as compared with pancreatic cancer with 12 months as defined by Groot et al. or liver cancer with 8 months as defined by Xing et al. , For other gastrointestinal tumor entities, similar or higher cutoffs are described in literature: 16 months for colon cancer, 21 months for esophageal squamous cell carcinoma, and 24 months for rectal cancer after neoadjuvant treatment. – Early recurrence is often perceived as failure of surgery, and therefore the value of surgical resection is questioned by both the patient and the physician. Consequently, predictors of early recurrence could aid in treatment decisions, especially in those patients with questionable fitness to endure the surgical procedure and expected prolonged postoperative recovery. , However, the only preoperative available predictor of early recurrence is pretreatment clinical nodal positivity, with a positive predictive value of 95%, meaning that, if present, early recurrence is likely. However, for clinical decision-making, some limitations of this measure must be considered. First, the negative predictive value is poor at only 18%. Second, as shown in other analyses, clinical nodal staging with computer tomography is imprecise. , , Enlargement of lymph nodes beyond the cutoff of 1 cm is deemed as suspicious for lymph node involvement of cancer. Since enlargement of lymph nodes can also be caused by peritumoral inflammation, while metastatic lymph nodes can measure below the cutoff value of 10 mm, poor performance of CT scans in terms of diagnostic accuracy (38–77%) and sensitivity (14–24%) is seen in clinical practice. Therefore, false-negative results are often present when comparing clinical with pathological nodal positivity. In the future, more adequate lymph node examinations, for example, through biopsies or more frequent use of PET-CT scans and liquid biopsies as a marker for systemic disease, may fill the gap and could influence the necessity of cytotoxic treatment or guide decision-making in borderline fit patients with high probability of early systemic recurrence. Importantly, the surgical approach, mode of neoadjuvant treatment, and use of PET-CT for clinical staging had no influence on the occurrence of early recurrence or recurrence overall. Also, there was no difference in recurrence patterns for early versus late recurrence or within the treatment regimens studies. In the absence of valid serum tumor markers for surveillance after esophagectomy, many institutions alternate CT imaging with upper endoscopy every 3 months within the first 2 years and every 6 months thereafter until the fifth postoperative year for every patient. , While unadjusted for tumor biology, some patients may not profit from such close follow-up that may cause a psychological burden for the patient and healthcare costs for society. On the other hand, those with high propensity for early recurrence could benefit from close follow-up and early change of treatment strategies. Early detection of recurrence could especially result in more curatively intended treatments before further progression of the disease. In this analysis, postoperatively advanced T stage, nodal positive disease, minor histopathological response, and no adjuvant treatment were predictors for early recurrence. In addition, positive resection margin and poor tumor differentiation can be found as predictors for recurrence in literature. , , In such patients, an intensified adjuvant therapy or even regimen switch compared with the preoperative neoadjuvant therapy may be considered together with close follow-up. As this was a retrospective analysis, it is associated with potential bias corresponding to the study design. While the aforementioned poor prognostic factors are indicative of unfavorable outcomes, it is imperative not to adopt an exclusively pessimistic outlook if one or multiple factors are present. On the other hand, their absence does not preclude unfavorable outcomes, as shown by some early recurrences despite complete pathologic response of the primary tumor in this study. Furthermore, selection of neoadjuvant therapy was based on institutional preference for FLOT in one center and radiochemotherapy in the other center. Clear trends are shown favoring less locoregional but more peritoneal recurrences in CROSS. While not the primary goal of this study, the limited number of patients treated with CROSS may have produced a false-negative result. However, large studies comparing CROSS versus FLOT also found equal recurrence rates and no clear differences for better local control via radiotherapy nor better systemic control via aggressive systemic treatment with FLOT. – Contrarily, although no recurrence rates and sites have been published yet, the recently presented ESOPEC trial suggests improved survival for FLOT compared with CROSS. Furthermore, the exclusion of patients without sufficient follow-up or who received neoadjuvant chemotherapy regimens other than the currently used regimens of CROSS and FLOT may limit the real-world applicability to a referral center with ongoing trials, including ours. However, by doing so, we ensure a uniform patient cohort with reliable timepoint of recurrence and treatment generalizable to current clinical practice. Promising results have been shown for PD-1-directed immunotherapy in the neoadjuvant setting, with increased rates of major pathological treatment response and lower recurrence rates when given adjuvantly. , In the future, treatment with PDL-1 inhibitors may become standard of care and could alter the prognostic relevance of some established pathologic features. This is the first study to provide an evidence-based cutoff for early recurrence in patients after resection of esophageal adenocarcinoma. Early recurrence can be defined as recurrent disease within 18 months based on SAR. Hallmarks for early recurrence are poor response to neoadjuvant therapy with remaining advanced disease (ypT3/4, ypN+) indicative of unfavorable tumor biology. However, these factors cannot be reliably predicted before resection. Importantly, the type of neoadjuvant treatment has no influence on timing of recurrence and recurrence patterns. In patients with remaining advanced disease after resection, additional adjuvant therapy and shorter follow-up times should be considered. Below is the link to the electronic supplementary material. Supplementary file1 (DOCX 1071 KB) |
Is 2020 the golden year of Otolaryngology research? The impact of COVID-19 on the Italian academic production | 09edc9bc-4857-4f1a-9339-abb1389bfc2c | 8182618 | Otolaryngology[mh] | The Coronavirus disease 2019 (COVID-19) pandemic caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection has profoundly impacted the healthcare, the economy, and the entire society worldwide ( ). Consequently, 2020 has been a remarkably difficult year characterized by many fundamental and unpredictable changes. On the other side, this unprecedented situation has posed new challenges to researchers and the whole scientific community which are working non-stop to defeat COVID-19. Undoubtedly, Otolaryngologists have been among the specialist physicians at highest risk ( ) and have been playing a role of primary importance in the treatment of the symptoms of the upper respiratory tract due to SARS-CoV-2 [i.e., smell and taste dysfunction ( , )], and in the execution of life-saving procedures in COVID-19 patients [i.e., tracheostomies ( )]. According to the Guidelines released by the international Ear, Nose and Throat (ENT) societies, a wide reorganization involved the Head and Neck Departments in many Italian hospitals. Especially during the first wave of the COVID-19 pandemic, oncological and emergency surgical procedures and outpatients’ visits were prioritized to limit the spread of the infection among patients and healthcare workers ( , ). Due to the cancellation of elective activities and to the decrease of ENT consultations ( ), many Otolaryngologists were called to serve as frontline workers in COVID-19 wards or assigned to perform nasopharyngeal swabs ( ). In this unprecedented scenario, also the ENT residency training had to be reorganized and wide space was given to virtual didactic education to overcome the reduction of clinical and surgical activities ( ). Given these premises, Otolaryngologists-Head and Neck Surgeons had the chance to collect valuable data for a better understanding of the disease and to provide their experience in the execution of invasive examinations/aerosol-generating procedures (AGPs) in COVID-19 patients ( ). In a relatively short time, scientific journals released an enormous quantity of articles pertaining the ENT field and, apparently, there was a renewed interest for the specialty during the pandemic. Also, Italy was one of the European countries on the frontlines of COVID-19 research, being the first one to experience a large-scale outbreak. Therefore, the aim of this research was to systematically review the articles published by the ENT departments of the Italian University Hospitals in 2019 and in 2020 to analyze the impact of the COVID-19 pandemic on the academic production.
Search strategy and inclusion criteria The National Library of Medicine through PubMed and Scopus database were searched for scientific papers published by the ENT departments of the Italian University Hospitals respectively in 2019 and in 2020. The Universities with accredited residency programs were included in this research: Bari, Bologna, Brescia, Cagliari, “Campus Biomedico” University of Rome, Catania, Catholic University of the Sacred Heart, Ferrara, Firenze, Foggia, Genova, Humanitas University, Varese Insubria, Milan, Modena and Reggio Emilia, Padova, Palermo, Pavia, Perugia, Piemonte Orientale, Pisa, Rome “La Sapienza”, Rome “Tor Vergata”, Siena, Turin, Trieste, Verona, “Vita-Salute San Raffaele” University . The main eligibility criteria were English-language articles published in peer reviewed scientific journals at any time in 2019 and in 2020, while there were no exclusion criteria related to topic, type of paper or study design. Papers published on preprint platforms were not included in this review. Moreover, the number of citations obtained in 2018, 2019 and 2020 was collected by using Scopus and Web of Science databases. Data extraction Two authors (E.M.C.T., and P.G.M.) independently screened the full-text version of each publication and conducted data extraction. Articles were grouped according to the year and University Hospital. Additionally, articles published in 2020 were classified into “COVID-19 articles” and “Non COVID-19 articles” and subclassified into seven domains corresponding to the main ENT subspecialties: rhinology, otology, laryngology, pediatric otolaryngology, sleep medicine, head and neck surgery and miscellaneous. When agreement could not be reached, the senior authors (M.G., M.C., and F.L.) were consulted.
The National Library of Medicine through PubMed and Scopus database were searched for scientific papers published by the ENT departments of the Italian University Hospitals respectively in 2019 and in 2020. The Universities with accredited residency programs were included in this research: Bari, Bologna, Brescia, Cagliari, “Campus Biomedico” University of Rome, Catania, Catholic University of the Sacred Heart, Ferrara, Firenze, Foggia, Genova, Humanitas University, Varese Insubria, Milan, Modena and Reggio Emilia, Padova, Palermo, Pavia, Perugia, Piemonte Orientale, Pisa, Rome “La Sapienza”, Rome “Tor Vergata”, Siena, Turin, Trieste, Verona, “Vita-Salute San Raffaele” University . The main eligibility criteria were English-language articles published in peer reviewed scientific journals at any time in 2019 and in 2020, while there were no exclusion criteria related to topic, type of paper or study design. Papers published on preprint platforms were not included in this review. Moreover, the number of citations obtained in 2018, 2019 and 2020 was collected by using Scopus and Web of Science databases.
Two authors (E.M.C.T., and P.G.M.) independently screened the full-text version of each publication and conducted data extraction. Articles were grouped according to the year and University Hospital. Additionally, articles published in 2020 were classified into “COVID-19 articles” and “Non COVID-19 articles” and subclassified into seven domains corresponding to the main ENT subspecialties: rhinology, otology, laryngology, pediatric otolaryngology, sleep medicine, head and neck surgery and miscellaneous. When agreement could not be reached, the senior authors (M.G., M.C., and F.L.) were consulted.
The ENT Departments of the Italian University Hospitals released 312 and 540 articles respectively in 2019 and in 2020. Therefore, there was an overall increase in the number of published articles by the 42.2% during the year of pandemic. Out of the articles published in 2020, 116 (21.5%) dealt with COVID-19 and ENT related topics, while 424 (78.5%) were non-COVID-19 articles. The most debated ENT fields were head and neck surgery, rhinology and otology both in 2019 and in 2020 ( ). However, publications pertaining Rhinology and Laryngology increased more significantly in 2020, respectively by the 50.5% and 72.2%. Additionally, there was an increase in the number of miscellaneous articles, which concerned topics of general interest (i.e., department organization, guidelines, telemedicine and residency training), by the 95.2% in 2020. Otology and Head and Neck articles increased respectively by the 32.9% and 32.5%, while the categories of pediatric otolaryngology and sleep medicine remained relatively stable. Also, the citation trends of the articles published by the Italian University Hospitals in 2018, 2019 and 2020 increased remarkably, respectively by the 13.8% from 2018 to 2019 and by the 22.5% from 2019 to 2020 ( ).
Although 2020 is going to be remembered as a year to forget from multiple points of view, it should be acknowledged for being the “golden year of research”. Paradoxically, the COVID-19 pandemic led to a global mobilization of scientific and human resources across countries and culminated in the development of several vaccines in less than one year ( ). Also, the Italian ENT researchers were deeply invested in this scientific mission and gave their valuable contribution to the fight against COVID-19. In fact, our results evidenced that there was an overall increase in the number of articles published by the ENT Departments of the Italian University Hospitals during the COVID-19 pandemic. Additionally, the increase in the scientific production was in proportion to the increase in the number of citations compared to the previous two years ( ). Moreover, it is worth noticing that a considerable percentage (21.5%) of the ENT articles published in 2020 focused on COVID-19. Surely, SARS-CoV-2 infection has been a topic of great interest, but it should also be considered that clinical trials going on before the pandemic have been negatively impacted by the current global crisis. A possible explanation for the recent surge of interest for the ENT specialty might be given by the central role played by Otolaryngologists in treating the upper airways symptoms of SARS-CoV-2 infection ( ) and in performing life-saving procedures (i.e., tracheostomies) in COVID-19 patients ( , ). Additionally, Head and Neck Departments all over Italy were characterized by a reduction in the number of elective activities and a reallocation of the personnel ( – ) and could consequently dedicate more time to research. Another explanation for the impulse to the ENT academic production can be given by the reorganization of the residents training which were redirected to didactic and scientific activities, especially during the first wave of pandemic ( ). Moreover, a fundamental contribution to COVID-19 research was given by the societies of young Otolaryngologists (i.e., Young Otolaryngologists of the International Federation of Otolaryngology Societies [Yo-IFOS], Italian Society of Otolaryngology-Head and Neck Surgery [SIOeChCF]) who produced a considerable number of high-quality scientific papers involving researchers from several Italian Universities ( – ). Interestingly, with 659 citations, the first multicenter European study about chemosensory dysfunction in COVID-19 patients promoted by the YO-IFOS ( ) was the most cited ENT COVID-19 article. A common feature of the articles released by the societies of young Otolaryngologists is the use of web surveys that allowed to collect a huge quantity of data in a relatively short time and to connect different communities of researchers. Although a criticism made by some Authors to COVID-19 literature is given by the redundancy of publications which often lack of rigorous data analysis and solid scientific evidence ( ), the use of online platforms, social media, and mobile applications ( ) can be very useful to collect pilot data in the era of ‘social distancing’. Efforts should be put by the international Societies worldwide to validate these instruments and to understand how to improve the quality of research. The scientific contributions published by the Italian University Hospitals in 2020 covered several topics concerning all the ENT subspecialties. However, the most remarkable increase compared to 2019 was in the number of Rhinology and Laryngology articles. The great interest for Rhinology can be easily explained by olfactory and gustatory dysfunctions as key symptoms of SARS-CoV-2 infection ( , , , ). In fact, the higher number of patients affected by chemosensory impairments seeking medical assistance induced Otolaryngologists to investigate smell and taste further to provide the best care options. A survey of UK-based consultants in 2007 revealed that only the 12% of Otolaryngologists routinely test for chemosensory disorder ( ). Therefore, “every cloud has a silver lining” and after 13 years we can say that the COVID-19 pandemic is having the merit to lead both clinicians and researchers to a major awareness of these diseases. The topic of major interest in the Laryngology literature was that of tracheostomies in COVID-19 patients. First, tracheostomy is an AGP that pose a high risk of contagion to surgeons; therefore, the development of guidelines and safety measures has been very important for the scientific community ( ). Indications and timing of tracheostomy in COVID-19 acute respiratory distress syndrome (ARDS) remain another controversial issue and many articles have been discussing it ( , ). Finally, laryngological sequalae (i.e., granulomas, tracheomalacia, fistulae etc.) and airway stenosis due to prolonged intubation and high rate of tracheostomies in COVID-19 patients are one of the main worries for the future and one of the most debated topics in the current literature ( – ). However, the category that evidenced the highest increase in the number of articles in 2020 was that of miscellaneous articles, which included new and emerging topics that became of global interest with the COVID-19 outbreak. Telehealth systems have become increasingly popular during the COVID-19 pandemic as they provide care to patients while minimizing the risk of SARS-CoV-2 infection among patients and personnel ( ). Given the decrease in the number of ENT consultations ( ), other topics of great interest were the presentation of strategies to reorganize the ENT department ( ), to minimize the risk management ( ) and new solutions to improve the residency training during the COVID-19 pandemic ( ). Lastly, other articles presented novel aerosol mitigation devices to limit the droplets diffusion during AGPs ( ). Also, articles pertaining Otology and Head and Neck surgery increased remarkably, although less than the other subspecialties previously mentioned. The treatment of head and neck cancer patients is not deferrable and surgical oncology treatments have never been cancelled even during the COVID-19 pandemic. Therefore, scientific journals gave special attention to articles that evidenced the importance of head and neck counselling and contained instructions on how to avoid delays in the diagnosis and treatment of head and neck tumors ( , ). Besides the release of guidelines to provide otologic and neurotologic care during the COVID-19 pandemic ( ), another research area of interest was that of the difficulties experienced by patients affected by hearing impairment in understanding and communicating with other individuals wearing facemasks in the era of social distancing ( , ). Lastly, the relationship between sudden sensorineural hearing loss and SARS-CoV-2 infection is still being investigated and remains a controversial issue ( ). Limitations The choice to include only the ENT departments of the Italian University Hospitals with accredited residency programs considerably limited the value of this review, because in Italy there are also ENT departments of non-University Hospitals and Research Institutes which annually give an important contribution to the scientific production. However, given the high number and the heterogeneity of these hospitals, it would have been difficult to meet strict criteria and systematically review their work. Additionally, this study assessed only the ENT literature, while many other specialties were on the frontlines of scientific research and should deserve further evaluation, as well.
The choice to include only the ENT departments of the Italian University Hospitals with accredited residency programs considerably limited the value of this review, because in Italy there are also ENT departments of non-University Hospitals and Research Institutes which annually give an important contribution to the scientific production. However, given the high number and the heterogeneity of these hospitals, it would have been difficult to meet strict criteria and systematically review their work. Additionally, this study assessed only the ENT literature, while many other specialties were on the frontlines of scientific research and should deserve further evaluation, as well.
The ENT academic production at the time of COVID-19 is emblematic of the potentialities endowed by the ENT researchers and the interest that the scientific community has for the ENT specialty. At the same time, it evidenced the necessity of a better organization of the academic research at the Italian Universities, which was probably limited by the clinical and surgical activities in the pre-COVID-19 era and should deserve more space both in the department workload and residency training. Also, the COVID-19 pandemic revealed the necessity to discuss more emerging topics and to provide new solutions to a society in constant evolution and that is not going to be the same in the post-COVID-19 world. Finally, the massive release of scientific articles has not always been correlated with standards of excellence in research. However, the scientific community needs quality more than quantity and it is now more important than ever to provide optimal medical indications and clear answers to patients’ questions.
|
Impact of a multi-pronged cholera intervention in an endemic setting | bf093ae5-a572-493d-85a2-65aa40d77b8f | 11838873 | Vaccination[mh] | Cholera is a bacterial water-borne diarrheal disease transmitted through the fecal-oral route. Since the beginning of the 7th cholera pandemic, cholera has been endemic in sub-Saharan Africa (SSA) which now experiences the highest morbidity and mortality globally , excluding major epidemic events that occurred in Haiti and Yemen. Typical cholera symptoms include vomiting and diarrhea with rice-water stools, potentially leading to severe dehydration. Individual symptoms can range from asymptomatic infections, to mild infections with symptoms that are hardly distinguishable from other diarrheal diseases, to the typical severe watery diarrhea . The case fatality rate (CFR) can reach 70% among severe cases without appropriate treatment, mainly rehydration . As many as 80% of infections can be asymptomatic in endemic areas , resulting in underestimates of cholera burden. Cholera’s causal agent, Vibrio cholerae ( V . cholerae ), specifically serogroups O1 and O139, survives in aquatic environments and is present in the excreta (stools and vomit) of infected individuals. Infection is acquired by ingesting a sufficient bacterial load from the environment (indirect transmission), or contact with infectious excreta (direct transmission). V . cholerae abundance in aquatic reservoirs varies through interactions with biotic and abiotic factors. Elements of aquatic flora and fauna are associated with V . cholerae abundance . Concomitantly, environmental parameters including water temperature and salinity also influence the V . cholerae life cycle in its aquatic reservoir . Viable V . cholerae can persist in the environment in suboptimal conditions for over 15 months in a non-culturable state , from which it can revert to a culturable state in favorable conditions. Inappropriate waste management can introduce V . cholerae in natural or manmade water reservoirs and trigger outbreaks through consumption of contaminated water. An outbreak can then be fueled by both direct and indirect transmission as the increased prevalence of the infection can result in contamination of additional water reservoirs. The dominant transmission routes can be hard to disentangle but their identification is critical to control cholera. The Global Task Force on Cholera Control (GTFCC) has set a road map to eliminate cholera in 20 endemic countries by 2030 , defining SSA as an important target. Generally, diseases or pathogens are considered endemic in an area when they display persistent local transmission for an extended period of time. For cholera, the World Health Organization (WHO) defines an area as endemic when local transmission caused confirmed cases in the previous three years . This definition encompasses a wide variety of transmission patterns, which could cause the same intervention to have different impacts in different endemic areas. In non-endemic areas, the environmental contribution to cholera transmission is often low, but in endemic areas the relative contribution of direct and indirect transmission routes is often unknown. The benefits expected from cholera interventions, as traditionally implemented in outbreak response, become less clear in endemic settings because they do not necessarily target the dominant transmission route. Cholera transmission can be prevented by improving water and sanitation infrastructures and with vaccination. Water, sanitation, and hygiene (WASH) improvements have historically been the primary prevention tool. WASH improvements are resource- and time-intensive to implement . They are extremely effective; waste management and water infrastructures have largely prevented cholera transmission in high income countries . Large scale WASH improvements are necessary to control cholera , however resource scarcities limit such improvements in the countries carrying most of the global burden: SSA nations have some of the poorest access to clean water and improved toilets in the world . In comparison, implementing a vaccination campaign is fast and can reduce cholera transmission quickly. The empirical results of the reactive use of oral cholera vaccines (OCV) in 2012 in Guinea and theoretical results from modeling studies demonstrated the utility of vaccination as a tool to control cholera . A quick vaccine rollout leads to a rapid increase in population immunity that can mitigate cholera transmission, but it is a short term solution because the acquired protection declines after about 2–3 years . The increasing stockpile of OCV allowed for more frequent use of vaccines in outbreak response and its novel use in endemic areas . Both OCV and WASH improvements are important components of the multisectoral interventions required to control cholera in areas with high burden . While the benefit of OCV is straightforward in epidemic settings , it might be narrow in an endemic setting. The impact of OCV on transmission correlates with the increase in population immunity but immunity may always be high if cholera exposure is frequent and widespread, which can be the case in endemic settings. Quantifying the impact of interventions using OCV in endemic settings could provide valuable information to inform control strategies and achieve the ambitious goals set by the GTFCC. The Democratic Republic of Congo (DRC) has consistently carried one of the highest cholera burdens in the African Great Lakes region . Cholera is endemic in the Congolese city of Kalemie, in Tanganyika Province, which lies on the shore of Lake Tanganyika ( ). The area displays annual peaks of cholera cases, typically during rainy seasons ( ), and reports suspected cholera cases all year. Lake Tanganyika could act as an environmental reservoir providing frequent exposure. In parallel, the local population is highly mobile with 24.7% of the residents of Tanganyika Province reporting travelling at least once in the previous 12 months for a duration of at least 1 month . The strong fishing activity, with fishermen moving seasonally and experiencing exposure to the lake and low sanitation conditions, may be a potential source of reintroduction . Such mobility could also promote cholera persistence through metapopulation dynamics. The city of Kalemie received a cholera intervention in 2013–2016 that included both an OCV campaign and limited WASH improvements. The health system in DRC is organized around nested geographical units: Provinces, health zones (HZ), and health areas (HA). Public health interventions are often organized and implemented at least at HZ level. The city of Kalemie spreads across two HZ, Kalemie and Nyemba ( ). The vaccination campaign targeted HA that were in Kalemie city, where attack rates had historically been the highest as of November 2013. The vaccination campaign originally targeted about 120,000 people in four HA with two doses of Shanchol, but was interrupted after three days due to security issues. It resumed in July 2014 and the expiration of vaccine doses led to reducing the target population to about 52,000 people in two HA. Ultimately, 81.2% of the target population received at least one dose . The WASH component of the intervention focused on improving access to clean water. Although it was not acting on every dimension of WASH, we simply refer to it as “WASH intervention” below. Doctors Without Borders ( Médecins Sans Frontières , MSF) extended access to tap water in the northern part of the city by laying pipes, building water reservoirs, distributing water filters, and setting up public drinking fountains in collaboration with Solidarites International . In addition, sand filters were installed on paths where people draw water from the lake, and chlorination activities were performed during outbreaks. The WASH intervention incurred delays in the aftermath of the security issue that delayed the OCV intervention. Its first milestone, extending access to tap water was achieved in October 2014 and the remaining components were completed incrementally until early 2016. We fit a group of deterministic compartmental models that included interhuman cholera transmission with and without environmental contribution and seasonal migration. We used the model with the best fit to assess the short-term impact of this multi-pronged intervention in the city of Kalemie while considering the potential influence of environmental drivers and their contributions to local transmission.
Ethics statement The Ethical Review Board of the University of Lubumbashi approved the study protocol to assess the impact of the vaccination campaign (study protocol ethical number: UNILU/CEM/028/2013) and its extension (study protocol ethical number: UNILU/CEM/050/2015). Individuals provided oral informed consent to be part of the vaccine coverage survey. If the participants were minor, oral consent was obtained from the parent/guardian. Pennsylvania State University’s Institutional Review Board determined the post-intervention handling and analyses of these anonymized data was not Human Research (STUDY00015621). Method overview We fit a group of Susceptible-Infected-Recovered-Susceptible models with a compartment, B, for the bacterial population in the environmental reservoir (SIRB), Lake Tanganyika . We explored the influence of seasonal migration on cholera transmission by fitting models with different structures: with both susceptible and infected (in bold in Eqs , , , and below), or only susceptible individuals migrating, or no migration. We fit the SIRB models to the reported suspected cholera cases presenting at the only cholera treatment center in the city of Kalemie from November 2013 to February 2016. For this period of time only, detailed surveillance data were gathered in an electronic register with support from MSF as part of a study to assess the impact of the intervention. Only residents of the city of Kalemie were included in the analysis. The structure of the full model is as follows: d S d t = δ R − β h S I N − β e B κ + B ( 1 + λ e f ( r a i n t ) ) S − η S + f S ( r a d t ) (1) d I d t = β h S I N + β e B κ + B ( 1 + λ e f ( r a i n t ) ) S − γ I + f I ( r a d t ) (2) d R d t = γ I − δ R + η S (3) d B d t = μ ( 1 + λ c f ( r a i n t ) ) I − B ( ε − φ t ) (4) with f S ( r a d t ) = α 1 r a d t (5) f I ( r a d t ) = α I r a d t (6) r a t i o S / I = α 1 α I (7) φ t = e α 2 + α 3 s s t t + α 4 c h l o r t (8) η = ( σ 1 V C 1 t + σ 2 V C 2 t ) ϴ (9) f ( r a i n t ) = r a i n t m a x ( r a i n t ) (10) The interpretation of all the parameters of the full model is presented in and is described in more details below. Susceptible individuals become infected through exposure to the environmental reservoir, β e , or through interhuman transmission, β h . The WASH intervention decreased the environmental exposure rate β e to β e − β WASH by the end of the study period. β e was assumed to decrease linearly from β e to β e − β WASH from the time the first component of the WASH improvements was completed (ISO week 40 in 2014). The model did not allow the environmental contamination to vary because the intervention did not target waste management. The infection probability from an exposure to the environment followed a dose-effect relationship, with the half saturation constant κ . Infected individuals transitioned to the recovered compartment at rate γ . Susceptible individuals could gain immunity through vaccination, η , 1 week after receiving the vaccine . This was included through a step function of the number of people who received 1 or 2 doses ( VC 1 t and VC 2 t ) of Shanchol. We estimated the number of vaccinated individuals from vaccine coverage estimates from a survey performed by MSF and the associated population size estimates (see ). We considered a range of values for vaccine effectiveness for one and two dose regimens ( σ 1 and σ 2 ), including estimates from studies done in the aftermath of reactive vaccination campaigns performed in Zambia and Guinea (see Table B in ). Our models assumed an all-or-nothing effect of vaccination, implying optimistic estimates of its impact, but we also fit an alternative model structure with a leaky vaccine as sensitivity analysis (see Table H, and Fig O and P in ). Considering the wide age range of the target population (everyone older than 1 year), we assumed that the proportions of susceptible, infected, and recovered among the vaccinated individuals were the same as the general population when the doses were distributed. Immunity waned at rate δ , returning immune individuals to the susceptible compartment. We did not include booster effects on immune individuals receiving vaccine. Booster effects are unlikely to be detected in the study period of 118 weeks (most doses were distributed on the 32 nd and 35 th week), because the study period is shorter than the average period of immunity, whether acquired through infection or vaccination . We also assumed that vaccination had no impact on those who were infected at the time of vaccination. We added a penalty term (ϴ) to account for the spatially targeted nature of the vaccination campaign, which focused on HA in the city of Kalemie with historically high attack rates, where residents had experienced more cholera exposure, further decreasing the proportion of susceptibles. We considered a range of possible values for ϴ (between 0.7 and 1) (see ). Population size was allowed to vary through seasonal migration ( f S ( rad t ) and f I ( rad t )), which can influence local cholera transmission through regular reintroductions from areas with ongoing transmission. We included migration by quantifying the seasonal variation of contemporaneous anthropogenic nighttime radiance, extracted from Visible Infrared Imaging Radiometer Suite (VIIRS) data (see ). We assumed that the net migration flow varied linearly with the first derivative of the nighttime radiance data in the area ( rad t ) . We first fit a generalized additive model with a cyclical spline to the radiance data and then extracted its first derivative (see ). We did not consider the mobility of immune individuals, because they do not actively contribute to transmission. We considered a range of values for the ratio of susceptible and infectious individuals among the mobile population ( ratio S/I ) (between 10 and 100) (see ). We explored alternative model structures allowing only susceptible individuals to be mobile ( f I ( rad t ) = 0) or removing seasonal mobility ( f S ( rad t ) = 0 and f I ( rad t ) = 0) (see ). We considered the influence of water temperature, with lake surface temperature (SST t ), and phytoplankton, with chlorophyll-a (chlor t ), as environmental drivers on aquatic bacterial growth . We extracted these values from Moderate Resolution Imaging Spectroradiometer data (see ). Precipitation (rain t ) could also increase exposure to environmental reservoir and its contamination with infectious human excreta by respectively contaminating drinking water sources ( λ e f ( rain t )) and flooding defecation sites ( λ c f ( rain t )). We extracted precipitation estimates from meteorological forcing data . The bacterial population in the environment increased with contamination of the lake from the excreta of infected individuals ( μ ), and a time dependent bacterial growth rate ( φ t ) that varied with SST t and chlor t . Conversely, it decreased through constant bacterial decay ( ε ). The models did not include births, deaths, or the age structure of the host population because of the short study period of 118 weeks. Based on case management and a CFR of 0.3% during this 118 week-period (5 deaths reported among the 1634 resident suspected cholera cases), we did not include cholera specific mortality. We used a negative binomial process to link the predicted number of weekly incident cases ( C t ) and the weekly reported suspected cases ( A t ) : A t ∼ N e g B i n o m ( C t r , C t ψ ) , with r, a combination of reporting rate and the portion of true cases captured by the suspected case definition (see ), assumed constant, and C t ψ , an overdispersion parameter scaling with the predicted number of new cases. The negative binomial distribution can handle overdispersion and its scaling overdispersion parameter allows variance estimates to better scale with fast and large variations of the incidence. Using different assumptions regarding ratio S / I , σ 1 , σ 2 , and ϴ , we fit a group of 96 models: 64 variations of the full model, 16 variations of the model with only susceptible individuals migrating, and 16 variations of the model without seasonal migration (see Table C and Fig D in ). We assessed model fit with the widely applicable information criteria (WAIC) and selected the best performing model presented here with the lowest WAIC or with fewer parameters for similar WAIC. We also performed a sensitivity analysis of the best performing model by removing the possibility for bacterial growth ( φ t = 0) or the environmental compartment and indirect transmission ( β e = 0) (see Fig E in ). We estimated the parameters β h , β e , β WASH , α 1 , α 2 , α 3 , α 4 , λ e , λ c , δ , μ , ε , r, and ψ , and the initial conditions S 0 , I 0 , B 0 through Markov chain Monte Carlo sampling using the Metropolis-Hastings algorithm. All the estimates presented are the mean values over the posterior distribution and their 95% credible interval (95% CrI) using the highest density interval. We assessed the short-term impact of each arm of the intervention separately and both arms together by estimating the number of additional cases in their absence. We fixed η to 0 while keeping β WASH unchanged, simulating WASH improvements without vaccination, did not allow β e to decrease ( β WASH = 0) while keeping η unchanged, simulating vaccination without WASH improvements, and then fixed both η and β WASH to 0, simulating no vaccination and no WASH improvements. We sampled 10,000 sets of parameters from the posterior distribution and calculated the number of additional cases in each of the alternative scenarios compared to the intervention as it happened. We explored alternative vaccination strategies by varying the timing and the size of the target population, between 50,000 to 200,000 (19.0–76.1% of the population of the city of Kalemie), assuming one campaign during the 118-week period with a two-dose regimen (without WASH). The maximum target population size considered is within the MSF vaccination capacity observed in other settings . We estimated the number of cases avoided for each scenario by calculating the reduction in cholera cases compared to no intervention for each of 10,000 set of parameters sampled from the posterior distribution. We considered 84 combinations of alternative timing and target population size. We sampled 500 sets of parameters for each combination, computational intensity prohibited more. We investigated the relative contributions of environmental exposure and contamination to transmission assuming no intervention by simulating scenarios with no environmental exposure ( β e = 0), or no environmental contamination ( μ = 0) and calculating the number of additional cases compared to having them both ( β e , and μ unchanged) for each of 10,000 sets of parameters sampled from the posterior distribution.
The Ethical Review Board of the University of Lubumbashi approved the study protocol to assess the impact of the vaccination campaign (study protocol ethical number: UNILU/CEM/028/2013) and its extension (study protocol ethical number: UNILU/CEM/050/2015). Individuals provided oral informed consent to be part of the vaccine coverage survey. If the participants were minor, oral consent was obtained from the parent/guardian. Pennsylvania State University’s Institutional Review Board determined the post-intervention handling and analyses of these anonymized data was not Human Research (STUDY00015621).
We fit a group of Susceptible-Infected-Recovered-Susceptible models with a compartment, B, for the bacterial population in the environmental reservoir (SIRB), Lake Tanganyika . We explored the influence of seasonal migration on cholera transmission by fitting models with different structures: with both susceptible and infected (in bold in Eqs , , , and below), or only susceptible individuals migrating, or no migration. We fit the SIRB models to the reported suspected cholera cases presenting at the only cholera treatment center in the city of Kalemie from November 2013 to February 2016. For this period of time only, detailed surveillance data were gathered in an electronic register with support from MSF as part of a study to assess the impact of the intervention. Only residents of the city of Kalemie were included in the analysis. The structure of the full model is as follows: d S d t = δ R − β h S I N − β e B κ + B ( 1 + λ e f ( r a i n t ) ) S − η S + f S ( r a d t ) (1) d I d t = β h S I N + β e B κ + B ( 1 + λ e f ( r a i n t ) ) S − γ I + f I ( r a d t ) (2) d R d t = γ I − δ R + η S (3) d B d t = μ ( 1 + λ c f ( r a i n t ) ) I − B ( ε − φ t ) (4) with f S ( r a d t ) = α 1 r a d t (5) f I ( r a d t ) = α I r a d t (6) r a t i o S / I = α 1 α I (7) φ t = e α 2 + α 3 s s t t + α 4 c h l o r t (8) η = ( σ 1 V C 1 t + σ 2 V C 2 t ) ϴ (9) f ( r a i n t ) = r a i n t m a x ( r a i n t ) (10) The interpretation of all the parameters of the full model is presented in and is described in more details below. Susceptible individuals become infected through exposure to the environmental reservoir, β e , or through interhuman transmission, β h . The WASH intervention decreased the environmental exposure rate β e to β e − β WASH by the end of the study period. β e was assumed to decrease linearly from β e to β e − β WASH from the time the first component of the WASH improvements was completed (ISO week 40 in 2014). The model did not allow the environmental contamination to vary because the intervention did not target waste management. The infection probability from an exposure to the environment followed a dose-effect relationship, with the half saturation constant κ . Infected individuals transitioned to the recovered compartment at rate γ . Susceptible individuals could gain immunity through vaccination, η , 1 week after receiving the vaccine . This was included through a step function of the number of people who received 1 or 2 doses ( VC 1 t and VC 2 t ) of Shanchol. We estimated the number of vaccinated individuals from vaccine coverage estimates from a survey performed by MSF and the associated population size estimates (see ). We considered a range of values for vaccine effectiveness for one and two dose regimens ( σ 1 and σ 2 ), including estimates from studies done in the aftermath of reactive vaccination campaigns performed in Zambia and Guinea (see Table B in ). Our models assumed an all-or-nothing effect of vaccination, implying optimistic estimates of its impact, but we also fit an alternative model structure with a leaky vaccine as sensitivity analysis (see Table H, and Fig O and P in ). Considering the wide age range of the target population (everyone older than 1 year), we assumed that the proportions of susceptible, infected, and recovered among the vaccinated individuals were the same as the general population when the doses were distributed. Immunity waned at rate δ , returning immune individuals to the susceptible compartment. We did not include booster effects on immune individuals receiving vaccine. Booster effects are unlikely to be detected in the study period of 118 weeks (most doses were distributed on the 32 nd and 35 th week), because the study period is shorter than the average period of immunity, whether acquired through infection or vaccination . We also assumed that vaccination had no impact on those who were infected at the time of vaccination. We added a penalty term (ϴ) to account for the spatially targeted nature of the vaccination campaign, which focused on HA in the city of Kalemie with historically high attack rates, where residents had experienced more cholera exposure, further decreasing the proportion of susceptibles. We considered a range of possible values for ϴ (between 0.7 and 1) (see ). Population size was allowed to vary through seasonal migration ( f S ( rad t ) and f I ( rad t )), which can influence local cholera transmission through regular reintroductions from areas with ongoing transmission. We included migration by quantifying the seasonal variation of contemporaneous anthropogenic nighttime radiance, extracted from Visible Infrared Imaging Radiometer Suite (VIIRS) data (see ). We assumed that the net migration flow varied linearly with the first derivative of the nighttime radiance data in the area ( rad t ) . We first fit a generalized additive model with a cyclical spline to the radiance data and then extracted its first derivative (see ). We did not consider the mobility of immune individuals, because they do not actively contribute to transmission. We considered a range of values for the ratio of susceptible and infectious individuals among the mobile population ( ratio S/I ) (between 10 and 100) (see ). We explored alternative model structures allowing only susceptible individuals to be mobile ( f I ( rad t ) = 0) or removing seasonal mobility ( f S ( rad t ) = 0 and f I ( rad t ) = 0) (see ). We considered the influence of water temperature, with lake surface temperature (SST t ), and phytoplankton, with chlorophyll-a (chlor t ), as environmental drivers on aquatic bacterial growth . We extracted these values from Moderate Resolution Imaging Spectroradiometer data (see ). Precipitation (rain t ) could also increase exposure to environmental reservoir and its contamination with infectious human excreta by respectively contaminating drinking water sources ( λ e f ( rain t )) and flooding defecation sites ( λ c f ( rain t )). We extracted precipitation estimates from meteorological forcing data . The bacterial population in the environment increased with contamination of the lake from the excreta of infected individuals ( μ ), and a time dependent bacterial growth rate ( φ t ) that varied with SST t and chlor t . Conversely, it decreased through constant bacterial decay ( ε ). The models did not include births, deaths, or the age structure of the host population because of the short study period of 118 weeks. Based on case management and a CFR of 0.3% during this 118 week-period (5 deaths reported among the 1634 resident suspected cholera cases), we did not include cholera specific mortality. We used a negative binomial process to link the predicted number of weekly incident cases ( C t ) and the weekly reported suspected cases ( A t ) : A t ∼ N e g B i n o m ( C t r , C t ψ ) , with r, a combination of reporting rate and the portion of true cases captured by the suspected case definition (see ), assumed constant, and C t ψ , an overdispersion parameter scaling with the predicted number of new cases. The negative binomial distribution can handle overdispersion and its scaling overdispersion parameter allows variance estimates to better scale with fast and large variations of the incidence. Using different assumptions regarding ratio S / I , σ 1 , σ 2 , and ϴ , we fit a group of 96 models: 64 variations of the full model, 16 variations of the model with only susceptible individuals migrating, and 16 variations of the model without seasonal migration (see Table C and Fig D in ). We assessed model fit with the widely applicable information criteria (WAIC) and selected the best performing model presented here with the lowest WAIC or with fewer parameters for similar WAIC. We also performed a sensitivity analysis of the best performing model by removing the possibility for bacterial growth ( φ t = 0) or the environmental compartment and indirect transmission ( β e = 0) (see Fig E in ). We estimated the parameters β h , β e , β WASH , α 1 , α 2 , α 3 , α 4 , λ e , λ c , δ , μ , ε , r, and ψ , and the initial conditions S 0 , I 0 , B 0 through Markov chain Monte Carlo sampling using the Metropolis-Hastings algorithm. All the estimates presented are the mean values over the posterior distribution and their 95% credible interval (95% CrI) using the highest density interval. We assessed the short-term impact of each arm of the intervention separately and both arms together by estimating the number of additional cases in their absence. We fixed η to 0 while keeping β WASH unchanged, simulating WASH improvements without vaccination, did not allow β e to decrease ( β WASH = 0) while keeping η unchanged, simulating vaccination without WASH improvements, and then fixed both η and β WASH to 0, simulating no vaccination and no WASH improvements. We sampled 10,000 sets of parameters from the posterior distribution and calculated the number of additional cases in each of the alternative scenarios compared to the intervention as it happened. We explored alternative vaccination strategies by varying the timing and the size of the target population, between 50,000 to 200,000 (19.0–76.1% of the population of the city of Kalemie), assuming one campaign during the 118-week period with a two-dose regimen (without WASH). The maximum target population size considered is within the MSF vaccination capacity observed in other settings . We estimated the number of cases avoided for each scenario by calculating the reduction in cholera cases compared to no intervention for each of 10,000 set of parameters sampled from the posterior distribution. We considered 84 combinations of alternative timing and target population size. We sampled 500 sets of parameters for each combination, computational intensity prohibited more. We investigated the relative contributions of environmental exposure and contamination to transmission assuming no intervention by simulating scenarios with no environmental exposure ( β e = 0), or no environmental contamination ( μ = 0) and calculating the number of additional cases compared to having them both ( β e , and μ unchanged) for each of 10,000 sets of parameters sampled from the posterior distribution.
Models with no seasonal migration had a comparable fit to the ones with only susceptible individuals migrating or seasonal migration of both susceptible and infected individuals (see Table C and Fig D in ). This suggested that mobility had minimal influence on the observed cholera dynamics. We selected the model with the lowest WAIC among the ones without seasonal migration, which had fewer parameters. It reproduced the reported weekly cholera cases well, with 98.3% (116/118) of the observed data in the model prediction’s envelope of the 95% CrI of weekly reported suspected cases ( ). The model suggested high local immunity, fluctuating between 88.8% and 99.9% ( ). This high immunity would be the likely consequence of annual outbreaks and persistent environmental exposure, which we explain further below. Based on our model, the targeted vaccinations occurred when population immunity was high: 97.8% (95% CrI: 96.7–98.6) in November 2013, 89. 0% (95% CrI: 76.7–96.7) in July 2014, and 89.1% (95% CrI: 77.2–96.6) during the catch-up in August 2014. Both the scenarios omitting vaccination (WASH only, and no WASH and no vaccination) visibly lacked a reduction in the susceptible proportion of the population in July 2014 ( , bottom panel). Over this 118 week period, we estimated: 3,702 (mean: 3,702.3, 95% CrI: 1,302.5–7,542.0) additional cases, a 2.56% increase (mean: 2.56%, 95% CrI: 1.79%-3.30%), when removing vaccination alone (scenario with WASH only), 1,585 (mean: 1,585.5, 95% CrI: 1,321.9–5,108.8) cases avoided, 1.03% (mean: 1.03%, 95% CrI: 0.01%-2.85%), by WASH alone (scenario with vaccination only), and 5,259 (mean: 5,258.6, 95% CrI: 1,576.6–11,337.8) cases avoided, 3.57% (mean: 3.57%, 95% CrI: 2.02%-5.72%), by implementing both vaccination and WASH (scenario with no vaccination and no WASH improvements) ( ). Our model suggested that vaccination campaigns with small target population sizes would have a limited impact in populations with high immunity ( ). However, the timing of a pulse of vaccination could substantially influence the impact of vaccination campaigns. Specifically, timing the vaccination to occur at the lowest point of population immunity and before an outbreak began increased its impact. The best performing vaccination scenario (darkest cell of the heatmap in ) avoided 12,777 cases (mean: 12,776.7, 95%CrI: 4,681.0.7–26,019.5) over 118 weeks for 200,000 vaccinated people. However, the high level of local immunity would result in vaccinating a large proportion of immune individuals, reducing the impact of the vaccination. We estimated that removing environmental exposure or contamination would have a critical impact on cholera dynamics. These strategies avoided 142,518 cases (mean: 142,518.3, 95% CrI: 36,670.0–303,068.3) and 134,373 cases (mean:134,372.8, 95% CrI: 30,921.5–266,103.8), respectively. In each of these scenarios local cholera transmission was virtually interrupted ( ). Environmental contamination appeared necessary to maintain a bacterial load sufficient to support environmentally-driven transmission because the fluctuation of V . cholerae population averaged towards net decay ( , right). The high immunity inferred by the model was maintained through annual flare-ups and constant environmental exposure. The environmental component of the force of infection ( Φ e = β e B κ + B ( 1 + λ e f ( r a i n t ) ) ) was consistently greater than the interhuman transmission component ( Φ h = β h I N ) despite β h being greater than β e ( ). Φ h remained low because epidemic flare-ups did not lead to a high prevalence of infection, the way they would in a mostly susceptible population. Conversely, Φ e strongly increased with pulses of net bacterial growth due to environmental drivers, despite an overall trend favoring net decay ( left and right).
Based on our model, the impact of the intervention performed in Kalemie was modest when measured by cases avoided, preventing an estimated 5,259 cases (mean: 5,258.6, 95% CrI: 1,576.6–11,337.8) for both intervention arms combined. The reduction of the target population size following the interruption of planned vaccination activities, the limited scale and the incremental implementation of the WASH improvements, and the high level of population immunity likely all contributed to mitigating the impact of the intervention. Benefitting from vaccination in endemic cholera settings, as defined by WHO, requires an understanding of dominant local transmission routes. Our model suggests that the impact of vaccination is small in settings where an environmental reservoir provides constant exposure and maintains high immunity, despite an optimistic assumption of an all-or-nothing vaccine. However, endemicity is more nuanced than the current WHO definition suggests and OCV could still play an important role in some endemic settings. The inability to identify and target the susceptible individuals would lead to vaccinating a majority of immune individuals in this situation. Achieving very high vaccine coverage would immunize a greater number of susceptible individuals, but at the cost of giving many additional doses to immune individuals. This cost could be reduced by targeting the age group most represented among susceptibles or by guiding vaccination with serosurveys. The age profile of the suspected cholera cases residing in Kalemie (median age of 15 years, and interquartile range (IQR) of 3–34 years) during this period would support restricting the maximum age of the target population to increase the impact of the vaccination campaign. However, defining a meaningful age group target would require high resolution historical epidemiological data, and those would only provide information on symptomatic cases, a portion of infected cases. Similarly, guiding vaccination efforts with serosurveys to target susceptibles would incur a substantial additional cost in addition to the difficulty of applying a binary interpretation to serosurvey results. Our estimates of average immunity duration (1/ δ = 3.7 years, 95% CrI: 1.8–8.0 years) and the cumulative incidence converted into an average yearly incidence rate (24.1%, 95% CrI: 11.7–47.5) are consistent with current knowledge of post-infection immunity and other incidence rate estimates in another well studied cholera endemic area, Bangladesh. Challenge studies have demonstrated that immunity lasts at least 3 years after natural infection . National incidence rate in late 2015 in Bangladesh was estimated at 17.3% based on a representative survey and analyses of vibriocidal titres . Our model suggests that a well-timed large-scale vaccination could improve the impact of vaccination in the city of Kalemie, potentially avoiding an average of 12,777 cases (95%CrI: 4,681.0–26,019.5) for 200,000 vaccinated individuals. However, this requires implementing a large vaccination campaign with precise timing. It would be logistically challenging and costly to implement vaccination campaigns of this scale with very precise timing, dictated by the need to vaccinate when immunity is at its lowest and before environmental drivers trigger a pulse of force of infection. This approach would still achieve only short term and small-scale benefits. On the other hand, our findings suggest that WASH improvements on a scale large enough to prevent environmental exposure and contamination for the whole population could have a dramatic long-term impact. Although we estimated that the WASH improvements in Kalemie prevented a modest number of cases, this is likely partially due to the short period of time considered to assess the impact of this part of the intervention. The main components of this WASH intervention consisted of extending the pipe network and building a water reservoir, and they were completed incrementally during the 118-week period. While extending access to the pipe network is an important step, it does not guarantee reliable and consistent access to chlorinated tap water . The magnitude of the improvements required to ensure both access to safe water and efficient waste management, not only in Kalemie but throughout the cholera-affected nation of DRC, appears immense but necessary to control cholera. Implementing WASH improvements should be considered a priority not only to control cholera, but also to prevent the transmission of other water-borne and fecal-oral pathogens that contribute to the disease burden in DRC . This approach will also help achieve the 6 th goal of the Sustainable Development Goals , to ensure availability and sustainable management of water and sanitation for all, in a country where WASH improvements are critically needed . Kalemie is not unique regarding a potentially strong environmental driver of cholera transmission. Substantial environmental contributions for cholera cases have been reported in Haiti and Zimbabwe, areas where the basic reproduction number was estimated to rely mostly on its environmental component . Environmental drivers are also important drivers in other endemic settings like Bangladesh and India, although they act differently: flooding in the early and late phase of the monsoon is strongly associated with higher cholera incidence , while the peak of the monsoon is associated with a cholera lull due the “dilution” of V . cholerae in its reservoir . Although mobility does not appear to be necessary for local cholera persistence in the city of Kalemie, movement could make Kalemie a source of cholera that can seed outbreaks in surrounding areas that lack an environmental source and where exposure is less frequent. The older ages of the suspected cholera cases residing outside of Kalemie (median age of 24.5 years, and IQR: 5.75–39.25 years) are consistent with lower exposure rates and source-sink dynamics. Our estimates supporting a major role of environmentally driven transmission in Kalemie’s local cholera dynamics appear plausible. Sensitivity analysis showed that removing the environmental component or bacterial growth of the model significantly decreased its ability to fit the observed data (see Fig E in ). Confirming our estimates regarding population immunity and the dominant source of bacterial infection would require further research including serosurveys and substantial microbiological monitoring of the lake water in the area. Serological data would allow us to support or disprove our findings, however such data are currently lacking in DRC. Evidence of environmental presence of toxigenic V . cholerae is scarce in the area. Extensive water sampling in Lake Tanganyika from October 22 nd to 26 th 2018 did not detect toxigenic V . cholera e . However, it was detected in ten environmental samples, in fish and water, also collected from Lake Tanganyika from October 2018 to March 2019 and there is some evidence of increased positive samples during rainy seasons in other environmental sampling studies [ – ]. We estimated that the natural variation of the V . cholerae population in the lake leans in favor of net decay. Previous modeling studies assumed bacterial growth rates to consistently be in favor of net decay, whether they varied over time or not . More recent studies considered the possibility for complex bacterial growth patterns but were entirely theoretical . Our model allowed environmental bacterial abundance to vary based on environmental inputs, leading to temporary switches to net bacterial growth. These were important in creating pulses of high environmentally-driven force of infection. Improving the quality of consumed water (reducing environmental exposure) had a large impact in our simulations and removing environmental contamination had an impact almost as large. The overall trend toward net bacterial decay in our model highlights that regularly replenishing local bacterial population through environmental contamination is potentially a critical component of local persistence. This emphasizes the potential compounded benefits of comprehensive improvements to sanitary infrastructures and access to clean water. We did not consider cholera-induced mortality because of the low number of cholera-induced deaths in this population and the local experience in managing cholera infections. However, there is evidence that a substantial portion of cholera mortality occurs in the community , so we cannot rule out that some cholera-induced mortality is not captured in the reported data. The lack of data on mortality in the community prevented us from estimating the number of deaths avoided by the intervention. Our model made simplifying assumptions regarding immunity: we did not account for the various levels of protection acquired after an infection with or without symptoms and did not consider a booster effect of the vaccination on already immune individuals. While immunity is likely shorter after an infection with no or mild symptoms, very little is known about these dynamics because they are difficult to measure . Our model likely presents a transmission pattern averaged across several infectious states that contribute variably to the force of infection; we did not attempt to explicitly separate them due to the absence of data to guide the necessary assumptions. Not considering the booster effect of vaccination on already immune individuals could have led us to slightly underestimate the duration of immunity but this is unlikely to impact our estimates considering the short study period (118 weeks) compared to our estimated average immunity period (3.7 years). We also assumed that immunity wanes at similar rates for susceptible individuals who were successfully vaccinated and following natural infection, but vaccine-induced immunity likely wanes faster . This would have little impact on our estimates considering the small proportion of susceptible individuals in the population when doses were distributed in our model as well as short study period . However, such additions in the model structure would be necessary for a longer time series as the impact of these simplifications would increase. We included the potential impact of the WASH intervention in a simplistic way, assuming a linear variation of the environmental transmission rate. This approach required the fewest additional assumptions; a more refined and detailed method would likely improve the validity of our estimates but the data required to do this do not exist. To estimate environmental drivers, we used measurements of chlorophyll-a and surface water temperature in the lake in addition to the influence of rain. The interactions between V . cholerae and other elements of its aquatic reservoir are only vaguely understood . We modeled one common environmental reservoir, as is customary in SIRB models, which implies that our estimates could hide spatially heterogeneous environmental exposures. Ultimately, we cannot assess how accurately we captured the main fluctuations of the environmental bacterial population in the absence of thorough environmental sampling in the area. However, we considered only environmental drivers that have been associated with V . cholerae environmental abundance or exposure to the environmental reservoir. Phytoplankton growth, indirectly measured through chlorophyll-a, has been associated with cholera outbreaks in several studies, and specifically cyanobacteria are a credible reservoir for V . cholerae . Water temperature influences phytoplankton growth , and the consequence of rainfall on environmental exposure and environmental contamination to/from V . cholerae is credible in this setting along a lake with low access to water and sanitation infrastructures . We explicitly included a direct proxy of human presence, anthropogenic nighttime radiance, in the models considering seasonal mobility. Nighttime radiance is a reliable indicator of human presence and has been used to infer population mobility in both high and low-income countries [ , , ]. We are confident that we robustly captured the seasonal migration and the environmental components in our model, though there might be limitations in the spatiotemporal resolution and availability of remote sensing data, particularly for chlorophyll-a. Impact assessments of cholera interventions are scarce in endemic settings, particularly beyond estimates of vaccine effectiveness and vaccine coverage. Studies like this one are crucial to guide cholera elimination. OCV and WASH improvements are core components of the toolbox to control or eliminate cholera. However, the value of OCV for reactive vaccination in epidemic settings has not been clear in areas with various patterns of endemicities. The assumption that most of the target population is susceptible becomes less accurate as transmission is increasingly environmentally driven. Reducing cholera transmission in endemic areas will require a location-specific understanding of the transmission routes to tailor a strategy; a “one size fits all” approach is unlikely to achieve satisfying results. Geographically-coordinated strategies that target location-specific transmission dynamics might also be necessary to achieve regional cholera control.
S1 Text Additional analyses on some of the assumptions of the model, additional information on the methods and the results, and details on the convergence checks. (DOCX)
|
Global mapping of randomized controlled trials in dentistry | 7f248581-9c3a-481f-a59f-8f469980fbd4 | 11654338 | Dentistry[mh] | The American Dental Association defines evidence-based dentistry as “an approach to oral health care that requires the judicious integration of systematic assessments of clinically relevant scientific evidence, relating to the patient’s oral and medical condition and history, with the dentist’s clinical expertise and the patient’s treatment needs and preferences” . Recently, this concept has become more popular and has gained importance. The need for predictable and effective dental treatments has led clinicians to seek validated therapeutic approaches to support their clinical decisions. One of the most powerful types of evidence used in decision-making originates from randomized controlled trials (RCTs) since they are considered the “gold standard” for evaluating health interventions , . Every year, an impressive number of systematic reviews and RCTs are published, which suggests that the value of these methods is being recognized , . A recent study that evaluated trends in clinical research literature from 1991 to 2020 found that the annual RCT growth rate maintained a steady upward trend until 2017 but with slight fluctuation over the last three years of evaluation. In 1991, 2037 RCTs were published, while in 2020, there were 17,415, highlighting a substantial increase in this type of study over time . Despite the recent increase in published RCTs, the number of RCTs in dentistry is still considerably lower than in the medical field, demonstrating a deficit in evidence-based research in dentistry. In 2017, only 533 RCTs in dentistry were indexed in PubMed . The publication of an RCT does not guarantee its quality. Additionally, studies suggest that articles with a high level of evidence are not consistently associated with the impact factor of the scientific journal in which they are published , . High-quality RCTs with reliable results and an impact on clinical practice must be well planned, conducted, and reported to avoid serious harm to patients, dentists, and the academic and scientific community. Thus, in recent years, initiatives have been developed to improve the quality of these studies, such as encouraging the protocol registration of RCTs, reporting guidelines such as CONSORT (Consolidated Standards of Reporting Trials Statement), and tools for assessing the risk of bias such as RoB 2 , , . Among these initiatives, CONSORT is the most endorsed, widely cited, and recognized as one of the main milestones of the last century in health research methods , . CONSORT has been available since 1996 to guide researchers in reporting RCTs systematically through a checklist of essential items that should be included in RCT reports to make them as complete and transparent as possible . In addition, the checklist serves as a method of evaluating the report and interpreting the study critically. Today, CONSORT has several extensions, and it was last updated in 2010 , . Over time, evidence of the impact of CONSORT has accumulated, and studies have already shown that its endorsement by journals improves the quality of RCT reports in dentistry ; however, several studies in different dental specialties indicate a need for improvement , , , , . Our initial published analysis identified a gender gap in RCTs in dentistry, which is present in study authorship and collaboration between authors . However, considering the relevance of RCTs for evidence-based dentistry, it is necessary to understand the current characteristics of these studies, and how they were conducted and reported. Identifying improvements and gaps in this type of study is essential for the advancement of quality scientific knowledge and for the process of "transforming" evidence into clinical practice. Thus, this study aimed to evaluate the conduct, reporting, and main characteristics of recently published RCTs in dentistry.
The meta-research study protocol was registered on the Open Science Framework and is available at the following link: https://osf.io/qbg9n/ Eligibility criteria As an inclusion criterion, the study needed to be an RCT as described by Friedman et al. . Furthermore, the RCTs should be in the dental field, that is, related to the evaluation, diagnosis, prevention, and/or treatment of diseases, disorders, and/or conditions of the oral, maxillofacial, and/or adjacent area and associated structures or that discussed educational aspects. Articles indexed from 31 December 2016 to 31 December 2021 were included, regardless of the topic (e.g., epidemiological, therapeutic, or diagnostic), methods, or level of detail reported. However, as an exclusion criterion, articles published in languages other than English were excluded due to a lack of funding for article translation. Search presents the search strategy used. We performed searches in PubMed, based on MeSH terms, for RCTs indexed from 31 December 2016 to 31 December 2021. Screening Search results were transferred to DistillerSR (DistillerSR, Evidence Partners Incorporated, ON, CA), an online software that automates screening and data extraction. A pilot test with 20 randomly selected studies was performed using the screening form. First, two researchers independently evaluated titles and abstracts for the presence of eligibility criteria. Articles were classified as “include,” “exclude,” or “uncertain.” Second, an additional eligibility screening was performed using the full text of the records classified as “include” and “uncertain.” This screening was performed independently by the same two reviewers. Discrepancies in evaluating titles, abstracts, and full texts were resolved by consensus or, in the absence of consensus, the opinion of a third reviewer. Sample The sample in this study is part of a larger project, and it is the same as that used in our previous study, which evaluated women's participation in science . We anticipated that around 2,500 RCTs would be identified based on a previous study ; thus, the minimum sample size to find associations, considering an error probability of 5% (α = 0.05), power (1-β) of 80%, an equal proportion of exposed and unexposed (women and men), and an estimated effect size of odds ratio (OR) = 1.5 based on a previous study of female team contribution , was 844 studies. We used an Excel list of random numbers containing all articles classified as included to randomly selected 844 studies considering the proportion of articles indexed per year. Only the most recent report was used if multiple reports from the same study were identified. Data extraction A standardized data extraction form was created in DistillerSR. A pilot test was conducted through discussion among the three reviewers to ensure consistency in interpreting the items. Twenty of the 844 included studies were selected for the pilot test using a list of random numbers in Microsoft Excel. Two reviewers extracted half of all included studies, and another reviewer verified the data extraction and consistency of interpretation. In cases of doubt or inconsistency, the data were re-extracted. All collected data are in our protocol available on the Open Science Framework. However, for this study, we extracted the following data: journal, impact factor (year 2022), type of journal access, subject of article (based on dental specialties recognized by the Federal Council of Dentistry of Brazil ), number of authors, country of corresponding author, country of the first ten authors of each article, total number of citations, and weighted relative citation ratio (wRCR) as reported by the iCite tool (https://icite.od.nih.gov) of each article included. The wRCR was considered a measure of influence, with higher values representing the most cited publications . In addition, we collected the following information on study reporting and conduct from the main points of the CONSORT statement: whether the use of the CONSORT guideline was reported and whether it was reported appropriately (i.e., as a tool to guide study reporting, not to assess the methodological quality of studies or determine how to design and conduct studies); the presence or absence of the term “randomized” in the title; the type of study design and whether it was reported; how many centers were involved and whether this information was reported; number of study groups; and the type of randomization, methods used to generate and implement the allocation sequence, and type of blinding and whether this information was reported. In other words, the reporting of RCTs was assessed based on whether or not the authors mentioned the aspects assessed in our studies. The conduct in RCTs was assessed based on what the authors reported about the methodological aspects of the study. Data analysis All descriptive analyses were performed in Microsoft Excel using frequency for categorical data and mean and standard deviation for continuous data. Using Microsoft Excel, we prepared a map depicting the number of RCTs by country of the corresponding authors. The darker the color of a country in the chart, the more RCTs were assigned to that country. Network graphs were generated in the bibliometric software VOSviewer (version 1.6.19) from an Excel spreadsheet detailing the scientific collaboration among countries based on the corresponding author of each of the articles. Only the first ten authors of each article were included in the analysis. We defined a cross-country collaboration as an article for which the country of the corresponding author differed from that of any of the other authors. The sizes of the circles are proportional to the total link strength between a given country and other countries. The colors of the circles represent the continents to which the countries belong (America: blue, Asia: yellow, Africa: purple, Europe: green, Oceania: red). The lines represent links between countries, and their thickness represents the strength of the connection.
As an inclusion criterion, the study needed to be an RCT as described by Friedman et al. . Furthermore, the RCTs should be in the dental field, that is, related to the evaluation, diagnosis, prevention, and/or treatment of diseases, disorders, and/or conditions of the oral, maxillofacial, and/or adjacent area and associated structures or that discussed educational aspects. Articles indexed from 31 December 2016 to 31 December 2021 were included, regardless of the topic (e.g., epidemiological, therapeutic, or diagnostic), methods, or level of detail reported. However, as an exclusion criterion, articles published in languages other than English were excluded due to a lack of funding for article translation.
presents the search strategy used. We performed searches in PubMed, based on MeSH terms, for RCTs indexed from 31 December 2016 to 31 December 2021.
Search results were transferred to DistillerSR (DistillerSR, Evidence Partners Incorporated, ON, CA), an online software that automates screening and data extraction. A pilot test with 20 randomly selected studies was performed using the screening form. First, two researchers independently evaluated titles and abstracts for the presence of eligibility criteria. Articles were classified as “include,” “exclude,” or “uncertain.” Second, an additional eligibility screening was performed using the full text of the records classified as “include” and “uncertain.” This screening was performed independently by the same two reviewers. Discrepancies in evaluating titles, abstracts, and full texts were resolved by consensus or, in the absence of consensus, the opinion of a third reviewer.
The sample in this study is part of a larger project, and it is the same as that used in our previous study, which evaluated women's participation in science . We anticipated that around 2,500 RCTs would be identified based on a previous study ; thus, the minimum sample size to find associations, considering an error probability of 5% (α = 0.05), power (1-β) of 80%, an equal proportion of exposed and unexposed (women and men), and an estimated effect size of odds ratio (OR) = 1.5 based on a previous study of female team contribution , was 844 studies. We used an Excel list of random numbers containing all articles classified as included to randomly selected 844 studies considering the proportion of articles indexed per year. Only the most recent report was used if multiple reports from the same study were identified.
A standardized data extraction form was created in DistillerSR. A pilot test was conducted through discussion among the three reviewers to ensure consistency in interpreting the items. Twenty of the 844 included studies were selected for the pilot test using a list of random numbers in Microsoft Excel. Two reviewers extracted half of all included studies, and another reviewer verified the data extraction and consistency of interpretation. In cases of doubt or inconsistency, the data were re-extracted. All collected data are in our protocol available on the Open Science Framework. However, for this study, we extracted the following data: journal, impact factor (year 2022), type of journal access, subject of article (based on dental specialties recognized by the Federal Council of Dentistry of Brazil ), number of authors, country of corresponding author, country of the first ten authors of each article, total number of citations, and weighted relative citation ratio (wRCR) as reported by the iCite tool (https://icite.od.nih.gov) of each article included. The wRCR was considered a measure of influence, with higher values representing the most cited publications . In addition, we collected the following information on study reporting and conduct from the main points of the CONSORT statement: whether the use of the CONSORT guideline was reported and whether it was reported appropriately (i.e., as a tool to guide study reporting, not to assess the methodological quality of studies or determine how to design and conduct studies); the presence or absence of the term “randomized” in the title; the type of study design and whether it was reported; how many centers were involved and whether this information was reported; number of study groups; and the type of randomization, methods used to generate and implement the allocation sequence, and type of blinding and whether this information was reported. In other words, the reporting of RCTs was assessed based on whether or not the authors mentioned the aspects assessed in our studies. The conduct in RCTs was assessed based on what the authors reported about the methodological aspects of the study.
All descriptive analyses were performed in Microsoft Excel using frequency for categorical data and mean and standard deviation for continuous data. Using Microsoft Excel, we prepared a map depicting the number of RCTs by country of the corresponding authors. The darker the color of a country in the chart, the more RCTs were assigned to that country. Network graphs were generated in the bibliometric software VOSviewer (version 1.6.19) from an Excel spreadsheet detailing the scientific collaboration among countries based on the corresponding author of each of the articles. Only the first ten authors of each article were included in the analysis. We defined a cross-country collaboration as an article for which the country of the corresponding author differed from that of any of the other authors. The sizes of the circles are proportional to the total link strength between a given country and other countries. The colors of the circles represent the continents to which the countries belong (America: blue, Asia: yellow, Africa: purple, Europe: green, Oceania: red). The lines represent links between countries, and their thickness represents the strength of the connection.
Through the PubMed search, we identified 5,557 studies, and 3,512 met the eligibility criteria. Of these, 844 studies were included in analyses, as suggested by the sample size calculation. More details of these steps can be found in our previous study . presents the main characteristics of the included studies. The 844 articles were published in 195 journals. The journal with the most published articles included in our study was Clinical Oral Investigations (63, 7.46%), followed by two journals in implant dentistry, Clinical Oral Implants Research and Clinical Implant Dentistry and Related Research (34, 4.03%, and 30, 3.55%, respectively). Most journals had hybrid-type access (531, 62.91%), and a minority had subscription access (85, 10.07%). The impact factors of the journals in our study ranged from 0.863 to 24.897, with a mean of 2.980 (± 0.856). The main specialties were periodontology (138 articles, 16.35%) and oral and maxillofacial surgery (135, 16.00%). The average number of citations per article was 7.20 (± 7.40), and the average wRCR was 1.91 (± 1.75). The number of authors ranged from 1-47, with an average of 6.5 (± 2.12). presents data on the reporting and conduct of the included RCTs. Most studies did not report the use of CONSORT (573, 67.89%), and of those that did, the majority reported its use inadequately (155, 18.37%). The term “randomized” was in the title of 71.92% (607) of included studies. However, most studies did not describe the trial design (399, 47.27%). Of those that did, the most frequent design was parallel (263, 31.16%). Additionally, 75.83% (640) of trials involved two study groups performed at a single center (542, 64.22%). Most studies did not report the type of randomization (585, 69.31%). When it was reported, block randomization was the most frequent (144, 17.06%). A computer program or website was the most frequently used method for generating a random allocation sequence (394, 46.68%). Many studies reported using opaque, sealed envelopes or containers (296, 35.07%) to implement the randomized allocation sequence, but most did not report the mechanism used (442, 52.37%). Of blinding techniques, single blinding was the most used in the included RCTs (327, 38.74%). depicts a map of the number of RCTs by country of the corresponding author. We identified 59 countries. The most significant number of trials was attributed to Brazil (140, 16.59%), followed by India (72, 8.53%), the USA (63, 7.46%), and Turkey (62, 7.35%). shows the cross-country collaboration network, according to the authors’ countries, for all RCTs included in our study. There were 62 countries represented by circles and 472 established connections. Authors from the USA established the most links with other countries (138, 29.24%), followed by Italy (87, 18.43%), Brazil (85, 18.01%), and Saudi Arabia (67, 14.19%). Similarly, authors from the USA collaborated the most with other countries , followed by Italy , the United Kingdom , and Brazil . The most frequent connections were between the USA and Brazil (28, 5.93%) and the USA and Saudi Arabia (24, 5.08%).
This is the first study to map recent dental RCTs globally. Our results highlight Brazil as the main source of RCTs in dentistry and the USA as a primary collaborator with other countries. In addition, we found that the reporting and conduct of RCTs are variable. Some practices, such as including “randomized” in the title, type of blinding, and method for generating the allocation sequence, seem to be commonly implemented. However, other important information, such as the use of CONSORT, type of randomization, and mechanism for implementing the allocation sequence, is often not appropriately reported, potentially jeopardizing the understanding of the article. Our study is important because it contributes to understanding the main characteristics of RCTs while highlighting those that deserve further exploration and development to improve this type of study. Moreover, publicly documenting shortcomings in reporting, conduct, and inequities in RCT production and collaboration across countries provides a means of assessing progress over the years. Unsurprisingly, few RCTs and researchers were from countries with low and lower-middle incomes. Studies have shown that the productivity of biomedical research worldwide largely depends on each country’s gross national product per capita and expenditures allocated to research and development , . However, we highlight the role of Brazil, a country with a medium-high income but the highest number of published dental RCTs. Other studies have also highlighted Brazil as an important source of scientific production in dentistry, especially for systematic reviews . Moreover, recently published data from 2022 in the Scimago Journal & Country Rank shows Brazil was the country that published the most international scientific articles in dentistry . Brazil is one of the countries with the fastest growth in academic dental production due to the large number of postgraduate programs in dentistry, in which evaluation processes, for a long time, encouraged maximization of article publication , . Although most included studies were conducted at a single center, which may result in a more significant treatment effect and risk of bias , we identified many cross-country collaborations. In 844 articles, 472 connections were established between authors from different countries. The map of collaborations is predominantly green, indicating a substantial presence of European countries. The USA played a central role in the collaborations identified in this study, corroborating a previous study in the biomedical field . Historically, medical research has had the most support in the USA, where, for many decades, more than half of the world’s funding has been generated. This advantage attracted many researchers to a better academic and scientific environment for many years. Also, many research projects in the USA are funded by private initiatives, unlike most other countries , , . Many studies have attempted to evaluate the reporting of RCTs. A recent study analyzed 20,571 RCTs from biomedical research and found improvements in reporting since 1990. However, in 2015, 30.2% of RCTs were still poorly reported . In dentistry, many authors agree that reporting of RCTs is still suboptimal , , , . Our findings corroborate this, particularly regarding details about randomization. The mechanism for implementing the allocation sequence was not reported in approximately 52% of the studies in our sample. Similar percentages of non-reporting were also observed in areas such as endodontics (40%) , pediatric dentistry (64%) , and restorative dentistry (60% to 83%) , . The benefits of using CONSORT and the improvement that its endorsement and implementation provide to the reporting of RCTs have already been documented in the literature . In a recent study, 85% of studies did not report using CONSORT . Unfortunately, in our study, most also RCTs did not report using CONSORT, which could explain the insufficient reporting of many important items. The word limit stipulated by some journals may also restrict authors from detailing all the methodological characteristics used in a study and impair proper reporting based on CONSORT recommendations. In addition, many authors report certain information only at the request of journal reviewers or model their reporting on similar studies, even if they lack complete knowledge of a given method or type of study. Our result indicated that most studies (71.92%) used the term “randomized” in the title. Similar results were reported in the evaluation of dental RCTs (64.2%) and RCTs on deep caries management (49.6%) , . However, the term's presence in the title is insufficient to indicate that the study was designed, developed, and reported as an adequate RCT. Previous research found that only 39.6% of articles titled RCTs in high-impact dental journals were, in fact, RCTs . These data reaffirm that reporting an item does not always constitute adequate conduct. Most studies we analyzed that reported using CONSORT misreported its use, often by stating it was a guide for conduct rather than for reporting. Methodological quality depends primarily on the degree to which a study’s design, conduct, and analysis meet the highest possible standards and reduce multiple potential biases . Nearly 60% of RCTs in biomedical research used inappropriate methods, according to a recent study . In dentistry, studies have shown that 40% to 52% of the RCTs included in their analyses presented a high risk of bias, indicating inappropriate conduct , , . In our sample, in addition to the reported inappropriate conduct regarding CONSORT, simple randomization was reported in approximately 6% of studies. However, the unpredictability of simple randomization can be disadvantageous. Simple randomization can produce highly disparate treatment arms in small sample sizes. In addition, simple randomization does not guarantee the control of important variables to be considered for both treatment groups, as occurs in stratified randomization , . Furthermore, randomization through the flip of a coin was also observed in our sample. Because of the lack of randomness, difficulties of implementation in larger samples, and absence of an audit, it is recommended that researchers avoid using coin flips as a randomization mechanism . Limitations of this study include the use of only one database, the non-evaluation of gray literature, and the inclusion of only articles in English, which may limit the generalization of the results. However, the database used includes the main journals in the field, most of which are published only in English. Data collection was not performed in duplicate, but we ensured data consistency by conducting a pilot test and involving a third reviewer in checking the collected data. The evaluation of the study reports was based on the main items present in the CONSORT statement; however, some items were not verified, and consequently, the completeness of the CONSORT was also not verified. Still, the evaluation of some RCT conduct characteristics was based only on the authors’ reports, and it was impossible to distinguish studies conducted with greater methodological rigor that were not reported or those studies that reported certain conduct but did not carry it out. Finally, the articles analyzed in our study were published until December 31, 2021, and the change in scenario to the present day must be considered. The authors believe that the data presented still reflect the current scenario and reinforce that it is commonly observed, in meta-research studies, a longer time between the methodological process of project execution and the publication of the article , . Despite the progress observed, considerable improvements in dental RCTs are necessary, and possible, and should be prioritized in future studies. These improvements are essential to ensure the quality of RCTs, avoid resource waste, and ensure that studies have a positive impact on global dental public health. Based on our findings, we suggest some future directions: 1) implement continuous training in courses and postgraduate programs to equip researchers with the best methodological practices to conduct good RCTs; 2) require authors to follow the CONSORT guidelines, including submitting the complete checklist with the manuscripts. The requirement can come from funders, journals, and reviewers and has the role of increasing knowledge and use of the guide, completeness of the report, and increasing the prevalence of reporting items that require more attention, such as randomization details, so essential in this type of study; 3) promote a more balanced distribution in the production of RCTs and scientific collaborations across different regions, especially underrepresented ones. This equity can be facilitated by greater financial support from public and private sources, as well as by encouraging the promotion of open science; and 4) continuing to strengthen the production of RCTs, ensuring that these increasingly higher-quality studies form the basis of clinical decision-making. In conclusion, our global analysis of RCTs in dentistry in recent years identified essential characteristics of these studies, such as the most frequent journals, the most studied subjects, and citation metrics. We highlight Brazil as the country that produces the most RCTs and the USA as a main collaborator. However, there are few studies and few identified collaborations in countries with low and lower-middle incomes. Finally, we emphasize the variability in the reporting and conduct of studies, with the report of CONSORT use and important randomization data being suboptimal. Attention should be focused on strengthening researchers' knowledge of the appropriate methods for conducting good RCTs and the correct way to report these studies by requiring the use of CONSORT.
|
Implementation of Telemedicine in a Laryngology Practice During the COVID-19 Pandemic: Lessons Learned, Experiences Shared | b0fb9246-c026-46b2-8fd2-bd834e3c1644 | 7309798 | Otolaryngology[mh] | The novel coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has created an unprecedented challenge for healthcare delivery. This crisis has placed extraordinary demands on hospitals, emergency departments, and healthcare offices both nationally and worldwide. The Centers for Disease Control and Prevention has recommended healthcare systems reduce face-to-face contact to promote physical distancing, slow disease transmission, and preserve supplies and personal protective equipment. As the pandemic unfolded in March 2020, the American Academy of Otolaryngology-Head and Neck Surgery recommended otolaryngologists limit both inpatient and outpatient care to individuals with time-sensitive, urgent, and emergent medical conditions. As a result, healthcare providers have adopted telehealth models to provide ongoing access to healthcare. Telemedicine is defined as the provision of healthcare services from a distance. Technological advances enable patients to access medical services without the provider being in the same room, for both urgent and nonurgent complaints. In general, telehealth improves patient care access and may improve compliance due to greater convenience of follow-up. , , , However, telemedicine carries clear limitations related to the ability to conduct thorough physical exam and other investigations (eg, laryngoscopy). , , , , There may be reduced personal connection between provider and patient. Technical challenges may limit access to care by the elderly or socioeconomically disadvantaged. While televisits allow healthcare providers an opportunity to offer ongoing specialty care during this pandemic, there are challenges in translating a subspecialty care model from the office to the virtual realm. For laryngology, medical history can be accurately obtained through telehealth platforms, but the ability to complete a comprehensive head and neck exam, stroboscopy, instrumental swallow exam and acoustic/aerodynamic voice analyses is markedly limited. , Furthermore, many laryngology centers employ a multidisciplinary assessment approach by including a speech-language pathologist (SLP) for most encounters. This model requires precise schedule coordination and may be more challenging to continue via telehealth in the COVID-19 era. Herein, we reflect on our early experience of using telehealth to provide laryngology subspecialty care during the COVID-19 pandemic. Collective knowledge and best practices surrounding COVID-19 continue to evolve rapidly. We hope that the following information provides useful insight to other otolaryngology providers to maximize effectiveness of telehealth visits and optimize patient care through a new “routine” healthcare access model.
In preparation for this review, four laryngologists and a voice-specialized SLP engaged in a structured group consensus conference in early April 2020. All participants are routinely involved in the delivery of laryngology subspecialty care at the UCSF Voice and Swallowing Center (VSC), a tertiary-care laryngology practice. All participants actively contributed to sharing and discussing their experiences and practice patterns employed via tele-medicine (via telephone or video-communications) during the early COVID-19 era. Barriers and challenges experienced to date were explored, and potential solutions for these difficulties proposed. Of note, the UCSF VSC has been offering video-telemedicine visits since June 2017, representing nearly three years of experience with the video-visit telemedicine model. These visits were, however, limited to follow-up therapy sessions; all initial joint laryngologist-SLP patient visits and initial SLP therapy sessions were previously completed in-person. Prior to March 2020 (ie, COVID-19 era), laryngologists at the UCSF VSC did not offer telemedicine services.
Based on the input and discussion from UCSF VSC's telemedicine consensus conference, several key areas of consideration were identified for implementing and adopting telemedicine in a multidisciplinary, tertiary-care laryngology practice. These key areas included (1) how to set up and structure a telemedicine visit and maintain patient confidentiality, (2) patient examination and treatment initiation, (3) optimization of the tele-visit, (4) limitations and recognition of when a tele-visit is insufficient for patient care needs, and (5) billing/reimbursement considerations. These topics are discussed individually in the following discussion.
Setting up a telemedicine visit Telemedicine video visits are provided through the Zoom for Healthcare platform (Zoom Video Communications, Inc., Version 4.6.10) at the UCSF Voice and Swallowing Center. Other vendors that provide HIPAA-compliant video communication products include Skype for Business, Updox, VSee, Doxy.me and Google G Suite Hangouts meet. In order for these platforms to meet HIPAA-compliance, several general requirements must be in place for video conferencing. These obligations include: (1) ensuring the confidentiality, integrity, and availability of all electronic protected health information the covered entity creates, receives, maintains, or transmits; (2) protecting against any reasonably anticipated threats or hazards to the security or integrity of such information; (3) protecting against any reasonably anticipated uses or disclosures of such information that are not permitted or required under the privacy regulations; and (4) ensuring compliance by its workforce. Specifics on how these requirements are met and maintained are well detailed online within the Zoom for Healthcare HIPAA-Compliance Guide. Institutions and clinician providers should check with the specific vendors that allow video-communication to ensure the above HIPAA-compliance measures are met as well for the specifics on how the requirements are guaranteed. Platforms like Zoom offer virtual “waiting rooms” from which the clinician “admits” the patient; this safety measure protects patient privacy as the clinician controls who is allowed to enter the visit. Multiple people can be admitted simultaneously (eg, interpreters, additional family members, or scribes) as authorized by the clinician. A flowchart to demonstrate our protocol at the UCSF VSC for arranging, preparing for, and conducting telemedicine visits is shown in . Patients are contacted by telephone in advance to confirm their interest, willingness, and ability to participate in a telemedicine video visit. Patients are informed about the benefits and limitations of a telehealth visit. Importantly, the patient is also notified that the video visit is a billable encounter, and thus insurance co-pay may be applied. A previsit check-in is performed by the clinical nursing team 1-3 days before the appointment. Medications, allergies, and personal information are reviewed, and relevant forms (eg, review of system and patient-reported outcome measure [PROM] questionnaires) are completed. All patients at the UCSF VSC complete standard laryngology PROMs (eg, Voice Handicap Index-10, reflux symptom index, cough severity index (CSI), dyspnea index, and eating assessment tool-10 ) prior to their clinic visit. A Zoom for Healthcare link is provided electronically through the secure patient portal system or via email. The clinical team confirms the patient is able to load and use the platform successfully before their scheduled visit. If possible, having the patient complete a “trial run” with the clinical staff team may enable early identification of technical issues and help the patient feel comfortable with the steps needed to connect to the virtual visit. At the start of the video-visit or telephone encounter, patient identity is verified; typically, name and date of birth can be sufficient. For video visits, the patient must connect to the virtual platform on their own accord, either by clicking on a provided link or by accessing a downloaded application. Verbal consent to utilize telemedicine for the encounter is obtained by the provider and documented within the patient's clinical encounter note. If patients have any concerns about HIPAA-compliance or confidentiality, these questions are addressed by the clinician at this time. UCSF institutional policy prohibits recording of the televisit. Patients verbally confirm their understanding of and agreement to abide by this policy, which further protects patient confidentiality. The remainder of the telehealth visit should be conducted in the same manner as an in-person visit. For referring providers, good communication in the referral regarding patient complaint, duration of symptoms, and suspected urgency are especially important during this time for appropriate patient triage. A review of the referral and existing medical history can help clinicians anticipate patient's needs, and identify those who would especially benefit from SLP involvement. Patients with complex laryngologic complaints may be best served by a joint visit with both laryngologist and SLP. This allows a detailed history to be obtained concurrently rather than separately, thereby saving time. A concurrent visit also allows for real-time discussion between the laryngologist and SLP, which facilitates efficiency and treatment decision-making. Virtual joint laryngologist/SLP visits may be possible by logging into one meeting room with the patient. After joint initial patient evaluation, the laryngologist can maximize efficiency while the SLP spends time assessing the patient for therapy candidacy (ie, stimulability , ) by completing other patient-care-related tasks. This may include reviewing previous testing (eg, pulmonary function testing, pH-impedance, manometry, or imaging), working on chart documentation, initiating referrals, or ordering medications or additional testing. After initial evaluation, in the Zoom platform, the patient may be returned to the “waiting room” temporarily so that the laryngologist and SLP can discuss the patient's case privately, as providers might typically do outside of the patient's examination room. Subsequently, the patient may be “brought back in” to the joint visit for discussion of suspected differential diagnosis and recommended next step(s) in evaluation or treatment. New patients present a particular challenge with respect to timing and coordination. History-taking can be time-consuming, especially if complex. A full SLP evaluation including voice stimulability testing, , acoustic evaluation, and/or clinical swallow evaluation requires adequate time allotment. During in-person visits, the laryngologist may go into another exam room with another patient during this time. For virtual visits, this is technically feasible with many telemedicine platforms but again requires appropriate coordination and planning. This may be most easily accomplished by the provider having more than one Zoom account or access to breakout rooms that can be utilized to run “multiple rooms” with multiple patients. Clearly, this time management is more challenging to coordinate during a video visit with multiple concurrent providers. Alternatively, it may be more advantageous for some new patients to be assessed independently by the laryngologist and SLP depending on the specific practice flow, patient complaint or anticipated patient needs. For follow-up visits, when care has already been established, a joint laryngologist/SLP session may be more feasible and efficient. Such situations include patient progress updates, re-evaluation after completion or plateau of voice/swallow therapy, evaluation of medication efficacy (eg, allergy or reflux treatment), review of test results (eg, modified barium swallow [MBSS]), and/or patient counseling. A joint session to determine the role for additional therapy, addition of medications or surgical intervention, or review of trialed devices (such as an Expiratory Muscle Strength Trainer or ProTrach) can be more easily accomplished. The visit duration for follow-up is usually shorter which facilitates schedule coordination. Examination and treatment Despite advances in telehealth, otolaryngologists and SLPs depend highly on the oral/pharyngeal/laryngeal examination for diagnosis and treatment. A limited physical examination is possible via video visit. A cursory view of the oral cavity and oropharynx can be obtained with appropriate intraoral lighting in some cooperative patients. , Perceptual voice analysis with sustained and dynamic vocalization (ie, varying pitch and loudness) tasks in addition to reading standardized passages provide important auditory diagnostic information. Clinical swallowing evaluations are frequently employed as an initial screening tool in dysphagia patients. While these assessments have well-described limitations, direct observation of the patient swallowing various consistencies and subsequently phonating during the tele-visit could provide helpful preliminary information about global swallowing function, voice quality after swallow, and strength of cough. For patients with breathing complaints, quiet respiration at baseline can be observed and rapid breathing exercises may elicit stridor. While a telemedicine model of laryngoscopy-videostroboscopy has been described, where a remotely performed exam is relayed for review through either real-time or cloud-based (store and review) technologies, the reliance on an in-office visit with someone able to perform endoscopy remains. , , During the peak of the pandemic, avoidance of nonurgent in-person visits and aerosol-generating procedures were recommended. , Nonurgent laryngoscopy was therefore deferred. Of course, any urgent laryngoscopy should still be performed as deemed medically necessary. It is imperative that patients are counseled about the diagnostic and treatment limitations in the absence of an endoscopic laryngeal examination. Furthermore, it must be emphasized to patients that such an examination is vital to complete as soon as it is deemed safe to do so. However, proceeding with therapeutic intervention in the absence of the above comprehensive evaluations is nonetheless possible. For example, in suspected laryngitis or vocal fold hemorrhage, a trial of voice rest could be considered, with reassessment for response in 5 to 7 days. Additionally, there are many medical therapies that laryngologists can initiate with acceptable benefit-to-risk ratios without confirmatory laryngoscopy, including proton pump inhibitors, H2 blockers, alginates, steroids, antitussives, mucolytics, and antibiotics, based on appropriate history and suspected diagnosis. Many clinical diagnoses are confirmed or supported by laryngoscopy, but may safely be treated with trials of medications. In these instances, laryngoscopy may be delayed without undue patient risk. For patients with known subglottic stenosis, use of a home peak flow meter to monitor airway symptoms may guide need for in-person visit or surgery. , , Results may be relayed during tele-video visits, and if significant decline in values are noted, a prompt in-office visit should be considered. Again, providers should proceed with caution, carefully monitoring progress and ensuring completion of full evaluation when safely possible. Once clinics resume safe instrumental examination, timely completion should be accomplished. Optimizing the success of the video visit Multiple steps can be taken on the part of both provider and patient in order to maximize telehealth efficiency and effectiveness. The office should create a virtual workflow for telehealth visits to maximize success prior to starting the visit. This can include a previsit call from office staff to ensure the patient is aware of and knows how to set up the video visit. Technical support or troubleshooting during this call can be helpful. For the provider, we recommend performing the telehealth visit in a quiet room and using headphones to hear the patient clearly. Similarly, it is helpful for the patient to participate in a quiet room using headphones. In this way, both provider and patient can be heard without excessive vocal strain or effort. For certain patients (eg, elderly, hard of hearing, or those with disabilities), having family members or other care providers present for the video visit is anecdotally beneficial. Reliable Internet connection is critical. Proximity to the router or being directly connected to Ethernet can be helpful to avoid disruptions of video or audio feed. A readily accessible informational technology hotline number can help providers and patients troubleshoot technical issues. Flexibility and patience are also important to ensuring success. In the event a patient does not show on time for their visit, a simple phone call can help the provider understand the reason for that delay. This may be due to a forgotten visit time, difficulty accessing the tele-visit platform, or another easily corrected technology problem. If the patient is having difficulty with use of a downloaded application, a simple workaround may be use of the website via the internet (eg, zoom.us) to join the meeting. For the clinician, having your Personal Meeting Identification number readily on hand to give to the patient to input manually on the website can expedite patient connection to the virtual meeting. Patient-reported outcome measures (eg, VHI-10, reflux symptom index, cough severity index, dyspnea index, and eating assessment tool-10 ) can be completed by patients prior to the visit, and submitted by email or directly through the electronic medical record. These results are then available to the providers prior to the visit for review. It is helpful to create a database of resources that can easily be shown during the video visit or sent to the patient via secure link for easy access. These resources serve as virtual replacement for the typical paper informational handouts given in the office. The Zoom for Healthcare platform allows the provider to share their screen. This feature facilitates reviewing results, imaging, previously archived endoscopic examinations, or diagrams for explaining proposed surgeries or procedures. Sound can sometimes be a challenge on a virtual visit. For example, sustained sounds during voice evaluation or voice therapy may be clipped on Zoom. Providers may change settings to allow original sound; if the issue continues to persist, the patient can also be instructed to make these changes ( ). When is a video visit not enough? Some voice and swallowing complaints can be safely managed remotely. However, the majority of patients will ultimately require laryngoscopy for evaluation. If on initial telehealth evaluation, the provider identifies medical complaints needing urgent evaluation, the patient should come in for an urgent office visit. Within laryngology, urgent issues include suspected malignancy, symptomatic airway obstruction, aspiration, and severe dysphagia without alternative nutritional intake means (ie, nasogastric or PEG tube). There are also situations when patients themselves are distressed and would be more reassured by in-person patient evaluation. These cases should be addressed more urgently at the provider's discretion. In-person SLP evaluation with instrumentation including MBSS should be limited to urgent cases only during peak pandemic. This includes patients with no means of nutrition and high risk of aspiration. Alternatively, MBSS can be performed at a facility closer to home and then reviewed by the provider to determine urgency and in-office assessment needs. In cases where urgent MBSS is required and the patient is unknown to the SLP, it is advantageous for the SLP to complete a video visit prior to the MBSS to obtain relevant history and complete a clinical swallow evaluation. Then, the in-person time in the fluoroscopy suite during the MBSS can be utilized most efficiently. For patients with a tracheoesophageal voice prosthesis, in-person visit during peak pandemic may be necessary. However, video visits can be used for troubleshooting, as the patient may be able to apply a plug to the tracheoesophageal voice prosthesis or use other strategies to reduce urgency or even delay the need for in-person visit. Laryngology tele-health in the post-COVID era Even after the risks of COVID-19 have declined sufficiently to allow resumption of more routine clinical practice, telehealth will likely remain a useful tool for providers. We envision that telehealth could be used as prescreening for patients to identify needs and optimize resource utilization. For example, patients with swallowing complaints could be screened prior to coming into clinic for need for an instrumental swallow evaluation (ie, flexible endoscopic evaluation of swallowing or MBSS). If deemed necessary, the instrumental swallow evaluation could be scheduled on the same day as an in-person office visit. This would be especially advantageous for those patients coming from a long distance or in centers where the wait time to schedule MBSS may be prolonged. Determination of who would be best to do this screening and how to be reimbursed for this practice remains unknown. Remote voice and swallow therapy have been performed for several years to increase patient access to care and have been shown to be cost-effective and efficient. , Patients benefit from both cost (eg, gas, toll roads, parking, and public transportation) and time (eg, driving, time off from work) savings with telehealth visits. However, not all patients are appropriate for nor desirous of remote therapy, and some patients may prefer or require hands-on, in-person work. Patients who require manual therapies and certain voice patients with subtle sound variations or who are less stimulable for vocal change are better seen in person. Telemedicine billing considerations Retroactive to January 27, 2020, the federal government has approved certain federal regulatory flexibilities and blanket waivers to expand access to telehealth for physicians. This includes: • Telehealth provided for any reason, even non-COVID-19-related care. • Allowing providers located out of state/territory to provide care to another state's Medicaid enrollees impacted by the emergency. • Temporary waiver of the requirement that physicians and other healthcare professionals be licensed in the state in which they are providing services, so long as they have equivalent licensing in another state. • Waiver of prior authorizations in fee-for-service programs. • Allowing physician or other practitioner to either reduce or waive cost-sharing obligations (ie, coinsurance and deductibles) that a beneficiary may owe for telehealth services furnished. It is important to be aware of potential state-to-state variability in these laws and regulations. For example, in California, a pay parity law requires that if a service is covered in person, that service must be reimbursed at the same rate for a video visit. Individual providers are encouraged to investigate the specific laws in their state. Physician billing may be determined based on visit complexity or by time spent. Criteria to meet various levels of billing for new and return patient video visits are similar to those for in-person visits, as specified by the American Medical Association. Often, limitations of physical exam performance related to telehealth restricts the clinician's ability to bill above level 3. In some situations (eg, discussion of test results with a follow-up patients), time-based billing may be more appropriate ( ). For patients requiring extensive record search and interpretation prior to or after the visit, use of the code 99358 may be appropriate for that time spent in addition to the actual visit. It is important to reference the date of the visit in the note documenting the non-direct service (ie, chart review). If the video visit fails, current guidelines allow for conversion to a telephone visit, which remains billable. If the video visit was >50% completed, then a video visit may be billed, but if <50% of the visit was performed, then this should be considered a telephone visit. Under current guidelines, the telephone visit may be billed as shown in .
Telemedicine video visits are provided through the Zoom for Healthcare platform (Zoom Video Communications, Inc., Version 4.6.10) at the UCSF Voice and Swallowing Center. Other vendors that provide HIPAA-compliant video communication products include Skype for Business, Updox, VSee, Doxy.me and Google G Suite Hangouts meet. In order for these platforms to meet HIPAA-compliance, several general requirements must be in place for video conferencing. These obligations include: (1) ensuring the confidentiality, integrity, and availability of all electronic protected health information the covered entity creates, receives, maintains, or transmits; (2) protecting against any reasonably anticipated threats or hazards to the security or integrity of such information; (3) protecting against any reasonably anticipated uses or disclosures of such information that are not permitted or required under the privacy regulations; and (4) ensuring compliance by its workforce. Specifics on how these requirements are met and maintained are well detailed online within the Zoom for Healthcare HIPAA-Compliance Guide. Institutions and clinician providers should check with the specific vendors that allow video-communication to ensure the above HIPAA-compliance measures are met as well for the specifics on how the requirements are guaranteed. Platforms like Zoom offer virtual “waiting rooms” from which the clinician “admits” the patient; this safety measure protects patient privacy as the clinician controls who is allowed to enter the visit. Multiple people can be admitted simultaneously (eg, interpreters, additional family members, or scribes) as authorized by the clinician. A flowchart to demonstrate our protocol at the UCSF VSC for arranging, preparing for, and conducting telemedicine visits is shown in . Patients are contacted by telephone in advance to confirm their interest, willingness, and ability to participate in a telemedicine video visit. Patients are informed about the benefits and limitations of a telehealth visit. Importantly, the patient is also notified that the video visit is a billable encounter, and thus insurance co-pay may be applied. A previsit check-in is performed by the clinical nursing team 1-3 days before the appointment. Medications, allergies, and personal information are reviewed, and relevant forms (eg, review of system and patient-reported outcome measure [PROM] questionnaires) are completed. All patients at the UCSF VSC complete standard laryngology PROMs (eg, Voice Handicap Index-10, reflux symptom index, cough severity index (CSI), dyspnea index, and eating assessment tool-10 ) prior to their clinic visit. A Zoom for Healthcare link is provided electronically through the secure patient portal system or via email. The clinical team confirms the patient is able to load and use the platform successfully before their scheduled visit. If possible, having the patient complete a “trial run” with the clinical staff team may enable early identification of technical issues and help the patient feel comfortable with the steps needed to connect to the virtual visit. At the start of the video-visit or telephone encounter, patient identity is verified; typically, name and date of birth can be sufficient. For video visits, the patient must connect to the virtual platform on their own accord, either by clicking on a provided link or by accessing a downloaded application. Verbal consent to utilize telemedicine for the encounter is obtained by the provider and documented within the patient's clinical encounter note. If patients have any concerns about HIPAA-compliance or confidentiality, these questions are addressed by the clinician at this time. UCSF institutional policy prohibits recording of the televisit. Patients verbally confirm their understanding of and agreement to abide by this policy, which further protects patient confidentiality. The remainder of the telehealth visit should be conducted in the same manner as an in-person visit. For referring providers, good communication in the referral regarding patient complaint, duration of symptoms, and suspected urgency are especially important during this time for appropriate patient triage. A review of the referral and existing medical history can help clinicians anticipate patient's needs, and identify those who would especially benefit from SLP involvement. Patients with complex laryngologic complaints may be best served by a joint visit with both laryngologist and SLP. This allows a detailed history to be obtained concurrently rather than separately, thereby saving time. A concurrent visit also allows for real-time discussion between the laryngologist and SLP, which facilitates efficiency and treatment decision-making. Virtual joint laryngologist/SLP visits may be possible by logging into one meeting room with the patient. After joint initial patient evaluation, the laryngologist can maximize efficiency while the SLP spends time assessing the patient for therapy candidacy (ie, stimulability , ) by completing other patient-care-related tasks. This may include reviewing previous testing (eg, pulmonary function testing, pH-impedance, manometry, or imaging), working on chart documentation, initiating referrals, or ordering medications or additional testing. After initial evaluation, in the Zoom platform, the patient may be returned to the “waiting room” temporarily so that the laryngologist and SLP can discuss the patient's case privately, as providers might typically do outside of the patient's examination room. Subsequently, the patient may be “brought back in” to the joint visit for discussion of suspected differential diagnosis and recommended next step(s) in evaluation or treatment. New patients present a particular challenge with respect to timing and coordination. History-taking can be time-consuming, especially if complex. A full SLP evaluation including voice stimulability testing, , acoustic evaluation, and/or clinical swallow evaluation requires adequate time allotment. During in-person visits, the laryngologist may go into another exam room with another patient during this time. For virtual visits, this is technically feasible with many telemedicine platforms but again requires appropriate coordination and planning. This may be most easily accomplished by the provider having more than one Zoom account or access to breakout rooms that can be utilized to run “multiple rooms” with multiple patients. Clearly, this time management is more challenging to coordinate during a video visit with multiple concurrent providers. Alternatively, it may be more advantageous for some new patients to be assessed independently by the laryngologist and SLP depending on the specific practice flow, patient complaint or anticipated patient needs. For follow-up visits, when care has already been established, a joint laryngologist/SLP session may be more feasible and efficient. Such situations include patient progress updates, re-evaluation after completion or plateau of voice/swallow therapy, evaluation of medication efficacy (eg, allergy or reflux treatment), review of test results (eg, modified barium swallow [MBSS]), and/or patient counseling. A joint session to determine the role for additional therapy, addition of medications or surgical intervention, or review of trialed devices (such as an Expiratory Muscle Strength Trainer or ProTrach) can be more easily accomplished. The visit duration for follow-up is usually shorter which facilitates schedule coordination.
Despite advances in telehealth, otolaryngologists and SLPs depend highly on the oral/pharyngeal/laryngeal examination for diagnosis and treatment. A limited physical examination is possible via video visit. A cursory view of the oral cavity and oropharynx can be obtained with appropriate intraoral lighting in some cooperative patients. , Perceptual voice analysis with sustained and dynamic vocalization (ie, varying pitch and loudness) tasks in addition to reading standardized passages provide important auditory diagnostic information. Clinical swallowing evaluations are frequently employed as an initial screening tool in dysphagia patients. While these assessments have well-described limitations, direct observation of the patient swallowing various consistencies and subsequently phonating during the tele-visit could provide helpful preliminary information about global swallowing function, voice quality after swallow, and strength of cough. For patients with breathing complaints, quiet respiration at baseline can be observed and rapid breathing exercises may elicit stridor. While a telemedicine model of laryngoscopy-videostroboscopy has been described, where a remotely performed exam is relayed for review through either real-time or cloud-based (store and review) technologies, the reliance on an in-office visit with someone able to perform endoscopy remains. , , During the peak of the pandemic, avoidance of nonurgent in-person visits and aerosol-generating procedures were recommended. , Nonurgent laryngoscopy was therefore deferred. Of course, any urgent laryngoscopy should still be performed as deemed medically necessary. It is imperative that patients are counseled about the diagnostic and treatment limitations in the absence of an endoscopic laryngeal examination. Furthermore, it must be emphasized to patients that such an examination is vital to complete as soon as it is deemed safe to do so. However, proceeding with therapeutic intervention in the absence of the above comprehensive evaluations is nonetheless possible. For example, in suspected laryngitis or vocal fold hemorrhage, a trial of voice rest could be considered, with reassessment for response in 5 to 7 days. Additionally, there are many medical therapies that laryngologists can initiate with acceptable benefit-to-risk ratios without confirmatory laryngoscopy, including proton pump inhibitors, H2 blockers, alginates, steroids, antitussives, mucolytics, and antibiotics, based on appropriate history and suspected diagnosis. Many clinical diagnoses are confirmed or supported by laryngoscopy, but may safely be treated with trials of medications. In these instances, laryngoscopy may be delayed without undue patient risk. For patients with known subglottic stenosis, use of a home peak flow meter to monitor airway symptoms may guide need for in-person visit or surgery. , , Results may be relayed during tele-video visits, and if significant decline in values are noted, a prompt in-office visit should be considered. Again, providers should proceed with caution, carefully monitoring progress and ensuring completion of full evaluation when safely possible. Once clinics resume safe instrumental examination, timely completion should be accomplished.
Multiple steps can be taken on the part of both provider and patient in order to maximize telehealth efficiency and effectiveness. The office should create a virtual workflow for telehealth visits to maximize success prior to starting the visit. This can include a previsit call from office staff to ensure the patient is aware of and knows how to set up the video visit. Technical support or troubleshooting during this call can be helpful. For the provider, we recommend performing the telehealth visit in a quiet room and using headphones to hear the patient clearly. Similarly, it is helpful for the patient to participate in a quiet room using headphones. In this way, both provider and patient can be heard without excessive vocal strain or effort. For certain patients (eg, elderly, hard of hearing, or those with disabilities), having family members or other care providers present for the video visit is anecdotally beneficial. Reliable Internet connection is critical. Proximity to the router or being directly connected to Ethernet can be helpful to avoid disruptions of video or audio feed. A readily accessible informational technology hotline number can help providers and patients troubleshoot technical issues. Flexibility and patience are also important to ensuring success. In the event a patient does not show on time for their visit, a simple phone call can help the provider understand the reason for that delay. This may be due to a forgotten visit time, difficulty accessing the tele-visit platform, or another easily corrected technology problem. If the patient is having difficulty with use of a downloaded application, a simple workaround may be use of the website via the internet (eg, zoom.us) to join the meeting. For the clinician, having your Personal Meeting Identification number readily on hand to give to the patient to input manually on the website can expedite patient connection to the virtual meeting. Patient-reported outcome measures (eg, VHI-10, reflux symptom index, cough severity index, dyspnea index, and eating assessment tool-10 ) can be completed by patients prior to the visit, and submitted by email or directly through the electronic medical record. These results are then available to the providers prior to the visit for review. It is helpful to create a database of resources that can easily be shown during the video visit or sent to the patient via secure link for easy access. These resources serve as virtual replacement for the typical paper informational handouts given in the office. The Zoom for Healthcare platform allows the provider to share their screen. This feature facilitates reviewing results, imaging, previously archived endoscopic examinations, or diagrams for explaining proposed surgeries or procedures. Sound can sometimes be a challenge on a virtual visit. For example, sustained sounds during voice evaluation or voice therapy may be clipped on Zoom. Providers may change settings to allow original sound; if the issue continues to persist, the patient can also be instructed to make these changes ( ).
Some voice and swallowing complaints can be safely managed remotely. However, the majority of patients will ultimately require laryngoscopy for evaluation. If on initial telehealth evaluation, the provider identifies medical complaints needing urgent evaluation, the patient should come in for an urgent office visit. Within laryngology, urgent issues include suspected malignancy, symptomatic airway obstruction, aspiration, and severe dysphagia without alternative nutritional intake means (ie, nasogastric or PEG tube). There are also situations when patients themselves are distressed and would be more reassured by in-person patient evaluation. These cases should be addressed more urgently at the provider's discretion. In-person SLP evaluation with instrumentation including MBSS should be limited to urgent cases only during peak pandemic. This includes patients with no means of nutrition and high risk of aspiration. Alternatively, MBSS can be performed at a facility closer to home and then reviewed by the provider to determine urgency and in-office assessment needs. In cases where urgent MBSS is required and the patient is unknown to the SLP, it is advantageous for the SLP to complete a video visit prior to the MBSS to obtain relevant history and complete a clinical swallow evaluation. Then, the in-person time in the fluoroscopy suite during the MBSS can be utilized most efficiently. For patients with a tracheoesophageal voice prosthesis, in-person visit during peak pandemic may be necessary. However, video visits can be used for troubleshooting, as the patient may be able to apply a plug to the tracheoesophageal voice prosthesis or use other strategies to reduce urgency or even delay the need for in-person visit.
Even after the risks of COVID-19 have declined sufficiently to allow resumption of more routine clinical practice, telehealth will likely remain a useful tool for providers. We envision that telehealth could be used as prescreening for patients to identify needs and optimize resource utilization. For example, patients with swallowing complaints could be screened prior to coming into clinic for need for an instrumental swallow evaluation (ie, flexible endoscopic evaluation of swallowing or MBSS). If deemed necessary, the instrumental swallow evaluation could be scheduled on the same day as an in-person office visit. This would be especially advantageous for those patients coming from a long distance or in centers where the wait time to schedule MBSS may be prolonged. Determination of who would be best to do this screening and how to be reimbursed for this practice remains unknown. Remote voice and swallow therapy have been performed for several years to increase patient access to care and have been shown to be cost-effective and efficient. , Patients benefit from both cost (eg, gas, toll roads, parking, and public transportation) and time (eg, driving, time off from work) savings with telehealth visits. However, not all patients are appropriate for nor desirous of remote therapy, and some patients may prefer or require hands-on, in-person work. Patients who require manual therapies and certain voice patients with subtle sound variations or who are less stimulable for vocal change are better seen in person.
Retroactive to January 27, 2020, the federal government has approved certain federal regulatory flexibilities and blanket waivers to expand access to telehealth for physicians. This includes: • Telehealth provided for any reason, even non-COVID-19-related care. • Allowing providers located out of state/territory to provide care to another state's Medicaid enrollees impacted by the emergency. • Temporary waiver of the requirement that physicians and other healthcare professionals be licensed in the state in which they are providing services, so long as they have equivalent licensing in another state. • Waiver of prior authorizations in fee-for-service programs. • Allowing physician or other practitioner to either reduce or waive cost-sharing obligations (ie, coinsurance and deductibles) that a beneficiary may owe for telehealth services furnished. It is important to be aware of potential state-to-state variability in these laws and regulations. For example, in California, a pay parity law requires that if a service is covered in person, that service must be reimbursed at the same rate for a video visit. Individual providers are encouraged to investigate the specific laws in their state. Physician billing may be determined based on visit complexity or by time spent. Criteria to meet various levels of billing for new and return patient video visits are similar to those for in-person visits, as specified by the American Medical Association. Often, limitations of physical exam performance related to telehealth restricts the clinician's ability to bill above level 3. In some situations (eg, discussion of test results with a follow-up patients), time-based billing may be more appropriate ( ). For patients requiring extensive record search and interpretation prior to or after the visit, use of the code 99358 may be appropriate for that time spent in addition to the actual visit. It is important to reference the date of the visit in the note documenting the non-direct service (ie, chart review). If the video visit fails, current guidelines allow for conversion to a telephone visit, which remains billable. If the video visit was >50% completed, then a video visit may be billed, but if <50% of the visit was performed, then this should be considered a telephone visit. Under current guidelines, the telephone visit may be billed as shown in .
COVID-19 has presented unprecedented challenge to providing clinical care. Telehealth has been quickly adapted into both otolaryngology and subspecialty laryngology care. As with any new endeavor, there are challenges to adaptation; however, with appropriate planning and flexibility, the telehealth model can be optimized to provide high-quality, multidisciplinary, laryngologic care. Telehealth is anticipated to remain an important component of laryngology care in the post-COVID-19 era. Therefore, ongoing refinement of telemedicine techniques should continue to optimize future patient care.
|
Enhanced Cholesterol-Lowering and Antioxidant Activities of Soymilk by Fermentation with | 0f0ba39b-90c3-4959-8129-37af58d05f29 | 10699276 | Microbiology[mh] | Increased dietary cholesterol intake can lead to hypercholesterolemia, defined as high blood cholesterol levels. Hypercholesterolemia is accompanied by an increase in low-density lipoprotein cholesterol and triglycerides and a decrease in high-density lipoprotein cholesterol . However, soybean and soy-based products are widely used worldwide because of their abundance of bioactive compounds, such as proteins or peptides, isoflavones, saponins, and protease inhibitors . Particularly, soy isoflavones, non-steroidal phytoestrogenic flavonoid molecules, have been determined the preventive effects on heart disease as well as metabolic diseases, menopause symptoms, osteoporosis, and some cancers through their hormonal and antioxidant activities [ - ]. Soy isoflavones are present in various forms such as β-glycosides, acetyl-glycosides, malonyl-glycosides, and aglycones. In particular, the bioconversion of isoflavones from glycoside forms to their aglycones, which is performed by β-glucosidase, is important for improving biological activities, as glycoside forms have lower absorption (in the small intestine), bioavailability, and bioactivity than the aglycone forms . Many studies have reported that lactic acid bacteria (LAB) exhibit β-glucosidase activity, and it has been suggested that the fermentation of soybeans using LAB may improve their biological activities for human health [ - ]. LAB have been used as probiotics to reduce reactive oxygen species (ROS) and oxidative stress and to improve associated diseases owing to high antioxidant activity [ - ]. It is also well known that LAB not only inhibits cholesterol synthesis by oxidative stress but can also help reduce cholesterol content in the body by inhibiting cholesterol synthesis through 3-hydroxy-3-methylglutaryl-coenzyme A reductase activity and increased cholesterol excretion . Therefore, this study aimed to investigate the enhancement effect of fermented soymilk with specific Lactobacillaceae strain on cholesterol-lowering and antioxidant activities. Subsequently, the bioactive compounds regarding soy isoflavone were identified.
Bacterial Strains Previously, 10 Lactobacillaceae strains were isolated from Korean kimchi, and their probiotic properties were estimated in a gastrointestinal tract model, which showed great resistance to acid and bile salts and adhesion to the intestine in HT-29 intestinal cells (data not shown). These strains were identified using the V3 and V4 regions of 16S rRNA sequencing ( ). All Lactobacillaceae strains were preserved in de Man, Rogosa and Sharpe (MRS) (Difco, Detroit, USA) broth with 50% (v/v) glycerol as a cryoprotectant at -80°C. The strains were incubated aerobically at 37°C in MRS broth and subcultured thrice before conducting all experiments. Determination of Antioxidant Activity The radical scavenging activities against 2-diphenyl-1-picrylhydrazyl (DPPH) and hydroxyl radicals and their ability to reduce iron ions were measured. The DPPH and hydroxyl radical scavenging activities were estimated according to the method described by Oh et al . and Li et al . , respectively. Reducing power was examined using the ferric reducing antioxidant power (FRAP) assay . Determination of Cholesterol Assimilation The cholesterol-lowering activity was determined using a cholesterol assimilation assay described by Choi and Chang . Briefly, samples were incubated in MRS broth containing 0.5% (w/v) bovine bile (Sigma-Aldrich, USA) and 0.1 g/l cholesterol (Kanto Chemical Co., Inc., Japan). The amount of residual cholesterol was estimated by ortho-phthalaldehyde (OPA) method. Determination of β-Glucosidase Activity The β-glucosidase activity was evaluated based on the hydrolysis rate of p -nitrophenyl β-D-glucopyranoside (p-NPG), following the method of Rekha and Vijayalakshmi , with a slight modification. Briefly, the samples were incubated with p-NPG, and the amount of p-nitrophenol (p-NP) released was determined at 405 nm using a spectrophotometer (Bio-Tek Instruments, USA). One unit (U) of the enzyme activity was defined as the amount of enzyme that released 1 μmol of p-NP from the substrate per milliliter per minute at 37°C under assay conditions. Whole Genome Sequencing and Comparative Genomic Analysis Whole genome sequencing of the selected Lactobacillaceae strain, Lactiplantibacillus plantarum KML06, was performed de novo using a Pacific Biosciences RSII sequencer (PacBio, USA). A 20 kb library was prepared using the PacBio DNA Template Prep Kit 1.0 (Pacific Biosciences, USA). The SMRTbell™ templates were annealed using a PacBio DNA/Polymerase Binding Kit P6. Sequencing analysis was performed by Macrogen, Inc. (Korea). The resulting contigs were scaffolded using Illumina Hiseq-X (Illumina, USA). For comparative genomic analysis, the genomes of four reference strains of other Lactobacillaceae strains ( L. plantarum BLS41 (GCF_002116955.1), L. plantarum LP3 (GCF_002286275.1), L. plantarum B21 (GCF_000931425.2), and L. plantarum pc-26 (GCF_006770485.1)) were obtained from the NCBI database ( https://www.ncbi.nlm.nih.gov/genome/ ) and were used to carry out comparative genomic analysis using pan-genome analysis. Roary v1.007001 was used for pan-genome analysis. A phylogenetic tree was constructed using FastTreeMP v2.1.11, using the core gene alignment results generated by Roary. Dendrograms and Venn diagrams were constructed based on gene content (presence or absence) and pan-genome orthologous groups (POGs), respectively. Fermentation of Soymilk with Selected Lactobacillaceae Strain Soymilk was prepared from a mixture of soybean [ Glycine max (L.) Merr.] and water in a ratio of 1:3 w/w (500 g of soybeans in 1,500 g of water). The mixture was boiled for 10 min and ground for 15 min. Following grinding, the resultant slurry was filtered through a 130 mesh nylon filter cloth and sterilized at 90°C for 15 min. The sterilized soymilk was cooled to 40°C and inoculated with the selected strain of L. plantarum KML06, which was sub-cultured three times, washed twice, and resuspended in saline (0.85% NaCl), approximately 7 log CFU/ml. Fermentation of the mixture was performed at 37°C for 48 h. Fermented soymilk was stored in a deep freezer at -80°C until use. The chemical compositions of non-fermented and fermented soymilk samples are listed in . Determination of Viable Cell Counts, pH, and Degree of Hydrolysis Changes in viable cell counts, pH, and degree of hydrolysis during 48 h of fermentation were determined at 6 h intervals. The proteolytic activity of fermented soymilk was tested using OPA, according to the method described by Church et al . . The OPA reagent was prepared by mixing 25 ml of 0.1 M sodium tetraborate, 2.5 ml of 20%(w/w) sodium dodecyl sulfate, 1 ml of OPA (40 mg OPA/ml of methanol), and 100 μl of β-mercaptoethanol and adding distilled water to a final volume of 50 ml. Fermented soymilk was centrifuged (7,800 × g , 10 min, 4°C) and 180 μl of the OPA reagent was added to 10 μl of the supernatant. The mixture was reacted for 2 min at room temperature and the absorbance was measured at 340 nm using a spectrophotometer. Isoflavone Analysis Identification of isoflavone using HPLC-ESI-MS/MS. HPLC-ESI-MS/MS analysis was performed using 2695 HPLC (Waters Corp., USA) coupled with a Micromass Quattro Micro API benchtop triple quadrupole mass spectrometer (Waters Corp.). The analysis was carried out using ESI in the positive ion mode, and Masslynx™ software v4.1 was used to control the instrument and acquire the data. A Capcell Pak C18 reversed-phase column (250 mm × 4.6 mm id, 5 μm, Shiseido, Japan) was used to separate the isoflavones. The column temperature was maintained at 25°C, and the injection volume of the standard and sample was 20 μl. Gradient systems with 0.1%(v/v) acetic acid in distilled water (solvent A) and 0.1% (v/v) acetic acid in ACN (solvent B) were used. The flow rate of the mobile phase was 0.8 ml/min, and the mobile phase for the HPLC was as following:0–2.5 min 80% A, 2.5–10 min 70% A, 10–20 min 65% A, 20–25 min 60% A, 25–30 min 60% A, 30–32 min 80% A, and 32–42 min 80% A. The source and desolvation temperatures were set to 110 and 250°C, respectively. Argon was used as the collision gas. Quantitative analysis of isoflavone. Six isoflavones (daidzin, genistin, glycitin, daidzein, genistein, and glycitein) were extracted according to the method described by Zhang and Schwartz . Isoflavone β-glycosides and aglycones of non-fermented and fermented soymilk were quantified by 2695 HPLC (Waters Corp.) equipped with a C18 reversed-phase column (250 mm × 4.6 mm id, 5 μm, Shiseido, Japan) and PDA (Waters Corp.), according to the method of Marazza et al . with some modification. Briefly, the mobile phase was 80% A (20%B), maintained for 2.5 min and decreased to 70% A for 7.5 min, and 65% A for 10 min. Subsequently, it was reduced to 60% A for 5 min, held for 5 min, increased to 80% A for 2 min, maintained for 10 min, and the analysis was terminated. The analysis was conducted for 42 min at 250 nm for daidzin and daidzein, or at 260 nm for genistin, genistein, glycitin, and glycitein. Statistical Analysis All experimental data are presented as the mean ± standard deviation (SD) of triplicate measurements. SPSS software (version 25.0; IBM, USA) was used for statistical analysis, and a one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test was used to evaluate the statistical significance between the groups. Statistical significance was set at p < 0.05.
Previously, 10 Lactobacillaceae strains were isolated from Korean kimchi, and their probiotic properties were estimated in a gastrointestinal tract model, which showed great resistance to acid and bile salts and adhesion to the intestine in HT-29 intestinal cells (data not shown). These strains were identified using the V3 and V4 regions of 16S rRNA sequencing ( ). All Lactobacillaceae strains were preserved in de Man, Rogosa and Sharpe (MRS) (Difco, Detroit, USA) broth with 50% (v/v) glycerol as a cryoprotectant at -80°C. The strains were incubated aerobically at 37°C in MRS broth and subcultured thrice before conducting all experiments.
The radical scavenging activities against 2-diphenyl-1-picrylhydrazyl (DPPH) and hydroxyl radicals and their ability to reduce iron ions were measured. The DPPH and hydroxyl radical scavenging activities were estimated according to the method described by Oh et al . and Li et al . , respectively. Reducing power was examined using the ferric reducing antioxidant power (FRAP) assay .
The cholesterol-lowering activity was determined using a cholesterol assimilation assay described by Choi and Chang . Briefly, samples were incubated in MRS broth containing 0.5% (w/v) bovine bile (Sigma-Aldrich, USA) and 0.1 g/l cholesterol (Kanto Chemical Co., Inc., Japan). The amount of residual cholesterol was estimated by ortho-phthalaldehyde (OPA) method.
The β-glucosidase activity was evaluated based on the hydrolysis rate of p -nitrophenyl β-D-glucopyranoside (p-NPG), following the method of Rekha and Vijayalakshmi , with a slight modification. Briefly, the samples were incubated with p-NPG, and the amount of p-nitrophenol (p-NP) released was determined at 405 nm using a spectrophotometer (Bio-Tek Instruments, USA). One unit (U) of the enzyme activity was defined as the amount of enzyme that released 1 μmol of p-NP from the substrate per milliliter per minute at 37°C under assay conditions.
Whole genome sequencing of the selected Lactobacillaceae strain, Lactiplantibacillus plantarum KML06, was performed de novo using a Pacific Biosciences RSII sequencer (PacBio, USA). A 20 kb library was prepared using the PacBio DNA Template Prep Kit 1.0 (Pacific Biosciences, USA). The SMRTbell™ templates were annealed using a PacBio DNA/Polymerase Binding Kit P6. Sequencing analysis was performed by Macrogen, Inc. (Korea). The resulting contigs were scaffolded using Illumina Hiseq-X (Illumina, USA). For comparative genomic analysis, the genomes of four reference strains of other Lactobacillaceae strains ( L. plantarum BLS41 (GCF_002116955.1), L. plantarum LP3 (GCF_002286275.1), L. plantarum B21 (GCF_000931425.2), and L. plantarum pc-26 (GCF_006770485.1)) were obtained from the NCBI database ( https://www.ncbi.nlm.nih.gov/genome/ ) and were used to carry out comparative genomic analysis using pan-genome analysis. Roary v1.007001 was used for pan-genome analysis. A phylogenetic tree was constructed using FastTreeMP v2.1.11, using the core gene alignment results generated by Roary. Dendrograms and Venn diagrams were constructed based on gene content (presence or absence) and pan-genome orthologous groups (POGs), respectively.
Lactobacillaceae Strain Soymilk was prepared from a mixture of soybean [ Glycine max (L.) Merr.] and water in a ratio of 1:3 w/w (500 g of soybeans in 1,500 g of water). The mixture was boiled for 10 min and ground for 15 min. Following grinding, the resultant slurry was filtered through a 130 mesh nylon filter cloth and sterilized at 90°C for 15 min. The sterilized soymilk was cooled to 40°C and inoculated with the selected strain of L. plantarum KML06, which was sub-cultured three times, washed twice, and resuspended in saline (0.85% NaCl), approximately 7 log CFU/ml. Fermentation of the mixture was performed at 37°C for 48 h. Fermented soymilk was stored in a deep freezer at -80°C until use. The chemical compositions of non-fermented and fermented soymilk samples are listed in .
Changes in viable cell counts, pH, and degree of hydrolysis during 48 h of fermentation were determined at 6 h intervals. The proteolytic activity of fermented soymilk was tested using OPA, according to the method described by Church et al . . The OPA reagent was prepared by mixing 25 ml of 0.1 M sodium tetraborate, 2.5 ml of 20%(w/w) sodium dodecyl sulfate, 1 ml of OPA (40 mg OPA/ml of methanol), and 100 μl of β-mercaptoethanol and adding distilled water to a final volume of 50 ml. Fermented soymilk was centrifuged (7,800 × g , 10 min, 4°C) and 180 μl of the OPA reagent was added to 10 μl of the supernatant. The mixture was reacted for 2 min at room temperature and the absorbance was measured at 340 nm using a spectrophotometer.
Identification of isoflavone using HPLC-ESI-MS/MS. HPLC-ESI-MS/MS analysis was performed using 2695 HPLC (Waters Corp., USA) coupled with a Micromass Quattro Micro API benchtop triple quadrupole mass spectrometer (Waters Corp.). The analysis was carried out using ESI in the positive ion mode, and Masslynx™ software v4.1 was used to control the instrument and acquire the data. A Capcell Pak C18 reversed-phase column (250 mm × 4.6 mm id, 5 μm, Shiseido, Japan) was used to separate the isoflavones. The column temperature was maintained at 25°C, and the injection volume of the standard and sample was 20 μl. Gradient systems with 0.1%(v/v) acetic acid in distilled water (solvent A) and 0.1% (v/v) acetic acid in ACN (solvent B) were used. The flow rate of the mobile phase was 0.8 ml/min, and the mobile phase for the HPLC was as following:0–2.5 min 80% A, 2.5–10 min 70% A, 10–20 min 65% A, 20–25 min 60% A, 25–30 min 60% A, 30–32 min 80% A, and 32–42 min 80% A. The source and desolvation temperatures were set to 110 and 250°C, respectively. Argon was used as the collision gas. Quantitative analysis of isoflavone. Six isoflavones (daidzin, genistin, glycitin, daidzein, genistein, and glycitein) were extracted according to the method described by Zhang and Schwartz . Isoflavone β-glycosides and aglycones of non-fermented and fermented soymilk were quantified by 2695 HPLC (Waters Corp.) equipped with a C18 reversed-phase column (250 mm × 4.6 mm id, 5 μm, Shiseido, Japan) and PDA (Waters Corp.), according to the method of Marazza et al . with some modification. Briefly, the mobile phase was 80% A (20%B), maintained for 2.5 min and decreased to 70% A for 7.5 min, and 65% A for 10 min. Subsequently, it was reduced to 60% A for 5 min, held for 5 min, increased to 80% A for 2 min, maintained for 10 min, and the analysis was terminated. The analysis was conducted for 42 min at 250 nm for daidzin and daidzein, or at 260 nm for genistin, genistein, glycitin, and glycitein.
All experimental data are presented as the mean ± standard deviation (SD) of triplicate measurements. SPSS software (version 25.0; IBM, USA) was used for statistical analysis, and a one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test was used to evaluate the statistical significance between the groups. Statistical significance was set at p < 0.05.
Antioxidant, Cholesterol Reduction, and β-Glucosidase Activities of the Lactobacillaceae Strains The antioxidant activities of the ten selected Lactobacillaceae strains were estimated by determining their reducing power and radical scavenging activity ( and ). The DPPH radical scavenging activity and reducing power of KML06 were significantly higher than those of the other strains. Moreover, the cholesterol reduction and β-glucosidase activities of KML06 were the highest among the strains with values of 69.5% and 2.88 U/ml, respectively ( and ). As strain KML06 showed the highest values in all assays, KML06 was selected for the fermentation of soymilk. Genomic Property of Lactiplantibacillus plantarum KML06 Whole genome sequencing and comparative genomic analyses of KML06 were performed to confirm the functionality and novelty of this strain, respectively. General genomic information of KML06 is shown in and the circular contig is shown in . A total of 72,105 reads with an average length of 10,343 bp (745,787,272 total subread bases) were obtained, and the genome contained 3,319,595 bp with a G + C content of 44.44%. Moreover, the KML06 genome consisted of four contigs with N50 values of 3,213,056 bp. The genome of KML06 was composed of 3,077 coding DNA sequences (CDSs), 16 rRNA genes, and 69 tRNA genes. The distribution of the Clusters of Orthologous Group (COG) categories is shown in . The most common COG category was S (unknown function), followed by R (general function prediction only), G (carbohydrate transport and metabolism), K (transcription), E (amino acid transport and metabolism), and M (cell wall/membrane/envelope biogenesis). In addition, the genome of KML06 was compared with those of four different L. plantarum strains pan-genome analysis ( ). The phylogenetic tree constructed based on the ortho-ANI value indicated that KML06 was a novel genomic strain, although it is located close to the reference strains. Growth Kinetics and Changes in pH and Lactic Acid Content To evaluate changes in microbiological properties during soymilk fermentation with KML06, the viable cell count, degree of hydrolysis, pH, and lactic acid content were measured ( ). The number of KML06 cells exponentially increased immediately after soymilk inoculation. The viable cell count increased from 7.26 log CFU/ml to 9.20 log CFU/ml until 12 h of fermentation but decreased slowly until the end of the fermentation, reaching a level similar to that at the beginning of fermentation. Soy proteins were degraded by KML06 and the number of liberated peptides increased during the exponential phase. Similar to the growth of KML06, lactic acid content increased rapidly during the exponential phase and showed a relatively small change after 24 h of fermentation. However, the pH rapidly decreased from 6.15 to 3.90 after 18 hours of fermentation. Changes in Antioxidant and Cholesterol Reduction Activities of Fermented Soymilk during Fermentation Changes in the antioxidant and cholesterol reduction activities of fermented soymilk were measured during fermentation ( ). DPPH and hydroxyl radical scavenging activities and reducing power increased by fermentation of soymilk with KML06. In particular, the hydroxyl radical scavenging activity increased exponentially upto 12 h of fermentation but showed a decreasing trend until the end of fermentation, whereas the DPPH radical scavenging activity and reducing power steadily increased until the end of fermentation. In addition, the cholesterol-reducing ability rapidly increased during 6 h of fermentation, but there was no change until 24 h of fermentation and then decreased until the end of fermentation as a result of hydroxyl radical scavenging activity. Identification and Quantification of Isoflavones Changes in the amounts of the six isoflavone glycosides and aglycones in soymilk during fermentation by KML06 are shown in and . Before soymilk fermentation, the amounts of isoflavone β-glycosides, such as genistin, daidzin, and glycitin, were higher than those of their aglycones, including genistein, daidzein, and glycitein. Among the β-glycosides, the amount of genistin (91.1 μg/ml) was the highest, followed by that of daidzin (52.5 μg/ml). During the first 12 h of fermentation, the amounts of genistin and daidzin rapidly decreased. In contrast, genistein and daidzein levels increased until 12 h of fermentation. Daidzin was not detected from 12 h until the end of fermentation. Additionally, the amount of glycitin steadily decreased until the end of fermentation. In addition, six isoflavones were identified using HPLC-MS/MS. shows the retention times, mass spectral characteristics, and multiple reaction monitoring transitions for each isoflavone. The mass spectra of the isoflavones are shown in .
Lactobacillaceae Strains The antioxidant activities of the ten selected Lactobacillaceae strains were estimated by determining their reducing power and radical scavenging activity ( and ). The DPPH radical scavenging activity and reducing power of KML06 were significantly higher than those of the other strains. Moreover, the cholesterol reduction and β-glucosidase activities of KML06 were the highest among the strains with values of 69.5% and 2.88 U/ml, respectively ( and ). As strain KML06 showed the highest values in all assays, KML06 was selected for the fermentation of soymilk.
Lactiplantibacillus plantarum KML06 Whole genome sequencing and comparative genomic analyses of KML06 were performed to confirm the functionality and novelty of this strain, respectively. General genomic information of KML06 is shown in and the circular contig is shown in . A total of 72,105 reads with an average length of 10,343 bp (745,787,272 total subread bases) were obtained, and the genome contained 3,319,595 bp with a G + C content of 44.44%. Moreover, the KML06 genome consisted of four contigs with N50 values of 3,213,056 bp. The genome of KML06 was composed of 3,077 coding DNA sequences (CDSs), 16 rRNA genes, and 69 tRNA genes. The distribution of the Clusters of Orthologous Group (COG) categories is shown in . The most common COG category was S (unknown function), followed by R (general function prediction only), G (carbohydrate transport and metabolism), K (transcription), E (amino acid transport and metabolism), and M (cell wall/membrane/envelope biogenesis). In addition, the genome of KML06 was compared with those of four different L. plantarum strains pan-genome analysis ( ). The phylogenetic tree constructed based on the ortho-ANI value indicated that KML06 was a novel genomic strain, although it is located close to the reference strains.
To evaluate changes in microbiological properties during soymilk fermentation with KML06, the viable cell count, degree of hydrolysis, pH, and lactic acid content were measured ( ). The number of KML06 cells exponentially increased immediately after soymilk inoculation. The viable cell count increased from 7.26 log CFU/ml to 9.20 log CFU/ml until 12 h of fermentation but decreased slowly until the end of the fermentation, reaching a level similar to that at the beginning of fermentation. Soy proteins were degraded by KML06 and the number of liberated peptides increased during the exponential phase. Similar to the growth of KML06, lactic acid content increased rapidly during the exponential phase and showed a relatively small change after 24 h of fermentation. However, the pH rapidly decreased from 6.15 to 3.90 after 18 hours of fermentation.
Changes in the antioxidant and cholesterol reduction activities of fermented soymilk were measured during fermentation ( ). DPPH and hydroxyl radical scavenging activities and reducing power increased by fermentation of soymilk with KML06. In particular, the hydroxyl radical scavenging activity increased exponentially upto 12 h of fermentation but showed a decreasing trend until the end of fermentation, whereas the DPPH radical scavenging activity and reducing power steadily increased until the end of fermentation. In addition, the cholesterol-reducing ability rapidly increased during 6 h of fermentation, but there was no change until 24 h of fermentation and then decreased until the end of fermentation as a result of hydroxyl radical scavenging activity.
Changes in the amounts of the six isoflavone glycosides and aglycones in soymilk during fermentation by KML06 are shown in and . Before soymilk fermentation, the amounts of isoflavone β-glycosides, such as genistin, daidzin, and glycitin, were higher than those of their aglycones, including genistein, daidzein, and glycitein. Among the β-glycosides, the amount of genistin (91.1 μg/ml) was the highest, followed by that of daidzin (52.5 μg/ml). During the first 12 h of fermentation, the amounts of genistin and daidzin rapidly decreased. In contrast, genistein and daidzein levels increased until 12 h of fermentation. Daidzin was not detected from 12 h until the end of fermentation. Additionally, the amount of glycitin steadily decreased until the end of fermentation. In addition, six isoflavones were identified using HPLC-MS/MS. shows the retention times, mass spectral characteristics, and multiple reaction monitoring transitions for each isoflavone. The mass spectra of the isoflavones are shown in .
Soy isoflavones are important because of their biological activities in the prevention of heart diseases, metabolic diseases, menopause symptoms, osteoporosis, and certain cancers [ - ]. They predominantly exist as glucosides, -glycosides, acetyl-glycosides, malonyl-glycosides, and aglycones . In particular, fermentation enhances the presence of aglycone forms in soy products through the action of bacterial β-glucosidases, which are recognized as a crucial mechanism leading to higher bioavailability of the aglycone forms compared to the glucoside forms . Lim et al . found that daidzein and genistein, the main soy isoflavone aglycones, were produced through soybean fermentation. They also observed that these compounds decreased serum total cholesterol levels and increased high-density lipoprotein cholesterol levels in mice fed a high-cholesterol diet. In addition, dietary soy isoflavones have been reported to have regulatory functions in cholesterol and fatty acid metabolism, and these functions are more effective in isoflavone aglycones than in isoflavone glucosides . In this study, we determined the cholesterol-lowering and antioxidant activities of soymilk fermented with a specific probiotic Lactobacillaceae strain. The probiotic strain L. plantarum KML06 was selected based on its highest antioxidant, cholesterol-lowering, and β-glucosidase activities among the candidate strains isolated from Korean kimchi. Several studies have shown that Lactobacillaceae species as probiotics have protective effects against various diseases through their health benefits , including lowering ROS and oxidative stress, thus alleviating oxidative stress-related diseases through high antioxidant activity [ - ]. It is also well known that probiotic Lactobacillaceae species not only inhibit cholesterol synthesis but also help reduce cholesterol content [ , , ]. A previous study reported that the cell-free supernatant of L. plantarum inhibited cholesterol synthesis through the phosphorylation of AMP protein kinase, which reduced the expression of 3-hydroxy-3-methylglutaryl-coenzyme A reductase . Moreover, supplementation of L. plantarum diet in patients with hypercholesterolemia significantly contributes to a reduction in serum cholesterol . Additionally, soy proteins are hydrolyzed into peptides and free amino acids through fermentation with probiotic strains, resulting in improved preventive effects against various diseases, including antioxidant, antihypertensive, hypocholesterolemic, and anti-diabetic effects on the host [ , , ]. In the present study, soymilk fermented using the probiotic strain KML06 exhibited significantly higher viable cell counts, degree of hydrolysis, antioxidant activities in radical scavenging and FRAP assays, and cholesterol-reducing activity than non-fermented soymilk. In particular, the values were the highest at 12 h of fermentation, and the levels of soy isoflavone aglycones daidzein and genistein, which are the two major aglycones produced by bacterial fermentation , were also detected to be the highest at 12 h of fermentation. These results are consistent with those of previous studies that showed that the bioconversion of glycosides to aglycones tends to increase and then decrease with increasing fermentation time . This could be due to the β-glucosidase produced by KML06, as the viable cell count of KML06 was the highest at 12 h of fermentation. During fermentation, the bioconversion of isoflavones from their glycoside forms to their aglycones is performed by microbial enzymes, including β-glucosidase, which hydrolyzes β-1,4 glycosidic bonds. However, since these enzymes do not exist in humans, microorganisms that represent β-glucosidase activity play an important role in the biotransformation of soy isoflavones in the human body . Isoflavone aglycones are absorbed faster into the mucosa of the small intestine and have higher antioxidant activity than their glycoside forms . In addition, their lipolytic effects have been previously reported in rats fed a high-fat diet . According to previous studies, the estrogenic activities of soy isoflavones are responsible for the lipolytic effects of soy proteins, leading to decreased serum cholesterol and triglyceride levels, thereby reducing the incidence of cardiovascular diseases . The metabolites of dietary isoflavones are believed to modulate hepatic cholesterol 7a-hydroxylase mRNA levels; the mRNA levels of cytochrome p450-2A and phosphoribosyl pyrophosphate synthase-associated protein were upregulated in gerbil liver after the consumption of soy isoflavones . The viable cells of the probiotic KML06 and soy isoflavone aglycones in fermented soymilk might help increase their antioxidant and cholesterol-lowering activities. In conclusion, soymilk fermented with the probiotic strain L. plantarum KML06 as a new functional ingredient could increase the therapeutic potential effect through its antioxidant- and cholesterol-assimilating activities derived from isoflavone aglycones. However, it is important to note that the regulative function of soy isoflavones on lipid metabolism still needs to be determined.
Supplementary data for this paper are available on-line only at http://jmb.or.kr .
|
Tele-neuro-oncology: Current Practices and Future Directions | 7fc161e4-1e8a-4282-bf17-2f80ddddd8a8 | 8773390 | Internal Medicine[mh] | Telemedicine is defined the remote diagnosis, treatment, and management of patients through the use of communication technologies and electronic information . A telemedicine encounter typically occurs using a secure telephone or video conferencing platform when patients and providers are separated by distance. The use of telemedicine has grown over the past decade across multiple medical specialties, initially as a means of addressing geographic and social barriers to healthcare and improving access to subspeciality care . Within medical oncology, telemedicine has been used to facilitate multidisciplinary care, access second opinions, and expand healthcare access for patients with rare tumor types . Telehealth has been utilized for nearly two decades to provide care for neurologic patients and has evolved significantly since its initial use for acute stroke management . Today, the use of telemedicine has expanded across most clinical neurologic subspecialties and has been shown to be feasible and satisfactory for patients and providers alike [ – ]. Prior to the COVID-19 pandemic, limited published data existed regarding the use and experience of telemedicine within the field of neuro-oncology. The COVID-19 pandemic beginning in 2020 resulted in rapid expansion of telehealth across the globe. In this new era of physical distancing, telemedicine has proven to be a useful tool for providing high-quality clinical care while limiting patients’ exposure and physical touchpoints within a healthcare system . This review article will address the current state of telehealth within neuro-oncology and its expansion during the COVID-19 pandemic. The delivery of telehealth and clinical considerations of tele-neuro-oncologic visits will be discussed in addition to the benefits, limitations, and opportunities for growth of this technology.
Tele-neuro-oncologic care should be delivered over a secure and HIPPA-compliant telephone connection or video conferencing platform. Virtual visits are treated with the same degree of professionalism, security, and privacy as in-person evaluations. Providing clear instructions and training to clinical and administrative staff and patients is imperative to ensure a successful telehealth visit. For patients with cognitive or visual impairment, the addition of a patient support person should be considered. All telemedicine encounters should begin with obtaining patient consent for the visit. For children with brain tumors, a caregiver or legal representative should provide consent and be present for the duration of the visit. All telemedicine encounters should be documented within the appropriate patient record. Video encounters are considered the gold standard and recommended for direct audio-visual communication and patient evaluation. Video encounters have several advantages over telephone visits when evaluating a patient with neuro-oncologic disease. A neurologic exam can be feasibly and accurately completed during a video visit, with some limitations such as examination of reflexes or sensation. Skill and mastery of the virtual neurologic examination by clinicians are necessary [ •]. Systemic findings such as oral candidiasis, lower extremity edema, and skin lesions can also successfully be assessed using a video encounter . A video encounter also allows for the clinician to share their screen to review neuroimaging and laboratory results with patients [ ••]. Importantly, during a video encounter, providers are able to witness patients' function, move, and ambulate within their home environments. Virtual consent for cancer therapies, genetic testing, or clinical trials may also be obtained during the video encounter and should be clearly recorded in the patient’s medical record . Policies regarding obtaining virtual consent for treatment or research vary by institution, and local guidance should be followed. Telephone encounters are also used in telemedicine and may be appropriate for certain patients when visual evaluation and neurologic examination are not absolutely necessary or when technologic or connectivity issues preclude the use of video conferencing. Video encounters have been shown to be barriers to care for older patients and those with limited Internet accesss . Telephone encounters may be suitable for stable patients such as long-term brain tumor survivors, those with low-grade or benign lesions, stable neurologic complications of cancer (such as chemotherapy induced neuropathy), or those with mild symptoms. Telephone encounters may also be used for urgent evaluation if a video or in-person visit is not feasible in a timely manner .
The use of telemedicine has many benefits for neuro-oncologic patients. First and foremost, telehealth allows for patients to be physically and geographically distanced from their providers. Often this results in providing care within a patient’s own home. Many neuro-oncology centers are located within urban areas that may be difficult for patients to access due to distance and resources. Telemedicine removes the need for long-distance travel for some patients which improves access to specialized neuro-oncologic care, particularly for patients in regions without established neuro-oncology providers. This also creates opportunities for expedited second opinion consultations, which may occur at centers outside of a patient’s home city or state and are not always feasible with traditional in-person visits. Additionally, many patients with neuro-oncologic diseases have cognitive or neurologic symptoms that limit their ability to travel for in-person clinical care. Patients at high risk of treatment toxicities, such as older patients, may benefit from increased frequency of telemedicine visits for clinical assessment and toxicity monitoring . For at-risk patient populations, longitudinal monitoring using telehealth may allow for improved symptom management and earlier detection of toxicities or disease progression. Additionally, telemedicine visits can often be coordinated more quickly than in-person visits, allowing for expedited evaluation and management of new or progressive symptoms. Another advantage of virtual visits is the ability to have multiple members of a patient’s family or support network be present during the visit. This can be particularly useful for patients with memory and cognitive deficits, where information obtained from the patient may be limited. Additionally, the presence of key caregivers ensures accurate information is received from the clinician, communication is streamlined, and that concerns from all parties are appropriately addressed. This enhanced communication helps to build trust with the medical team and allows for expedited evaluation of patient and family concerns. Telemedicine in medical oncology has been successful in expanding access to additional services including genetic counselling, survivorship care, palliative care, and support services . Virtual visits allow for multiple specialty consultations, often in a timelier fashion than in-person visits, due to decreased constraints by clinic scheduling and availability. Enhanced services provided via telemedicine may also include speech therapy, physical therapy, and integrated and lifestyle counselling. Additionally, providers are able to share and review images and documents with patients in real time during the visit which may improve communication and enhance patient education. This is particularly useful when completing patient education regarding medications and cancer therapies and when reviewing consent for clinical trials.
While telemedicine has many advantages, there are limitations and potential barriers to more widespread use and implementation. The primary barriers to telehealth are technology specific including issues with connectivity, communication devices, and the technologic proficiency of users . Telehealth requires training, technological skill, and proficiency in the use of virtual conferencing platforms for medical staff, clinicians, and patient alike. Appropriate resources, training, and familiarity with workflows are necessary for clinical staff to properly create and execute virtual encounters. Additionally, staff and clinicians should be trained on technical troubleshooting for common connectivity issues to assist patients in real time during a telemedicine visit. Virtual conferencing platforms should be secure, easy to use, cost-effective, and integrated with the electronic medical record . Patients should receive clear instructions and guidelines on the appropriate use of virtual platforms. Patients with lower technology literacy such as some older patients, those with language barriers, lower levels of education, and cognitive impairment may require additional support prior to a telemedicine visit . Equitable access and distribution of telehealth services must also be considered. Telehealth services may not be accessible for patients who lack access to affordable or reliable broadband Internet or communication devices . Some clinicians and practices may not have the resources or infrastructure to support costly audiovisual equipment, computer software, and staff required for telehealth. Additionally, some institutions may have internal policies that do not support or encourage the use of telemedicine. During the COVID-19 pandemic, reimbursement for telemedicine service was supported by Medicare, Medicaid, and most health insurance companies. This coverage extended to both video and telephone encounters, although reimbursement for telephone encounters has been limited . As the public health emergency evolves, coverage and reimbursement for telemedicine is expected to change. Governmental policies supporting the continued use and expansion of telemedicine services will play a key role in determining the future of telehealth. Competitive rates for telemedicine visits are needed for flexibility and sustainability of healthcare systems . The integration of telehealth services within neuro-oncology will also depend on local, institutional, and state guidelines. A frequent criticism of telemedicine is the impact it may have on the patient-clinician relationship. Studies across multiple specialties have demonstrated that telehealth visits are at least equivalent for patient and provider satisfaction when compared to in-person visits [ – ]. Additional care should be taken by providers having difficult conversations or providing bad news. Lastly, not all aspects of a comprehensive neurologic or physical exam can be adequately completed over video. Patients with new or unexplained symptoms which require a detailed examination or diagnostic examination maneuvers that cannot be completed via video should be evaluated in-person.
Telemedicine presents an opportunity to evolve the way in which clinical and supportive care, research including clinical trials, and education are delivered. Multidisciplinary care is paramount for neuro-oncologic patients, often involving providers from neurosurgery, radiation oncology, medical oncology, neurologic subspecialties, and palliative care. Advanced practice providers, nurses, social workers, and support staff are also key members of the patient care team. As oncologic patients become more complex, care teams will grow in number and specialty and may include palliative care, genetics, psychiatry, or mental health providers and others. The use of an online video conferencing platform provides the opportunity for simultaneous or sequential care by multiple clinicians and members of the care team. This allows for streamlined multidisciplinary care in real time without the constraints of provider schedules, physical clinic space, or geography. Multidisciplinary virtual visits provide patients the ability to have a robust discussion about their care, limit the cost and resources required for multiple in-person specialist appointments, and help to build trust and consensus. From a provider standpoint, multidisciplinary visits allow for real-time discussion and development of a treatment plan with colleagues, ideally leading to more expedited patient care and management . These multidisciplinary visits may be of particular importance for advanced care planning, complex patients, those with rare tumor types and patients near the end of life. Patient and provider education is another opportunity for telemedicine. Telemedicine has been successfully used to provide patient education and peer support for patients with various medical conditions [ – ]. The use of a virtual format is conducive for broad brain tumor–specific education of multiple patients simultaneously from geographically disparate locations. Facilitated group sessions also allow for patients and caregivers with similar diagnoses to learn from and support each other, fostering a sense of community. These patient-centered forums present a unique opportunity to highlight specific programs, ongoing research, and clinical trials. Additionally, multidisciplinary discussion may be provided in a virtual tumor board setting which allows multiple providers to discuss the diagnosis, treatment, and management of patients with brain tumors, regardless of their geographic location . Tumor boards are an essential component of comprehensive oncologic care and require the presence and input of clinicians from multiple specialties . The virtual format allows for the expansion of the local tumor board and may include regional or national experts and may lead to improved attendance and timely discussion by clinicians due to reduced travel time . Virtual tumors boards allow for bypassing of geographic barriers, which may access to neuro-oncologic care, and for collaboration with other neuro-oncologic providers in the region. Addition of other providers may also raise awareness for new technologies, treatments, and clinical trials. Integration of telehealth within clinical trials represents a significant opportunity for the advancement of neuro-oncologic research. In a field where treatment options are limited and outcomes for most primary brain tumors remain poor, enrollment of patients on clinical trials is paramount . In addition to patient- and tumor-specific factors, socioeconomic factors such as employment status, level of education, and treatment facility location have been shown to influence enrollment onto glioblastoma clinical trials . Additionally, older patients with brain tumors have been historically underrepresented on interventional clinical trials [ •]. Currently, most primary brain tumor clinical trials are located at major brain tumor programs in urban centers. The geographic location of these centers alone may prohibit study enrollment for rural and low socioeconomic status patients and those unable to drive due to neurologic deficits [ •]. Telemedicine provides an opportunity for patients to receive information and screening for clinical trials at brain tumor programs around the country. This tool also allows centers to potentially increase enrollment on studies as the number of potential subjects increases. Some study visits may be able to be safely completed via telemedicine, decreasing the burden on patients and potentially decreasing study cost. The development of appropriate criteria and clinical assessments for virtual study encounters are needed for telemedicine to further develop within the field.
The COVID-19 pandemic has catalyzed the use of telemedicine which has significant implications for the care of neuro-oncologic patients. Virtual clinical, specialized, and multidisciplinary care can feasibly be provided for this patient population. Through video conferencing, providers can complete a neurologic examination, evaluate treatment toxicities, review and share imaging to assess treatment response and provide patient education and supportive care interventions, while involving additional members of the patient’s family and care team. The use of telemedicine is expected to grow as patients and providers become more accustomed to this technology and provides an opportunity to increase access to subspecialty neuro-oncologic care. Clinical trials represent a major utilization for telehealth as a means of decreasing the burden associated with travelling for clinical trials while improving access to research and potentially impacting enrollment on studies. Thoughtful consideration and support from clinicians, study sponsors, regulatory agencies, and payers is needed for virtual neuro-oncologic care to evolve to fit the needs of patients and the field.
|
The rhizosphere Microbiome of | bcb01b4b-0554-4b18-980f-ec9179eba18f | 9862814 | Microbiology[mh] | Malus sieversii (Ldb.) Roem. was an important germplasm resource in China and is the main ancestral origin of cultivated apples . Belonging to Rosaceae ( Malus Mill.), it is a Tertiary relict that has been included in the National Second-Class Protected Plant List and the China Plant Red Book . This species was mainly distributed in the Tianshan Mountains across Central Asia. In China, it was found mostly in Xinyuan, Yining, Huocheng, and Gongliu counties in the Yili region of Xinjiang, as well as Emin and Tuoli counties in the Tacheng region . The climatic environment of these regions is complex. Xinyuan, Gongliu, and Huocheng counties belonged to the trans-temperate continental and alpine climate. The Tacheng region, on the other hand, was in the mid-temperate arid and semiarid climate zones. In recent years, the pressure on the ecosystem was increasing due to soil nutrient loss, human grazing activities, climate change, and outbreaks of pest and disease . This led to the ecological fragility of the wild apple forestland, the loss of its original adaptive capacity, its large-scale degradation and even mortality. Bacteria, archaea and eukaryotes in the rhizosphere microbial community were essential bioactive components of forest ecosystems . It was found that maintaining the dynamic balance and diversity of the rhizosphere microbial community structure will helped to protect the co-ordination mechanisms of the whole ecosystem and buffered negative impacts . Philippot et al. confirmed that the rate of the denitrification process decreases with the loss of soil microorganisms . Many studies have shown that ecosystem processes are determined by the balance among species identity, community composition and species richness . In addition, environmental factors that affect microbial community diversity and composition also affect ecosystem processes and functions . Therefore, understanding the response patterns of wild apple rhizosphere microorganisms in different regions to complex environmental changes from the aspects of community composition, functional groups, and ecosystems is not only helpful for the protection of wild apple germplasm resources, but also can provide the theoretical basis for optimizing rhizosphere ecological environmental conditions of wild apple growth and decreasing the occurrence of soil-borne diseases. There were many studies on the apple rhizosphere microbial community that mainly focus on biological control. Jiang et al. analyzed the rhizosphere microbial communities of apple trees around the Bohai Sea, and used high-throughput sequencing technology to explore the rhizosphere microbial communities of perennial and replanted apple trees in the Bohai Bay area, including bacteria and fungi . The microbiota of the Lebanese wild apple ( Malus trilobata ) is a rich source of potential biocontrol agents for postharvest fungal pathogens in apples. Elie et al. explored the microbiota of wild apples ( M. trilobata ) as a potential source of two novel biocontrol agents for postharvest ( Botrytis cinerea and Penicillium expansum ) affecting commercial apples . However, there are relatively fewer studies on the community structural analysis of rhizosphere bacteria and eukaryotes of wild apples in Xinjiang and the protection of germplasm resources. In this study, 16S/18S rDNA analysis were conducted on the rhizosphere soil of wild apples in eight regions of Yili, Xinjiang, China. Further, the interaction of microbial community composition and diversity in the wild apple rhizosphere with geographic and environmental distance was investigated by mantel analysis. The effects of geographic and environmental gradients on microorganisms in the rhizosphere of wild apples were also explored. In this paper, it was hypothesized that geographic and environmental gradients would influence the diversity of microbial communities in the wild apple rhizosphere. To validate this, it was investigated using mantel analysis. In summary, this study offers theoretical guidance for the sustainable management and ecological construction of wild apple forests China.
Site description and sampling Eight sampling sites (Daxigou in Huocheng County, Wild Fruit Forest in Xinyuan County, Resource Nursery, Damorhu in Gongliu County, Xiaomorhu in Gongliu County, and Gongliu County Nazi Work Team, Laofengkou Guozigou in Tuoli County, and Yeguolin Scenic Area in Emin County) were located in the Yili River Valley and Tacheng region of Xinjiang, China (Fig. ). The climatic environment of the sample plot was relatively complex. The Ili River Valley has a temperate continental climate, whereas the Tacheng area is in the mid-temperate arid and semiarid regions. Some areas are inhabited by herdsmen year-round, affecting the wild fruit forest due to grazing to varying degrees. Wild apple rhizosphere soil sampling was carried out in May. Within the sampling area, apple trees with a diameter at breast height of 90–100 cm were randomly selected as sample trees. The topsoil was removed with a sterilized spade. The fibrous roots, together with the lateral roots in the 10–20 cm soil layer, were dig out with the soil around them and placed in sterile bags. Samples were collected in three replicates, with multiple points sampled for each replicate. Stones and other debris were picked out and only the 2–3 mm of rhizosphere soil attached to the roots was collected and equal amounts of soil samples were mixed evenly. After collecting the samples, they were divided into two. One was immediately stored in liquid nitrogen for DNA extraction, and the other was used for the determination of the physical and chemical properties of the soil. Soil nutrient analysis Soil moisture content was measured by the drying method. An acidity meter was used to measure soil pH (NYT 1121.2–2006). Soil AP (Available Phosphorus) content was determined by molybdenum-antimony resistance colorimetry (NY/T1121.7–2014). AN (Available Nitrogen) was determined by the diffusion method (DB13/T 843–2007). Soil AK (Available Potassium) was determined by ammonium acetate extraction (NYT 889–2004). The potassium dichromate external heating method (NYT 1121.6–2006) was used to measure soil organic matter (OM). Soil catalase (EA) was measured. Temp (Temperature) and RH(Relative Humidity) were measured with a hygrothermograph (TA621A). DNA extraction and high-throughput sequencing Total genomic DNA of the soil samples was extracted using the CTAB method. DNA concentration was determined by NanoDrop. Purity and completeness were evaluated using 1% agarose gel electrophoresis. DNA was diluted to 1 ng/μl with sterile water. The 16S/18S rDNA genes were amplified using the specific primers with barcodes. The V3 + V4 of 16S and V4 region of 18S were selected amplification and sequencing. The primers used for 16S rDNA gene were 341F (CCTAYGGGRBGCASCAG) and 806R (GGACTCANNGGGTATCTAAT) . The primers for 18S rDNA gene were 528F (GCGGTAATTCCAGCTCCAA) and 706R (AATCCRAGAATTTCACCTCT) . PCR was performed in a 30 μL reaction with 15 μL Master Mix (New England Biolabs), 0.2 μM each of forward and reverse primers and 10 ng of template DNA. The thermocycling conditions were as follows: predegeneration at 98℃ for 1 min, 30 cycles of denaturation at 98 °C for 10 s, annealing at 50 °C for 30 s, and extension at 72℃ for 30 s, followed by a final extension at 72℃ for 5 min. Electrophoretic detection was performed on a 2% agarose gel. Samples with bright main band between 400 and 450 bp were selected for subsequent experiments. PCR products were pooled equally and purified using Kit (Tiangen Biotech). The purified products were used to prepare the library. Sequencing libraries were generated using TIANSeq Fast DNA Library Prep Kit (Tiangen Biotech). Library quality was assessed on a Qubit 2.0 fluorometer (Thermo Scientific) and an Agilent 2100 Bioanalyzer. Finally, the libraries were sequenced on the Illumina platform using a 2 × 250 bp paired-end protocol. Amplicon data processing Bioinformatics analysis of high-throughput sequencing data was performed using QIIME 2, with slight modifications according to the official tutorials. Briefly, the raw sequence data were demultiplexed using the demux plugin. Primers cutting was then performed with the cutadapt plugin . Sequences were quality-filtered, denoised, merged, and chimera-removed using the DADA2 plugin . This was followed by species annotation. Statistical analysis Α-diversity analysis was used to reflect the complexity of the species diversity for the sample by observed-species, Chao1, Shannon, Simpson, ACE, and Good-coverage indices. All indices were calculated by QIIME 2 and displayed with R (version 3.6.2) . β-Diversity analysis was used to assess sample differences in species complexity. β-diversity was calculated for weighted and unweighted Unifrac by QIIME 2. Principal component analysis (PCoA) was performed prior to cluster analysis. The dimension of the original variables was decreased using the “statpackage” and “ggbiplot” packages. The three-dimensional PCoA results were displayed using QIIME 2, while the two-dimensional results were displayed using the “ade” and “ggplot2” packages in R, Metastats and STAMP were utilized to confirm differences in the abundance of individual taxonomic or functional annotations between the groups. Linear discriminate analysis (LDA) effect size (LEfSe) was used the potentially enriched microbial lineages with significant difference ( P < 0.05) between different groups at various taxonomic levels with LDA threshold set ≥ 4. Analysis of variance (ANOVA) was performed the differences between the two groups of microbial communities based on the Bray–Curtis dissimilarity distance matrix . In addition, the correlations between un Weighted or Weighted UniFrac distance matrices and the spatial distance matrices were measured by using partial Mantel test in R.
Eight sampling sites (Daxigou in Huocheng County, Wild Fruit Forest in Xinyuan County, Resource Nursery, Damorhu in Gongliu County, Xiaomorhu in Gongliu County, and Gongliu County Nazi Work Team, Laofengkou Guozigou in Tuoli County, and Yeguolin Scenic Area in Emin County) were located in the Yili River Valley and Tacheng region of Xinjiang, China (Fig. ). The climatic environment of the sample plot was relatively complex. The Ili River Valley has a temperate continental climate, whereas the Tacheng area is in the mid-temperate arid and semiarid regions. Some areas are inhabited by herdsmen year-round, affecting the wild fruit forest due to grazing to varying degrees. Wild apple rhizosphere soil sampling was carried out in May. Within the sampling area, apple trees with a diameter at breast height of 90–100 cm were randomly selected as sample trees. The topsoil was removed with a sterilized spade. The fibrous roots, together with the lateral roots in the 10–20 cm soil layer, were dig out with the soil around them and placed in sterile bags. Samples were collected in three replicates, with multiple points sampled for each replicate. Stones and other debris were picked out and only the 2–3 mm of rhizosphere soil attached to the roots was collected and equal amounts of soil samples were mixed evenly. After collecting the samples, they were divided into two. One was immediately stored in liquid nitrogen for DNA extraction, and the other was used for the determination of the physical and chemical properties of the soil.
Soil moisture content was measured by the drying method. An acidity meter was used to measure soil pH (NYT 1121.2–2006). Soil AP (Available Phosphorus) content was determined by molybdenum-antimony resistance colorimetry (NY/T1121.7–2014). AN (Available Nitrogen) was determined by the diffusion method (DB13/T 843–2007). Soil AK (Available Potassium) was determined by ammonium acetate extraction (NYT 889–2004). The potassium dichromate external heating method (NYT 1121.6–2006) was used to measure soil organic matter (OM). Soil catalase (EA) was measured. Temp (Temperature) and RH(Relative Humidity) were measured with a hygrothermograph (TA621A).
Total genomic DNA of the soil samples was extracted using the CTAB method. DNA concentration was determined by NanoDrop. Purity and completeness were evaluated using 1% agarose gel electrophoresis. DNA was diluted to 1 ng/μl with sterile water. The 16S/18S rDNA genes were amplified using the specific primers with barcodes. The V3 + V4 of 16S and V4 region of 18S were selected amplification and sequencing. The primers used for 16S rDNA gene were 341F (CCTAYGGGRBGCASCAG) and 806R (GGACTCANNGGGTATCTAAT) . The primers for 18S rDNA gene were 528F (GCGGTAATTCCAGCTCCAA) and 706R (AATCCRAGAATTTCACCTCT) . PCR was performed in a 30 μL reaction with 15 μL Master Mix (New England Biolabs), 0.2 μM each of forward and reverse primers and 10 ng of template DNA. The thermocycling conditions were as follows: predegeneration at 98℃ for 1 min, 30 cycles of denaturation at 98 °C for 10 s, annealing at 50 °C for 30 s, and extension at 72℃ for 30 s, followed by a final extension at 72℃ for 5 min. Electrophoretic detection was performed on a 2% agarose gel. Samples with bright main band between 400 and 450 bp were selected for subsequent experiments. PCR products were pooled equally and purified using Kit (Tiangen Biotech). The purified products were used to prepare the library. Sequencing libraries were generated using TIANSeq Fast DNA Library Prep Kit (Tiangen Biotech). Library quality was assessed on a Qubit 2.0 fluorometer (Thermo Scientific) and an Agilent 2100 Bioanalyzer. Finally, the libraries were sequenced on the Illumina platform using a 2 × 250 bp paired-end protocol.
Bioinformatics analysis of high-throughput sequencing data was performed using QIIME 2, with slight modifications according to the official tutorials. Briefly, the raw sequence data were demultiplexed using the demux plugin. Primers cutting was then performed with the cutadapt plugin . Sequences were quality-filtered, denoised, merged, and chimera-removed using the DADA2 plugin . This was followed by species annotation.
Α-diversity analysis was used to reflect the complexity of the species diversity for the sample by observed-species, Chao1, Shannon, Simpson, ACE, and Good-coverage indices. All indices were calculated by QIIME 2 and displayed with R (version 3.6.2) . β-Diversity analysis was used to assess sample differences in species complexity. β-diversity was calculated for weighted and unweighted Unifrac by QIIME 2. Principal component analysis (PCoA) was performed prior to cluster analysis. The dimension of the original variables was decreased using the “statpackage” and “ggbiplot” packages. The three-dimensional PCoA results were displayed using QIIME 2, while the two-dimensional results were displayed using the “ade” and “ggplot2” packages in R, Metastats and STAMP were utilized to confirm differences in the abundance of individual taxonomic or functional annotations between the groups. Linear discriminate analysis (LDA) effect size (LEfSe) was used the potentially enriched microbial lineages with significant difference ( P < 0.05) between different groups at various taxonomic levels with LDA threshold set ≥ 4. Analysis of variance (ANOVA) was performed the differences between the two groups of microbial communities based on the Bray–Curtis dissimilarity distance matrix . In addition, the correlations between un Weighted or Weighted UniFrac distance matrices and the spatial distance matrices were measured by using partial Mantel test in R.
Diversity of 16S/18S rDNA genes in different regions 16S/18S rDNA genes sequencing was used to study the population characteristics of rhizosphere bacteria and eukaryotes in wild apples from different regions of Xinjiang. After filtering, denoising, and removing chimeras, the average number of bacterial sequences per sample was 54,992, with an effective rate of 79.26%. The average number of eukaryotic sequences was 108,548, with an effective rate of 84.79%. The bacterial sequences were clustered into 15,789 operational taxonomic units (OTUs), and eukaryotic sequences were clustered into 16,853 OTUs. The coverage of all samples was > 0.99, and with the rarefaction and Shannon curves converged. This indicates that sampling was reasonable, and that the sequencing was sufficient to characterize the diversity of the study sites. A plot shows (Fig. ) that there were 47 core OTUs of bacteria and 184 core OTUs of eukaryotes, respectively, in wild apples rhizosphere soil samples from eight regions in Xinjiang. The number of bacterial OTUs in regions C, D, G, and H was greater than that of eukaryotic OTUs, whereas the number of bacteria in regions A, B, E, and F was lesser than that of eukaryotic OTUs. The results showed that the number of OTUs in each soil sample varied across the eight regions. The bacterial Chao 1, ACE, Shannon and Simpson indices in regions C, D, G, and H were higher than the eukaryotic biodiversity indices, whereas the bacterial diversity indices in regions A, B, E, and F were lower than the eukaryotic biodiversity indices (Fig. ). The α-diversity analysis showed that species diversity in different regions was quite different. The bacterial species diversity in regions C, D, G, and H was higher than in other regions whereas regions A, B, E, and F had higher species diversity than in other regions. PCoA was used to rank at OTU level to reveal similarities or differences in community structure between different regional groups. The first and second axes of the bacterial community structure contributed 41.3% and 13.4% of the explanation, respectively, and 16.4% and 13.7% for eukaryotes, respectively. (Fig. a,b). In general, the majority of samples from each group clustered together, indicating significant differences in the community composition of bacterial and eukaryotic species. Rhizosphere bacteria species composition in different regions Compared to the Silva138 database, the proportion of phylum, class, order, family, genus, and species in bacteria was 97.78%, 97.24%, 95.8%, 94.24%, 84.63%, and 32.39% respectively. The 16S rRNA gene sequences were divided into phyla, and the nine most abundant phyla were Firmicutes, Proteobacteria, Actinobacteriota, Bacteroidota, Acidobacteriota, Verrucomicrobiota, Chloroflexi, Planctomycetota, and Methylomirabilota, unclassified (Fig. a). Firmicutes were the most dominant bacterial phylum in regions of A (56.47%), C (24.46%), D (32.45%), E (57.58%), and F (56.43%), and Proteobacteria were the most dominant bacterial phylum in regions B (40.82%), G (54.70%), and H (46.23%). The community composition in regions A, E, and F was less; in particular, the relative abundance of Acidobacteriota, Verrucomicrobiota, Planctomycetota, Chloroflexi, and Methylomirabilot was very low. The relationships between the sequential datasets were visualized by non-metric multidimensional scale (NMDS), and the samples in each region were basically clustered together, but the distance between groups was large, with a P ≤ 0.001 (Fig. b), indicating differences in the composition of bacterial communities in different regions. Linear discriminant analysis effect size (LEfSe) analysis was used to identify microorganisms explicitly enriched in bacteria from the phyla-to-species level in different regions. The results showed that there were 41 phylum-to-species differences in the eight regions. In the bacterial species community, the various species in region A were Bifidobacteriaceae, Actinobacteria, and Coriobacteriales. The different communities in region B were Bacteroidaceae and uncultured_bacteria. Chloroflexi, Bacillus _ luciferensis , and uncultured bacteria were the different species in region C. The main communities in region D included Vicinamibacteraceae, Blastocatellia, and Pyrinomonadaceae. There were no apparent species differences in regions E and F, but uncultured_bacteria were different in region G. Prevotella and Bacteroidia were different in region H (Fig. c). Figure d shows that the rhizosphere bacterial family of wild apples co-occurred. The results showed that the rhizosphere bacterial communities of wild apples in different regions cooperated, and a few competed. Rhizosphere eukaryotic species composition in different regions The proportion of the eukaryotic phylum level was 83.17%, that of the class level was 75.23%, that of the order-level was 69.39%, that of the family level was 61.79%, that of the genus level was 53.09%, and that of the species-level was 30.05%. The 18S rRNA gene sequences were divided into phylum levels, among which the nine most abundant phyla were Ascomycota, Phragmoplastophyta, Basidiomycota, Cercozoa, Ochrophyta, Ciliophora, Mucoromycota, Chytridiomycota, and Chlorophyta, unclassified (Fig. a). Eukaryotes were dominated by Ascomycota, Phragmoplastophyta, and Basidiomycota. Ascomycota was the most dominant phylum in regions A (44.71%), B (23.23%), D (38.59%), E (39.98%), F (29.58%), G (37.94%) and H (42.70%), and Phragmoplastophyta was the most dominant phylum in region C (40.42%) regions. NMDS analysis showed that the samples in region C were far apart, and the samples in each area were clustered together, with a P ≤ 0.001 (Fig. b), indicating differences in the eukaryotic community composition in different regions. LEfSe analysis showed significant differences in the leading 41 phyla-to-species in the eight different regions. In the eukaryotic community, the main differential species in region A were Trichoderma , Nectriaceae, and Phallus_hadriani . The differential species in region B was Cercomonadidae, and there was no apparent difference in fungal species. In region C, there was also no significant difference in species. Pleosporales and Dothideomycetes were the differential species in region D. Solicoccozyma and Piskurozymaceae were the differential species in region E. The differential species in region F included Filobasidiales. Helotiales, Incertaesedis, Leotiomycetes, and Geminibasidium were the differential species in region G (Fig. c). Figure d shows that the rhizosphere eukaryotic community of wild apples co-occured, and the network structure was different from that of bacterial. The results showed that the rhizosphere eukaryotic communities of wild apples in different regions cooperated with each other, and a few competed. Relationship between β-diversity and environmental factors The impact of geographic distance, climatic distance and soil pH on the composition of rhizosphere bacterial and eukaryotic communities in wild apples was examined by mantel analysis. It was found that the correlation between climatic distance and β-diversity was greater than that between geographic distance and soil pH (Fig. a). Geographic and climatic distances correlate were more strongly correlated with bacterial β-diversity than with eukaryotic β-diversity. Geographic distance remained positively correlated with eukaryotic β-diversity even after controlling for climatic distance and/or soil pH (Fig. b). The results indicated that geographic and climatic differences were important predictors for microbial β-diversity. Soil pH and climatic distance were positively with bacterial β-diversity and insignificantly with eukaryotic β-diversity when geographic distance was controlled (Fig. b). Climate distance was negatively with microbial β-diversity when the soil pH was controlled. The converse was also true. In summary, there was a correlation between geographical distance, climatic distance and soil pH with microbial β-diversity.
16S/18S rDNA genes sequencing was used to study the population characteristics of rhizosphere bacteria and eukaryotes in wild apples from different regions of Xinjiang. After filtering, denoising, and removing chimeras, the average number of bacterial sequences per sample was 54,992, with an effective rate of 79.26%. The average number of eukaryotic sequences was 108,548, with an effective rate of 84.79%. The bacterial sequences were clustered into 15,789 operational taxonomic units (OTUs), and eukaryotic sequences were clustered into 16,853 OTUs. The coverage of all samples was > 0.99, and with the rarefaction and Shannon curves converged. This indicates that sampling was reasonable, and that the sequencing was sufficient to characterize the diversity of the study sites. A plot shows (Fig. ) that there were 47 core OTUs of bacteria and 184 core OTUs of eukaryotes, respectively, in wild apples rhizosphere soil samples from eight regions in Xinjiang. The number of bacterial OTUs in regions C, D, G, and H was greater than that of eukaryotic OTUs, whereas the number of bacteria in regions A, B, E, and F was lesser than that of eukaryotic OTUs. The results showed that the number of OTUs in each soil sample varied across the eight regions. The bacterial Chao 1, ACE, Shannon and Simpson indices in regions C, D, G, and H were higher than the eukaryotic biodiversity indices, whereas the bacterial diversity indices in regions A, B, E, and F were lower than the eukaryotic biodiversity indices (Fig. ). The α-diversity analysis showed that species diversity in different regions was quite different. The bacterial species diversity in regions C, D, G, and H was higher than in other regions whereas regions A, B, E, and F had higher species diversity than in other regions. PCoA was used to rank at OTU level to reveal similarities or differences in community structure between different regional groups. The first and second axes of the bacterial community structure contributed 41.3% and 13.4% of the explanation, respectively, and 16.4% and 13.7% for eukaryotes, respectively. (Fig. a,b). In general, the majority of samples from each group clustered together, indicating significant differences in the community composition of bacterial and eukaryotic species.
Compared to the Silva138 database, the proportion of phylum, class, order, family, genus, and species in bacteria was 97.78%, 97.24%, 95.8%, 94.24%, 84.63%, and 32.39% respectively. The 16S rRNA gene sequences were divided into phyla, and the nine most abundant phyla were Firmicutes, Proteobacteria, Actinobacteriota, Bacteroidota, Acidobacteriota, Verrucomicrobiota, Chloroflexi, Planctomycetota, and Methylomirabilota, unclassified (Fig. a). Firmicutes were the most dominant bacterial phylum in regions of A (56.47%), C (24.46%), D (32.45%), E (57.58%), and F (56.43%), and Proteobacteria were the most dominant bacterial phylum in regions B (40.82%), G (54.70%), and H (46.23%). The community composition in regions A, E, and F was less; in particular, the relative abundance of Acidobacteriota, Verrucomicrobiota, Planctomycetota, Chloroflexi, and Methylomirabilot was very low. The relationships between the sequential datasets were visualized by non-metric multidimensional scale (NMDS), and the samples in each region were basically clustered together, but the distance between groups was large, with a P ≤ 0.001 (Fig. b), indicating differences in the composition of bacterial communities in different regions. Linear discriminant analysis effect size (LEfSe) analysis was used to identify microorganisms explicitly enriched in bacteria from the phyla-to-species level in different regions. The results showed that there were 41 phylum-to-species differences in the eight regions. In the bacterial species community, the various species in region A were Bifidobacteriaceae, Actinobacteria, and Coriobacteriales. The different communities in region B were Bacteroidaceae and uncultured_bacteria. Chloroflexi, Bacillus _ luciferensis , and uncultured bacteria were the different species in region C. The main communities in region D included Vicinamibacteraceae, Blastocatellia, and Pyrinomonadaceae. There were no apparent species differences in regions E and F, but uncultured_bacteria were different in region G. Prevotella and Bacteroidia were different in region H (Fig. c). Figure d shows that the rhizosphere bacterial family of wild apples co-occurred. The results showed that the rhizosphere bacterial communities of wild apples in different regions cooperated, and a few competed.
The proportion of the eukaryotic phylum level was 83.17%, that of the class level was 75.23%, that of the order-level was 69.39%, that of the family level was 61.79%, that of the genus level was 53.09%, and that of the species-level was 30.05%. The 18S rRNA gene sequences were divided into phylum levels, among which the nine most abundant phyla were Ascomycota, Phragmoplastophyta, Basidiomycota, Cercozoa, Ochrophyta, Ciliophora, Mucoromycota, Chytridiomycota, and Chlorophyta, unclassified (Fig. a). Eukaryotes were dominated by Ascomycota, Phragmoplastophyta, and Basidiomycota. Ascomycota was the most dominant phylum in regions A (44.71%), B (23.23%), D (38.59%), E (39.98%), F (29.58%), G (37.94%) and H (42.70%), and Phragmoplastophyta was the most dominant phylum in region C (40.42%) regions. NMDS analysis showed that the samples in region C were far apart, and the samples in each area were clustered together, with a P ≤ 0.001 (Fig. b), indicating differences in the eukaryotic community composition in different regions. LEfSe analysis showed significant differences in the leading 41 phyla-to-species in the eight different regions. In the eukaryotic community, the main differential species in region A were Trichoderma , Nectriaceae, and Phallus_hadriani . The differential species in region B was Cercomonadidae, and there was no apparent difference in fungal species. In region C, there was also no significant difference in species. Pleosporales and Dothideomycetes were the differential species in region D. Solicoccozyma and Piskurozymaceae were the differential species in region E. The differential species in region F included Filobasidiales. Helotiales, Incertaesedis, Leotiomycetes, and Geminibasidium were the differential species in region G (Fig. c). Figure d shows that the rhizosphere eukaryotic community of wild apples co-occured, and the network structure was different from that of bacterial. The results showed that the rhizosphere eukaryotic communities of wild apples in different regions cooperated with each other, and a few competed.
The impact of geographic distance, climatic distance and soil pH on the composition of rhizosphere bacterial and eukaryotic communities in wild apples was examined by mantel analysis. It was found that the correlation between climatic distance and β-diversity was greater than that between geographic distance and soil pH (Fig. a). Geographic and climatic distances correlate were more strongly correlated with bacterial β-diversity than with eukaryotic β-diversity. Geographic distance remained positively correlated with eukaryotic β-diversity even after controlling for climatic distance and/or soil pH (Fig. b). The results indicated that geographic and climatic differences were important predictors for microbial β-diversity. Soil pH and climatic distance were positively with bacterial β-diversity and insignificantly with eukaryotic β-diversity when geographic distance was controlled (Fig. b). Climate distance was negatively with microbial β-diversity when the soil pH was controlled. The converse was also true. In summary, there was a correlation between geographical distance, climatic distance and soil pH with microbial β-diversity.
Rhizosphere microbial diversity in different regions The microbial balance in the rhizosphere of wild apple forests in Xinjiang was one of the main driving forces shaping ecosystem function. It contributed to the “soil-microbe-plant” nutrient cycle, primary production and waste decomposition . Therefore, it was crucial to understand the structure, diversity and functional role of microbial communities in terms of community composition, functional groups and ecosystems when human activities, such as climate change, pest and disease outbreaks and grazing, affected the balance of microorganisms in the rhizosphere . This study explored the biodiversity, community structure and functional prediction of wild apples rhizosphere bacteria and eukaryotes in Xinjiang. The results showed that there were more bacterial OTUs than eukaryotic OTUs in zones C, D, G, and H, and the number of bacterial OTUs in the other zones decreased to varying degrees. While the number of eukaryotic OTUs in regions D, G, and H was lesser than that in other regions (Fig. ). Although the species diversity of different regions was quite different, overall, in the α-diversity study, the bacterial species diversity in regions C, D, G, and H was higher than that of eukaryotic species, and the evenness of species distribution was also greater than that of eukaryotic species. Regions A, B, E, and F showed the opposite results. Blaire Steven et al. (2021) suggested that the fir tree rhizosphere microbial 16S rRNA gene dataset was more diverse than the 18S rRNA gene dataset, and the bacterial community was more varied than the eukaryotes community . The diversity study results were consistent with the rhizosphere microbial diversity conclusions in regions C, D, G, and H. However, they were in contrast to the diversity study results in regions A, B, E, and F. According to Yurong Yang et al. (2021), regarding the microbial diversity and communities in the rhizosphere and inner compartments of grassland dominant perennials, grazing decreased interactions among bacterial genera, but there was no difference in interactions among fungal genera . Their results were similar to those of regions A, B, E, and F. Therefore, the differences in microbial diversity in the rhizosphere of wild apples in different regions of Xinjiang might be due to the decrease in the number of microorganisms in regions A, B, E, and F, owing to different climates, grazing, and loss of soil and water nutrients, resulting in greater regional eukaryotes than bacteria. In this study, the dominant bacteria taxa in wild apples rhizosphere soil included Firmicutes, Proteobacteria, Actinobacteriota, Bacteroidota, Acidobacteriota, Verrucomicrobiota, Chloroflexi, Planctomycetota and Methylomirabilota (Fig. a). The bacterial community composition in regions A, E, and F was less, and the relative abundance of Acidobacteriota, Verrucomicrobiota, Chloroflexi, Planctomycetota, and Methylomirabilota was low. The bacterial community composition of region B was similar to that of regions C, D, G, and H, but its OTU was lower. The bacterial community composition differed significantly in the eight regions. Firmicutes are primarily endospore-forming bacteria, and one of the most abundant and ubiquitous bacterial groups in the environment, surviving grazing, fertilization , and harsh environments . Alexandre suggested that the conversion of forest to pasture increased the abundance of microbial taxa associated with nitrogen fixation, including Bacteroidetes and Firmicutes . The relative abundance of Firmicutes in the rhizosphere soil of regions A and C to F regions was relatively high, indicating that the number of Firmicutes in regions A and C to F may have increased due to grazing and adverse climatic conditions. Proteobacteria are one of the rhizosphere trophic bacteria, that promote nitrogen cycling . The relative abundance of Proteobacteria in regions B, G, and H was high, but the relative abundance in other regions was low. It was speculated that the number of Proteobacteria was high in these three regions due to the lack of grazing or abandonment of grazing, consistent with the results obtained by Alexandre Pedrinho . Verrucomicrobia and Acidobacteriota were thought to survive in nutrient-limited soil environments . The relative abundance of the two bacteria in regions D, G, and H was higher, but some soil nutrients in these regions were higher than those in other regions (Table ). Therefore, the local climatic environment and other nutrients in the soil should be considered comprehensively. The eukaryotic community composition included Ascomycota, Phragmoplastophyta, Basidiomycota, Cercozoa, Ochrophyta, Ciliophora, Mucoromycota, Chytridiomycota, and Chlorophyta (Fig. a). The eukaryotic community composition was more uniform and less variable than the bacterial community composition. Ascomycota and Basidiomycota are essential decomposers in the rhizosphere; they promote nutrient absorption, have a high tolerance to environmental stress , and were distributed in the eight regions of our study. Cercozoa is a bacterial predator with a higher relative abundance in regions B and F. Blaire Steven et al. also found Cercozoa in the fir tree rhizosphere and was sensitive to changes in soil pH . Chytridiomycota was associated with cellulose degradation , was distributed in all the eight regions of our study, and was more abundant in region E. Overall, the changes in bacterial community composition in different regions were more significant than those in eukaryotic community composition. In addition, NMDS analysis, LEfSe analysis, and network analysis indicated changes in the rhizosphere microbial community structure in different regions. LEfSe analysis showed significant differences among the rhizospheres in different regions. These differences were probably caused by regional climate change, human activities, and various environmental factors on microorganisms , indicating that the rhizosphere soil environment of 10 to 20 cm was affected by regional ecological changes. The stability of the rhizosphere ecological environment can also be analyzed through the symbiotic network. The number of bacterial community connections was higher than that of eukaryotic community connections, and the microbial communities cooperated with each other, indicating that the rhizosphere environment composed of the overall microbial community was relatively stable. Factors affecting β-diversity in wild apples of different regions Many studies demonstrated that environmental factors had important impacts on the abundance, structure and function of microbial communities . In this study, correlations between differences in geographic distance, climatic distance and soil pH with bacterial as well as eukaryotic community composition were tested by mantel analysis. Meanwhile, the effects of geographic and environmental gradients on wild apple rhizosphere microbes were explored. A study by Shi et al. on Fritillaria thunbergii showed that soil pH had an effect on bacterial species composition . This was consistent with the results of the present work, that is, soil pH was positively with bacterial β-diversity, controlling for geographic distance. However, in this study, there was no significant correlation between soil pH and eukaryotic β-diversity. In contrast, a study by Ren et al. found that the abundance and community structure of eukaryotic OTUs in lake sediments correlated with soil pH. Differences in microbial communities were due to a combination of plant and ecological influences and it was difficult to identify a single factor that had an impact. Therefore, it was normal for the results of individual studies to differ. In addition, Jing et al. and Na et al. confirmed that geographic and climatic distance were significantly associated with microbial community. This was consistent with the findings of this study. That is, rhizosphere microbial β-diversity was correlated with geographic and climatic distance. Among these, bacterial β-diversity showed a greater correlation with geographic and climatic distance than eukaryotes. It will help to maintain the stability of the wild apple rhizosphere and conserve its germplasm resources by understanding the interactions between microbial communities and geographic and climatic distances. In conclusion, this study revealed the structure and diversity rhizosphere microbial communities from eight sampling regions where wild apples were mainly distributed, and their relationship with geographic and environmental gradients. The results demonstrated that the rhizosphere microbial community were affected to certain extent by geographic distance, climate change and grazing. These factors reduced the stability of the rhizosphere. For future research, the effect of grazing on rhizosphere microorganisms at different depths of as well as the interaction mechanism between wild apples and rhizosphere microorganisms can be further explored. It is a way of conserving wild apple germplasm resources.
The microbial balance in the rhizosphere of wild apple forests in Xinjiang was one of the main driving forces shaping ecosystem function. It contributed to the “soil-microbe-plant” nutrient cycle, primary production and waste decomposition . Therefore, it was crucial to understand the structure, diversity and functional role of microbial communities in terms of community composition, functional groups and ecosystems when human activities, such as climate change, pest and disease outbreaks and grazing, affected the balance of microorganisms in the rhizosphere . This study explored the biodiversity, community structure and functional prediction of wild apples rhizosphere bacteria and eukaryotes in Xinjiang. The results showed that there were more bacterial OTUs than eukaryotic OTUs in zones C, D, G, and H, and the number of bacterial OTUs in the other zones decreased to varying degrees. While the number of eukaryotic OTUs in regions D, G, and H was lesser than that in other regions (Fig. ). Although the species diversity of different regions was quite different, overall, in the α-diversity study, the bacterial species diversity in regions C, D, G, and H was higher than that of eukaryotic species, and the evenness of species distribution was also greater than that of eukaryotic species. Regions A, B, E, and F showed the opposite results. Blaire Steven et al. (2021) suggested that the fir tree rhizosphere microbial 16S rRNA gene dataset was more diverse than the 18S rRNA gene dataset, and the bacterial community was more varied than the eukaryotes community . The diversity study results were consistent with the rhizosphere microbial diversity conclusions in regions C, D, G, and H. However, they were in contrast to the diversity study results in regions A, B, E, and F. According to Yurong Yang et al. (2021), regarding the microbial diversity and communities in the rhizosphere and inner compartments of grassland dominant perennials, grazing decreased interactions among bacterial genera, but there was no difference in interactions among fungal genera . Their results were similar to those of regions A, B, E, and F. Therefore, the differences in microbial diversity in the rhizosphere of wild apples in different regions of Xinjiang might be due to the decrease in the number of microorganisms in regions A, B, E, and F, owing to different climates, grazing, and loss of soil and water nutrients, resulting in greater regional eukaryotes than bacteria. In this study, the dominant bacteria taxa in wild apples rhizosphere soil included Firmicutes, Proteobacteria, Actinobacteriota, Bacteroidota, Acidobacteriota, Verrucomicrobiota, Chloroflexi, Planctomycetota and Methylomirabilota (Fig. a). The bacterial community composition in regions A, E, and F was less, and the relative abundance of Acidobacteriota, Verrucomicrobiota, Chloroflexi, Planctomycetota, and Methylomirabilota was low. The bacterial community composition of region B was similar to that of regions C, D, G, and H, but its OTU was lower. The bacterial community composition differed significantly in the eight regions. Firmicutes are primarily endospore-forming bacteria, and one of the most abundant and ubiquitous bacterial groups in the environment, surviving grazing, fertilization , and harsh environments . Alexandre suggested that the conversion of forest to pasture increased the abundance of microbial taxa associated with nitrogen fixation, including Bacteroidetes and Firmicutes . The relative abundance of Firmicutes in the rhizosphere soil of regions A and C to F regions was relatively high, indicating that the number of Firmicutes in regions A and C to F may have increased due to grazing and adverse climatic conditions. Proteobacteria are one of the rhizosphere trophic bacteria, that promote nitrogen cycling . The relative abundance of Proteobacteria in regions B, G, and H was high, but the relative abundance in other regions was low. It was speculated that the number of Proteobacteria was high in these three regions due to the lack of grazing or abandonment of grazing, consistent with the results obtained by Alexandre Pedrinho . Verrucomicrobia and Acidobacteriota were thought to survive in nutrient-limited soil environments . The relative abundance of the two bacteria in regions D, G, and H was higher, but some soil nutrients in these regions were higher than those in other regions (Table ). Therefore, the local climatic environment and other nutrients in the soil should be considered comprehensively. The eukaryotic community composition included Ascomycota, Phragmoplastophyta, Basidiomycota, Cercozoa, Ochrophyta, Ciliophora, Mucoromycota, Chytridiomycota, and Chlorophyta (Fig. a). The eukaryotic community composition was more uniform and less variable than the bacterial community composition. Ascomycota and Basidiomycota are essential decomposers in the rhizosphere; they promote nutrient absorption, have a high tolerance to environmental stress , and were distributed in the eight regions of our study. Cercozoa is a bacterial predator with a higher relative abundance in regions B and F. Blaire Steven et al. also found Cercozoa in the fir tree rhizosphere and was sensitive to changes in soil pH . Chytridiomycota was associated with cellulose degradation , was distributed in all the eight regions of our study, and was more abundant in region E. Overall, the changes in bacterial community composition in different regions were more significant than those in eukaryotic community composition. In addition, NMDS analysis, LEfSe analysis, and network analysis indicated changes in the rhizosphere microbial community structure in different regions. LEfSe analysis showed significant differences among the rhizospheres in different regions. These differences were probably caused by regional climate change, human activities, and various environmental factors on microorganisms , indicating that the rhizosphere soil environment of 10 to 20 cm was affected by regional ecological changes. The stability of the rhizosphere ecological environment can also be analyzed through the symbiotic network. The number of bacterial community connections was higher than that of eukaryotic community connections, and the microbial communities cooperated with each other, indicating that the rhizosphere environment composed of the overall microbial community was relatively stable.
Many studies demonstrated that environmental factors had important impacts on the abundance, structure and function of microbial communities . In this study, correlations between differences in geographic distance, climatic distance and soil pH with bacterial as well as eukaryotic community composition were tested by mantel analysis. Meanwhile, the effects of geographic and environmental gradients on wild apple rhizosphere microbes were explored. A study by Shi et al. on Fritillaria thunbergii showed that soil pH had an effect on bacterial species composition . This was consistent with the results of the present work, that is, soil pH was positively with bacterial β-diversity, controlling for geographic distance. However, in this study, there was no significant correlation between soil pH and eukaryotic β-diversity. In contrast, a study by Ren et al. found that the abundance and community structure of eukaryotic OTUs in lake sediments correlated with soil pH. Differences in microbial communities were due to a combination of plant and ecological influences and it was difficult to identify a single factor that had an impact. Therefore, it was normal for the results of individual studies to differ. In addition, Jing et al. and Na et al. confirmed that geographic and climatic distance were significantly associated with microbial community. This was consistent with the findings of this study. That is, rhizosphere microbial β-diversity was correlated with geographic and climatic distance. Among these, bacterial β-diversity showed a greater correlation with geographic and climatic distance than eukaryotes. It will help to maintain the stability of the wild apple rhizosphere and conserve its germplasm resources by understanding the interactions between microbial communities and geographic and climatic distances. In conclusion, this study revealed the structure and diversity rhizosphere microbial communities from eight sampling regions where wild apples were mainly distributed, and their relationship with geographic and environmental gradients. The results demonstrated that the rhizosphere microbial community were affected to certain extent by geographic distance, climate change and grazing. These factors reduced the stability of the rhizosphere. For future research, the effect of grazing on rhizosphere microorganisms at different depths of as well as the interaction mechanism between wild apples and rhizosphere microorganisms can be further explored. It is a way of conserving wild apple germplasm resources.
In summary, the stability of the rhizosphere environment wild apples in Xinjiang was closely related to geographic distance, climate change, and human activities. The bacterial OTUs numbers, Shannon values, and community composition were significantly lower in regions A, E, and F. The abundance of Acidobacteriota, Verrucomicrobiota, Planctomycetota, Chloroflexi, and Methylomirabilota was also low in these regions. The community composition of region B was similar to that of regions C, D, G, and H, but the number of OTUs and Shannon values were lower than in these four regions with higher values than in regions A, E, and F. The results showed that the abundance and diversity of wild apples rhizosphere bacteria in regions A, B, E, and F were relatively low and were greatly affected by the external environment. However, there were differences in the relative abundance of eukaryotic communities, and they were less than those of bacterial communities. In addition, mantel analysis showed that geographic and climatic distance were key factors influencing the composition and diversity of wild apple rhizosphere microbial communities. While these two factors correlated more with bacterial diversity than that of eukaryotes. Furthermore, this study confirmed the relevance of the structure and diversity of wild apple rhizosphere microbial community to geographic and environmental gradients. The results of this study provide a theoretical basis for subsequent research on microbial function mining and biofertilizer. Further, it can serve as a theoretical guide for regulating the stable balance of wild apple rhizosphere microbial communities and the sustainable management of wild apple forests.
|
Global, regional, and national burden of neuroblastoma and peripheral nervous system tumours in individuals aged over 60 from 1990 to 2021: a trend analysis of global burden of disease study | 88da8542-ef6a-4425-b212-8acd94069354 | 11916991 | Medicine[mh] | Although neuroblastoma and peripheral nervous system tumours are relatively rare, their impact on the global cancer burden cannot be ignored, particularly among children . However, the influence of these tumours on the elderly population has often been overlooked. Neuroblastoma is an embryonal tumour of the sympathetic nervous system that typically occurs in early childhood, while peripheral nervous system tumours encompass a diverse range of conditions, including schwannomas, neurofibromas, and other malignancies that can affect individuals across various age groups . In the context of elderly individuals, these tumours are rarely studied, and the available data on their epidemiology in this age group is limited . As global ageing accelerates, the healthcare demands of older populations are increasing, making it critical to assess the burden of rare diseases such as neuroblastoma and peripheral nervous system tumours in this demographic . The elderly population faces unique challenges, including a higher prevalence of comorbidities, a weakened immune system, and the need for specialised care . These factors can complicate the diagnosis, treatment, and prognosis of neuroblastoma and peripheral nervous system tumours in older adults, which warrants closer attention. Although these diseases are rare in this age group, their occurrence in elderly individuals requires focused investigation due to the potential for more severe outcomes and the specific healthcare challenges they present. Additionally, gender differences in malignancies are observed among senior populations, and this may influence the presentation and outcomes of neuroblastoma and peripheral nervous system tumours in older adults . These differences, though less pronounced, can still provide important insights into the disease burden and healthcare needs of elderly men and women. Despite ongoing demographic changes, studies focusing on the burden of neuroblastoma and peripheral nervous system tumours in those aged 60 and above are relatively limited. Existing research has largely concentrated on paediatric cases, given the higher incidence of these tumours in children . However, the incidence of neuroblastoma and peripheral nervous system tumours in the elderly is not well documented, and the impact of these conditions on older individuals has been understudied. This represents a critical research gap. Understanding how these diseases affect elderly populations is crucial for effective healthcare planning, particularly in resource-limited settings where specialized care may be less accessible. Addressing this gap can lead to better-targeted healthcare policies, improved resource allocation, and better health outcomes for elderly individuals living with these complex diseases. The Global Burden of Disease (GBD) study offers a comprehensive dataset that provides valuable insights into the epidemiological trends of various cancers and other diseases across different age groups and regions. However, specific data on neuroblastoma and peripheral nervous system tumours in the elderly has yet to be thoroughly investigated. This study seeks to address this gap by offering a global, regional, and national examination of the epidemiological trends of neuroblastoma and peripheral nervous system tumours in individuals aged 60 and above from 1990 to 2021. By using data from the 2021 GBD study, we will examine temporal trends, regional variations, and the epidemiological characteristics of these tumours in older adults. This will allow us to identify regions with the highest disease burden and to provide evidence-based recommendations for healthcare strategies, resource allocation, and future research directions. Ultimately, the results of this research will provide insights into improving healthcare services and health outcomes for the ageing global population.
Study population and data collection This study used data sourced from the GBD 2021 dataset, which offers detailed insights into the global and regional impacts of 371 diseases and injuries, in addition to 88 risk factors, across 204 nations and territories spanning from 1990 to 2021. According to the International Classification of Diseases, 10th Revision (ICD-10), neuroblastoma and peripheral nervous system tumours are defined as malignant neoplasms of the adrenal gland, unspecified (C74.9), malignant neoplasms of peripheral nerves and autonomic nervous systems (C47), and benign neoplasms of peripheral and autonomic nervous systems (C36.1). The elderly population was defined as individuals aged 60 and above, in accordance with the World Health Organization (WHO) . This study specifically focused on patients aged 60 and over with neuroblastoma and peripheral nervous system tumours and further segmented them into seven age ranges (60–64, 65–69, 70–74, 75–79, 80–84, 85–89, and 90–94) to more accurately depict the burden of these tumours within this population. This segmentation helps to capture more specific age-related trends, given the heterogeneity in health and disease presentation among older adults. The age cutoff of 60+ was chosen based on WHO’s classification and demographic studies indicating a higher incidence of age-related diseases starting at this age. We accessed and downloaded data on prevalence, incidence, mortality, and DALYs related to neuroblastoma and peripheral nervous system tumours among the elderly from the Global Health Data Exchange (GHDx) platform ( http://ghdx.healthdata.org/gbd-results-tool ). Data were available at global, regional (21 regions), and national (204 countries) levels. We also obtained Socio-Demographic Index (SDI) data to evaluate the impact of socioeconomic factors on the disease burden. The data used in this study are publicly available; therefore, the Ethics Committee of the First Affiliated Hospital of Nanchang University determined that ethical approval was not required. This study adhered to the guidelines for accurate and transparent reporting of cross-sectional studies as outlined in health estimation reporting standards . Statistical analysis R (version 4.3.2) was employed for the statistical analysis, employing the ‘ggplot2’ and ‘sf’ packages for data visualisation, and the ‘broom’ and ‘dplyr’ packages for regression output processing. The 95% Uncertainty Intervals (UIs) were calculated using Monte Carlo simulations. This method was chosen due to its robustness in estimating uncertainty in complex models by repeatedly drawing random samples from the data. These simulations allow for a more comprehensive understanding of the variability in disease burden estimates. The EAPC was calculated using linear regression models. Both univariate and multivariate regression models were used. The univariate model analyzed each condition independently, while the multivariate model included adjustments for potential confounders such as geographic region, SDI, and age group. This approach ensures that the results reflect the true trends in disease burden, accounting for external factors that could influence the data. The multivariate model was particularly important for controlling for the impact of socioeconomic and demographic factors, which may vary across regions and populations. Given the extensive data and potential for multiple comparisons, Bonferroni correction was used to adjust for multiple testing. This method is particularly useful when conducting multiple comparisons across different regions or subgroups, reducing the risk of Type I errors and ensuring the statistical rigor of the findings. Systematic bias adjustment and regression analyses To assess systematic biases and ensure the validity of our findings, we implemented regression analyses with adjustments for multiple confounding variables. Specifically, the analysis was conducted for each disease and injury type, with multivariate regression models applied to adjust for factors such as geographic region, SDI, and age group. The process for adjusting for systematic bias was as follows: Geographic adjustment: Geographic regions were included as independent variables in the regression models to control for regional differences in disease burden. This adjustment allowed us to account for disparities in disease incidence that could arise from location-specific factors. SDI adjustment: The Socio-Demographic Index (SDI), which measures the socioeconomic development of a region, was included as a covariate to account for variations in disease burden among regions with different levels of economic and social conditions. Age group adjustment: Age-specific trends were considered by stratifying data into various age groups (e.g., 60–64, 65–69, 70–74) to account for the potential impact of age on disease burden, especially given that older populations may exhibit different disease trends compared to younger groups. For each disease and injury, the following procedure was followed to ensure the estimates were as accurate and reliable as possible: Initial estimation: We first derived the raw estimates of disease burden (prevalence, incidence, mortality, and DALYs) from the GBD dataset at global, regional, and national levels. Systematic bias adjustment: Using multivariate regression models, we adjusted the raw estimates for factors such as geographic region, SDI, and age group. This adjustment process mitigated the effects of systematic biases and confounding variables that might skew the results. EAPC calculation: The EAPC was calculated by applying linear regression models to the adjusted estimates, providing a measure of trend direction and rate of change in disease burden over time. Uncertainty assessment: We incorporated Monte Carlo simulations to calculate 95% Uncertainty Intervals (UIs) for each estimate, reflecting the inherent uncertainty in the data and the modeling process. These steps ensured that the final estimates and trends were adjusted for systematic biases, accurately reflecting the true patterns of disease burden across different regions and populations. Prevalence and incidence calculation In the statistical analysis, prevalence and incidence were calculated based on the detailed information in the GBD dataset. For each disease and injury, we first calculated the number of prevalent and incident cases per 1000 people, stratified by age group, sex, and geographic region. These calculations rely on the total case counts and population base provided in the dataset, using the following formulas: [12pt]{minimal}
$$ & {} = \, ( {{}/{}} ) \, *{ 1}000 \\ & {} = \, ( {{}/{}} ) \, *{ 1}000 \\ $$ Prevalence = Total cases of a specific disease or injury / Total population ∗ 1000 Incidence = New cases of a specific disease or injury / Total population ∗ 1000 These data were aggregated and standardized at the global, regional, and national levels to ensure comparability across different regions. Exclusion criteria No exclusion criteria were used in selecting the data, as we employed the complete and publicly available GBD dataset for all regions and conditions. This ensures that the analysis reflects the most comprehensive and inclusive dataset available. Software and reproducibility Finally, the reason R was chosen for statistical analysis lies in its flexibility, reproducibility, and the extensive set of packages available for complex statistical modeling and visualisation. R ensures transparency in data processing and result reporting, and the codes used for statistical analyses will be made available for full reproducibility.
This study used data sourced from the GBD 2021 dataset, which offers detailed insights into the global and regional impacts of 371 diseases and injuries, in addition to 88 risk factors, across 204 nations and territories spanning from 1990 to 2021. According to the International Classification of Diseases, 10th Revision (ICD-10), neuroblastoma and peripheral nervous system tumours are defined as malignant neoplasms of the adrenal gland, unspecified (C74.9), malignant neoplasms of peripheral nerves and autonomic nervous systems (C47), and benign neoplasms of peripheral and autonomic nervous systems (C36.1). The elderly population was defined as individuals aged 60 and above, in accordance with the World Health Organization (WHO) . This study specifically focused on patients aged 60 and over with neuroblastoma and peripheral nervous system tumours and further segmented them into seven age ranges (60–64, 65–69, 70–74, 75–79, 80–84, 85–89, and 90–94) to more accurately depict the burden of these tumours within this population. This segmentation helps to capture more specific age-related trends, given the heterogeneity in health and disease presentation among older adults. The age cutoff of 60+ was chosen based on WHO’s classification and demographic studies indicating a higher incidence of age-related diseases starting at this age. We accessed and downloaded data on prevalence, incidence, mortality, and DALYs related to neuroblastoma and peripheral nervous system tumours among the elderly from the Global Health Data Exchange (GHDx) platform ( http://ghdx.healthdata.org/gbd-results-tool ). Data were available at global, regional (21 regions), and national (204 countries) levels. We also obtained Socio-Demographic Index (SDI) data to evaluate the impact of socioeconomic factors on the disease burden. The data used in this study are publicly available; therefore, the Ethics Committee of the First Affiliated Hospital of Nanchang University determined that ethical approval was not required. This study adhered to the guidelines for accurate and transparent reporting of cross-sectional studies as outlined in health estimation reporting standards .
R (version 4.3.2) was employed for the statistical analysis, employing the ‘ggplot2’ and ‘sf’ packages for data visualisation, and the ‘broom’ and ‘dplyr’ packages for regression output processing. The 95% Uncertainty Intervals (UIs) were calculated using Monte Carlo simulations. This method was chosen due to its robustness in estimating uncertainty in complex models by repeatedly drawing random samples from the data. These simulations allow for a more comprehensive understanding of the variability in disease burden estimates. The EAPC was calculated using linear regression models. Both univariate and multivariate regression models were used. The univariate model analyzed each condition independently, while the multivariate model included adjustments for potential confounders such as geographic region, SDI, and age group. This approach ensures that the results reflect the true trends in disease burden, accounting for external factors that could influence the data. The multivariate model was particularly important for controlling for the impact of socioeconomic and demographic factors, which may vary across regions and populations. Given the extensive data and potential for multiple comparisons, Bonferroni correction was used to adjust for multiple testing. This method is particularly useful when conducting multiple comparisons across different regions or subgroups, reducing the risk of Type I errors and ensuring the statistical rigor of the findings.
To assess systematic biases and ensure the validity of our findings, we implemented regression analyses with adjustments for multiple confounding variables. Specifically, the analysis was conducted for each disease and injury type, with multivariate regression models applied to adjust for factors such as geographic region, SDI, and age group. The process for adjusting for systematic bias was as follows: Geographic adjustment: Geographic regions were included as independent variables in the regression models to control for regional differences in disease burden. This adjustment allowed us to account for disparities in disease incidence that could arise from location-specific factors. SDI adjustment: The Socio-Demographic Index (SDI), which measures the socioeconomic development of a region, was included as a covariate to account for variations in disease burden among regions with different levels of economic and social conditions. Age group adjustment: Age-specific trends were considered by stratifying data into various age groups (e.g., 60–64, 65–69, 70–74) to account for the potential impact of age on disease burden, especially given that older populations may exhibit different disease trends compared to younger groups. For each disease and injury, the following procedure was followed to ensure the estimates were as accurate and reliable as possible: Initial estimation: We first derived the raw estimates of disease burden (prevalence, incidence, mortality, and DALYs) from the GBD dataset at global, regional, and national levels. Systematic bias adjustment: Using multivariate regression models, we adjusted the raw estimates for factors such as geographic region, SDI, and age group. This adjustment process mitigated the effects of systematic biases and confounding variables that might skew the results. EAPC calculation: The EAPC was calculated by applying linear regression models to the adjusted estimates, providing a measure of trend direction and rate of change in disease burden over time. Uncertainty assessment: We incorporated Monte Carlo simulations to calculate 95% Uncertainty Intervals (UIs) for each estimate, reflecting the inherent uncertainty in the data and the modeling process. These steps ensured that the final estimates and trends were adjusted for systematic biases, accurately reflecting the true patterns of disease burden across different regions and populations.
In the statistical analysis, prevalence and incidence were calculated based on the detailed information in the GBD dataset. For each disease and injury, we first calculated the number of prevalent and incident cases per 1000 people, stratified by age group, sex, and geographic region. These calculations rely on the total case counts and population base provided in the dataset, using the following formulas: [12pt]{minimal}
$$ & {} = \, ( {{}/{}} ) \, *{ 1}000 \\ & {} = \, ( {{}/{}} ) \, *{ 1}000 \\ $$ Prevalence = Total cases of a specific disease or injury / Total population ∗ 1000 Incidence = New cases of a specific disease or injury / Total population ∗ 1000 These data were aggregated and standardized at the global, regional, and national levels to ensure comparability across different regions.
No exclusion criteria were used in selecting the data, as we employed the complete and publicly available GBD dataset for all regions and conditions. This ensures that the analysis reflects the most comprehensive and inclusive dataset available.
Finally, the reason R was chosen for statistical analysis lies in its flexibility, reproducibility, and the extensive set of packages available for complex statistical modeling and visualisation. R ensures transparency in data processing and result reporting, and the codes used for statistical analyses will be made available for full reproducibility.
Global trends Globally, the number of elderly individuals (aged 60 and above) diagnosed with neuroblastoma and peripheral nervous system tumours increased from 281.36 (95% UI 221.73, 352.95) per 100,000 in 1990 to 1166.65 (95% UI 918.18, 1413.19) per 100,000 in 2021, representing an increase of 314.65%. Simultaneously, the age-standardised prevalence rate increased from 0.06 per 100,000 (95% UI 0.05, 0.08) in 1990 to 0.11 per 100,000 (95% UI 0.09, 0.13) in 2021, with an EAPC of 1.74 (95% CI 1.57, 1.91) (Fig. A, Table ). Regarding incidence, the number of new cases rose from 562.72 (95% UI 443.46, 705.91) per 100,000 in 1990 to 2333.31 (95% UI 1836.36, 2826.39) per 100,000 in 2021, also marking a 314.65% increase. The age-standardised incidence rate showed a similar EAPC of 1.74 (95% CI 1.57, 1.91), increasing from 0.12 per 100,000 (95% UI 0.09, 0.15) in 1990 to 0.22 per 100,000 (95% UI 0.17, 0.26) in 2021 (Fig. B, Supplementary Table 1). Regarding mortality, the death count in 2021 was 1593.92 (95% UI 1329.94, 1781.63) per 100,000, a significant rise from 424.42 (95% UI 361.82, 491.51) per 100,000 in 1990, representing an increase of 275.55%. The age-standardised mortality rate similarly showed a notable increase, rising from 0.09 per 100,000 (95% UI 0.08, 0.11) in 1990 to 0.15 per 100,000 (95% UI 0.12, 0.17) in 2021, accompanied by an EAPC of 1.42 (95% CI 1.29, 1.56) (Fig. C, Supplementary Table 2). Regarding DALYs, the total number of DALYs in 2021 resulting from neuroblastoma and peripheral nervous system tumours in the elderly was 32,050.19 (95% UI 26,840.23, 35,736.83), an increase of 262.53% compared to 8,840.78 (95% UI 7,492.67, 10,292.56) in 1990. The age-standardised DALY rate also rose from 1.82 per 100,000 (95% UI 1.55, 2.12) in 1990 to 2.95 per 100,000 (95% UI 2.47, 3.29) in 2021, with an EAPC of 1.40 (95% CI 1.27, 1.53) (Fig. D, Supplementary Table 3). Global trends by gender Between 1990 and 2021, the age-standardised prevalence, incidence, mortality, and DALY rates for neuroblastoma and peripheral nervous system tumours among the elderly population showed a marked upward trend for both males and females globally, an even more noticeable rise was seen in males (Fig. A–D). Notably, the trends in age-standardised prevalence and incidence were consistent, with males having an EAPC of 2.01 (95% CI 1.85, 2.18) compared to 1.39 (95% CI 1.21, 1.58) for females (Fig. A, B, Table , Supplementary Table 1). For age-standardised mortality rates, the EAPC for males was 1.69 (1.56, 1.82), while for females, it was 1.08 (95% CI 0.93, 1.24). Similarly, in terms of age-standardised DALY rates, males had an EAPC of 1.62 (95% CI 1.49, 1.75), compared to 1.11 (95% CI 0.97, 1.25) for females (Fig. C, D, Supplementary Table 2 and Table 3). Form 1990 to 2021, the disease burden of neuroblastoma and peripheral nervous system tumours among elderly males has consistently been greater than that observed in females. Global trends by age groups Globally, in 2021, the age-specific prevalence, incidence, and mortality rates of neuroblastoma and peripheral nervous system tumours among the elderly exhibited a gradual increase with advancing age, peaking within the 90–94 age group. The rates for this age group reached 0.23 per 100,000 (95% UI 0.28, 0.17) for prevalence, 0.46 per 100,000 (95% UI 0.56, 0.34) for incidence, and 3.05 per 100,000 (95% UI 3.49, 2.43) for mortality (Fig. A–C, Table ). The number of cases, including prevalence, incidence, and mortality, peaked predominantly within the 65–69 age group, with figures of 41.08 (95% UI 50.04, 30.84), 82.16 (95% UI 100.08, 61.68), and 61.17 (95% UI 70.02, 48.56), respectively. Following this peak, the numbers showed a consistent decline as age increased (Fig. A–C). In contrast, the age-specific DALY rate demonstrated a relatively stable trend, with its highest point observed in the 70–74 age group at 3.32 per 100,000 (95% UI 2.74, 3.71). The peak number of DALYs was concentrated in the 60–64 age group, reaching 8766.73 (95% UI 7257.25, 9675.18), and subsequently declined gradually with age (Fig. D, Table ). Global trends by SDI quintiles Between 1990 and 2021, the age-standardised prevalence, incidence, mortality, and DALY rates for neuroblastoma and peripheral nervous system tumours among the elderly across all SDI regions demonstrated broadly similar trends. Low SDI regions consistently maintained the lowest levels, while all other SDI regions, except for high SDI regions which showed minimal change, exhibited a gradual upward trend. The middle SDI regions experienced the highest growth rate throughout the period (Fig. A–D). In terms of age-standardised prevalence and incidence rates, high SDI regions held the highest levels until 2017, after which high-middle SDI regions surpassed them and took the lead. Regarding the rate of increase, middle SDI regions showed the highest EAPC at 4.22 (95% CI 4.09, 4.35), followed by middle-low SDI regions and then low SDI regions (Fig. A, B, Table , Supplementary Table 1). For age-standardised mortality rates, high SDI regions remained at the highest level until 2014, after which high-middle SDI regions took over. The highest growth rate in age-standardised mortality was observed in middle SDI regions, with an EAPC of 3.46 (95% CI 3.36, 3.56), followed by middle-low SDI regions and low SDI regions (Fig. C, Supplementary Table 2). With respect to age-standardised DALY rates, high SDI regions maintained the highest levels until 2013, after which high-middle SDI regions consistently led. The middle SDI regions again displayed the highest growth rate, with an EAPC of 3.42 (95% CI 3.32, 3.53), followed by middle-low SDI regions and low SDI regions (Fig. D, Supplementary Table 3). Regional trends The GBD regional classification includes 204 countries and territories, grouped into 21 regions . An analysis of the burden of neuroblastoma and peripheral nervous system tumours among the elderly across 21 global regions in 2021 revealed that Central Europe exhibited the highest age-standardized rates for prevalence, incidence, and mortality, recorded at 0.16 per 100,000 (95% UI 0.13, 0.20), 0.32 per 100,000 (95% UI 0.26, 0.40), and 0.21 per 100,000 (95% UI 0.18, 0.25), respectively. Meanwhile, Eastern Europe showed the highest age-standardised DALY rate globally, reaching 4.58 per 100,000 (95% UI 3.91, 5.29) (Fig. A, Supplementary Fig. 1A, Supplementary Fig. 2A, Supplementary Fig. 3A, Table , Supplementary Tables 1–3). In terms of case numbers, East Asia recorded the highest numbers for prevalence, incidence, mortality, and DALYs in 2021 among the 21 regions, with figures of 399.83 (95% UI 262.84, 522.80), 799.66 (95% UI 525.67, 1045.61), 532.57 (95% UI 367.41, 670.62), and 10,742.81 (95% UI 7,395.77, 13,606.59), respectively (Fig. B, Supplementary Figs. 1B, 2B, 3B, Table , Supplementary Tables 1–3). Between 1990 and 2021, most regions exhibited an upward trend in age-standardised prevalence and incidence rates, except for Australasia and Oceania, where no significant increases were observed. East Asia demonstrated the fastest growth, with EAPCs of 5.89 (95% CI 5.52, 6.25) for both prevalence and incidence rates. Similarly, in terms of age-standardised mortality and DALY rates, most regions showed a marked increase, with East Asia again exhibiting the highest growth. The EAPCs for age-standardised mortality and DALY rates in East Asia were 4.86 (95% CI 4.55, 5.18) and 4.82 (95% CI 4.49, 5.15), respectively (Fig. A–D). National trends On a national scale, in 2021, the Republic of Estonia recorded the highest age-standardised prevalence, incidence, and mortality rates globally, at 0.42 per 100,000 (95% UI 0.28, 0.63), 0.85 per 100,000 (95% UI 0.56, 1.25), and 0.51 per 100,000 (95% UI 0.34, 0.72), respectively. In contrast, the Republic of Croatia had the highest age-standardised DALY rate at 9.14 per 100,000 (95% UI 6.36, 12.73) (Fig. A, Supplementary Figs. 4A, 5A, Supplementary Tables 4–8). In 2021, the People's Republic of China reported the highest numbers for prevalence, incidence, mortality, and DALYs among the 204 countries studied, with figures of 389.19 (253.48, 511.35), 778.39 (95% UI 506.97, 1022.69), 518.98 (95% UI 354.83, 656.37), and 10,464.64 (95% UI 7,140.82, 13,312.97), respectively (Fig. B, Supplementary Figs. 4B, 5B, 6B, Supplementary Tables 4–8). Examining trends from 1990 to 2021 across 204 countries, more than 90% exhibited an increasing pattern in age-standardised prevalence and incidence rates. Georgia exhibited the highest growth rate globally, with EAPCs for age-standardised prevalence and incidence rates of 13.35 (95% CI 11.93, 14.80). In terms of age-standardised mortality and DALY rates, over 85% of countries demonstrated an increasing trend, with Georgia again showing the highest growth. The EAPCs for age-standardised mortality and DALY rates in Georgia were 13.34 (95% UI 11.89, 14.80) and 13.26 (95% CI 11.81, 14.73), respectively (Fig. C, Supplementary Figs. 4C, 5C, 6C).
Globally, the number of elderly individuals (aged 60 and above) diagnosed with neuroblastoma and peripheral nervous system tumours increased from 281.36 (95% UI 221.73, 352.95) per 100,000 in 1990 to 1166.65 (95% UI 918.18, 1413.19) per 100,000 in 2021, representing an increase of 314.65%. Simultaneously, the age-standardised prevalence rate increased from 0.06 per 100,000 (95% UI 0.05, 0.08) in 1990 to 0.11 per 100,000 (95% UI 0.09, 0.13) in 2021, with an EAPC of 1.74 (95% CI 1.57, 1.91) (Fig. A, Table ). Regarding incidence, the number of new cases rose from 562.72 (95% UI 443.46, 705.91) per 100,000 in 1990 to 2333.31 (95% UI 1836.36, 2826.39) per 100,000 in 2021, also marking a 314.65% increase. The age-standardised incidence rate showed a similar EAPC of 1.74 (95% CI 1.57, 1.91), increasing from 0.12 per 100,000 (95% UI 0.09, 0.15) in 1990 to 0.22 per 100,000 (95% UI 0.17, 0.26) in 2021 (Fig. B, Supplementary Table 1). Regarding mortality, the death count in 2021 was 1593.92 (95% UI 1329.94, 1781.63) per 100,000, a significant rise from 424.42 (95% UI 361.82, 491.51) per 100,000 in 1990, representing an increase of 275.55%. The age-standardised mortality rate similarly showed a notable increase, rising from 0.09 per 100,000 (95% UI 0.08, 0.11) in 1990 to 0.15 per 100,000 (95% UI 0.12, 0.17) in 2021, accompanied by an EAPC of 1.42 (95% CI 1.29, 1.56) (Fig. C, Supplementary Table 2). Regarding DALYs, the total number of DALYs in 2021 resulting from neuroblastoma and peripheral nervous system tumours in the elderly was 32,050.19 (95% UI 26,840.23, 35,736.83), an increase of 262.53% compared to 8,840.78 (95% UI 7,492.67, 10,292.56) in 1990. The age-standardised DALY rate also rose from 1.82 per 100,000 (95% UI 1.55, 2.12) in 1990 to 2.95 per 100,000 (95% UI 2.47, 3.29) in 2021, with an EAPC of 1.40 (95% CI 1.27, 1.53) (Fig. D, Supplementary Table 3).
Between 1990 and 2021, the age-standardised prevalence, incidence, mortality, and DALY rates for neuroblastoma and peripheral nervous system tumours among the elderly population showed a marked upward trend for both males and females globally, an even more noticeable rise was seen in males (Fig. A–D). Notably, the trends in age-standardised prevalence and incidence were consistent, with males having an EAPC of 2.01 (95% CI 1.85, 2.18) compared to 1.39 (95% CI 1.21, 1.58) for females (Fig. A, B, Table , Supplementary Table 1). For age-standardised mortality rates, the EAPC for males was 1.69 (1.56, 1.82), while for females, it was 1.08 (95% CI 0.93, 1.24). Similarly, in terms of age-standardised DALY rates, males had an EAPC of 1.62 (95% CI 1.49, 1.75), compared to 1.11 (95% CI 0.97, 1.25) for females (Fig. C, D, Supplementary Table 2 and Table 3). Form 1990 to 2021, the disease burden of neuroblastoma and peripheral nervous system tumours among elderly males has consistently been greater than that observed in females.
Globally, in 2021, the age-specific prevalence, incidence, and mortality rates of neuroblastoma and peripheral nervous system tumours among the elderly exhibited a gradual increase with advancing age, peaking within the 90–94 age group. The rates for this age group reached 0.23 per 100,000 (95% UI 0.28, 0.17) for prevalence, 0.46 per 100,000 (95% UI 0.56, 0.34) for incidence, and 3.05 per 100,000 (95% UI 3.49, 2.43) for mortality (Fig. A–C, Table ). The number of cases, including prevalence, incidence, and mortality, peaked predominantly within the 65–69 age group, with figures of 41.08 (95% UI 50.04, 30.84), 82.16 (95% UI 100.08, 61.68), and 61.17 (95% UI 70.02, 48.56), respectively. Following this peak, the numbers showed a consistent decline as age increased (Fig. A–C). In contrast, the age-specific DALY rate demonstrated a relatively stable trend, with its highest point observed in the 70–74 age group at 3.32 per 100,000 (95% UI 2.74, 3.71). The peak number of DALYs was concentrated in the 60–64 age group, reaching 8766.73 (95% UI 7257.25, 9675.18), and subsequently declined gradually with age (Fig. D, Table ).
Between 1990 and 2021, the age-standardised prevalence, incidence, mortality, and DALY rates for neuroblastoma and peripheral nervous system tumours among the elderly across all SDI regions demonstrated broadly similar trends. Low SDI regions consistently maintained the lowest levels, while all other SDI regions, except for high SDI regions which showed minimal change, exhibited a gradual upward trend. The middle SDI regions experienced the highest growth rate throughout the period (Fig. A–D). In terms of age-standardised prevalence and incidence rates, high SDI regions held the highest levels until 2017, after which high-middle SDI regions surpassed them and took the lead. Regarding the rate of increase, middle SDI regions showed the highest EAPC at 4.22 (95% CI 4.09, 4.35), followed by middle-low SDI regions and then low SDI regions (Fig. A, B, Table , Supplementary Table 1). For age-standardised mortality rates, high SDI regions remained at the highest level until 2014, after which high-middle SDI regions took over. The highest growth rate in age-standardised mortality was observed in middle SDI regions, with an EAPC of 3.46 (95% CI 3.36, 3.56), followed by middle-low SDI regions and low SDI regions (Fig. C, Supplementary Table 2). With respect to age-standardised DALY rates, high SDI regions maintained the highest levels until 2013, after which high-middle SDI regions consistently led. The middle SDI regions again displayed the highest growth rate, with an EAPC of 3.42 (95% CI 3.32, 3.53), followed by middle-low SDI regions and low SDI regions (Fig. D, Supplementary Table 3).
The GBD regional classification includes 204 countries and territories, grouped into 21 regions . An analysis of the burden of neuroblastoma and peripheral nervous system tumours among the elderly across 21 global regions in 2021 revealed that Central Europe exhibited the highest age-standardized rates for prevalence, incidence, and mortality, recorded at 0.16 per 100,000 (95% UI 0.13, 0.20), 0.32 per 100,000 (95% UI 0.26, 0.40), and 0.21 per 100,000 (95% UI 0.18, 0.25), respectively. Meanwhile, Eastern Europe showed the highest age-standardised DALY rate globally, reaching 4.58 per 100,000 (95% UI 3.91, 5.29) (Fig. A, Supplementary Fig. 1A, Supplementary Fig. 2A, Supplementary Fig. 3A, Table , Supplementary Tables 1–3). In terms of case numbers, East Asia recorded the highest numbers for prevalence, incidence, mortality, and DALYs in 2021 among the 21 regions, with figures of 399.83 (95% UI 262.84, 522.80), 799.66 (95% UI 525.67, 1045.61), 532.57 (95% UI 367.41, 670.62), and 10,742.81 (95% UI 7,395.77, 13,606.59), respectively (Fig. B, Supplementary Figs. 1B, 2B, 3B, Table , Supplementary Tables 1–3). Between 1990 and 2021, most regions exhibited an upward trend in age-standardised prevalence and incidence rates, except for Australasia and Oceania, where no significant increases were observed. East Asia demonstrated the fastest growth, with EAPCs of 5.89 (95% CI 5.52, 6.25) for both prevalence and incidence rates. Similarly, in terms of age-standardised mortality and DALY rates, most regions showed a marked increase, with East Asia again exhibiting the highest growth. The EAPCs for age-standardised mortality and DALY rates in East Asia were 4.86 (95% CI 4.55, 5.18) and 4.82 (95% CI 4.49, 5.15), respectively (Fig. A–D).
On a national scale, in 2021, the Republic of Estonia recorded the highest age-standardised prevalence, incidence, and mortality rates globally, at 0.42 per 100,000 (95% UI 0.28, 0.63), 0.85 per 100,000 (95% UI 0.56, 1.25), and 0.51 per 100,000 (95% UI 0.34, 0.72), respectively. In contrast, the Republic of Croatia had the highest age-standardised DALY rate at 9.14 per 100,000 (95% UI 6.36, 12.73) (Fig. A, Supplementary Figs. 4A, 5A, Supplementary Tables 4–8). In 2021, the People's Republic of China reported the highest numbers for prevalence, incidence, mortality, and DALYs among the 204 countries studied, with figures of 389.19 (253.48, 511.35), 778.39 (95% UI 506.97, 1022.69), 518.98 (95% UI 354.83, 656.37), and 10,464.64 (95% UI 7,140.82, 13,312.97), respectively (Fig. B, Supplementary Figs. 4B, 5B, 6B, Supplementary Tables 4–8). Examining trends from 1990 to 2021 across 204 countries, more than 90% exhibited an increasing pattern in age-standardised prevalence and incidence rates. Georgia exhibited the highest growth rate globally, with EAPCs for age-standardised prevalence and incidence rates of 13.35 (95% CI 11.93, 14.80). In terms of age-standardised mortality and DALY rates, over 85% of countries demonstrated an increasing trend, with Georgia again showing the highest growth. The EAPCs for age-standardised mortality and DALY rates in Georgia were 13.34 (95% UI 11.89, 14.80) and 13.26 (95% CI 11.81, 14.73), respectively (Fig. C, Supplementary Figs. 4C, 5C, 6C).
The probability of developing neurological tumours increases significantly in the elderly due to factors such as reduced cellular regeneration capacity, the build-up of genetic alterations, the decline of immune system function, and prolonged exposure to environmental and lifestyle risk factors. Neuroblastoma and peripheral nervous system tumours exemplify this phenomenon, exhibiting distinctive patterns of onset and clinical presentation [ – ]. This is the first comprehensive study to detail the global, regional, and national prevalence, incidence, mortality, and DALY rates of neuroblastoma and peripheral nervous system tumours among the elderly between 1990 and 2021. Our analysis segments the distribution of these tumors across regions with different SDI levels, areas, countries, genders, and age groups, providing crucial insights into the status of neuroblastoma and peripheral nervous system tumour burden among older adults worldwide. This study underscores the pressing demand for more powerful and targeted health interventions to mitigate the burden of these tumours among the elderly . Increasing global burden Our findings indicate that the burden of neuroblastoma and peripheral nervous system tumours has shown a significant upward trend from 1990 to 2021. The age-standardised prevalence and incidence rates increased from 0.06 per 100,000 (95% UI 0.05, 0.08) and 0.12 per 100,000 (95% UI 0.09, 0.15) to 0.11 per 100,000 (95% UI 0.09, 0.13) and 0.22 per 100,000 (95% UI 0.17, 0.26), respectively. Similarly, the age-standardised mortality and DALY rates also displayed a notable increase, rising from 0.09 per 100,000 (95% UI 0.08, 0.11) and 1.82 per 100,000 (95% UI 1.55, 2.12) to 2.95 per 100,000 (95% UI 2.47, 3.29). This upward trend highlights the growing impact of neuroblastoma and peripheral nervous system tumours on the global elderly population (Fig. ). The study suggests that this trend may be closely linked to a combination of external and internal factors, including genetic susceptibility, environmental exposure, and lifestyle changes. For instance, exposure to certain chemicals and pollutants is believed to be associated with tumour development, and changes in dietary patterns may also influence disease risk. With a rise in the consumption of processed foods and greater exposure to harmful substances in daily life, these cumulative factors merit further attention [ – ]. Additionally, the ageing population is more susceptible to these factors due to physiological decline and alterations in immune function, which contribute to the rising incidence of neuroblastoma and peripheral nervous system tumours [ – ]. Future research should focus on investigating the potential links between these external environmental factors and neurological tumours, as well as the unique role that the elderly population may play in this process. Additionally, studies should explore the relationship between neurodegenerative diseases and these tumours to determine whether there are shared pathogenic mechanisms, thereby paving the way for more effective preventive and therapeutic strategies . Regional disparities Our findings reveal significant regional disparities in the burden of neuroblastoma and peripheral nervous system tumours among the elderly population. In this study, we observed that Central Europe exhibited the highest age-standardised prevalence, incidence, and mortality rates, while Eastern Europe had the highest age-standardised DALY rate, indicating that these two regions bear a heavy disease burden. In terms of the rate of increase in age-standardised prevalence, incidence, mortality, and DALY rates, East Asia showed the highest levels, suggesting that the burden in East Asia has been escalating rapidly over the past three decades. These regional differences may be attributed to a combination of genetic, environmental, healthcare, and lifestyle factors . The disease burden in Central and Eastern Europe may stem from multiple interacting factors. Genetic susceptibility plays a key role, with certain genetic mutations or familial aggregation linked to higher disease rates in these populations. Additionally, early industrialisation exposed these regions to greater environmental pollutants, such as heavy metals and chemicals, which contribute to an elevated risk of neuroblastoma and other neurological tumours. The accessibility and quality of healthcare resources also contribute, with uneven distribution between urban and rural areas hindering early diagnosis and treatment, which increases mortality rates [ – ]. Behavioral factors including elevated smoking rates, alcohol intake, and poor eating habits further contribute to the cancer burden in these regions. Research indicates that lifestyle changes have directly impacted cancer incidence, particularly in populations with limited health education and resources [ – ]. Regarding the situation in East Asia, we observed that the region exhibited the highest growth in age-standardised prevalence, incidence, mortality, and DALY rates. This trend may be related to rapid economic development, accelerated urbanisation, and lifestyle changes in the region . In the last thirty years, shifts in eating habits and lifestyle behaviors and environmental factors in East Asia may have contributed to the increased disease burden . Socio-demographic variations Our study found that in 2021, regions with medium–high SDI had the highest age-standardised prevalence, incidence, mortality, and DALY rates for neuroblastoma and peripheral nervous system tumours among the elderly, followed by high SDI regions. This highlights that regions with higher SDI carry a greater disease burden. While high SDI regions often have more advanced medical facilities, enabling better cancer screening and early diagnosis, this may also result in higher reported incidence rates, as more cases are detected. Thus, the higher disease burden in these regions may reflect improved detection rather than higher true incidence. According to an analysis from the GLOBOCAN database, elderly populations in high-income countries are typically more efficient in early screening and diagnosis of neurological tumours . However, this efficiency in healthcare does not necessarily translate into lower disease rates but may rather highlight the more effective detection of cases that may otherwise go undiagnosed. Additionally, these regions often exhibit more diverse lifestyles, including unhealthy dietary habits, a lack of physical activity, and high smoking rates, all of which are closely linked to cancer incidence. Research indicates that lifestyle changes in high-income countries have led to an increase in cancer rates . Furthermore, the rapid industrialisation and urbanisation seen in high SDI regions have led to increased environmental pollution, including air and water contamination, which are known cancer risk factors . Conversely, medium SDI regions showed the most significant growth in disease burden. This may be due to economic transitions and rapid urbanisation, which exacerbate environmental factors like pollution. These regions still face challenges in healthcare access and early diagnosis, leading to higher incidence and mortality rates. Furthermore, socio-economic inequalities may impede residents’ ability to access quality healthcare, further exacerbating the disease burden . The ageing process in medium SDI regions may also be accelerating faster than in other SDI categories, which makes the disease burden more pronounced in elderly populations. Moreover, in medium SDI regions, the westernisation of lifestyles driven by rapid economic development has resulted in widespread unhealthy eating habits and an increased risk of obesity and related diseases, directly contributing to the rise in cancer incidence . Trends in different age groups Our study analysed age-related disease trends in neuroblastoma and peripheral nervous system tumours among the elderly, indicating that the incidence and impact patterns vary across different age groups. Among these, people aged 90–94 had the highest rates of prevalence, incidence, and mortality, whereas individuals between 70 and 74 years old had the highest DALY rates for their age group. These results emphasize that, with increasing age, the disease burden of neuroblastoma and peripheral nervous system tumours shows a clear upward trend in the elderly population. Some literature suggests that the incidence of neuroblastoma and peripheral nervous system tumours rises significantly among the oldest age groups, particularly those aged over 90 . Earlier studies have shown that the decline in immune surveillance is vital to the progression of neurological tumours as individuals age. Factors such as immunosenescence, reduced T-cell function, chronic inflammation, and cumulative DNA damage contribute to the increased susceptibility of the elderly to neuroblastoma and peripheral nervous system tumours . The progressive loss of immune system efficiency with aging leads to a diminished capacity to detect and eliminate malignant cells, resulting in an increased risk of tumour development. Additionally, genetic mutations accumulated over a lifetime may compound the risk, as older individuals are more likely to have mutations that drive tumour growth. These biological changes, in conjunction with environmental exposures, lead to more pronounced diagnoses and mortality rates in the oldest age groups [ – ]. We also observed that individuals aged 70–74 experienced a notable burden in terms of DALYs. The deterioration of survival and quality of life associated with peripheral nervous system tumours is particularly prominent in older age groups. Earlier studies have shown that, although the overall mortality rate for those aged 70–74 may not be as significant as in older age groups, the disability burden remains substantial. This may be due to a higher prevalence of chronic conditions or multimorbidity within this age group. Additionally, elderly individuals diagnosed with neuroblastoma and peripheral nervous system tumours may face psychological issues such as depression and anxiety. Functional loss and dependence on others as a result of these diseases can lead to feelings of loneliness, helplessness, and even a loss of interest in life, leading to a marked reduction in the quality of life for these individuals [ – ].
Our findings indicate that the burden of neuroblastoma and peripheral nervous system tumours has shown a significant upward trend from 1990 to 2021. The age-standardised prevalence and incidence rates increased from 0.06 per 100,000 (95% UI 0.05, 0.08) and 0.12 per 100,000 (95% UI 0.09, 0.15) to 0.11 per 100,000 (95% UI 0.09, 0.13) and 0.22 per 100,000 (95% UI 0.17, 0.26), respectively. Similarly, the age-standardised mortality and DALY rates also displayed a notable increase, rising from 0.09 per 100,000 (95% UI 0.08, 0.11) and 1.82 per 100,000 (95% UI 1.55, 2.12) to 2.95 per 100,000 (95% UI 2.47, 3.29). This upward trend highlights the growing impact of neuroblastoma and peripheral nervous system tumours on the global elderly population (Fig. ). The study suggests that this trend may be closely linked to a combination of external and internal factors, including genetic susceptibility, environmental exposure, and lifestyle changes. For instance, exposure to certain chemicals and pollutants is believed to be associated with tumour development, and changes in dietary patterns may also influence disease risk. With a rise in the consumption of processed foods and greater exposure to harmful substances in daily life, these cumulative factors merit further attention [ – ]. Additionally, the ageing population is more susceptible to these factors due to physiological decline and alterations in immune function, which contribute to the rising incidence of neuroblastoma and peripheral nervous system tumours [ – ]. Future research should focus on investigating the potential links between these external environmental factors and neurological tumours, as well as the unique role that the elderly population may play in this process. Additionally, studies should explore the relationship between neurodegenerative diseases and these tumours to determine whether there are shared pathogenic mechanisms, thereby paving the way for more effective preventive and therapeutic strategies .
Our findings reveal significant regional disparities in the burden of neuroblastoma and peripheral nervous system tumours among the elderly population. In this study, we observed that Central Europe exhibited the highest age-standardised prevalence, incidence, and mortality rates, while Eastern Europe had the highest age-standardised DALY rate, indicating that these two regions bear a heavy disease burden. In terms of the rate of increase in age-standardised prevalence, incidence, mortality, and DALY rates, East Asia showed the highest levels, suggesting that the burden in East Asia has been escalating rapidly over the past three decades. These regional differences may be attributed to a combination of genetic, environmental, healthcare, and lifestyle factors . The disease burden in Central and Eastern Europe may stem from multiple interacting factors. Genetic susceptibility plays a key role, with certain genetic mutations or familial aggregation linked to higher disease rates in these populations. Additionally, early industrialisation exposed these regions to greater environmental pollutants, such as heavy metals and chemicals, which contribute to an elevated risk of neuroblastoma and other neurological tumours. The accessibility and quality of healthcare resources also contribute, with uneven distribution between urban and rural areas hindering early diagnosis and treatment, which increases mortality rates [ – ]. Behavioral factors including elevated smoking rates, alcohol intake, and poor eating habits further contribute to the cancer burden in these regions. Research indicates that lifestyle changes have directly impacted cancer incidence, particularly in populations with limited health education and resources [ – ]. Regarding the situation in East Asia, we observed that the region exhibited the highest growth in age-standardised prevalence, incidence, mortality, and DALY rates. This trend may be related to rapid economic development, accelerated urbanisation, and lifestyle changes in the region . In the last thirty years, shifts in eating habits and lifestyle behaviors and environmental factors in East Asia may have contributed to the increased disease burden .
Our study found that in 2021, regions with medium–high SDI had the highest age-standardised prevalence, incidence, mortality, and DALY rates for neuroblastoma and peripheral nervous system tumours among the elderly, followed by high SDI regions. This highlights that regions with higher SDI carry a greater disease burden. While high SDI regions often have more advanced medical facilities, enabling better cancer screening and early diagnosis, this may also result in higher reported incidence rates, as more cases are detected. Thus, the higher disease burden in these regions may reflect improved detection rather than higher true incidence. According to an analysis from the GLOBOCAN database, elderly populations in high-income countries are typically more efficient in early screening and diagnosis of neurological tumours . However, this efficiency in healthcare does not necessarily translate into lower disease rates but may rather highlight the more effective detection of cases that may otherwise go undiagnosed. Additionally, these regions often exhibit more diverse lifestyles, including unhealthy dietary habits, a lack of physical activity, and high smoking rates, all of which are closely linked to cancer incidence. Research indicates that lifestyle changes in high-income countries have led to an increase in cancer rates . Furthermore, the rapid industrialisation and urbanisation seen in high SDI regions have led to increased environmental pollution, including air and water contamination, which are known cancer risk factors . Conversely, medium SDI regions showed the most significant growth in disease burden. This may be due to economic transitions and rapid urbanisation, which exacerbate environmental factors like pollution. These regions still face challenges in healthcare access and early diagnosis, leading to higher incidence and mortality rates. Furthermore, socio-economic inequalities may impede residents’ ability to access quality healthcare, further exacerbating the disease burden . The ageing process in medium SDI regions may also be accelerating faster than in other SDI categories, which makes the disease burden more pronounced in elderly populations. Moreover, in medium SDI regions, the westernisation of lifestyles driven by rapid economic development has resulted in widespread unhealthy eating habits and an increased risk of obesity and related diseases, directly contributing to the rise in cancer incidence .
Our study analysed age-related disease trends in neuroblastoma and peripheral nervous system tumours among the elderly, indicating that the incidence and impact patterns vary across different age groups. Among these, people aged 90–94 had the highest rates of prevalence, incidence, and mortality, whereas individuals between 70 and 74 years old had the highest DALY rates for their age group. These results emphasize that, with increasing age, the disease burden of neuroblastoma and peripheral nervous system tumours shows a clear upward trend in the elderly population. Some literature suggests that the incidence of neuroblastoma and peripheral nervous system tumours rises significantly among the oldest age groups, particularly those aged over 90 . Earlier studies have shown that the decline in immune surveillance is vital to the progression of neurological tumours as individuals age. Factors such as immunosenescence, reduced T-cell function, chronic inflammation, and cumulative DNA damage contribute to the increased susceptibility of the elderly to neuroblastoma and peripheral nervous system tumours . The progressive loss of immune system efficiency with aging leads to a diminished capacity to detect and eliminate malignant cells, resulting in an increased risk of tumour development. Additionally, genetic mutations accumulated over a lifetime may compound the risk, as older individuals are more likely to have mutations that drive tumour growth. These biological changes, in conjunction with environmental exposures, lead to more pronounced diagnoses and mortality rates in the oldest age groups [ – ]. We also observed that individuals aged 70–74 experienced a notable burden in terms of DALYs. The deterioration of survival and quality of life associated with peripheral nervous system tumours is particularly prominent in older age groups. Earlier studies have shown that, although the overall mortality rate for those aged 70–74 may not be as significant as in older age groups, the disability burden remains substantial. This may be due to a higher prevalence of chronic conditions or multimorbidity within this age group. Additionally, elderly individuals diagnosed with neuroblastoma and peripheral nervous system tumours may face psychological issues such as depression and anxiety. Functional loss and dependence on others as a result of these diseases can lead to feelings of loneliness, helplessness, and even a loss of interest in life, leading to a marked reduction in the quality of life for these individuals [ – ].
Our findings revealed that among patients aged 60 and above, the global age-standardised prevalence rate for males was 1.56 times greater than for females, while the age-standardised incidence, mortality, and DALY rates were 1.65, 1.58, and 1.59 times higher, respectively. Form 1990 to 2021, the EAPC for neuroblastoma and peripheral nervous system tumours in males significantly surpassed that of females, indicating a much higher disease burden among older men globally. This observed gender disparity may result from a mixture of age-related biological and socio-behavioral factors. Hormonal shifts and immune system changes in older males may contribute to their heightened susceptibility to these tumours . Specifically, the weakening of immune function as individuals age could reduce the body's capacity to detect and fight tumour cells, while hormonal changes, particularly in testosterone levels, may affect tumour development and progression [ – ]. Additionally, gender-related delays in seeking medical care and differences in treatment approaches may further exacerbate the observed disparity [ – ]. Understanding these age-modified gender differences is crucial for developing tailored healthcare strategies to address the unique needs of elderly male patients. Need for effective interventions With the intensifying trend of global ageing, the tumour burden among the elderly, particularly neuroblastoma and peripheral nervous system tumours, has become a significant public health challenge for the future. According to the World Population Prospects report published by the United Nations in 2020, the number of people aged 60 and older worldwide is projected to surpass 2 billion by 2050, projecting a more severe burden of neuroblastoma and peripheral nervous system tumours. Although these tumours are more commonly seen in adolescents and children, they are also showing a steady increase in the elderly population as society ages. While current treatments have shown some effectiveness in certain cases, traditional methods often have considerable limitations due to the biological characteristics of these tumours and the unique physiological conditions of older patients, highlighting the urgent need for new, more effective intervention measures. For the elderly population, conventional treatments are often poorly tolerated due to declining immune and metabolic functions, leading to suboptimal outcomes and significant adverse effects. This is particularly true for neuroblastoma, a tumour originating in the sympathetic nervous system, where traditional treatments are often insufficient in addressing issues related to metastasis and recurrence. Therefore, we recommend personalised precision treatment strategies involving the detection of specific gene mutations or related molecular markers in tumour tissue. For instance, drugs targeting ALK (Anaplastic Lymphoma Kinase) mutations can be selectively used to inhibit tumour growth and reduce the side effects associated with traditional chemotherapy and radiotherapy . We suggest that high-SDI regions accelerate research into targeted therapies and immunotherapies, particularly focusing on the needs of elderly patients. Meanwhile, in lower-SDI regions, there is a need to enhance the development of imaging technologies (e.g., CT, MRI) and biomarkers, especially in identifying tumour-specific markers that can provide more tools for early clinical screening. In resource-limited areas, strengthening training for primary care physicians and promoting health check-ups can help increase the likelihood of early detection. Limitations There are several limitations to this study. First, each iteration of the GBD data differs from previous generations, making them incomparable. These discrepancies are likely due to improvements in algorithms and models over time . Second, an accurate global assessment of the burden of neuroblastoma and peripheral nervous system tumours in the elderly requires more precise epidemiological data, as the databases for neuroblastoma and peripheral nervous system tumours in some countries are still incomplete , which may lead to inaccuracies in estimates based on complex statistical models. Third, the GBD does not differentiate between subtypes of neuroblastoma and peripheral nervous system tumours or the severity of the condition, limiting our ability to analyse the specific burden of these subtypes on a global scale and provide detailed epidemiological insights based on the severity of patients' conditions. Fourth, in regions with lower levels of healthcare, the lack of expertise in medical examinations and advanced neuroimaging technologies may hinder the accurate differentiation of neuroblastoma and peripheral nervous system tumours from other diseases (such as kidney tumours, liver tumours, and metastatic cancers), increasing diagnostic uncertainty in these areas.
With the intensifying trend of global ageing, the tumour burden among the elderly, particularly neuroblastoma and peripheral nervous system tumours, has become a significant public health challenge for the future. According to the World Population Prospects report published by the United Nations in 2020, the number of people aged 60 and older worldwide is projected to surpass 2 billion by 2050, projecting a more severe burden of neuroblastoma and peripheral nervous system tumours. Although these tumours are more commonly seen in adolescents and children, they are also showing a steady increase in the elderly population as society ages. While current treatments have shown some effectiveness in certain cases, traditional methods often have considerable limitations due to the biological characteristics of these tumours and the unique physiological conditions of older patients, highlighting the urgent need for new, more effective intervention measures. For the elderly population, conventional treatments are often poorly tolerated due to declining immune and metabolic functions, leading to suboptimal outcomes and significant adverse effects. This is particularly true for neuroblastoma, a tumour originating in the sympathetic nervous system, where traditional treatments are often insufficient in addressing issues related to metastasis and recurrence. Therefore, we recommend personalised precision treatment strategies involving the detection of specific gene mutations or related molecular markers in tumour tissue. For instance, drugs targeting ALK (Anaplastic Lymphoma Kinase) mutations can be selectively used to inhibit tumour growth and reduce the side effects associated with traditional chemotherapy and radiotherapy . We suggest that high-SDI regions accelerate research into targeted therapies and immunotherapies, particularly focusing on the needs of elderly patients. Meanwhile, in lower-SDI regions, there is a need to enhance the development of imaging technologies (e.g., CT, MRI) and biomarkers, especially in identifying tumour-specific markers that can provide more tools for early clinical screening. In resource-limited areas, strengthening training for primary care physicians and promoting health check-ups can help increase the likelihood of early detection.
There are several limitations to this study. First, each iteration of the GBD data differs from previous generations, making them incomparable. These discrepancies are likely due to improvements in algorithms and models over time . Second, an accurate global assessment of the burden of neuroblastoma and peripheral nervous system tumours in the elderly requires more precise epidemiological data, as the databases for neuroblastoma and peripheral nervous system tumours in some countries are still incomplete , which may lead to inaccuracies in estimates based on complex statistical models. Third, the GBD does not differentiate between subtypes of neuroblastoma and peripheral nervous system tumours or the severity of the condition, limiting our ability to analyse the specific burden of these subtypes on a global scale and provide detailed epidemiological insights based on the severity of patients' conditions. Fourth, in regions with lower levels of healthcare, the lack of expertise in medical examinations and advanced neuroimaging technologies may hinder the accurate differentiation of neuroblastoma and peripheral nervous system tumours from other diseases (such as kidney tumours, liver tumours, and metastatic cancers), increasing diagnostic uncertainty in these areas.
In conclusion, since the 1990s, the incidence of neuroblastoma and peripheral nervous system tumours among the elderly has steadily increased in almost every country, with notable variations observed among regions, genders, and age categories. The treatment of neuroblastoma and peripheral nervous system tumours continues to present considerable challenges, highlighting the critical demand for more efficient and focused interventions to alleviate the burden of these central nervous system tumours. To address the disparities in disease incidence by region, gender, and age, a comprehensive approach integrating medical, psychosocial, and public health strategies is essential. Future studies should prioritize identifying possible risk factors and creating innovative treatment methods to address this prevalent and significant disease.
Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Supplementary Material 5 Supplementary Material 6 Supplementary Material 7 Supplementary Material 8 Supplementary Material 9 Supplementary Material 10 Supplementary Material 11 Supplementary Material 12 Supplementary Material 13 Supplementary Material 14
|
Acute Kidney Injury among Patients Visiting the Nephrology Unit in a Tertiary Care Centre: A Descriptive Cross-sectional Study | fd703715-8d87-4a27-9976-4b83e4241872 | 9795132 | Internal Medicine[mh] | Acute kidney injury (AKI) is an abrupt decline in renal function often associated with a decrease in urine output. AKI is a leading cause of in-hospital mortality worldwide. It is related to a prolonged hospital stay and mechanical ventilation with severe forms requiring short-term dialysis. Improving Global Outcomes (KDIGO) group has published a consensus definition and classification for AKI. KDIGO covers both acute kidney injury network (AKIN) and Risk, Injury, Failure, Loss of Kidney Function, and End-Stage Kidney Disease (RIFLE) criteria, taking into account changes in creatinine within 48 hours or a decline in glomerular filtration rate (GFR) over 7 days. Etiology and incidence of AKI differ between high-income and low-to-middle-income countries. Epidemiologic studies of AKI are sparse, making incidence, prevalence and particularly outcomes difficult to compare. The aim of the study was to find out the prevalence of acute kidney injury among patients visiting the Nephrology unit in a tertiary care centre.
This descriptive cross-sectional study was conducted from 9 February 2022 to 21 October 2022 in the Nephrology Unit, Department of Internal Medicine at Universal College of Medical Sciences (UCMS). Ethical approval was taken from Institutional Review Committee (Reference number: UCMS/IRC/047/22). Data was collected from hospital records and the outcome was recorded in terms of in-hospital mortality and the requirement for renal replacement therapy. Patients admitted under the Nephrology Unit in the Department of Internal Medicine were included in the study. Patients with incomplete hospital data, underlying chronic kidney disease, glomerular disease and structural damage to the kidneys were excluded from the study. Convenience sampling was done and the sample size was calculated using the formula: n = Z 2 × p × q e 2 = 1.96 2 × 0.08 × 0.92 0.02 2 = 707 Where, n = minimum required sample size, Z = 1.96 at 95% Confidence Interval (CI), p = prevalence of patients with AKI taken from previous study, 8% q = I-P e = margin of error, 2% The calculated sample size was found to be 707. On doubling, the sample size becomes 1414. However, we took 1848 samples. Patients were classified based on KDIGO staging for AKI and the outcome was recorded in terms of inhospital all-cause mortality and requirement of renal replacement therapy (RRT). Patients were provided with either conventional hemodialysis or sustained low-efficiency dialysis. Data were analysed using IBM SPSS Statistics version 17.0. Point estimate and 95% CI were calculated.
Among 1848 patients, 113 (6.12%) (5.03-7.21, 95% Confidence Interval) had acute kidney injury. The mean age was 47.09±16.86 years with a male: female ratio of 1.21:1 . Ninety-one (80.53%) patients had infective causes as the aetiology of acute kidney injury and 1 (0.89%) had an electric shock as the aetiology . Among infective causes, 33 (36.26%) had pneumonia followed by GI sepsis in 21 (23.08%) . Among the comorbidities, 35 (31.00%) had diabetes mellitus, 36 (31.90%) had hypertension and 5 (4.40%) had ischemic heart disease. The mean serum creatinine level at admission was 3.50±2.24 mg/dl and the mean estimated glomerular filtration rate on admission calculated using chronic kidney disease epidemiology collaboration equation was 35.04±20.36 ml/min/1.73 m . The average duration of hospital stay was 6.27±2.56 days. Forty-seven (41.59%) of the patients had a KDIGO stage 3 whereas 21 (18.58%) had stage 1 . About 38 (33.62%) required inotropes whereas 10 (8.85%) required mechanical ventilation. In-hospital all-cause mortality was seen in 14 (12.40%) of the study population. About 20 (17%) of the study population required RRT.
Among 1848 patients, 113 (6.12%) had acute kidney injury. Our study of 113 patients with AKI showed infection as the major contributor seen in more than two third of the study population which is similar to a study at Bir Hospital. Among the infections, pneumonia (36%) followed by acute gastrointestinal (GI) infections (23%) was the major cause leading to AKI whereas in another study the leading cause of AKI was GI infections. , This implies that pneumonia as a cause of AKI is increasing as shown by another study in 2019. Mean serum creatinine at admission was 3.50±2.24 mg/dl which was similar to a study. About 20 (17%) of the study population required RRT implying that more than 80% were managed conservatively which is in accordance with the study from Nepal Medical College. Our study showed that 41% of the study population had KDIGO Stage III AKI and nearly half of them required RRT. Nearly 30% required ionotropic support to maintain mean arterial pressure and 10% required Mechanical ventilation. Mortality in patients with AKI was 12.40% which is similar to a study. Thus, our study had similar findings to previous studies and it used the revised consensus classification of AKI based on KDIGO guidelines. The limitation of our study is that serum creatinine in itself is an imperfect marker of kidney function. It is affected by the use of drugs, and muscle mass and also it lags behind the fall in GFR thus early detection is not possible. The use of novel biomarkers for AKI was not feasible in our setting.
The prevalence of acute kidney injury was found to be lower than the studies done in similar settings. Acute kidney injury is common in patients admitted with infection. It is responsible for in hospital mortality. Majority of the patients can be managed conservatively though few require short term dialysis.
|
Comparison of laparoscopic Heller myotomy and endoscopic balloon dilation in the treatment of achalasia: Effects on quality of life and patient satisfaction | 67d98b34-bb29-4acd-9be5-d76f989bf52e | 11843379 | Surgical Procedures, Operative[mh] | Achalasia is a rare esophageal disorder characterized by the degeneration of the myenteric plexus, leading to impaired esophageal peristalsis and incomplete relaxation of the lower esophageal sphincter (LES). This results in a constellation of symptoms including dysphagia, regurgitation, nocturnal cough, recurrent aspiration pneumonia, chest pain, and weight loss. Achalasia has an estimated prevalence of 1 per 100,000 individuals in the Western world. Diagnosis of achalasia is established based on a combination of clinical presentation, esophageal manometry, barium esophagogram, esophagogastroduodenoscopy, and esophageal motility testing. Manometric findings characteristic of achalasia include aperistalsis and impaired LES relaxation. Treatment options for achalasia include pharmacological therapy, endoscopic procedures, and minimally invasive surgical interventions. Among minimally invasive approaches, laparoscopic Heller myotomy (LHM), endoscopic balloon dilation (EBD), and peroral endoscopic myotomy (POEM) are the most commonly employed techniques. Peroral endoscopic myotomy has recently gained recognition as an attractive endoscopic treatment modality for achalasia. Laparoscopic Heller myotomy has remained a mainstay in the treatment of achalasia since its introduction. This procedure involves a myotomy extending at least 6 cm above the gastroesophageal junction and at least 3 cm distal to the gastric cardia, along with Dor fundoplication, following dissection of the phrenoesophageal ligament. Endoscopic balloon dilation, on the other hand, involves controlled disruption of the LES using an air-filled balloon dilator, effectively relieving distal esophageal obstruction and improving dysphagia; however, it is less effective in addressing reflux symptoms. Notably, a significant proportion of patients treated with EBD require repeat procedures. Compared to EBD, LHM offers a single-session treatment and may provide a more definitive solution for patients’ complaints. The primary aim of achalasia treatment is the elimination of dysphagia while preventing the development of reflux. Reflux can occur following either LHM or EBD. Therefore, LHM is often combined with fundoplication to minimize the risk of reflux. There is debate about the initial treatment for achalasia. Therefore, the aim of this study is to contribute to the clarification of this issue by comparing the effectiveness of LHM and EBD procedures in achalasia patients treated at our center, the long-term clinical outcomes, and their effects on quality of life and patient satisfaction. Study Design A retrospective study was conducted by reviewing the patient records of Bezmialem Vakıf University Hospital to identify patients who underwent procedures for achalasia between 2010 and 2023. Patients were divided into two groups based on the treatment modality received: LHM + Dor Fundoplication or EBD. Patient Selection Criteria Inclusion Criteria: Age ≥ 18 years Diagnosis of achalasia Treatment with LHM + Dor Fundoplication or EBD. Exclusion Criteria: Incomplete follow-up data LHM without fundoplication Malignancy-related achalasia Treatment with modalities other than LHM or EBD. Ethical Approval The study was approved by the Ethics Committee of Bezmialem Vakıf University Hospital (2023/106). Data Collection Demographic and clinical data were collected from patients’ medical records, including age, gender, comorbidities, symptom duration, pre-treatment esophageal manometry findings, treatment details, the Pre-operative Eckardt Score (Pre-ES), and the Post-operative Eckardt Score (Post-ES). Achalasia Symptom Quality of Life Questionnaire (ASQL) and Patient Satisfaction scores were assessed through telephone interviews. Surgical Procedures LHM + Dor Fundoplication After dissection of the phrenoesophageal ligament, a myotomy extending at least 6 cm above the gastroesophageal junction and at least 3 cm distal to the gastric cardia was performed, along with Dor fundoplication. EBD Under sedation, esophagogastroduodenoscopy was performed to identify the esophagogastric junction. A pneumatic balloon was inserted and positioned at the esophagogastric junction. Dilatations were performed using a 30 mm balloon. After device removal, careful upper endoscopy was performed to evaluate the patency of the esophagogastric junction. Outcome Measures Eckardt Score (ES) The ES assesses dysphagia, regurgitation, retrosternal pain, and weight loss . Each symptom is scored from 0 to 3. The maximum possible ES score is 12, and the lowest score is 0. Achalasia Symptom Quality of Life Questionnaire (ASQL) The ASQL is a 19-item questionnaire that evaluates patients’ quality of life before and after treatment for achalasia. The total possible score is 28, with a minimum score of 8. The questionnaire assesses the following aspects of dysphagia, food intake, and overall quality of life: Ability to eat and drink various foods Frequency of water intake to dislodge stuck food Frequency of chest pain during eating Impact of chest pain on daily activities Eating speed Restriction of lifestyle due to achalasia Restriction of food and liquid intake due to treatment Overall satisfaction with health after treatment. Statistical Analysis Data collected for the purpose of the study were organized in Microsoft Excel. The Student’s t-test was used to compare the mean ages of the two groups. Additionally, the Student’s t-test, Chi-square test, Fisher’s Exact Probability Test, and Odds Ratio were utilized to compare differences between the two groups. The Paired t-test was used to evaluate the quality of life of the patients, and the Mann-Whitney U Test was used to assess the patients’ satisfaction status. All P values less than 0.05 were considered statistically significant for the study. A retrospective study was conducted by reviewing the patient records of Bezmialem Vakıf University Hospital to identify patients who underwent procedures for achalasia between 2010 and 2023. Patients were divided into two groups based on the treatment modality received: LHM + Dor Fundoplication or EBD. Inclusion Criteria: Age ≥ 18 years Diagnosis of achalasia Treatment with LHM + Dor Fundoplication or EBD. Exclusion Criteria: Incomplete follow-up data LHM without fundoplication Malignancy-related achalasia Treatment with modalities other than LHM or EBD. The study was approved by the Ethics Committee of Bezmialem Vakıf University Hospital (2023/106). Demographic and clinical data were collected from patients’ medical records, including age, gender, comorbidities, symptom duration, pre-treatment esophageal manometry findings, treatment details, the Pre-operative Eckardt Score (Pre-ES), and the Post-operative Eckardt Score (Post-ES). Achalasia Symptom Quality of Life Questionnaire (ASQL) and Patient Satisfaction scores were assessed through telephone interviews. After dissection of the phrenoesophageal ligament, a myotomy extending at least 6 cm above the gastroesophageal junction and at least 3 cm distal to the gastric cardia was performed, along with Dor fundoplication. Under sedation, esophagogastroduodenoscopy was performed to identify the esophagogastric junction. A pneumatic balloon was inserted and positioned at the esophagogastric junction. Dilatations were performed using a 30 mm balloon. After device removal, careful upper endoscopy was performed to evaluate the patency of the esophagogastric junction. The ES assesses dysphagia, regurgitation, retrosternal pain, and weight loss . Each symptom is scored from 0 to 3. The maximum possible ES score is 12, and the lowest score is 0. The ASQL is a 19-item questionnaire that evaluates patients’ quality of life before and after treatment for achalasia. The total possible score is 28, with a minimum score of 8. The questionnaire assesses the following aspects of dysphagia, food intake, and overall quality of life: Ability to eat and drink various foods Frequency of water intake to dislodge stuck food Frequency of chest pain during eating Impact of chest pain on daily activities Eating speed Restriction of lifestyle due to achalasia Restriction of food and liquid intake due to treatment Overall satisfaction with health after treatment. Data collected for the purpose of the study were organized in Microsoft Excel. The Student’s t-test was used to compare the mean ages of the two groups. Additionally, the Student’s t-test, Chi-square test, Fisher’s Exact Probability Test, and Odds Ratio were utilized to compare differences between the two groups. The Paired t-test was used to evaluate the quality of life of the patients, and the Mann-Whitney U Test was used to assess the patients’ satisfaction status. All P values less than 0.05 were considered statistically significant for the study. Complications A total of 36 patients who underwent LHM (n=19) or EBD (n=17) for the treatment of achalasia were included in the study. The mean age of the 19 patients who underwent LHM was 49.37±10.48 years. All of these 19 patients underwent LHM with Dor fundoplication. The procedures were completed laparoscopically in all 19 patients. The mean age of the 17 patients who underwent EBD was 59.24±14.39 years. The Chicago Classification of achalasia type and the percentage treated with LHM were as follows: Type I, n=9 (47% LHM); Type II, n=8 (42% LHM); Type III, n=2 (11% LHM) and the percentage of achalasia type treated with the Chicago Classification and EBD was as follows: Type I, n=10 (59% EBD); Type II, n=7 (41% EBD). Intraoperative esophageal mucosal perforation was observed in one patient in the LHM group, while intraoperative complications were seen in four patients in the EBD group [bleeding in three patients (17.64%), esophageal perforation in one patient (5.88%)] . The comparison of complications between the two groups was statistically significant (p<0.05). According to the Fisher Exact Probability Test and Odds Ratio analysis, EBD increased the risk of bleeding and perforation, whereas LHM decreased these risks. In the postoperative follow-up period, gastroesophageal reflux developed in two patients (10.53%) in the LHM group and in eight patients (47.1%) in the EBD group. The comparison of late postoperative complications between the two groups was statistically significant. Since the chi-square test, Fisher’s Exact Probability Test, and p<0.05 indicated a statistically significant difference in the incidence of reflux between LHM and EBD procedures, the incidence of reflux was lower in the LHM group compared to the EBD group. These findings suggest that EBD has a higher risk of both intraoperative and postoperative complications. Preoperative Symptom Both groups reported similar symptoms before surgery. The most common symptom was dysphagia, which was present in 34 of the 36 patients (LHM: 17, EBD: 17). When asked about their primary symptom, 17 patients (89.47%) in the LHM group had dysphagia, four patients (21%) had regurgitation, one patient (5.26%) had chest pain, and six patients (31.58%) had weight loss. In the EBD group, all 17 patients (100%) had dysphagia, six patients (35.29%) had regurgitation, three patients (17.65%) had chest pain, and six patients (35.29%) had weight loss . Gastroesophageal reflux symptoms were not present in either group. Postoperative Outcomes The mean follow-up period was 89.1±43.19 months for the LHM group and 90.28±34.5 months for the EBD group. Preoperative Eckardt Scores were not different between the two groups (Student’s T-Test: LHM: 4.84, EBD: 4.68, p-value: 0.341). Postoperative Eckardt Scores were significantly lower in the LHM group compared to the EBD group. Quality of Life and Patient Satisfaction When comparing the results of the Achalasia Symptom Quality of Life Questionnaire and patient satisfaction between the LHM and EBD groups, it was concluded that LHM was more effective than EBD in improving patients’ quality of life and achieving patient satisfaction (p-value: 0.001). A total of 36 patients who underwent LHM (n=19) or EBD (n=17) for the treatment of achalasia were included in the study. The mean age of the 19 patients who underwent LHM was 49.37±10.48 years. All of these 19 patients underwent LHM with Dor fundoplication. The procedures were completed laparoscopically in all 19 patients. The mean age of the 17 patients who underwent EBD was 59.24±14.39 years. The Chicago Classification of achalasia type and the percentage treated with LHM were as follows: Type I, n=9 (47% LHM); Type II, n=8 (42% LHM); Type III, n=2 (11% LHM) and the percentage of achalasia type treated with the Chicago Classification and EBD was as follows: Type I, n=10 (59% EBD); Type II, n=7 (41% EBD). Intraoperative esophageal mucosal perforation was observed in one patient in the LHM group, while intraoperative complications were seen in four patients in the EBD group [bleeding in three patients (17.64%), esophageal perforation in one patient (5.88%)] . The comparison of complications between the two groups was statistically significant (p<0.05). According to the Fisher Exact Probability Test and Odds Ratio analysis, EBD increased the risk of bleeding and perforation, whereas LHM decreased these risks. In the postoperative follow-up period, gastroesophageal reflux developed in two patients (10.53%) in the LHM group and in eight patients (47.1%) in the EBD group. The comparison of late postoperative complications between the two groups was statistically significant. Since the chi-square test, Fisher’s Exact Probability Test, and p<0.05 indicated a statistically significant difference in the incidence of reflux between LHM and EBD procedures, the incidence of reflux was lower in the LHM group compared to the EBD group. These findings suggest that EBD has a higher risk of both intraoperative and postoperative complications. Both groups reported similar symptoms before surgery. The most common symptom was dysphagia, which was present in 34 of the 36 patients (LHM: 17, EBD: 17). When asked about their primary symptom, 17 patients (89.47%) in the LHM group had dysphagia, four patients (21%) had regurgitation, one patient (5.26%) had chest pain, and six patients (31.58%) had weight loss. In the EBD group, all 17 patients (100%) had dysphagia, six patients (35.29%) had regurgitation, three patients (17.65%) had chest pain, and six patients (35.29%) had weight loss . Gastroesophageal reflux symptoms were not present in either group. The mean follow-up period was 89.1±43.19 months for the LHM group and 90.28±34.5 months for the EBD group. Preoperative Eckardt Scores were not different between the two groups (Student’s T-Test: LHM: 4.84, EBD: 4.68, p-value: 0.341). Postoperative Eckardt Scores were significantly lower in the LHM group compared to the EBD group. When comparing the results of the Achalasia Symptom Quality of Life Questionnaire and patient satisfaction between the LHM and EBD groups, it was concluded that LHM was more effective than EBD in improving patients’ quality of life and achieving patient satisfaction (p-value: 0.001). Achalasia is a rare neurodegenerative disorder characterized by the loss of inhibitory neurons in the myenteric plexus, which is responsible for normalizing esophageal peristalsis and LES relaxation. While pharmacologic agents can be used in the treatment of achalasia, minimally invasive interventions such as LHM, EBD, and POEM are also commonly used. In our study, LHM and EBD methods were applied and ASQL long-term results of 36 patients included in the study were compared to determine the most appropriate treatment approach for achalasia. The Eckardt Score is used to evaluate the prevalence of achalasia symptoms, to monitor how the symptoms of patients change over time, and to evaluate the effectiveness of treatment. In our study, when the Pre-ES, which was used to evaluate the symptoms of the patients before the procedure, was calculated, it was found to be similar in both groups. In the post-procedure period, it was concluded that Post-ES decreased significantly in the LHM group compared to the EBD group. Although the relief of symptoms after the intervention is the main goal, achieving this success rate under safe conditions should also be a primary goal. This approach guides the selection of the method with the fewest complications that may occur during and after the intervention. As in our study, many studies have reported that LHM has a lower complication rate compared to EBD. These complications are frequently observed as bleeding and esophageal perforation. In our study, one patient in the LHM group experienced esophageal mucosal complications in the early period (5.26%), while bleeding was observed in three patients (17.65%) and esophageal perforation in one patient (5.88%) in the EBD group, which was statistically significant. This indicates that LHM should be the first choice in the treatment of achalasia. The main goal is to improve the quality of life of patients with achalasia. This is assessed using quality of life and satisfaction questionnaires. In the quality of life and satisfaction assessment conducted in our study, it was statistically significant that LHM was more effective than EBD. Many studies have reported similar results regarding this situation. In conclusion, compared to EBD, LHM appears to be effective in the treatment of achalasia in terms of symptom relief, improving quality of life, and patient satisfaction. However, LHM showed a lower complication rate than EBD. It shows that EBD increases the risk of bleeding and perforation, while LHM reduces these risks. In light of these findings, it was concluded that LHM is a safer and preferable intervention compared to EBD, considering the low risk of complications, improved quality of life, and high level of satisfaction. However, more studies with a larger sample size are needed to reach a more definite conclusion. |
Spatial organization of a soil cyanobacterium and its cyanosphere through GABA/Glu signaling to optimize mutualistic nitrogen fixation | 2527dc32-e8a9-4445-bfa4-09b7d7bc85a6 | 10881301 | Microbiology[mh] | Biological soil crusts (or biocrusts) are photosynthetic communities that play major roles in arid land soil stability and fertility, contributing significantly to the global C and N cycles [ - ]. Microcoleus vaginatus , a motile filamentous cyanobacterium, is the predominant pioneer of many biocrusts (see for reviews) and possibly the most abundant terrestrial cyanobacterium globally . One of its defining traits is the ability to aggregate into bundles of trichomes held within a common sheath through dynamic motility responses that help it attain macroscopic size and initially stabilize unconsolidated soils . While M. vaginatus cannot fix nitrogen, it does colonize N-limited bare soils due to its ability to spatially arrange other soil bacteria around it, repelling or avoiding competing cyanobacteria by reacting to their exudates and attracting a mutualistic “cyanosphere microbiome” that trades newly fixed nitrogen in exchange for its photosynthates [ - ]. Further, bundling seems to be an integral part of the mechanism for mutualistic interaction , as it is enhanced in culture by N limitation and by the presence of specific cyanosphere mutualists. But the mechanism by which these mutualistic partners find each other in the soil and optimize their spatial architecture is currently unknown. Theoretically, populations of cyanobacteria may need to remain as bundles close to populations of appropriate diazotrophs that remain sessile on the bundle’s common sheath in order to optimize C for N exchanges. As C for N symbioses often involve exchanges of nitrogenous compounds (such as ammonium or amino acids) via diffusion and active transport , we hypothesized that such compounds could play a secondary role in interspecies signaling to guide interactions, as happens in other cyanobacteria–heterotroph associations . To understand factors leading to bundle formation as a driver of biocrust microbiome self-organization, we investigated the effects of N source, nutrient status, and microbial interactions on the motile and bundling behavior of M. vaginatus , using both pedigreed, cultivated representative strains and naturally existing biocrust populations. This led us to uncover the signaling role of the gamma-aminobutyric acid/glutamate (GABA/Glu) system behind bundle formation and spatial organization of the biocrust microbiome.
Culturing and natural biocrust sourcing Microcoleus vaginatus (PCC 9802), an axenic cyanobacterial strain isolated from biocrusts in the US southwest and available through the Pasteur Culture Collection of Cyanobacteria (Paris, France), was maintained in 250-mL vented suspension culture flasks (Greiner Bio-One, Kremsmünster, Austria) in 50% BG11 medium at 23°C, 18–20 μE m −2 s −1 of white light illumination and 14 h light/10 dark cycle. To impose N-limitation, biomass was centrifugally washed (×3, 8000 rpm, 8 min) in N-free 50% BG11 0 medium and incubated for 14 days. The diazo-heterotrophic strains ( Bacillus sp. O64, Arthrobacter sp. O80, and Massilia sp. METH4), previously isolated from the cyanosphere of M. vaginatus were maintained on Burks +1% gellan gum medium in the dark. Biocrust used as a source of natural populations or for direct experimentation were sampled using 15-cm-diameter Petri dishes from the Jornada Basin LTER, NM (Chihuahuan Desert) or from Mesa, AZ (Sonoran Desert). Biocrusts were air dried and maintained inactive at 15% RH in darkness until experimentation. Biocrust chl a concentration, a proxy for photosynthetic biomass, was determined spectrophotometrically as previously described . Growth of M. vaginatus on different N-sources N-starved biomass was inoculated at 0.24 ± 0.04 mg chl a L −1 on 12-well plates containing 2 mL 50% BG11 0 supplemented with 10 mM (final) of different potential nitrogen sources ( n = 4 wells), unless otherwise stated. N-sources were nitrate, ammonium, urea, and amino acids representing all amino acid side groups (alanine, arginine, aspartate, cysteine, GABA, glutamine, glutamate, glycine, methionine, serine, tryptophan; Sigma-Aldrich, St. Louis, USA). All media were adjusted to pH 7 after the addition of amino acids or inorganic N, and NaCl was added to maintain equivalent ionic strength across treatments and controls. 100 μL homogenized aliquots from each well were taken initially and after 6 days, added to 900 μL of acetone in 2-mL microcentrifuge tubes containing 0.25 g of 0.5-mm zirconium beads, beat for 2 min at 30 s intervals, and then extracted in the dark for 24 h at 4°C. Chl a was determined from absorbance and absorption coefficients at 663 nm. Castenholz motility assays For “Castenholz Motility Assays” , M. vaginatus PCC 9802 suspensions were inoculated into 12-well plates containing 1 mL of appropriate liquid media and allowed to acclimate under standard culture conditions for 24 h before measurements were taken. The working volume of wells within the 12- and 24-well plates used for experimentation were 3 mL and 1 mL, respectively. All growth media were adjusted to pH 7 after the addition of amino acids or inorganic N and supplemented with NaCl to maintain equivalent ionic strength across treatments, when appropriate. Cultures were then mixed to homogeneity by repeated in-and-out pipetting, photographed, allowed to entangle for 60 min of contraction, and photographed again. Responses were quantified by image analyses in ImageJ as a percent change in areal cover of cyanobacterial biomass from dispersed to contracted states. Significance was tested with a one-way ANOVA using R , after arcsine square root transformation. Isolation and cleansing of environmental bundles To obtain natural bundles, biocrusts were wetted with sterile RO water for 1 h. Thereafter, individual cyanobacterial bundles were pulled under the dissecting scope using Watchmaker’s forceps, and either used directly or cleansed by dragging through sterile 2% agar plates to various degrees as needed for experiments and as previously video-documented . Repeated passages causes progressively larger losses of soil and heterotrophic populations attached to the bundle by simple sheer. In experiments for production of GABA and Glu, we conducted at least eight passes. Determination of trichome motility responses Trichome motility responses were measured on cultured material or in natural populations excised from biocrusts. Identification of the cyanobacteria in individual bundles as M. vaginatus was by microscopic observation. Once taxonomically identified, individual bundles were placed in 24-well plates containing 0.5 mL liquid medium under appropriate experimental treatments, and then, gliding speeds and direction reversal frequency of individual trichomes ( n = 12 for each) were measured under the compound microscope visually using an ocular micrometer and a stopwatch. Differences were tested for significance with a one-way ANOVA using R . Bundle stability assays Bundles excised from natural crusts and taxonomically identified were transferred to wells of 24-well plates containing 0.5 mL of 50% BG11 0 + 2% agar and the appropriate additions for the treatment at hand ( n = 12 for each), initially photographed under the dissecting scope, incubated for 24 h in standard culture conditions, and photographed again. Images were analyzed in ImageJ to determine the area occupied by trichomes spreading from the bundles, as well as the initial cross-sectional area of the bundle, from which an initial bundle volume of revolution was calculated. Stability was gauged as the final trichome spread area divided by the initial bundle volume. Differences in stability were tested with a one-way ANOVA using R . For experiments requiring glutamate decarboxylase (GAD), a 10 μM purified GAD solution in 50% BG11 0 was added. Heterologously expressed GAD from Bacteroides fragilis was purified in our laboratory. To visualize motile behavior in situ , intact biocrusts were fragmented into cm-sized pieces, distributed into six-well plates, supplemented with 1 mL of 50% BG11 0 supplemented as necessary ( n = 3), and incubated under standard culture conditions for 24 h. Photographs were taken initially and after 24 h for qualitative comparisons. Determination of GABA and glutamate We analyzed GABA and Glu concentrations in the biocrust pore water (soil solution) and in growth medium of cultures toward the end of exponential growth. For biocrusts, 10 g samples were placed on a polycarbonate filter in a Büchner funnel, saturated with 10 mL of sterilized RO water, incubated for 1 h, and then flushed with 35 mL of sterile RO water, which was collected through vacuum filtration ( n = 11). For heterotrophic cultures, 4-day-old spent medium was collected after centrifugation, and the pellet, which was cohesive, was weighed to determine biomass ( n = 4) after blotting over tissue paper to remove excess water. For cyanobacteria cultures, biomass was determined after 14 days through chl a concentrations ( n = 4) and converted to wet biomass using a chl a content of 1% of DW or 0.2% of wet weight . After collection, pore water or spent media were filter-sterilized and stored at 4°C for less than 48 h before filtration through 5 KDa MWCO 500-μL spin filters (Cytiva, Vivaspin) at 6000 g for 30 min and then stored at −20°C until analysis. GABA was measured spectrophotometrically by a coupled enzyme assay with a GABase mixture (Sigma) where GABA is converted to succinic semialdehyde and then to succinate with concurrent production of NADPH, whose absorbance can be measured at 340 nm and converted to concentrations using an extinction coefficient of 6220 M −1 cm −1 , and where [NADPH] = [GABA]. Reactions contained 25 μL of the sample and 75 μL of GABase assay mix (10 mM ßME, 2 mM α-ketoglutarate, 600 μM NADP+, 30 μg (0.015 U/mL) of GABase in 50 mM Tris–HCl, pH 8.6) and were incubated at room temperature for 90 min. For glutamate, samples were incubated overnight at 37°C with 5 μM of B. fragilis glutamate decarboxylase (GAD), heterologously expressed, and purified in our laboratory, in 50 mM sodium acetate, pH 4.7. GAD converts glutamate to GABA , which was measured by the GABase assay above. Sterile RO water or the appropriate uninoculated medium was used as controls. Initial GABA concentrations in samples without GAD addition were used as controls. Glu concentrations were calculated as: [Glu] = [GABA] with GAD – [GABA] without GAD.
Microcoleus vaginatus (PCC 9802), an axenic cyanobacterial strain isolated from biocrusts in the US southwest and available through the Pasteur Culture Collection of Cyanobacteria (Paris, France), was maintained in 250-mL vented suspension culture flasks (Greiner Bio-One, Kremsmünster, Austria) in 50% BG11 medium at 23°C, 18–20 μE m −2 s −1 of white light illumination and 14 h light/10 dark cycle. To impose N-limitation, biomass was centrifugally washed (×3, 8000 rpm, 8 min) in N-free 50% BG11 0 medium and incubated for 14 days. The diazo-heterotrophic strains ( Bacillus sp. O64, Arthrobacter sp. O80, and Massilia sp. METH4), previously isolated from the cyanosphere of M. vaginatus were maintained on Burks +1% gellan gum medium in the dark. Biocrust used as a source of natural populations or for direct experimentation were sampled using 15-cm-diameter Petri dishes from the Jornada Basin LTER, NM (Chihuahuan Desert) or from Mesa, AZ (Sonoran Desert). Biocrusts were air dried and maintained inactive at 15% RH in darkness until experimentation. Biocrust chl a concentration, a proxy for photosynthetic biomass, was determined spectrophotometrically as previously described .
M. vaginatus on different N-sources N-starved biomass was inoculated at 0.24 ± 0.04 mg chl a L −1 on 12-well plates containing 2 mL 50% BG11 0 supplemented with 10 mM (final) of different potential nitrogen sources ( n = 4 wells), unless otherwise stated. N-sources were nitrate, ammonium, urea, and amino acids representing all amino acid side groups (alanine, arginine, aspartate, cysteine, GABA, glutamine, glutamate, glycine, methionine, serine, tryptophan; Sigma-Aldrich, St. Louis, USA). All media were adjusted to pH 7 after the addition of amino acids or inorganic N, and NaCl was added to maintain equivalent ionic strength across treatments and controls. 100 μL homogenized aliquots from each well were taken initially and after 6 days, added to 900 μL of acetone in 2-mL microcentrifuge tubes containing 0.25 g of 0.5-mm zirconium beads, beat for 2 min at 30 s intervals, and then extracted in the dark for 24 h at 4°C. Chl a was determined from absorbance and absorption coefficients at 663 nm.
For “Castenholz Motility Assays” , M. vaginatus PCC 9802 suspensions were inoculated into 12-well plates containing 1 mL of appropriate liquid media and allowed to acclimate under standard culture conditions for 24 h before measurements were taken. The working volume of wells within the 12- and 24-well plates used for experimentation were 3 mL and 1 mL, respectively. All growth media were adjusted to pH 7 after the addition of amino acids or inorganic N and supplemented with NaCl to maintain equivalent ionic strength across treatments, when appropriate. Cultures were then mixed to homogeneity by repeated in-and-out pipetting, photographed, allowed to entangle for 60 min of contraction, and photographed again. Responses were quantified by image analyses in ImageJ as a percent change in areal cover of cyanobacterial biomass from dispersed to contracted states. Significance was tested with a one-way ANOVA using R , after arcsine square root transformation.
To obtain natural bundles, biocrusts were wetted with sterile RO water for 1 h. Thereafter, individual cyanobacterial bundles were pulled under the dissecting scope using Watchmaker’s forceps, and either used directly or cleansed by dragging through sterile 2% agar plates to various degrees as needed for experiments and as previously video-documented . Repeated passages causes progressively larger losses of soil and heterotrophic populations attached to the bundle by simple sheer. In experiments for production of GABA and Glu, we conducted at least eight passes.
Trichome motility responses were measured on cultured material or in natural populations excised from biocrusts. Identification of the cyanobacteria in individual bundles as M. vaginatus was by microscopic observation. Once taxonomically identified, individual bundles were placed in 24-well plates containing 0.5 mL liquid medium under appropriate experimental treatments, and then, gliding speeds and direction reversal frequency of individual trichomes ( n = 12 for each) were measured under the compound microscope visually using an ocular micrometer and a stopwatch. Differences were tested for significance with a one-way ANOVA using R .
Bundles excised from natural crusts and taxonomically identified were transferred to wells of 24-well plates containing 0.5 mL of 50% BG11 0 + 2% agar and the appropriate additions for the treatment at hand ( n = 12 for each), initially photographed under the dissecting scope, incubated for 24 h in standard culture conditions, and photographed again. Images were analyzed in ImageJ to determine the area occupied by trichomes spreading from the bundles, as well as the initial cross-sectional area of the bundle, from which an initial bundle volume of revolution was calculated. Stability was gauged as the final trichome spread area divided by the initial bundle volume. Differences in stability were tested with a one-way ANOVA using R . For experiments requiring glutamate decarboxylase (GAD), a 10 μM purified GAD solution in 50% BG11 0 was added. Heterologously expressed GAD from Bacteroides fragilis was purified in our laboratory. To visualize motile behavior in situ , intact biocrusts were fragmented into cm-sized pieces, distributed into six-well plates, supplemented with 1 mL of 50% BG11 0 supplemented as necessary ( n = 3), and incubated under standard culture conditions for 24 h. Photographs were taken initially and after 24 h for qualitative comparisons.
We analyzed GABA and Glu concentrations in the biocrust pore water (soil solution) and in growth medium of cultures toward the end of exponential growth. For biocrusts, 10 g samples were placed on a polycarbonate filter in a Büchner funnel, saturated with 10 mL of sterilized RO water, incubated for 1 h, and then flushed with 35 mL of sterile RO water, which was collected through vacuum filtration ( n = 11). For heterotrophic cultures, 4-day-old spent medium was collected after centrifugation, and the pellet, which was cohesive, was weighed to determine biomass ( n = 4) after blotting over tissue paper to remove excess water. For cyanobacteria cultures, biomass was determined after 14 days through chl a concentrations ( n = 4) and converted to wet biomass using a chl a content of 1% of DW or 0.2% of wet weight . After collection, pore water or spent media were filter-sterilized and stored at 4°C for less than 48 h before filtration through 5 KDa MWCO 500-μL spin filters (Cytiva, Vivaspin) at 6000 g for 30 min and then stored at −20°C until analysis. GABA was measured spectrophotometrically by a coupled enzyme assay with a GABase mixture (Sigma) where GABA is converted to succinic semialdehyde and then to succinate with concurrent production of NADPH, whose absorbance can be measured at 340 nm and converted to concentrations using an extinction coefficient of 6220 M −1 cm −1 , and where [NADPH] = [GABA]. Reactions contained 25 μL of the sample and 75 μL of GABase assay mix (10 mM ßME, 2 mM α-ketoglutarate, 600 μM NADP+, 30 μg (0.015 U/mL) of GABase in 50 mM Tris–HCl, pH 8.6) and were incubated at room temperature for 90 min. For glutamate, samples were incubated overnight at 37°C with 5 μM of B. fragilis glutamate decarboxylase (GAD), heterologously expressed, and purified in our laboratory, in 50 mM sodium acetate, pH 4.7. GAD converts glutamate to GABA , which was measured by the GABase assay above. Sterile RO water or the appropriate uninoculated medium was used as controls. Initial GABA concentrations in samples without GAD addition were used as controls. Glu concentrations were calculated as: [Glu] = [GABA] with GAD – [GABA] without GAD.
Effects of cyanosphere population size on bundle stability To test if cyanosphere populations help keep M. vaginatus trichomes within bundles, we excised bundles directly from biocrusts and subjected these natural populations to increasingly intense physical cleansing to remove heterotrophs attached to the bundle sheaths. We then incubated cleansed bundles on solid media under N limitation, which promotes bundle conformation , for 24 h and quantified the relative proportion of trichomes that had left the original bundle using the bundle stability assay (BSA; ). An inverse relationship between level of cleansing and bundle stability was evident and statistically significant (ANOVA, P < .01). Thus, the migration of trichomes away from bundles was positively correlated to a reduction of cyanosphere populations, which is consistent with the hypothesis that sheath-resident populations produce a diffusible cue that promotes bundle conformations in M. vaginatus. Nitrogen source effects on growth and aggregation of M. vaginatus PCC 9802 To establish potential nitrogenous metabolites that may support the C for N mutualism between M. vaginatus and its cyanosphere, we grew axenic M. vaginatus in nitrogen-free medium with equimolar additions of either inorganic compounds or amino acids as a sole source of N ( ). Microcoleus vaginatus grew best on nitrate and urea but could also grow well on ammonium, cysteine, glycine, and tryptophan, and less efficiently on alanine, arginine, and aspartate. Biomass yields on glutamate (Glu) and serine were no different from those in N-free controls (ANOVA, P = 1 and .43, respectively), indicating that they could neither serve as a N source for growth nor likely support the N transfer from heterotrophs to phototroph. To test if any of these compounds guide M. vaginatus motility responses, we used Castenholz’ clumping assay (see Materials and Methods and ) based on the quantitation of macroscopically visible motility-driven entanglement of trichomes in a liquid suspension. The assay represents an integrative assessment of several types of motility responses that can contribute to apparent aggregation. Aggregation dynamics were clearly influenced by the available N source ( ), with Glu having the strongest effect, followed by aspartate, both leading to significantly stronger aggregation than N-free controls (ANOVA, P < .001, for both). Other N sources either made no difference (alanine, glycine, methionine, urea) or significantly diminished aggregation over N-free controls (ANOVA, P < .05 for all remaining; ). Glu was the only compound tested that induced aggregation but could not be used as N source for growth, suggesting a specific role as a signaling molecule to motility behavior. The effect of aspartate, though significant, was not further studied here as aspartate could be used by M. vaginatus as a N source for growth. While aspartate is a common signaling compound in bacterial chemotaxis , the observed effect of aspartate could be potentially ascribed to a lack of specificity in transporters or signal processing proteins, given its molecular similarity with Glu. Responsivity to Glu after an initial pulse of exposure slowly faded away within 4–5 h of its removal ( ), indicating a tight regulation of sensitivity. Motility responses of M. vaginatus to glutamate To determine the basis for the macroscopic aggregation, we examined behavioral responses of M. vaginatus at the organismal (trichome) level. Gliding trichome speeds of N-starved strain PCC 9802 showed marked differences upon exposure to different N sources ( ). Glu elicited a positive chemokinetic response, with speeds almost doubling those of the N-free control (ANOVA, P < .01). Conversely, exposure to either ammonium or nitrate slowed trichomes to around one-third of controls (ANOVA, P < .01 for both, and nondifferent between them P = .97). While the motility responses of gliding cyanobacteria typically include modulations in the frequency of gliding direction reversal, model strain PCC 9802 apparently lost this capacity to any measurable extent through decades of laboratory cultivation. To compensate for this, and to investigate the universality of our culture-based findings, we extended assays to environmental bundles excised from biocrusts by micromanipulation. After 24-h incubation, environmental M. vaginatus trichomes also showed differing motility responses to various N sources. Gliding speeds of environmental M. vaginatus were typically higher than those measured in strain PCC 9802, but the differential responses were similar ( ). Compared to the N-free controls, trichomes exposed to Glu exhibited a slight, albeit nonsignificant (ANOVA, P = .26) positive chemokinesis, while trichomes exposed to ammonium and nitrate exhibited lower gliding speeds (also nonsignificant compared to controls; ANOVA, P = .48 and P = .60, respectively). However, differences between Glu and either ammonium or nitrate exposures were significant (ANOVA, P < .05 for both). Additionally, gliding reversal frequency was also differentially affected ( ). Glu doubled reversal frequency over controls (ANOVA, P < .001), whereas ammonium or nitrate halved the frequency, respectively (ANOVA, P < .05 for both). A different type of motility response (true chemotaxis, an ability to move in the direction of a gradient of attractant) can sometimes be ascribed to cyanobacterial responses. In our case, while we could demonstrate PCC 9802’s ability to respond phototactically, and while this strain responds chemotactically to other chemical cues , it did not respond chemotactically to Glu ( ). Glutamate as a cyanosphere match-maker One could predict from the previous results that exposure to Glu under N limitation will result in trichomes moving faster, but reversing much more often than in its absence. This will make the population stay in place, promoting trichome crowding within the bundle. Using the BSA ( ), we could consistently show that field trichomes exposed to Glu tended to remain in or near the bundle, while those exposed to nitrate or ammonium tended to leave the bundle and spread out, compared to N-free controls (ANOVA, P < .05). We could further test this directly in situ , taking advantage of observations that, if maintained under wet conditions for extended periods, motile biocrust cyanobacteria tend to leave their sheath “tracks” and spread as single trichomes over the soil surface , creating a thin greenish veil. We thus monitored the distribution of surface trichomes between bundled and free trichome conformations after incubation with solutions containing Glu, nitrate, or no additions. Indeed, differences were marked: glutamate enhanced the proportion of in-bundle trichomes, whereas the presence of nitrate promoted their spread as single trichomes ( ). This suggested that the unknown cyanosphere signaling factor from previous experiments could indeed be glutamate ( ). To test this strictly, we attempted to interrupt an alleged Glu-based interspecies communication by artificially lowering the extracellular levels of Glu in active mutualisms. To do this, we incubated natural bundles in the presence of purified glutamate decarboxylase (EC 4.1.1.15; GAD), which decarboxylates glutamate to gamma-aminobutyric acid (GABA). Theoretically, incubations with excess enzyme lasting 24 h should have brought Glu well below the concentrations that elicits responses (see below). Our prediction was that lowered extracellular concentrations would impede signaling and lead to eventual loss of bundle organization. However, we observed the exact opposite result: bundles subject to GAD activity (GAD+) were more stable than controls (GAD-) ( t -test, P = .003; ). Differential response of M. vaginatus to GABA and Glu A possible explanation for the unexpected result above is that GABA itself can elicit responses more strongly than Glu, so that GAD activity would in fact enhance the response. We then characterized aggregation responses of N-limited M. vaginatus PCC 9802 to a wide concentration range of both Glu and GABA. For Glu, maximal responses were attained between 100 μM and 10 mM ( ) and were barely detectable at 10 μM. GABA did indeed elicit swift aggregation as well, displaying a bimodal concentration dependence. A maximal response was attained at 100 nM GABA, with measurable responses down to 10 nM, and a second maximum detected around 10 mM ( ). As expected ( ), only N-starved M. vaginatus responded to Glu or GABA (ANOVA, P < .001 for either). To determine the motility mechanisms leading to aggregation and to extend our culture-based findings to natural populations, we compared motility responses in environmental M. vaginatus at low (1 μM) and high (10 mM) concentrations of GABA and Glu ( ). Gliding speeds increased slightly with exposure to either 1 μM GABA or 10 mM Glu over the N-free controls, but they did not with exposure to either 1 μM Glu or 10 mM GABA. However, these differences were not significant (ANOVA, P > .65 for all; ). The effects on reversal rates were more marked. One micromolar GABA or 10 mM Glu doubled reversal frequency and were different from the control (ANOVA, P < .001) but were not significantly different from each other (ANOVA, P = .997). Predictably, 1 μM Glu produced no significant changes in reversal frequency over controls (ANOVA, P = .94). However, 10 mM GABA did not increase reversal rates, but rather, more than halved its frequency (ANOVA, P < .03; ). That GABA increased reversal frequency like Glu did, but at 1000-fold lower concentrations, suggests that such low concentrations of GABA would also be sufficient to induce increased bundle stability as was observed with Glu. Indeed, we found that to be the case: 1 μM GABA had a similar bundle stabilizing effect as 10 mM Glu (ANOVA, P = .97), significantly increasing stability over control or 1 μM Glu (ANOVA, P < .08 and .11, respectively; ). The bundle destabilizing effect of GABA at high concentration that one could have predicted from the halving of reversal frequency did indeed also materialize experimentally as significant differences in trichome spread between 10 mM GABA, controls, and all other treatments ( ). This response divergence between GABA and Glu indicates that their sensing relies on somewhat differentiated mechanisms. Combining the information obtained from cultures ( ) and from environmental bundles ( ), it is possible to deduce the dependence of bundle stability as a function of Glu and GABA concentrations ( ). GABA/Glu production in bacterial cultures and natural communities The responsivity to Glu and GABA ( ) highlights their potential as a signaling metabolite for M. vaginatus under N limitation. If true, GABA or Glu should be produced by at least some of the mutualistic partners, released to the exometabolome under appropriate environmental conditions, and accumulate in the environment to concentrations capable of inducing motility responses. To determine this, we measured GABA and Glu concentrations in natural biocrusts varying in origin, collection season, and level of development. While concentrations were quite variable among samples, both Glu and GABA were detectable. Mean Glu concentrations in biocrusts reached 71 ± 5 μmol L −1 of soil ( n = 11; ranging from 25 to 138 μmol L −1 of soil) and GABA attained 11 ± 15 μmol L −1 of soil ( n = 20; ranging from undetectable to 64 μmol L −1 of soil). The range of concentrations measured in natural biocrusts ( ) does indeed make it possible for both compounds to play a role in motility responses, judging from the concentration ranges for GABA and Glu that induced motility responses in our experiments ( ). To test for potential sources of GABA and Glu, we analyzed the spent growth medium of the mutualistic partners in culture ( M. vaginatus PCC 9802 and three cyanosphere heterotrophs: Bacillus sp. O64, Arthrobacter sp. O80, and Massilia sp. METH4). Under N replete conditions, M. vaginatus itself produced barely detectable levels of GABA; around 0.08 μmol g −1 of biomass d −1 , reaching extracellular concentrations around 2 μM. However, production was enhanced more than 20-fold under N-limitation; around 1.73 μmol g −1 of biomass d −1 , reaching media concentrations around 52 μM ( ). None of the cyanosphere heterotrophs produced detectable GABA under carbon-replete conditions, but all three produced some under C limitation ( ). Arthrobacter sp. O80 produced 0.74 μmol GABA g −1 of biomass d −1 (reaching 3.8 ± 2.2 μM in the medium), Massilia sp. METH4 produced 3.82 μmol g −1 of biomass d −1 (reaching 17.2 ± 2.1 μM in the medium), and Bacillus sp. O64 produced 0.07 μmol g −1 of biomass d −1 (reaching 1.9 μM in the medium). Thus, not only members of the cyanosphere but also M. vaginatus itself are potential contributors to the GABA signaling pool when under conditions leading to mutualism (C or N limitation, respectively). Glutamate, by contrast, was only produced in measurable quantities by M. vaginatus PCC 9802 and Massilia sp. METH4 . The cyanobacterium produced it under N-replete and N-free conditions, though production was significantly higher under N limitation (2.01 μmol g −1 of biomass d −1 ; t -test, P = .006) than under N-replete growth (1.42 μmol g −1 of biomass d −1 ), reaching concentrations 10–20 μM in the medium. In Massilia sp. METH4, Glu production (2.94 μmol g −1 of biomass d −1 ) occurred only under C limitation, with media concentrations around 13 μM. Again here, both cyanosphere members and M. vaginatus are potential contributors to the Glu signaling pool in a manner that is enhanced under conditions leading to mutualism. In order to ascertain the relative role of heterotrophic vs. phototrophic mutualistic partners in driving signaling compound concentrations under natural conditions, we carried out an experiment in which environmental bundles of M. vaginatus were collected directly from soil, incubated in the light together with their surrounding cyanosphere bacterial population intact (i.e. uncleansed), or reduced (cleansed) to determine their production potential for Glu and GABA. In the uncleansed, mixed communities, we detected production of both Glu and GABA, Glu production exceeding that of GABA by about 30-fold ( ). Cleansing of the cyanosphere resulted in significant decreases ( t -test, P = .011; by about 2/3) of Glu production and also in a decrease in GABA to undetectable levels, although this decrease was not statistically significant ( t -test, P = .37) ( ) since we were operating close to the analytical limit of GABA detection of our assay.
To test if cyanosphere populations help keep M. vaginatus trichomes within bundles, we excised bundles directly from biocrusts and subjected these natural populations to increasingly intense physical cleansing to remove heterotrophs attached to the bundle sheaths. We then incubated cleansed bundles on solid media under N limitation, which promotes bundle conformation , for 24 h and quantified the relative proportion of trichomes that had left the original bundle using the bundle stability assay (BSA; ). An inverse relationship between level of cleansing and bundle stability was evident and statistically significant (ANOVA, P < .01). Thus, the migration of trichomes away from bundles was positively correlated to a reduction of cyanosphere populations, which is consistent with the hypothesis that sheath-resident populations produce a diffusible cue that promotes bundle conformations in M. vaginatus.
M. vaginatus PCC 9802 To establish potential nitrogenous metabolites that may support the C for N mutualism between M. vaginatus and its cyanosphere, we grew axenic M. vaginatus in nitrogen-free medium with equimolar additions of either inorganic compounds or amino acids as a sole source of N ( ). Microcoleus vaginatus grew best on nitrate and urea but could also grow well on ammonium, cysteine, glycine, and tryptophan, and less efficiently on alanine, arginine, and aspartate. Biomass yields on glutamate (Glu) and serine were no different from those in N-free controls (ANOVA, P = 1 and .43, respectively), indicating that they could neither serve as a N source for growth nor likely support the N transfer from heterotrophs to phototroph. To test if any of these compounds guide M. vaginatus motility responses, we used Castenholz’ clumping assay (see Materials and Methods and ) based on the quantitation of macroscopically visible motility-driven entanglement of trichomes in a liquid suspension. The assay represents an integrative assessment of several types of motility responses that can contribute to apparent aggregation. Aggregation dynamics were clearly influenced by the available N source ( ), with Glu having the strongest effect, followed by aspartate, both leading to significantly stronger aggregation than N-free controls (ANOVA, P < .001, for both). Other N sources either made no difference (alanine, glycine, methionine, urea) or significantly diminished aggregation over N-free controls (ANOVA, P < .05 for all remaining; ). Glu was the only compound tested that induced aggregation but could not be used as N source for growth, suggesting a specific role as a signaling molecule to motility behavior. The effect of aspartate, though significant, was not further studied here as aspartate could be used by M. vaginatus as a N source for growth. While aspartate is a common signaling compound in bacterial chemotaxis , the observed effect of aspartate could be potentially ascribed to a lack of specificity in transporters or signal processing proteins, given its molecular similarity with Glu. Responsivity to Glu after an initial pulse of exposure slowly faded away within 4–5 h of its removal ( ), indicating a tight regulation of sensitivity.
M. vaginatus to glutamate To determine the basis for the macroscopic aggregation, we examined behavioral responses of M. vaginatus at the organismal (trichome) level. Gliding trichome speeds of N-starved strain PCC 9802 showed marked differences upon exposure to different N sources ( ). Glu elicited a positive chemokinetic response, with speeds almost doubling those of the N-free control (ANOVA, P < .01). Conversely, exposure to either ammonium or nitrate slowed trichomes to around one-third of controls (ANOVA, P < .01 for both, and nondifferent between them P = .97). While the motility responses of gliding cyanobacteria typically include modulations in the frequency of gliding direction reversal, model strain PCC 9802 apparently lost this capacity to any measurable extent through decades of laboratory cultivation. To compensate for this, and to investigate the universality of our culture-based findings, we extended assays to environmental bundles excised from biocrusts by micromanipulation. After 24-h incubation, environmental M. vaginatus trichomes also showed differing motility responses to various N sources. Gliding speeds of environmental M. vaginatus were typically higher than those measured in strain PCC 9802, but the differential responses were similar ( ). Compared to the N-free controls, trichomes exposed to Glu exhibited a slight, albeit nonsignificant (ANOVA, P = .26) positive chemokinesis, while trichomes exposed to ammonium and nitrate exhibited lower gliding speeds (also nonsignificant compared to controls; ANOVA, P = .48 and P = .60, respectively). However, differences between Glu and either ammonium or nitrate exposures were significant (ANOVA, P < .05 for both). Additionally, gliding reversal frequency was also differentially affected ( ). Glu doubled reversal frequency over controls (ANOVA, P < .001), whereas ammonium or nitrate halved the frequency, respectively (ANOVA, P < .05 for both). A different type of motility response (true chemotaxis, an ability to move in the direction of a gradient of attractant) can sometimes be ascribed to cyanobacterial responses. In our case, while we could demonstrate PCC 9802’s ability to respond phototactically, and while this strain responds chemotactically to other chemical cues , it did not respond chemotactically to Glu ( ).
One could predict from the previous results that exposure to Glu under N limitation will result in trichomes moving faster, but reversing much more often than in its absence. This will make the population stay in place, promoting trichome crowding within the bundle. Using the BSA ( ), we could consistently show that field trichomes exposed to Glu tended to remain in or near the bundle, while those exposed to nitrate or ammonium tended to leave the bundle and spread out, compared to N-free controls (ANOVA, P < .05). We could further test this directly in situ , taking advantage of observations that, if maintained under wet conditions for extended periods, motile biocrust cyanobacteria tend to leave their sheath “tracks” and spread as single trichomes over the soil surface , creating a thin greenish veil. We thus monitored the distribution of surface trichomes between bundled and free trichome conformations after incubation with solutions containing Glu, nitrate, or no additions. Indeed, differences were marked: glutamate enhanced the proportion of in-bundle trichomes, whereas the presence of nitrate promoted their spread as single trichomes ( ). This suggested that the unknown cyanosphere signaling factor from previous experiments could indeed be glutamate ( ). To test this strictly, we attempted to interrupt an alleged Glu-based interspecies communication by artificially lowering the extracellular levels of Glu in active mutualisms. To do this, we incubated natural bundles in the presence of purified glutamate decarboxylase (EC 4.1.1.15; GAD), which decarboxylates glutamate to gamma-aminobutyric acid (GABA). Theoretically, incubations with excess enzyme lasting 24 h should have brought Glu well below the concentrations that elicits responses (see below). Our prediction was that lowered extracellular concentrations would impede signaling and lead to eventual loss of bundle organization. However, we observed the exact opposite result: bundles subject to GAD activity (GAD+) were more stable than controls (GAD-) ( t -test, P = .003; ).
M. vaginatus to GABA and Glu A possible explanation for the unexpected result above is that GABA itself can elicit responses more strongly than Glu, so that GAD activity would in fact enhance the response. We then characterized aggregation responses of N-limited M. vaginatus PCC 9802 to a wide concentration range of both Glu and GABA. For Glu, maximal responses were attained between 100 μM and 10 mM ( ) and were barely detectable at 10 μM. GABA did indeed elicit swift aggregation as well, displaying a bimodal concentration dependence. A maximal response was attained at 100 nM GABA, with measurable responses down to 10 nM, and a second maximum detected around 10 mM ( ). As expected ( ), only N-starved M. vaginatus responded to Glu or GABA (ANOVA, P < .001 for either). To determine the motility mechanisms leading to aggregation and to extend our culture-based findings to natural populations, we compared motility responses in environmental M. vaginatus at low (1 μM) and high (10 mM) concentrations of GABA and Glu ( ). Gliding speeds increased slightly with exposure to either 1 μM GABA or 10 mM Glu over the N-free controls, but they did not with exposure to either 1 μM Glu or 10 mM GABA. However, these differences were not significant (ANOVA, P > .65 for all; ). The effects on reversal rates were more marked. One micromolar GABA or 10 mM Glu doubled reversal frequency and were different from the control (ANOVA, P < .001) but were not significantly different from each other (ANOVA, P = .997). Predictably, 1 μM Glu produced no significant changes in reversal frequency over controls (ANOVA, P = .94). However, 10 mM GABA did not increase reversal rates, but rather, more than halved its frequency (ANOVA, P < .03; ). That GABA increased reversal frequency like Glu did, but at 1000-fold lower concentrations, suggests that such low concentrations of GABA would also be sufficient to induce increased bundle stability as was observed with Glu. Indeed, we found that to be the case: 1 μM GABA had a similar bundle stabilizing effect as 10 mM Glu (ANOVA, P = .97), significantly increasing stability over control or 1 μM Glu (ANOVA, P < .08 and .11, respectively; ). The bundle destabilizing effect of GABA at high concentration that one could have predicted from the halving of reversal frequency did indeed also materialize experimentally as significant differences in trichome spread between 10 mM GABA, controls, and all other treatments ( ). This response divergence between GABA and Glu indicates that their sensing relies on somewhat differentiated mechanisms. Combining the information obtained from cultures ( ) and from environmental bundles ( ), it is possible to deduce the dependence of bundle stability as a function of Glu and GABA concentrations ( ).
The responsivity to Glu and GABA ( ) highlights their potential as a signaling metabolite for M. vaginatus under N limitation. If true, GABA or Glu should be produced by at least some of the mutualistic partners, released to the exometabolome under appropriate environmental conditions, and accumulate in the environment to concentrations capable of inducing motility responses. To determine this, we measured GABA and Glu concentrations in natural biocrusts varying in origin, collection season, and level of development. While concentrations were quite variable among samples, both Glu and GABA were detectable. Mean Glu concentrations in biocrusts reached 71 ± 5 μmol L −1 of soil ( n = 11; ranging from 25 to 138 μmol L −1 of soil) and GABA attained 11 ± 15 μmol L −1 of soil ( n = 20; ranging from undetectable to 64 μmol L −1 of soil). The range of concentrations measured in natural biocrusts ( ) does indeed make it possible for both compounds to play a role in motility responses, judging from the concentration ranges for GABA and Glu that induced motility responses in our experiments ( ). To test for potential sources of GABA and Glu, we analyzed the spent growth medium of the mutualistic partners in culture ( M. vaginatus PCC 9802 and three cyanosphere heterotrophs: Bacillus sp. O64, Arthrobacter sp. O80, and Massilia sp. METH4). Under N replete conditions, M. vaginatus itself produced barely detectable levels of GABA; around 0.08 μmol g −1 of biomass d −1 , reaching extracellular concentrations around 2 μM. However, production was enhanced more than 20-fold under N-limitation; around 1.73 μmol g −1 of biomass d −1 , reaching media concentrations around 52 μM ( ). None of the cyanosphere heterotrophs produced detectable GABA under carbon-replete conditions, but all three produced some under C limitation ( ). Arthrobacter sp. O80 produced 0.74 μmol GABA g −1 of biomass d −1 (reaching 3.8 ± 2.2 μM in the medium), Massilia sp. METH4 produced 3.82 μmol g −1 of biomass d −1 (reaching 17.2 ± 2.1 μM in the medium), and Bacillus sp. O64 produced 0.07 μmol g −1 of biomass d −1 (reaching 1.9 μM in the medium). Thus, not only members of the cyanosphere but also M. vaginatus itself are potential contributors to the GABA signaling pool when under conditions leading to mutualism (C or N limitation, respectively). Glutamate, by contrast, was only produced in measurable quantities by M. vaginatus PCC 9802 and Massilia sp. METH4 . The cyanobacterium produced it under N-replete and N-free conditions, though production was significantly higher under N limitation (2.01 μmol g −1 of biomass d −1 ; t -test, P = .006) than under N-replete growth (1.42 μmol g −1 of biomass d −1 ), reaching concentrations 10–20 μM in the medium. In Massilia sp. METH4, Glu production (2.94 μmol g −1 of biomass d −1 ) occurred only under C limitation, with media concentrations around 13 μM. Again here, both cyanosphere members and M. vaginatus are potential contributors to the Glu signaling pool in a manner that is enhanced under conditions leading to mutualism. In order to ascertain the relative role of heterotrophic vs. phototrophic mutualistic partners in driving signaling compound concentrations under natural conditions, we carried out an experiment in which environmental bundles of M. vaginatus were collected directly from soil, incubated in the light together with their surrounding cyanosphere bacterial population intact (i.e. uncleansed), or reduced (cleansed) to determine their production potential for Glu and GABA. In the uncleansed, mixed communities, we detected production of both Glu and GABA, Glu production exceeding that of GABA by about 30-fold ( ). Cleansing of the cyanosphere resulted in significant decreases ( t -test, P = .011; by about 2/3) of Glu production and also in a decrease in GABA to undetectable levels, although this decrease was not statistically significant ( t -test, P = .37) ( ) since we were operating close to the analytical limit of GABA detection of our assay.
GABA/Glu as microbial signaling molecules GABA and glutamate are important signaling molecules in animals, acting as potent neurotransmitters , as well as in plants where the role of GABA signaling has also been recently established . In some cases, plant GABA influences interactions between plants and associated pathogenic or mutualistic microbes [ , , ], and some of the GABA-mediated control of microbial behavior described in the literature is exerted through interference with bacterial quorum sensing systems . GABA and Glu are commonly synthesized and metabolized by a large variety of bacteria as a source of C or N, and many bacteria can chemotactically respond to them , but there were until now no reports of either compound role in signaling among microbes. There is, however, much scientific interest in understanding microbial interactions with GABA and Glu as evidence suggests that the gut microbiome could influence human neurology through the gut–brain axis by their consumption or production . The research presented here clearly speaks for the signaling role of the amino acid pair in mutualistic interactions within soil microbiomes in that (i) they directly elicit behavioral responses in the cyanobacterial partner, (ii) they are produced and excreted by various community members to concentrations sufficiently high to elicit these responses, and (iii) this signaling occurs in a manner regulated in response to their informational value, rather than nutritional value ( ). It is possible that GABA constitutes the biochemically true signaling compound at the receiving end and that the much weaker responses to Glu are the consequence of bleed-through from extracellular decarboxylation of Glu to GABA (enzymatic or spontaneous) in the soil or in culture. In the end, however, this is immaterial for the phenomenon at hand. Likewise, it is as yet unknown if the signaling role of the Glu/GABA pair we established here could be generalizable to microbiomes other than biocrusts, but the potential for interference between fully microbial and “trans-kingdom” signaling would make this worth investigating in animal or plant host–associated microbiomes. GABA/Glu as cyanobacterial autoinducers for N starvation We demonstrate that both GABA and Glu have the characteristics of autoinducers in the motility responses of M. vaginatus to N starvation . Quorum-sensing, or autoinduction, is a form of intercellular, intraspecific signaling dependent on the density of the population and other factors that allow populations to behave in a coordinated fashion . It is widespread in bacteria and has been reported in some cyanobacteria . Glu excretion is enhanced by N starvation and always leads to filament aggregation (bundle formation), but M. vaginatus is relatively insensitive to it. By contrast, it is extremely sensitive to GABA, down to nM concentrations, which is a typical threshold in known quorum-sensing autoinducers . Among cyanobacteria, GABA biosynthesis relies on the conversion of glutamate through GAD , and GABA does accumulate to variable levels among strains intracellularly . While it is plausible that other biocrust cyanobacteria could exude these compounds under similar conditions, we could not test this hypothesis as other axenic strains of relevant pioneer cyanobacteria do not exist in culture collections. However, we found no prior reports of GABA excretion in cyanobacteria nor about its role as autoinducer in any bacterium in the literature. Consistent with this notion, the GAD gene is detectable in multiple publicly available M. vaginatus genomes , and genes related to quorum sensing, such as acyl-homoserine lactone acylase homologs, are consistently found in its close genomic neighborhood. Mechanism of action for autoinduced, size-constrained bundle formation Together, the phenomenology of M. vaginatus signal responses suggests how the system may work in a manner consistent with observations in nature and in culture. Under N-replete conditions, any effects of self-excreted Glu ( ) will be denied through motility responses elicited by inorganic N, as seen in the preference for single-trichome conformation under nitrogen-replete conditions ( ). Consistently, M. vaginatus is rarely seen as single trichomes in nature, as most biocrusts are N-limited . Under N-limitation, however, self-excreted GABA/Glu will start to accumulate around single trichomes, promoting the formation of bundles in a self-enhancing manner at random locations. Eventually, these bundles will grow to contain many trichomes, the local concentrations of self-produced GABA rising to its second, opposite activity peak, now counteracting the bundle-forming effects of Glu alone, and halting the process or resulting in active bundle deconstruction. The dual nature of the responses to Glu and GABA, thus enables a first-positive-then-negative feedback loop that will cause bundles to initially form and then reach an equilibrium at a certain maximal size. Role of the cyanosphere It is clear that bundle formation based on mere autoinduction as discussed in the previous section has no bearing on an eventual improvement of the N-limitation condition. To account for this, one must include interactions with diazotrophic members of the bundle-associated cyanosphere . Logically, this would be accomplished if M. vaginatus would form bundles preferentially where populations of heterotrophic mutualists exist and particularly where they exist “primed” for metabolic exchanges through C starvation. The cyanosphere depletion experiments presented direct evidence for this ( ), showing that the cyanosphere directly helps keep M. vaginatus in its bundled conformation. This can be traced back to the GABA/Glu mechanism by showing that the production of signals in environmental bundles was to a large extent ascribable to cyanosphere mutualists ( ). This conclusion is further supported by the direct measurement of either GABA and/or Glu excretion in pedigreed heterotrophic diazotrophs in culture ( ) and by the fact that heterotroph production is clearly dependent on a physiological C limitation of growth. The fact that C-limited bacteria excrete valuable amino acids can only be understood if this brings some fitness benefit in ending C limitation, which is consistent with the proven role of attracting M. vaginatus to form C-exuding bundles to their proximity. Logically, the combined effects of cyanosphere-produced Glu and GABA will attract trichomes only until the autoinducing cyanobacterial GABA slows further recruitment, likely so that the provision of N for the cyanobacteria is still sufficient. A second feedback loop is also likely at play when bundles become too large around heterotrophic populations: cyanobacterial C excretions could relieve the limitations status of the cyanosphere, thus decreasing Glu/GABA signaling locally and leading to bundle disassembly. In this manner, cyanobacterial bundles will develop preferentially where diazotrophic populations exist and attain sizes commensurate to the heterotroph population density. This represents a mechanistically simple means of optimizing the exchange of C for N spatially within the soil. The importance of chemokinesis If the only beneficial outcome of responses to GABA/Glu was to stay close to a sheath-bound cyanosphere, it would principally suffice to stop gliding, rather than to glide faster and reverse more often concurrently, as we found ( ). But if all trichomes within a bundle (easily tens to hundreds) were to remain static, those in the bundle core would receive considerably less, if any, benefits than those in the periphery. By maintaining active gliding, trichomes glide against each other in their typical ropelike conformation, randomizing their relative positions in time ( ), distributing resources more equally, while the negative chemophobic response ensures that the population as a whole remains in the neighborhood where resources are found.
GABA and glutamate are important signaling molecules in animals, acting as potent neurotransmitters , as well as in plants where the role of GABA signaling has also been recently established . In some cases, plant GABA influences interactions between plants and associated pathogenic or mutualistic microbes [ , , ], and some of the GABA-mediated control of microbial behavior described in the literature is exerted through interference with bacterial quorum sensing systems . GABA and Glu are commonly synthesized and metabolized by a large variety of bacteria as a source of C or N, and many bacteria can chemotactically respond to them , but there were until now no reports of either compound role in signaling among microbes. There is, however, much scientific interest in understanding microbial interactions with GABA and Glu as evidence suggests that the gut microbiome could influence human neurology through the gut–brain axis by their consumption or production . The research presented here clearly speaks for the signaling role of the amino acid pair in mutualistic interactions within soil microbiomes in that (i) they directly elicit behavioral responses in the cyanobacterial partner, (ii) they are produced and excreted by various community members to concentrations sufficiently high to elicit these responses, and (iii) this signaling occurs in a manner regulated in response to their informational value, rather than nutritional value ( ). It is possible that GABA constitutes the biochemically true signaling compound at the receiving end and that the much weaker responses to Glu are the consequence of bleed-through from extracellular decarboxylation of Glu to GABA (enzymatic or spontaneous) in the soil or in culture. In the end, however, this is immaterial for the phenomenon at hand. Likewise, it is as yet unknown if the signaling role of the Glu/GABA pair we established here could be generalizable to microbiomes other than biocrusts, but the potential for interference between fully microbial and “trans-kingdom” signaling would make this worth investigating in animal or plant host–associated microbiomes.
We demonstrate that both GABA and Glu have the characteristics of autoinducers in the motility responses of M. vaginatus to N starvation . Quorum-sensing, or autoinduction, is a form of intercellular, intraspecific signaling dependent on the density of the population and other factors that allow populations to behave in a coordinated fashion . It is widespread in bacteria and has been reported in some cyanobacteria . Glu excretion is enhanced by N starvation and always leads to filament aggregation (bundle formation), but M. vaginatus is relatively insensitive to it. By contrast, it is extremely sensitive to GABA, down to nM concentrations, which is a typical threshold in known quorum-sensing autoinducers . Among cyanobacteria, GABA biosynthesis relies on the conversion of glutamate through GAD , and GABA does accumulate to variable levels among strains intracellularly . While it is plausible that other biocrust cyanobacteria could exude these compounds under similar conditions, we could not test this hypothesis as other axenic strains of relevant pioneer cyanobacteria do not exist in culture collections. However, we found no prior reports of GABA excretion in cyanobacteria nor about its role as autoinducer in any bacterium in the literature. Consistent with this notion, the GAD gene is detectable in multiple publicly available M. vaginatus genomes , and genes related to quorum sensing, such as acyl-homoserine lactone acylase homologs, are consistently found in its close genomic neighborhood.
Together, the phenomenology of M. vaginatus signal responses suggests how the system may work in a manner consistent with observations in nature and in culture. Under N-replete conditions, any effects of self-excreted Glu ( ) will be denied through motility responses elicited by inorganic N, as seen in the preference for single-trichome conformation under nitrogen-replete conditions ( ). Consistently, M. vaginatus is rarely seen as single trichomes in nature, as most biocrusts are N-limited . Under N-limitation, however, self-excreted GABA/Glu will start to accumulate around single trichomes, promoting the formation of bundles in a self-enhancing manner at random locations. Eventually, these bundles will grow to contain many trichomes, the local concentrations of self-produced GABA rising to its second, opposite activity peak, now counteracting the bundle-forming effects of Glu alone, and halting the process or resulting in active bundle deconstruction. The dual nature of the responses to Glu and GABA, thus enables a first-positive-then-negative feedback loop that will cause bundles to initially form and then reach an equilibrium at a certain maximal size.
It is clear that bundle formation based on mere autoinduction as discussed in the previous section has no bearing on an eventual improvement of the N-limitation condition. To account for this, one must include interactions with diazotrophic members of the bundle-associated cyanosphere . Logically, this would be accomplished if M. vaginatus would form bundles preferentially where populations of heterotrophic mutualists exist and particularly where they exist “primed” for metabolic exchanges through C starvation. The cyanosphere depletion experiments presented direct evidence for this ( ), showing that the cyanosphere directly helps keep M. vaginatus in its bundled conformation. This can be traced back to the GABA/Glu mechanism by showing that the production of signals in environmental bundles was to a large extent ascribable to cyanosphere mutualists ( ). This conclusion is further supported by the direct measurement of either GABA and/or Glu excretion in pedigreed heterotrophic diazotrophs in culture ( ) and by the fact that heterotroph production is clearly dependent on a physiological C limitation of growth. The fact that C-limited bacteria excrete valuable amino acids can only be understood if this brings some fitness benefit in ending C limitation, which is consistent with the proven role of attracting M. vaginatus to form C-exuding bundles to their proximity. Logically, the combined effects of cyanosphere-produced Glu and GABA will attract trichomes only until the autoinducing cyanobacterial GABA slows further recruitment, likely so that the provision of N for the cyanobacteria is still sufficient. A second feedback loop is also likely at play when bundles become too large around heterotrophic populations: cyanobacterial C excretions could relieve the limitations status of the cyanosphere, thus decreasing Glu/GABA signaling locally and leading to bundle disassembly. In this manner, cyanobacterial bundles will develop preferentially where diazotrophic populations exist and attain sizes commensurate to the heterotroph population density. This represents a mechanistically simple means of optimizing the exchange of C for N spatially within the soil.
If the only beneficial outcome of responses to GABA/Glu was to stay close to a sheath-bound cyanosphere, it would principally suffice to stop gliding, rather than to glide faster and reverse more often concurrently, as we found ( ). But if all trichomes within a bundle (easily tens to hundreds) were to remain static, those in the bundle core would receive considerably less, if any, benefits than those in the periphery. By maintaining active gliding, trichomes glide against each other in their typical ropelike conformation, randomizing their relative positions in time ( ), distributing resources more equally, while the negative chemophobic response ensures that the population as a whole remains in the neighborhood where resources are found.
We provide evidence for the role of the Glu/GABA pair as interspecies signaling molecules responsible for spatial organization of biocrust microbiomes to optimize C for N mutualisms, and for their role as an autoinducer molecule in the regulation of cyanobacterial bundle formation and size under N limitation. Further investigation into the role of GABA/Glu as signaling compounds in other microbiomes may be of interest, given its role in animal and plants systems as well.
Corey Nelson (research concept, manuscript writing, experimental design, data analyses, figure preparation, results discussion, review manuscriot). Ferran Garcia-Pichel (research concept, manuscript writing, experimental design, data analyses, figure preparation, review manuscript). Pavandi Dadi (perform experiments an analyses, discuss results, review manuscript). Dhara Shah (perform experimenta an analyses, discuss results, review manuscript).
The authors declare no conflict of interest.
This work was supported in part by the Jornada Basin LTER Graduate Research Fellowship Program to C.N. (DEB 2025166) and by National Science Foundation (DEB 2129537) to F.G.-P.
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.
Supplementary_Information_wrad029 Supplementary_videos_wrad029
|
Extremozymes and compatible solute production potential of halophilic and halotolerant bacteria isolated from crop rhizospheric soils of Southwest Saurashtra Gujarat | 583d846e-6893-425a-ac0c-bce7bd3e4f8d | 11231302 | Microbiology[mh] | Extremophiles are organisms that can thrive in environments with extreme conditions such as extreme temperature, pressure, radiation, salinity, pH levels, etc. , that have traditionally been considered inhospitable for life . These organisms cope with harsh environments by adopting certain unique strategies for survival. For instance, the first reported extremophiles termed halophiles are known to have developed two classic strategies called ‘high-salt-in-cytoplasm’ and ‘salt-out-in-cytoplasm’ or ‘low-salt-high-compatible-solute-in-cytoplasm’ for osmoadaptation . The former mechanism involves the accumulation of intracellular KCl concentrations higher than the external NaCl concentration to maintain the turgor pressure . However, the resulting high ionic strength in their cytoplasm cost them a compensatory evolutionary change from normal to acidic proteome to keep the proteins soluble for maintaining normal functionality of key cellular activities thereby confining their adaptability to only hypersaline habitats without frequent fluctuations to typically 5 M NaCl or more, found mostly in extreme halophiles such as Halobacteriales (archaea), Salinibacter ruber (bacteria) and Haloanaerobiales (anaerobic moderate halophilic bacteria) , . Although energetically more expensive, the latter strategy employs a physiologically much more flexible mechanism involving either an accumulation of high concentrations of organic compatible solutes in the cytoplasm from external environments or their de-novo synthesis for osmotic adjustments thereby circumventing the long-lasting large-scale accumulation of ions . This strategy facilitates the adaptability of the organisms possessing them to a wide range of salinities (typically 0.5–3 M NaCl). It is found largely in halotolerant and moderate halophiles . These halophilic extremophiles also produce a special and unique enzyme called extremozymes to survive in intolerably hostile environments . These extremozymes are known for their promising capability to withstand unusual extreme conditions required in industrial product synthesis processes where the mesophilic enzymes usually precipitate or denature . They reportedly replace chemical catalysts in many industries, such as manufacturing chemicals, textiles, pharmaceuticals, detergents, food, paper, etc. , . Halophiles, reportedly one of the most important groups of such extremophiles are microorganisms that can flourish in environments with very high salt concentrations. They include members of all three domains of life, viz. Archaea, Bacteria, and Eukarya. In contrast, bacteria that can tolerate relatively high NaCl concentrations and grow regardless of salt's presence or absence are labeled halotolerant . Halophilic microorganisms have been reported as an excellent source of extremozyme called halozymes that can function stably under high salt concentrations and withstand high temperatures, alkaline pH, toxicants, etc. encountered in many industrial bioconversion processes . The polyextremophilic nature of their enzymes makes halophiles a potential candidate for meeting the current industrial enzyme demands. Many scientists have reported their stability and dynamic performance in multiple extreme conditions, such as low water activity environments, aqueous/organic and non-aqueous solvents or media, etc. , . Such novel halozymes reported in halophiles include proteases, amylases, lipases, xylanases, nucleases, cellulases, catalases, and esterases . The ability of these halophiles to produce beneficial lysis enzymes such as chitinase, cellulase, and protease implies their direct potential use as biocontrol agents for controlling phytopathogenic fungi, pests, nematodes and for rhizospheric soil decomposition for increasing plant nutrient availability for sustainable agriculture and as livestock feed additives, etc. . Besides, they also find applications in many industries such as food and feed, laundry and detergent, leather and textiles, pulp and paper, alcohol and beverages, medicine and pharmaceuticals, environmental bioremediations, biomass conversions for biofuel production, etc. So, given their profound agricultural and industrial significance, these enzymes viz., chitinase, cellulase, and protease were chosen in this study to unlock their production potential within the halophilic and halotolerant bacterial isolates to unveil and shed light on their vast potential applications. These industrial enzymes are found in various sources, such as plants, animals, and microorganisms. However, microbial sources are preferred due to their high stability, cost-effectiveness, less time and space requirement, high consistency, production and optimization ease, and increasing demand in many industries , as evidenced by their total contribution of more than 82% of revenue share in 2022 . However, despite these enzymes being widely studied in many organisms, only a few reports have been made on the extracellular enzymes of halophilic and halotolerant bacteria, especially Halomonas sp. . Besides, halophiles have been isolated and investigated for many other possible biotechnological applications, such as the production of compatible solutes, enhanced oil recovery, and the degradation of industrial pollutants in saline habitats and as potential agricultural bioinoculants for the recovery of saline soils , . Compatible solutes or osmoprotectants are low molecular weight organic molecules with a neutral charge and low toxicity at high concentrations either accumulated from external environments or secreted by halophiles in their cytoplasm to act as osmolytes for their survivability against the extreme osmotic stresses . They include several different classes, such as high water-soluble sugars, alcohols or polyols, betaines, amino acids, ectoine, and its derivatives, among which ectoine and glycine-betaine are reported as the most predominant ones. They are used in many biotech industries for stabilizing enzymes, DNA, and whole cells against freezing and thawing, drying and heating, and denaturants such as urea and salts, and as salt antagonists, stress-protective agents, moisturizers, therapeutics, and for increasing the freshness of foods in food industries , . In plants, their accumulation is said to increase survivability against various stresses such as salinity, heat, and drought. The genetic manipulation of these osmoprotectants' responsive genes has been suggested as one of the strategies to improve plant stress tolerance by enhancing their production – . In light of the above perspectives, the current research was conducted to study the production potentiality of extracellular halozymes viz. protease, cellulase, and chitinase, and ectoine compatible solute, and PCR based molecular detection of the biosynthetic gene of ectoine and glycine betaine of halophilic and halotolerant bacteria isolated from the crop rhizospheric soils of agricultural fields of southwest coastline of Saurashtra Gujarat.
Preliminary soil analysis The preliminary soil analysis results are summarized in Table . The physicochemical properties of the soil samples, including pH, electrical conductivity (E.C.), organic carbon content, and availability of phosphorous and potash ranged from 7.4 to 8.1, 0.76–1.59 dS m −1 , 4.03–7.47 g kg −1 , 29.57–54.33 kg ha −1 and 166.70–248.33 kg ha −1 respectively . Halophilic characterization The characterization of the isolates by NaCl tolerance test showed that the isolates could tolerate up to 25% with optimum growth between 10 and 15% NaCl concentrations . The isolates S 1 through S 9 and S 11 , all belonging to the Halomonas species, exhibited remarkable and robust growth despite challenging salt concentrations exceeding 10–15% NaCl. These Halomonas isolates surpassed their counterparts S 10 , S 12 , S 13 , S 14 , and S 15 belonging to the Oceanobacillus and Bacillus species with exceptional vigor. Intriguingly, the Halomonas species isolates maintained a consistently thriving growth pattern even at higher NaCl concentrations, with only a modest decline observed up to 25% NaCl. This remarkable resilience and adaptability demonstrated by the Halomonas isolates highlight their exceptional ability to thrive in extremely saline environments. In stark contrast, the Oceanobacillus and Bacillus species isolates, S 10 , S 12 , S 13 , S 14 , and S 15 , exhibited a stark decline in their growth trajectory, indicating their limited capacity for salt tolerance and a considerably less robust response to the challenging conditions presented by 25% NaCl concentration. Microscopic characterization The gram staining analysis of the isolates revealed that all the Halomonas species isolates exhibited gram-negative characteristics, while the Bacillus species isolates exhibited gram-positive characteristics at the microscopic level. This result was complemented by the identification of isolates based on 16S rRNA partial gene sequencing. The scanning electron microscopic characterization identified the isolates as short to thin rod-shaped bacteria in single or pairs to bunchy type organization (Fig. ) with sizes ranging from 0.38 to 0.83 μm by 0.75–6.78 μm . All the isolates were motile, as observed under the microscope. Molecular identification and phylogenetic analysis The analysis of the partial 16S rRNA gene sequence of the isolates to determine their genetic relatedness and taxonomic identity by comparing their sequences to the known reference strains in the curated 16S database of EzBioCloud revealed that isolates S 1 , S 3 , S 5 , S 6 , S 8 , and S 11 belonged to Halomonas pacifica with percent similarity of 99.15%, 99.45%, 99.52%, 99.45%, 99.01% and 99.52% with H. pacifica NBRC 102220 respectively (Table ). In contrast, isolate S2 belonged to Halomonas stenophila with a percent similarity of 99.12% with H. stenophila N12. Isolates S 4 and S 7 belonged to H. salifodinae with percent similarity of 99.34% and 99.22% respectively while isolate S9 belonged to Halomonas binhaiensis with 99.52% similarity with Halomonas binhaiensis Y2R2 (Table ). On the other hand, isolate S10 belonged to Oceanobacillus oncorhynchi with a percent similarity of 98.23% with Oceanobacillus oncorhynchi subsp. oncorhynchi R-2 while isolates S12, S13, S14 and S15 belonged to Bacillus paralicheniformis with percent similarity of 99.59%, 99.46%, 98.19% and 98.44% with B. paralicheniformis KJ-16 (Table ). The partial 16S rRNA gene sequence of isolates S1 and S2 were submitted to NCBI with accession numbers MK955347 and MK961217 re-designating the isolates as Halomonas pacifica HPSB1 and Halomonas stenophila HPSB2 respectively . The BLAST analysis demonstrated the highest level of genetic relations between the respective Halomonas , Oceanobacillus , and Bacillus isolates which features and supports their taxonomic uniformity within the genus. The phylogenetic analysis of these isolates, along with their genetically closest reference species, constructed using MEGA11 following the Minimum Evolution tree method using their multiple sequence alignment (MSA) aligned by ClustalW, and analyzed based on the maximum composite likelihood substitution model, distinctly elucidated the interconnectedness among various genera within their respective species, as visually depicted in Fig. . Quantification of protease, cellulase, and chitinase enzymes The results of the quantitative protease, cellulase, and chitinase enzyme assays and the respective enzymes’ specific activities of the isolates are presented in , respectively. The various standard curves viz . tyrosine, glucose, NAG, and BSA used to estimate the protease, cellulase, and chitinase enzymes, and protein content of the isolates in the respective enzyme production medium are shown in Supplementary Figs. , , , and respectively. The protease, cellulase, and chitinase activities of the isolates ranged from 6.90 to 35.38 U ml −1 min −1 , 0.004–0.042 U ml −1 min −1 , and 0.097–0.550 U ml −1 h −1 respectively, while their respective corresponding specific activities ranged from 7.23 to 36.21 U mg −1 min −1 , 0.007–0.062 U mg −1 min −1 , and 0.146–0.471 U mg −1 h −1 , respectively. The highest protease, cellulase, and chitinase activities were shown by isolate S13, while the lowest was shown by isolates S4 (Fig. ), S5 (Fig. ), and S7 (Fig. ), respectively. On the other hand, the highest protease, cellulase, and chitinase-specific activities were shown by isolates S15 (Fig. ), S12 (Fig. ), and S13 (Fig. ), respectively while the lowest was shown by isolates S4, S5, and S7, respectively. The protein content of the isolates in their respective protease, cellulase, and chitinase enzyme production medium ranged from 0.80 to 1.28 mg ml −1 , 0.450–0.908 mg ml −1 , and 0.550 to 1.166 mg ml −1 , respectively. Isolate S13 exhibited the highest protein content in the protease, cellulase, and chitinase enzyme production medium, while isolates S3 (Fig. ), S2 (Fig. ), and S3 (Fig. ), respectively, showed the lowest protein content. The statistical analysis of variance (ANOVA) for all the data of tyrosine, glucose, and NAG released, protease, cellulase, and chitinase activity, protein contents, and respective enzymes’ corresponding specific activities of the isolates indicated a high level of significance inferred from the greater value of calculated F than that of table F at both the 1% and 5% significance levels. Ectoine production potentiality The results of the ectoine production potentiality of isolates are presented in Table . The ectoine production of the isolates was determined according to the peak generated by the ectoine standard at varying concentrations in LCMS (Fig. ). The ectoine production ranged from 0.01 to 3.17 mgl −1 shown by the isolates S9 and S10 and S5, respectively (Fig. ). The chromatogram and mass spectrum profile of the highest ectoine production by the isolate S5 is shown in Fig. . However, five out of fifteen isolates showed no detectable ectoine production, perhaps due to deficient ectoine production below the detection threshold limit as confirmed by the presence of their ectoine biosynthetic gene described below in PCR based molecular detection of ectoine biosynthetic gene. PCR amplification of ectoine and glycine betaine biosynthetic genes The PCR amplification targeting the ectC and BADH1 genes confirmed their presence in all fifteen isolates, yielding amplicons of 370 bp (Fig. ) and 1473 bp (Fig. ), closely aligning with the sizes reported by Rajan et al . for ectC and Anburajan et al . for BADH1 . The PCR-based molecular detection of the ectC gene thus confirmed and validated the positive ectoine production result obtained by QTOF LCMS as described above.
The preliminary soil analysis results are summarized in Table . The physicochemical properties of the soil samples, including pH, electrical conductivity (E.C.), organic carbon content, and availability of phosphorous and potash ranged from 7.4 to 8.1, 0.76–1.59 dS m −1 , 4.03–7.47 g kg −1 , 29.57–54.33 kg ha −1 and 166.70–248.33 kg ha −1 respectively .
The characterization of the isolates by NaCl tolerance test showed that the isolates could tolerate up to 25% with optimum growth between 10 and 15% NaCl concentrations . The isolates S 1 through S 9 and S 11 , all belonging to the Halomonas species, exhibited remarkable and robust growth despite challenging salt concentrations exceeding 10–15% NaCl. These Halomonas isolates surpassed their counterparts S 10 , S 12 , S 13 , S 14 , and S 15 belonging to the Oceanobacillus and Bacillus species with exceptional vigor. Intriguingly, the Halomonas species isolates maintained a consistently thriving growth pattern even at higher NaCl concentrations, with only a modest decline observed up to 25% NaCl. This remarkable resilience and adaptability demonstrated by the Halomonas isolates highlight their exceptional ability to thrive in extremely saline environments. In stark contrast, the Oceanobacillus and Bacillus species isolates, S 10 , S 12 , S 13 , S 14 , and S 15 , exhibited a stark decline in their growth trajectory, indicating their limited capacity for salt tolerance and a considerably less robust response to the challenging conditions presented by 25% NaCl concentration.
The gram staining analysis of the isolates revealed that all the Halomonas species isolates exhibited gram-negative characteristics, while the Bacillus species isolates exhibited gram-positive characteristics at the microscopic level. This result was complemented by the identification of isolates based on 16S rRNA partial gene sequencing. The scanning electron microscopic characterization identified the isolates as short to thin rod-shaped bacteria in single or pairs to bunchy type organization (Fig. ) with sizes ranging from 0.38 to 0.83 μm by 0.75–6.78 μm . All the isolates were motile, as observed under the microscope.
The analysis of the partial 16S rRNA gene sequence of the isolates to determine their genetic relatedness and taxonomic identity by comparing their sequences to the known reference strains in the curated 16S database of EzBioCloud revealed that isolates S 1 , S 3 , S 5 , S 6 , S 8 , and S 11 belonged to Halomonas pacifica with percent similarity of 99.15%, 99.45%, 99.52%, 99.45%, 99.01% and 99.52% with H. pacifica NBRC 102220 respectively (Table ). In contrast, isolate S2 belonged to Halomonas stenophila with a percent similarity of 99.12% with H. stenophila N12. Isolates S 4 and S 7 belonged to H. salifodinae with percent similarity of 99.34% and 99.22% respectively while isolate S9 belonged to Halomonas binhaiensis with 99.52% similarity with Halomonas binhaiensis Y2R2 (Table ). On the other hand, isolate S10 belonged to Oceanobacillus oncorhynchi with a percent similarity of 98.23% with Oceanobacillus oncorhynchi subsp. oncorhynchi R-2 while isolates S12, S13, S14 and S15 belonged to Bacillus paralicheniformis with percent similarity of 99.59%, 99.46%, 98.19% and 98.44% with B. paralicheniformis KJ-16 (Table ). The partial 16S rRNA gene sequence of isolates S1 and S2 were submitted to NCBI with accession numbers MK955347 and MK961217 re-designating the isolates as Halomonas pacifica HPSB1 and Halomonas stenophila HPSB2 respectively . The BLAST analysis demonstrated the highest level of genetic relations between the respective Halomonas , Oceanobacillus , and Bacillus isolates which features and supports their taxonomic uniformity within the genus. The phylogenetic analysis of these isolates, along with their genetically closest reference species, constructed using MEGA11 following the Minimum Evolution tree method using their multiple sequence alignment (MSA) aligned by ClustalW, and analyzed based on the maximum composite likelihood substitution model, distinctly elucidated the interconnectedness among various genera within their respective species, as visually depicted in Fig. .
The results of the quantitative protease, cellulase, and chitinase enzyme assays and the respective enzymes’ specific activities of the isolates are presented in , respectively. The various standard curves viz . tyrosine, glucose, NAG, and BSA used to estimate the protease, cellulase, and chitinase enzymes, and protein content of the isolates in the respective enzyme production medium are shown in Supplementary Figs. , , , and respectively. The protease, cellulase, and chitinase activities of the isolates ranged from 6.90 to 35.38 U ml −1 min −1 , 0.004–0.042 U ml −1 min −1 , and 0.097–0.550 U ml −1 h −1 respectively, while their respective corresponding specific activities ranged from 7.23 to 36.21 U mg −1 min −1 , 0.007–0.062 U mg −1 min −1 , and 0.146–0.471 U mg −1 h −1 , respectively. The highest protease, cellulase, and chitinase activities were shown by isolate S13, while the lowest was shown by isolates S4 (Fig. ), S5 (Fig. ), and S7 (Fig. ), respectively. On the other hand, the highest protease, cellulase, and chitinase-specific activities were shown by isolates S15 (Fig. ), S12 (Fig. ), and S13 (Fig. ), respectively while the lowest was shown by isolates S4, S5, and S7, respectively. The protein content of the isolates in their respective protease, cellulase, and chitinase enzyme production medium ranged from 0.80 to 1.28 mg ml −1 , 0.450–0.908 mg ml −1 , and 0.550 to 1.166 mg ml −1 , respectively. Isolate S13 exhibited the highest protein content in the protease, cellulase, and chitinase enzyme production medium, while isolates S3 (Fig. ), S2 (Fig. ), and S3 (Fig. ), respectively, showed the lowest protein content. The statistical analysis of variance (ANOVA) for all the data of tyrosine, glucose, and NAG released, protease, cellulase, and chitinase activity, protein contents, and respective enzymes’ corresponding specific activities of the isolates indicated a high level of significance inferred from the greater value of calculated F than that of table F at both the 1% and 5% significance levels.
The results of the ectoine production potentiality of isolates are presented in Table . The ectoine production of the isolates was determined according to the peak generated by the ectoine standard at varying concentrations in LCMS (Fig. ). The ectoine production ranged from 0.01 to 3.17 mgl −1 shown by the isolates S9 and S10 and S5, respectively (Fig. ). The chromatogram and mass spectrum profile of the highest ectoine production by the isolate S5 is shown in Fig. . However, five out of fifteen isolates showed no detectable ectoine production, perhaps due to deficient ectoine production below the detection threshold limit as confirmed by the presence of their ectoine biosynthetic gene described below in PCR based molecular detection of ectoine biosynthetic gene.
The PCR amplification targeting the ectC and BADH1 genes confirmed their presence in all fifteen isolates, yielding amplicons of 370 bp (Fig. ) and 1473 bp (Fig. ), closely aligning with the sizes reported by Rajan et al . for ectC and Anburajan et al . for BADH1 . The PCR-based molecular detection of the ectC gene thus confirmed and validated the positive ectoine production result obtained by QTOF LCMS as described above.
The NaCl tolerance test revealed that the isolates were moderate halophiles and halotolerant in nature based on the concentration of salt required for optimum growth and their maximum salt tolerance capacity as per the most widely accepted classification by Kushner and Kamekura , . This finding was further confirmed and validated by the molecular identification of the isolates by 16S rRNA gene sequencing at genus and species levels. The 16S rRNA gene sequence of the isolates belonging to H. pacifica and H. stenophila were submitted to NCBI with accession numbers MK955347 and MK961217, respectively . The protease activity of the isolates belonging to Halomonas species was approximately twofold higher than that of H. meridian HC4321C1 reported by Anithajothi et al. . For those isolates belonging to Bacillus and Oceanobacillus species, the protease activity was in line with that of B. licheniformis P003 and Oceanobacillus aidingensis , reported by Sarker et al. and Kumar et al . , respectively. However, the findings of many reports suggest that B. licheniformis strains are capable of producing much higher protease activity ranging above 100 to more than a few 1000 U ml −1 – depending upon the degree of culture enrichment with different sources of nitrogen, carbon, and substrates, etc., and is supposedly reported as one of the industrial strain of choice for enzyme production. The protease activity of H. pacifica and H. stenophila is being reported for the first time in our study. The optimum enzyme production potentiality of any microorganism is said to depend on the degree of optimized conditions for many factors such as pH, temperature, incubation period, agitation rate, substrate type, sources of carbon, nitrogen, etc., therefore, subject to vary from genus to genus and species to species. These results thus suggest that the isolates under study present suitable candidates for producing protease enzymes for various biotechnological and industrial applications. The cellulase activity of the isolates belonging to Halomonas species was found to agree with the report of Shivanand et al. but much less than that of Halomonas sp. PV1 reported by Benit et al. . In contrast, the isolate belonging to Bacillus paralicheniformis showed slightly less than the lowest cellulase activity of the same species reported by da Silva et al. . However, the cellulase activity of the isolates belonging to Bacillus species was at least tenfold higher than that of isolates belonging to Halomonas species. These findings imply that Bacillus species have higher cellulase production potential than those Halomonas species. On the other hand, the cellulase activity of the isolate belonging to Oceanobacillus oncorhynchi was nearly tenfold lesser than that of Oceanobacillus profundus reported by Gbenro et al. . Nevertheless, the ability of the isolates to produce cellulase enzyme suggests their ability to degrade cellulose, thereby implying the need for further investigation to uncover their fullest potential for the production of the same on a larger scale by providing the best optimum production conditions and suitable infrastructure facilities. The highest chitinase activity shown by the isolate belonging to B. paralicheniformis was found to agree with the report of Akhir et al. . While it was approximately tenfold higher than that of B. licheniformis JP2 reported by Keliat et al . , Hussin and Majid have reported even a much lesser chitinase activity of similar species. However, the highest chitinase activity observed in our isolates is still much lesser than that of similar species reported by scientists such as Akeed et al. and Sasi et al. . On the other hand, the chitinase activity of halophilic bacterial species is very limited to date, specifically of Halomonas species, and is being reported for the first time in our study. Furthermore, few earlier reports on the chitinase activity of halophilic bacteria such as Virgibacillus marismortui M3-23 and Planococcus rifitoensis M2-26 do prove the chitinase production potentiality of halophilic bacteria. However, there are multiple reports on the chitinase activity of many halotolerant bacteria, particularly that of Bacillus species which is supposedly used as one of the commercial, industrial strains. Nonetheless, the ability of our isolates to produce chitinase enzyme does suggest their ability to degrade chitin compounds, thereby representing a potential biocontrol agent for sustainable agriculture besides various other applications in many industries. The ectoine production by all the isolates was found to be negligible to significantly less as compared to that of different Halomonas species reported by Zhang et al . , Van-Thuoc et al . , Van-Thuoc et al . , Chen et al . , Chen et al . where ectoine production ranged from 3.65 to 13.96 g l −1 . The ectoine production of the isolates belonging to H. pacifica, H. stenophila, H. salifodinae , H. binhaiensis , O. oncorhynchi, and B. paralicheniformis has not been reported earlier. It is being reported for the first time in our study. The lower amount of ectoine production by the isolates may be attributed to the unoptimized rate of agitation, which is reported to impact oxygen transfer in which a higher agitation rate for some halophiles results in higher dissolved oxygen (DO) level and thereby higher ectoine production. In contrast, a higher agitation rate has also been reported to result in a high shearing force of agitation, lowering microbial growth and ectoine production at an agitation rate higher than 200 rpm . Furthermore, Chen et al . achieved higher ectoine production potentiality after utilizing a well-optimized production system with the best carbon and nitrogen sources, optimum ratio, optimum NaCl concentration, and agitation rate. Nevertheless, the production of ectoine by the isolates, even though in smaller quantity, still implies their potentiality for the production of ectoine-compatible solute. The ectC gene amplified in the isolates is reported to encode putative proteins of 129 amino acids and codes for the L-ectoine synthase protein that catalyzes the final step of the ectoine biosynthetic pathway leading to the synthesis of ectoine osmolyte . On the other hand, the Betaine Aldehyde Dehydrogenase gene is reportedly encoded by a polynucleotide of 1473 bp (Fig. ) and polypeptides of 490 amino acids . The BADH1 gene catalyzes the conversion of betaine aldehyde to glycine betaine in the last step of the biosynthetic pathway that leads to the synthesis of the effective compatible solute glycine betaine, which maintains the fluidity of membranes and protects the biological structure of the organisms under salt stress conditions . Additionally, similar to ectoine, glycine betaine also aids in stabilizing key proteins like proteases, cellulases, and chitinases against salt-induced denaturation, which are essential for nutrient acquisition and energy metabolism from organic substrates in extreme environments. The presence of the BADH1 gene in the halophilic and halotolerant bacterial isolates suggests their potential to produce glycine betaine, expanding their repertoire of compatible solutes beyond ectoine thereby enabling osmolyte switch. This finding thus emphasizes the diverse adaptive strategies employed by halophilic and halotolerant bacteria to thrive in saline environments, aligning with previous reports of glycine betaine production in related halotolerant species such as B. halodurans SMBPL06 and B. subtilis MA04 thereby implying its importance as one of the reportedly most predominant solutes produced besides ectoine in true halophiles studied till date . The ability of ectoine-compatible solute production by halophilic bacteria belonging to Halomonas species is evidenced by comprehensive reports on ectoine production in many other Halomonas species, such as H. elongate – , H. boliviensis and halotolerant bacteria such as Bacillus halodurans . Likewise, the production of glycine betaine has also been reported in halotolerant bacteria such as B. halodurans , B. subtilis , etc. The accumulation of these compatible solutes has been reported to confer osmotolerance in plants. The hyperosmotic tolerance conferred by the genetic transformation of the ectoine biosynthetic gene has already been reported in many plants, such as tobacco , and tomato plants . Similarly, the salt tolerance conferred by transforming the glycine betaine biosynthetic gene has been reported in barley and wheat . The ability of the isolates to produce these compatible solutes thus showed their significance as a source of osmoprotectant responsive genes, which hold a tremendous potentiality for conferring osmotolerance to plants through their genetic transformations. The application of compatible solutes in various industries for stabilizing enzymes suggests a possible correlation that the ectoine-compatible solute produced by the isolates could be involved in aiding the production of extremozymes under saline conditions by preventing their denaturation from salinity and thereby maintaining their production. Enzymes like protease, cellulase, chitinase, etc. are produced and secreted by halophiles to acquire nutrients and energy from organic substrates present in their extreme environments. However, the high salt concentrations in their environments can disrupt the structure and function of these enzymes by interfering with their electrostatic interactions and hydrogen bonding. So, to counteract these denaturing effects of salt, halophiles have evolved to produce these compatible solutes not only as part of their adaptation to saline environments but also to protect and stabilize their metabolically and physiologically important enzymes for survivability. These compatible solutes are said to form protective hydration shells around the enzymes, shielding them from the disruptive effects of salt ions. This hydration stabilizes the enzyme's structure and allows it to remain active and functional in the presence of high salt. Notably, the observations made by Roberts and Detkova et al. affirm this correlation between compatible solutes and extremozymes production. They documented the multifaceted role of compatible solutes in halophilic bacteria, emphasizing not only their pivotal function in osmoregulation to maintain cellular osmotic equilibrium but also their significant capacity to serve as effective stabilizers of proteins and even whole cells. These findings collectively underscore the intimate connection between compatible solute accumulation and the production and stability of extremozymes, including proteases, cellulases, and chitinases, in halophilic microorganisms.
The investigation was carried out at the “Department of Biotechnology, College of Agriculture, Junagadh Agricultural University, Junagadh” during 2019–2022. Isolation of bacteria The halophilic and halotolerant bacteria were isolated from 15 different soil samples, each approximately 100 g in weight, collected from various crop rhizospheres in different agricultural fields lying along the southwest coastline of Saurashtra, Gujarat. Specifically, samples were obtained from Junagadh and Porbandar districts, located at coordinates 21.52° N 70.47° E and 21°37′48″ N 69°36′0″ E, respectively, as outlined in Table , as previously reported by Reang et al . . Following the streak plate method, the bacteria were isolated from 10 ml of soil suspensions (prepared from 1 g) by streaking a loopful of the 10 –5 dilution onto a freshly prepared autoclaved halophilic agar media supplemented with 10% NaCl, adjusted to pH 7.2 ± 0.2 (at 25 °C), and incubated at 37 °C for 5 days. The isolates were characterized for halophilic and halotolerant nature by subjecting them to a varying concentration of NaCl ranging from 5, 10, 15, 20, and 25% in halophilic broth for salt tolerance test. Pure culture plates of the above isolates were prepared on the same media and used to prepare primary inoculum seed culture. Preliminary soil analysis The preliminary analysis of soil samples was conducted to assess soil chemical properties, including soil pH using potentiometry and electrical conductivity via the conductometry method . Soil organic carbon content was determined using the back titration method , while available soil phosphorus was measured using a colorimetric method , and soil potash was analyzed via flame photometry . Preparation of inoculum A primary inoculum of the isolates was prepared by inoculating a single colony from each pure culture plate as prepared above on a freshly prepared autoclaved 10 ml halophilic broth in test tubes and incubated at 37 °C for 24 h. As described below, the primary inoculum was then used as seed culture for the extracellular enzymes and compatible solutes production potentiality experiments. Microscopic characterization of isolates Gram’s staining A thick smear of all cultures was prepared on a clean glass slide individually and subjected to the gram staining process. Subsequently, the slides were allowed to air dry at room temperature overnight and observed under the Zeiss Imager.Z2 optical microscope to ascertain the orientation of the isolates. Scanning electron microscopy A loopful of the bacterial isolates colony was picked from their respective fresh culture plates, and a light smear was made on the aluminium stub with the help of inoculating needle. The smeared stub was then flooded with 4% glutaraldehyde and kept in a fridge at 4 ℃ for 24 h. The following day the smeared samples were dehydrated by using a gradient dilution of acetone in a concentration ranging from 30, 50, 70, 80, 90, and 100% and treating each sample by dipping into dilution of each respective concentration in the order of 30–100% for 15 min. The samples treated by dipping in an acetone concentration of 100% were repeated for a second time for another 15 min for each sample. The dehydrated samples were then coated in a spotter coater with a gold–palladium mixture plate and observed under a scanning electron microscope (Zeiss EVO 18). Motility test The bacterial motility test was done by the hanging drop method. A few drops of liquid culture were placed onto the coverslip sterilely. A depression slide was taken, and the concave portion over the drop pressed the slide onto the cover slip. The slide was inverted quickly to keep from disrupting the drop. Then the motility was examined under the Zeiss Imager.Z2 optical microscope at 40× magnification. Molecular identification of bacterial isolates The identification of halophilic and halotolerant bacterial isolates was performed using BLAST analysis of their partial 16S rRNA gene sequences against known reference strains in the curated 16S database of EzBioCloud. Reference strains exhibiting a sequence similarity threshold above 96% with the query isolates were selected for further analysis. A phylogenetic tree was constructed using MEGA11, applying the maximum composite likelihood substitution model with 100 bootstrap replications. The tree was generated under the minimum evolution method to elucidate the genetic interconnectedness among various genera within their respective species. Protease enzyme production The protease enzyme production was carried out by inoculating 1% of the test isolates in a 250 ml Erlenmeyer flask containing 100 ml autoclaved protease production broth prepared by dissolving 3% nutrient gelatin, 0.8% nutrient broth, 0.5% casein, 0.01% MnCl 2 , 15% NaCl, and 1.2 ml of 20% glycerol and incubated at 37 ℃ for 72 h in a shaker incubator (150 rpm). After 72 h of growth, the cells were harvested at 10,000 rpm for 15 min, and the supernatant thus obtained was used as crude enzyme for quantitative assay. Protease enzyme quantification assay The protease enzyme assay followed Sigma's non-specific protease assay described by Cupp-Enyard . The assay was performed in triplicates in 15 ml test tubes using enzyme extracts from each isolate, with one tube serving as a blank control. In each set of four tubes, 5 ml of 0.65% casein solution was added and equilibrated at 37 °C for 5 min. Then, 1 ml of enzyme extract was introduced to three of the tubes (excluding the blank), mixed, and incubated at 37 °C for 10 min. Tyrosine release and protease activity were measured and compared among test isolates during this incubation period. After 10 min, 5 ml of 110 mM TCA reagent was added to stop the reaction and incubated for 30 min at 37 °C. Concurrently, a series of standard tyrosine dilutions were prepared in six test tubes, with incremental volumes of 1.1 mM tyrosine standard stock solution, precisely measuring 0.00, 0.05, 0.10, 0.20, 0.40, and 0.50 ml, respectively. Subsequently, each standard dilution was adjusted to a final volume of 2 ml by the addition of an appropriate volume of purified water. After a 30-min incubation, each test solution and blank was filtered using a 0.45 μm polyethersulfone syringe filter to eliminate insoluble components. The resulting 2 ml filtrate from both the test isolates and blanks was transferred to new test tubes. In all the test tubes containing standards and standard blank, 5 ml of sodium carbonate was added, followed immediately by 1 ml of Folin’s reagent to stabilize pH. After observing a cloudy appearance due to the reaction with free tyrosine, the tubes were gently mixed and incubated at 37 °C for 30 min. After incubation, the tubes with standards were seen with a gradation of color correlating with the amount of tyrosine added. This color change was also observed in tubes with test samples. Subsequently, 2 ml of these solutions were filtered into cuvettes using a 0.45 μm polyethersulfone syringe filter. The absorbance of the standards, standard blank, the different test isolates, and test blank were measured in a spectrophotometer at 660 nm wavelength with a 1 cm light path. The standard tyrosine curve was then constructed to determine the amount of tyrosine released and estimate the protease activity of the test isolates using the below-described formula. One unit of protease activity (U) was defined as the amount of enzyme capable of releasing 1.0 μ mol of tyrosine per min from the casein substrate under the described reaction conditions. The protein content was estimated by Lowry’s method, with bovine serum albumin (BSA) as the standard. The protease enzyme and specific activity were then determined by calculating using the following formula: [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = } { 11}}}{{( {{1 } { 1}0 \, { 2}} )}}$$ Protease activity U ml - 1 min - 1 = μ mol of tyrosine equivalents released × 11 1 × 10 × 2 where, 11 = Total volume of assay (ml). 10 = Time of assay (min) as per the unit definition. 1 = Volume of enzyme used (ml). 2 = Volume taken in cuvette for colorimetric determination. [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = {}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, /{}( {{}^{{ - {1}}} } )$$ Protease specific activity U mg - 1 min - 1 = Enzyme activity U ml - 1 min - 1 / Protein content mg ml - 1 Cellulase enzyme production The cellulase enzyme production was carried out by inoculating 1% of the bacterial isolates in a 250 ml Erlenmeyer flask containing 100 ml autoclaved cellulase production broth prepared by dissolving 1% CMC, 0.2% NaNO 3 , 0.05% MgSO 4 , 0.005% K 2 HPO 4 , 0.001% FeSO 4 , 0.002% CaCl 2 and MnSO 4 , 15% NaCl, and incubated on water bath shaker at 120 rpm at 37 °C for 5 days for cellulase enzyme production. After incubation, the bacterial cultures were harvested by centrifugation at 5000 rpm for 20 min. The culture supernatants were used for the quantification of the cellulase enzyme. Cellulase enzyme quantification assay The cellulase activity was assayed using the DNSA method, followed by Lay Mg Mg et al . . One milliliter of culture supernatant was mixed with 1 ml of 0.05 M citrate buffer (pH 4.8) solution in test tubes containing 1% cellulose substrate. The resulting reaction mixture was incubated at 50 °C for 60 min in a water bath shaker at 80–85 rpm. After the reaction time, 3 ml of DNSA reagent was added to the reaction mixture and boiled for exactly 5 min to terminate the reaction in a vigorously boiling water bath. The reaction mixture was then cooled in a cold water bath, and the absorbance was measured by a spectrophotometer at 540 nm against the blank without enzyme filtrate. Anhydrous glucose was the standard . One unit of cellulase activity (U) was defined as the amount of enzyme capable of releasing 1.0 mg of glucose per min from the cellulose substrate under the described reaction conditions. The protein content was estimated as described above. The cellulase enzyme activity and specific activity were then determined by calculating using the following formula: [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = } { 5 } \, 0.{5}} )}}{{( {{1 } { 2 } { 6}0} )}}$$ Cellulase activity U ml - 1 min - 1 = mg of glucose released × 5 × 0.5 1 × 2 × 60 where, 5 = Total volume of assay (ml). 0.5 = Dilution factor (DF). DF = (Vol m of enzyme extract/Vol m of enzyme + buffer). 1 = Volume of enzyme extract used (ml). 2 = Volume of reaction mixture taken in cuvette (ml). 60 = Incubation time (min) [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = {}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, /{}( {{}^{{ - {1}}} } )$$ Cellulase specific activity U mg - 1 min - 1 = Enzyme activity U ml - 1 min - 1 / Protein content mg ml - 1 Chitinase enzyme production The chitinase enzyme production was performed by modifying the standard method followed by Hsu and Lockwood . One milliliter of the bacterial isolates was inoculated in the 250 ml Erlenmeyer flask containing 100 ml autoclaved Minimal Medium (Designated as MM) broth prepared by dissolving 0.5% colloidal chitin, 0.05% MgSO 4 ∙7H 2 O, 0.03% KH 2 PO 4 , 0.07% K 2 HPO 4 , 0.0001% MnCl 2 , 0.001% FeSO 4 ∙7H 2 O and 0.0001% ZnSO 4 , 15% NaCl in 1000 ml distilled water and the final pH was adjusted to 7. The inoculated tubes were incubated in a shaker incubator (200 rpm) at 30 °C for 48 h. After incubation, the isolates' cultures were harvested and used for carrying out further quantitative chitinase enzyme assay. An uninoculated test tube containing the same liquid broth was kept as blank. Chitinase enzyme quantification assay Chitinase activity was determined by modifying a colorimetric method followed by Setia and Sohorjono in triplicates. The reaction mixture consisted of 1 ml of the crude enzyme and 2 ml of 1.25% (w/v) colloidal chitin substrate in a 200 mM potassium phosphate buffer (pH 6.0). The mixture was incubated at 30 °C for 2 h and boiled for 10 min to stop the reaction, then cooled to room temperature in a cold water bath, and 1 unit of β- N -Acetylglucosaminidase (NAGase) was added and then centrifuged at 8000 rpm for 20 min. The 1 ml of test supernatant obtained from the above centrifugation was added to 1.5 ml of freshly prepared color reagent solution prepared by mixing 96 mM DNSA (3,5-Dinitrosalicylic Acid) reagent in 5.3 M sodium potassium tartrate solution and diluted to 40 ml with deionized water. The test supernatant and color reagent solution mixture was then boiled for 5 min and cooled to room temperature. The concentration of GlcNAc ( N -acetylglucosamine) released was then measured at 540 nm. The standard curve of GlcNAc was plotted between GlcNAc concentration and GlcNAc absorbance. One unit of chitinase enzyme activity (U) was defined as the amount of enzyme capable of liberating 1.0 mg GlcNAc per hour from the chitin substrate under reaction conditions. The protein content in isolates was determined by the Folin-Lowry method using BSA as standard. Data on chitinase enzyme activity, specific activity, and protein content was analyzed with a single factorial CRD analysis of variance ( α = 0.05). [12pt]{minimal}
$$ {}: \\ {}\;{}: \\ {}0{}\;{} = {}0{} - {}0{} \\ {} {}0{}. \\ {}: \\ {}0{} = {}0{} - {}0{} \\ $$ Calculations : Standard curve : Δ A54 0 nm Standard = A54 0 nm Std - A54 0 nm Std Blank The Δ A54 0 nm of the standards was plotted against the milligrams of NAG released . Sample determination : Δ A54 0 nm Sample = A54 0 nm Test - A54 0 nm Test Blank The milligrams of NAG liberated were determined using the standard curve, and the chitinase enzyme activity (U ml −1 h −1 ) and its specific activity (U mg −1 h −1 ) defined per mg of protein estimated in isolates were then calculated using the following formula: [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = } ) \, ( {{3 } + {}^{{3}} } )}}{{( {{2 } { 1 } { 1}} )}}$$ Chitinase activity U ml - 1 h - 1 = mg NAG released 3 + Volume of NAGase 3 2 × 1 × 1 where, 3 = Initial reaction volume of assay. 2 = Conversion factor for converting 2 h to 1 h as per the unit definition. 1 = Volume (ml) of supernatant used in colorimetric determination. 1 = Volume (ml) of crude enzyme used. Volume of NAGase = 0.5 ml. [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = {}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, /{}( {{}^{{ - {1}}} } )$$ Chitinase specific activity U mg - 1 h - 1 = Enzyme activity U ml - 1 h - 1 / Protein content mg ml - 1 Ectoine production potentiality Culture medium The ectoine production of the bacterial isolates was carried out in an autoclaved freshly prepared culture medium consisting of yeast extract (86 gl −1 ), ammonium sulfate (28 gl −1 ), FeCl 2 .4H 2 O (0.5 mM), MnSO 4 ∙H 2 O (10 μM), KCl (2 gl −1 ), MgSO 4 .7H 2 O (100 mM), and sodium chloride (1 M) as per the modified method of Chen et al . . Growth conditions A primary inoculum seed culture of all the isolates was first prepared by inoculating a loopful of the respective halophilic bacterial cells to 10 ml of autoclaved freshly prepared halophilic broth and incubated in a shaker incubator at 37 ℃ and 180 rpm for 24 h. After 24 h cultivation, 1 ml of the seed culture was inoculated into a 250 ml Erlenmeyer flask containing 50 ml autoclaved freshly prepared ectoine production broth and incubated at 30 ℃ and 200 rpm for 24 h. The flask uninoculated with isolate culture served as the negative control. After cultivation, the samples were taken to analyze cell growth and ectoine concentration. Ectoine quantification assay The halophilic bacterial isolates culture cultivated above was harvested by centrifugation at 8000 rpm. The pellets were resuspended in 80% ethanol (v/v) (Sigma) with vigorous shaking for 30 min. The ethanol extracts were filtrated through a 0.45 mm filter to analyze ectoine production. The ectoine concentration was then determined from the filtrate obtained by LCMS as per the details in Table . Ectoine, purchased from Sigma, was used as the reference standard. PCR-based molecular detection of ectoine and glycine betaine biosynthetic genes The isolates' production potentiality of the ectoine-compatible solute was confirmed by PCR-based molecular detection of its biosynthetic gene ectC which encodes for the ectoine synthase enzyme. Besides, the PCR-based molecular screening also showed the presence of BADH1 gene encoding for betaine aldehyde dehydrogenase enzyme which is responsible for the biosynthesis of glycine betaine compatible solute. The primer sequences for PCR amplification of these genes were obtained from literature reported by Rajan et al. and Anburajan et al . for ectoine and glycine betaine, respectively. The isolates' genomic DNA was isolated by Qiagen's blood and tissue kit based on the manufacturer's instructions. The isolated genomic DNA was then analyzed on 0.8% agarose gel electrophoresis, quantified by nanodrop spectrophotometer, and used for further PCR reactions. The PCR reaction was carried out in an Applied Biosystems thermal cycler in a 20 ml reaction volume system containing 50 ng of genomic DNA, 0.5 mM of each primer, 100 mM of each dNTP, 5 U of KAPA Taq DNA polymerase, and 10X Taq A buffer supplemented with 25 mM MgCl 2 . A single PCR vial containing all the above PCR reaction mixture except the DNA was used as the no template control (NTC) to check the chances of amplification due to primer contamination with the template DNA. The thermal cycler amplification reaction conditions were set with an initial denaturation at 94 ℃ for 5 min, followed by 35 cycles of denaturation at 94 ℃ for 45 s, annealing temperature at 50 ℃ for 45 s, and primary extension at 72 ℃ for 1 min, followed by a final extension at 72 ℃ for 7 min and hold at 4 ℃. The PCR amplified products were checked by running on agarose gel electrophoresis at 90 V cm −1 in 1.5% low EEO agarose gel prepared in 1× TAE buffer and added with 3% of 1000 ppm ethidium bromide. The resulting electrophoresed amplicons of the respective genes were then scanned and captured by the gel documentation system. Statistical analysis The above experiments were carried out in triplicate replications. The data obtained from their mean values were used for statistical analysis of variance (ANOVA) using a Completely Randomized Design (CRD) for the interpretation of results.
The halophilic and halotolerant bacteria were isolated from 15 different soil samples, each approximately 100 g in weight, collected from various crop rhizospheres in different agricultural fields lying along the southwest coastline of Saurashtra, Gujarat. Specifically, samples were obtained from Junagadh and Porbandar districts, located at coordinates 21.52° N 70.47° E and 21°37′48″ N 69°36′0″ E, respectively, as outlined in Table , as previously reported by Reang et al . . Following the streak plate method, the bacteria were isolated from 10 ml of soil suspensions (prepared from 1 g) by streaking a loopful of the 10 –5 dilution onto a freshly prepared autoclaved halophilic agar media supplemented with 10% NaCl, adjusted to pH 7.2 ± 0.2 (at 25 °C), and incubated at 37 °C for 5 days. The isolates were characterized for halophilic and halotolerant nature by subjecting them to a varying concentration of NaCl ranging from 5, 10, 15, 20, and 25% in halophilic broth for salt tolerance test. Pure culture plates of the above isolates were prepared on the same media and used to prepare primary inoculum seed culture.
The preliminary analysis of soil samples was conducted to assess soil chemical properties, including soil pH using potentiometry and electrical conductivity via the conductometry method . Soil organic carbon content was determined using the back titration method , while available soil phosphorus was measured using a colorimetric method , and soil potash was analyzed via flame photometry .
A primary inoculum of the isolates was prepared by inoculating a single colony from each pure culture plate as prepared above on a freshly prepared autoclaved 10 ml halophilic broth in test tubes and incubated at 37 °C for 24 h. As described below, the primary inoculum was then used as seed culture for the extracellular enzymes and compatible solutes production potentiality experiments.
Gram’s staining A thick smear of all cultures was prepared on a clean glass slide individually and subjected to the gram staining process. Subsequently, the slides were allowed to air dry at room temperature overnight and observed under the Zeiss Imager.Z2 optical microscope to ascertain the orientation of the isolates. Scanning electron microscopy A loopful of the bacterial isolates colony was picked from their respective fresh culture plates, and a light smear was made on the aluminium stub with the help of inoculating needle. The smeared stub was then flooded with 4% glutaraldehyde and kept in a fridge at 4 ℃ for 24 h. The following day the smeared samples were dehydrated by using a gradient dilution of acetone in a concentration ranging from 30, 50, 70, 80, 90, and 100% and treating each sample by dipping into dilution of each respective concentration in the order of 30–100% for 15 min. The samples treated by dipping in an acetone concentration of 100% were repeated for a second time for another 15 min for each sample. The dehydrated samples were then coated in a spotter coater with a gold–palladium mixture plate and observed under a scanning electron microscope (Zeiss EVO 18). Motility test The bacterial motility test was done by the hanging drop method. A few drops of liquid culture were placed onto the coverslip sterilely. A depression slide was taken, and the concave portion over the drop pressed the slide onto the cover slip. The slide was inverted quickly to keep from disrupting the drop. Then the motility was examined under the Zeiss Imager.Z2 optical microscope at 40× magnification. Molecular identification of bacterial isolates The identification of halophilic and halotolerant bacterial isolates was performed using BLAST analysis of their partial 16S rRNA gene sequences against known reference strains in the curated 16S database of EzBioCloud. Reference strains exhibiting a sequence similarity threshold above 96% with the query isolates were selected for further analysis. A phylogenetic tree was constructed using MEGA11, applying the maximum composite likelihood substitution model with 100 bootstrap replications. The tree was generated under the minimum evolution method to elucidate the genetic interconnectedness among various genera within their respective species.
A thick smear of all cultures was prepared on a clean glass slide individually and subjected to the gram staining process. Subsequently, the slides were allowed to air dry at room temperature overnight and observed under the Zeiss Imager.Z2 optical microscope to ascertain the orientation of the isolates.
A loopful of the bacterial isolates colony was picked from their respective fresh culture plates, and a light smear was made on the aluminium stub with the help of inoculating needle. The smeared stub was then flooded with 4% glutaraldehyde and kept in a fridge at 4 ℃ for 24 h. The following day the smeared samples were dehydrated by using a gradient dilution of acetone in a concentration ranging from 30, 50, 70, 80, 90, and 100% and treating each sample by dipping into dilution of each respective concentration in the order of 30–100% for 15 min. The samples treated by dipping in an acetone concentration of 100% were repeated for a second time for another 15 min for each sample. The dehydrated samples were then coated in a spotter coater with a gold–palladium mixture plate and observed under a scanning electron microscope (Zeiss EVO 18).
The bacterial motility test was done by the hanging drop method. A few drops of liquid culture were placed onto the coverslip sterilely. A depression slide was taken, and the concave portion over the drop pressed the slide onto the cover slip. The slide was inverted quickly to keep from disrupting the drop. Then the motility was examined under the Zeiss Imager.Z2 optical microscope at 40× magnification.
The identification of halophilic and halotolerant bacterial isolates was performed using BLAST analysis of their partial 16S rRNA gene sequences against known reference strains in the curated 16S database of EzBioCloud. Reference strains exhibiting a sequence similarity threshold above 96% with the query isolates were selected for further analysis. A phylogenetic tree was constructed using MEGA11, applying the maximum composite likelihood substitution model with 100 bootstrap replications. The tree was generated under the minimum evolution method to elucidate the genetic interconnectedness among various genera within their respective species.
The protease enzyme production was carried out by inoculating 1% of the test isolates in a 250 ml Erlenmeyer flask containing 100 ml autoclaved protease production broth prepared by dissolving 3% nutrient gelatin, 0.8% nutrient broth, 0.5% casein, 0.01% MnCl 2 , 15% NaCl, and 1.2 ml of 20% glycerol and incubated at 37 ℃ for 72 h in a shaker incubator (150 rpm). After 72 h of growth, the cells were harvested at 10,000 rpm for 15 min, and the supernatant thus obtained was used as crude enzyme for quantitative assay.
The protease enzyme assay followed Sigma's non-specific protease assay described by Cupp-Enyard . The assay was performed in triplicates in 15 ml test tubes using enzyme extracts from each isolate, with one tube serving as a blank control. In each set of four tubes, 5 ml of 0.65% casein solution was added and equilibrated at 37 °C for 5 min. Then, 1 ml of enzyme extract was introduced to three of the tubes (excluding the blank), mixed, and incubated at 37 °C for 10 min. Tyrosine release and protease activity were measured and compared among test isolates during this incubation period. After 10 min, 5 ml of 110 mM TCA reagent was added to stop the reaction and incubated for 30 min at 37 °C. Concurrently, a series of standard tyrosine dilutions were prepared in six test tubes, with incremental volumes of 1.1 mM tyrosine standard stock solution, precisely measuring 0.00, 0.05, 0.10, 0.20, 0.40, and 0.50 ml, respectively. Subsequently, each standard dilution was adjusted to a final volume of 2 ml by the addition of an appropriate volume of purified water. After a 30-min incubation, each test solution and blank was filtered using a 0.45 μm polyethersulfone syringe filter to eliminate insoluble components. The resulting 2 ml filtrate from both the test isolates and blanks was transferred to new test tubes. In all the test tubes containing standards and standard blank, 5 ml of sodium carbonate was added, followed immediately by 1 ml of Folin’s reagent to stabilize pH. After observing a cloudy appearance due to the reaction with free tyrosine, the tubes were gently mixed and incubated at 37 °C for 30 min. After incubation, the tubes with standards were seen with a gradation of color correlating with the amount of tyrosine added. This color change was also observed in tubes with test samples. Subsequently, 2 ml of these solutions were filtered into cuvettes using a 0.45 μm polyethersulfone syringe filter. The absorbance of the standards, standard blank, the different test isolates, and test blank were measured in a spectrophotometer at 660 nm wavelength with a 1 cm light path. The standard tyrosine curve was then constructed to determine the amount of tyrosine released and estimate the protease activity of the test isolates using the below-described formula. One unit of protease activity (U) was defined as the amount of enzyme capable of releasing 1.0 μ mol of tyrosine per min from the casein substrate under the described reaction conditions. The protein content was estimated by Lowry’s method, with bovine serum albumin (BSA) as the standard. The protease enzyme and specific activity were then determined by calculating using the following formula: [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = } { 11}}}{{( {{1 } { 1}0 \, { 2}} )}}$$ Protease activity U ml - 1 min - 1 = μ mol of tyrosine equivalents released × 11 1 × 10 × 2 where, 11 = Total volume of assay (ml). 10 = Time of assay (min) as per the unit definition. 1 = Volume of enzyme used (ml). 2 = Volume taken in cuvette for colorimetric determination. [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = {}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, /{}( {{}^{{ - {1}}} } )$$ Protease specific activity U mg - 1 min - 1 = Enzyme activity U ml - 1 min - 1 / Protein content mg ml - 1
The cellulase enzyme production was carried out by inoculating 1% of the bacterial isolates in a 250 ml Erlenmeyer flask containing 100 ml autoclaved cellulase production broth prepared by dissolving 1% CMC, 0.2% NaNO 3 , 0.05% MgSO 4 , 0.005% K 2 HPO 4 , 0.001% FeSO 4 , 0.002% CaCl 2 and MnSO 4 , 15% NaCl, and incubated on water bath shaker at 120 rpm at 37 °C for 5 days for cellulase enzyme production. After incubation, the bacterial cultures were harvested by centrifugation at 5000 rpm for 20 min. The culture supernatants were used for the quantification of the cellulase enzyme.
The cellulase activity was assayed using the DNSA method, followed by Lay Mg Mg et al . . One milliliter of culture supernatant was mixed with 1 ml of 0.05 M citrate buffer (pH 4.8) solution in test tubes containing 1% cellulose substrate. The resulting reaction mixture was incubated at 50 °C for 60 min in a water bath shaker at 80–85 rpm. After the reaction time, 3 ml of DNSA reagent was added to the reaction mixture and boiled for exactly 5 min to terminate the reaction in a vigorously boiling water bath. The reaction mixture was then cooled in a cold water bath, and the absorbance was measured by a spectrophotometer at 540 nm against the blank without enzyme filtrate. Anhydrous glucose was the standard . One unit of cellulase activity (U) was defined as the amount of enzyme capable of releasing 1.0 mg of glucose per min from the cellulose substrate under the described reaction conditions. The protein content was estimated as described above. The cellulase enzyme activity and specific activity were then determined by calculating using the following formula: [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = } { 5 } \, 0.{5}} )}}{{( {{1 } { 2 } { 6}0} )}}$$ Cellulase activity U ml - 1 min - 1 = mg of glucose released × 5 × 0.5 1 × 2 × 60 where, 5 = Total volume of assay (ml). 0.5 = Dilution factor (DF). DF = (Vol m of enzyme extract/Vol m of enzyme + buffer). 1 = Volume of enzyme extract used (ml). 2 = Volume of reaction mixture taken in cuvette (ml). 60 = Incubation time (min) [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = {}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, /{}( {{}^{{ - {1}}} } )$$ Cellulase specific activity U mg - 1 min - 1 = Enzyme activity U ml - 1 min - 1 / Protein content mg ml - 1
The chitinase enzyme production was performed by modifying the standard method followed by Hsu and Lockwood . One milliliter of the bacterial isolates was inoculated in the 250 ml Erlenmeyer flask containing 100 ml autoclaved Minimal Medium (Designated as MM) broth prepared by dissolving 0.5% colloidal chitin, 0.05% MgSO 4 ∙7H 2 O, 0.03% KH 2 PO 4 , 0.07% K 2 HPO 4 , 0.0001% MnCl 2 , 0.001% FeSO 4 ∙7H 2 O and 0.0001% ZnSO 4 , 15% NaCl in 1000 ml distilled water and the final pH was adjusted to 7. The inoculated tubes were incubated in a shaker incubator (200 rpm) at 30 °C for 48 h. After incubation, the isolates' cultures were harvested and used for carrying out further quantitative chitinase enzyme assay. An uninoculated test tube containing the same liquid broth was kept as blank.
Chitinase activity was determined by modifying a colorimetric method followed by Setia and Sohorjono in triplicates. The reaction mixture consisted of 1 ml of the crude enzyme and 2 ml of 1.25% (w/v) colloidal chitin substrate in a 200 mM potassium phosphate buffer (pH 6.0). The mixture was incubated at 30 °C for 2 h and boiled for 10 min to stop the reaction, then cooled to room temperature in a cold water bath, and 1 unit of β- N -Acetylglucosaminidase (NAGase) was added and then centrifuged at 8000 rpm for 20 min. The 1 ml of test supernatant obtained from the above centrifugation was added to 1.5 ml of freshly prepared color reagent solution prepared by mixing 96 mM DNSA (3,5-Dinitrosalicylic Acid) reagent in 5.3 M sodium potassium tartrate solution and diluted to 40 ml with deionized water. The test supernatant and color reagent solution mixture was then boiled for 5 min and cooled to room temperature. The concentration of GlcNAc ( N -acetylglucosamine) released was then measured at 540 nm. The standard curve of GlcNAc was plotted between GlcNAc concentration and GlcNAc absorbance. One unit of chitinase enzyme activity (U) was defined as the amount of enzyme capable of liberating 1.0 mg GlcNAc per hour from the chitin substrate under reaction conditions. The protein content in isolates was determined by the Folin-Lowry method using BSA as standard. Data on chitinase enzyme activity, specific activity, and protein content was analyzed with a single factorial CRD analysis of variance ( α = 0.05). [12pt]{minimal}
$$ {}: \\ {}\;{}: \\ {}0{}\;{} = {}0{} - {}0{} \\ {} {}0{}. \\ {}: \\ {}0{} = {}0{} - {}0{} \\ $$ Calculations : Standard curve : Δ A54 0 nm Standard = A54 0 nm Std - A54 0 nm Std Blank The Δ A54 0 nm of the standards was plotted against the milligrams of NAG released . Sample determination : Δ A54 0 nm Sample = A54 0 nm Test - A54 0 nm Test Blank The milligrams of NAG liberated were determined using the standard curve, and the chitinase enzyme activity (U ml −1 h −1 ) and its specific activity (U mg −1 h −1 ) defined per mg of protein estimated in isolates were then calculated using the following formula: [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = } ) \, ( {{3 } + {}^{{3}} } )}}{{( {{2 } { 1 } { 1}} )}}$$ Chitinase activity U ml - 1 h - 1 = mg NAG released 3 + Volume of NAGase 3 2 × 1 × 1 where, 3 = Initial reaction volume of assay. 2 = Conversion factor for converting 2 h to 1 h as per the unit definition. 1 = Volume (ml) of supernatant used in colorimetric determination. 1 = Volume (ml) of crude enzyme used. Volume of NAGase = 0.5 ml. [12pt]{minimal}
$${}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, = {}( {{}^{{ - {1}}} {}^{{ - {1}}} } ) \, /{}( {{}^{{ - {1}}} } )$$ Chitinase specific activity U mg - 1 h - 1 = Enzyme activity U ml - 1 h - 1 / Protein content mg ml - 1
Culture medium The ectoine production of the bacterial isolates was carried out in an autoclaved freshly prepared culture medium consisting of yeast extract (86 gl −1 ), ammonium sulfate (28 gl −1 ), FeCl 2 .4H 2 O (0.5 mM), MnSO 4 ∙H 2 O (10 μM), KCl (2 gl −1 ), MgSO 4 .7H 2 O (100 mM), and sodium chloride (1 M) as per the modified method of Chen et al . . Growth conditions A primary inoculum seed culture of all the isolates was first prepared by inoculating a loopful of the respective halophilic bacterial cells to 10 ml of autoclaved freshly prepared halophilic broth and incubated in a shaker incubator at 37 ℃ and 180 rpm for 24 h. After 24 h cultivation, 1 ml of the seed culture was inoculated into a 250 ml Erlenmeyer flask containing 50 ml autoclaved freshly prepared ectoine production broth and incubated at 30 ℃ and 200 rpm for 24 h. The flask uninoculated with isolate culture served as the negative control. After cultivation, the samples were taken to analyze cell growth and ectoine concentration. Ectoine quantification assay The halophilic bacterial isolates culture cultivated above was harvested by centrifugation at 8000 rpm. The pellets were resuspended in 80% ethanol (v/v) (Sigma) with vigorous shaking for 30 min. The ethanol extracts were filtrated through a 0.45 mm filter to analyze ectoine production. The ectoine concentration was then determined from the filtrate obtained by LCMS as per the details in Table . Ectoine, purchased from Sigma, was used as the reference standard.
The ectoine production of the bacterial isolates was carried out in an autoclaved freshly prepared culture medium consisting of yeast extract (86 gl −1 ), ammonium sulfate (28 gl −1 ), FeCl 2 .4H 2 O (0.5 mM), MnSO 4 ∙H 2 O (10 μM), KCl (2 gl −1 ), MgSO 4 .7H 2 O (100 mM), and sodium chloride (1 M) as per the modified method of Chen et al . .
A primary inoculum seed culture of all the isolates was first prepared by inoculating a loopful of the respective halophilic bacterial cells to 10 ml of autoclaved freshly prepared halophilic broth and incubated in a shaker incubator at 37 ℃ and 180 rpm for 24 h. After 24 h cultivation, 1 ml of the seed culture was inoculated into a 250 ml Erlenmeyer flask containing 50 ml autoclaved freshly prepared ectoine production broth and incubated at 30 ℃ and 200 rpm for 24 h. The flask uninoculated with isolate culture served as the negative control. After cultivation, the samples were taken to analyze cell growth and ectoine concentration.
The halophilic bacterial isolates culture cultivated above was harvested by centrifugation at 8000 rpm. The pellets were resuspended in 80% ethanol (v/v) (Sigma) with vigorous shaking for 30 min. The ethanol extracts were filtrated through a 0.45 mm filter to analyze ectoine production. The ectoine concentration was then determined from the filtrate obtained by LCMS as per the details in Table . Ectoine, purchased from Sigma, was used as the reference standard.
The isolates' production potentiality of the ectoine-compatible solute was confirmed by PCR-based molecular detection of its biosynthetic gene ectC which encodes for the ectoine synthase enzyme. Besides, the PCR-based molecular screening also showed the presence of BADH1 gene encoding for betaine aldehyde dehydrogenase enzyme which is responsible for the biosynthesis of glycine betaine compatible solute. The primer sequences for PCR amplification of these genes were obtained from literature reported by Rajan et al. and Anburajan et al . for ectoine and glycine betaine, respectively. The isolates' genomic DNA was isolated by Qiagen's blood and tissue kit based on the manufacturer's instructions. The isolated genomic DNA was then analyzed on 0.8% agarose gel electrophoresis, quantified by nanodrop spectrophotometer, and used for further PCR reactions. The PCR reaction was carried out in an Applied Biosystems thermal cycler in a 20 ml reaction volume system containing 50 ng of genomic DNA, 0.5 mM of each primer, 100 mM of each dNTP, 5 U of KAPA Taq DNA polymerase, and 10X Taq A buffer supplemented with 25 mM MgCl 2 . A single PCR vial containing all the above PCR reaction mixture except the DNA was used as the no template control (NTC) to check the chances of amplification due to primer contamination with the template DNA. The thermal cycler amplification reaction conditions were set with an initial denaturation at 94 ℃ for 5 min, followed by 35 cycles of denaturation at 94 ℃ for 45 s, annealing temperature at 50 ℃ for 45 s, and primary extension at 72 ℃ for 1 min, followed by a final extension at 72 ℃ for 7 min and hold at 4 ℃. The PCR amplified products were checked by running on agarose gel electrophoresis at 90 V cm −1 in 1.5% low EEO agarose gel prepared in 1× TAE buffer and added with 3% of 1000 ppm ethidium bromide. The resulting electrophoresed amplicons of the respective genes were then scanned and captured by the gel documentation system.
The above experiments were carried out in triplicate replications. The data obtained from their mean values were used for statistical analysis of variance (ANOVA) using a Completely Randomized Design (CRD) for the interpretation of results.
In the present study, it was thus concluded that the halophilic and halotolerant bacteria isolated from the soils of agricultural fields lying along the southwest coastline of Saurashtra, Gujarat, showed a promising potentiality for production of the industrially important proteolytic, cellulolytic, and chitinolytic extremozymes and may have potential application in many industries especially cellulose and chitinous biomass conversion for biofuel production, etc. Besides, the isolates also represent a potent source of biocontrol agents. They may contribute to sustainable agriculture as an alternative to chemical pesticides against the control and management of fungal diseases, insect pests, and nematodes. However, the isolated halophilic and halotolerant bacteria's reported activities only hint at the possibility of such novel applications. Further study on the candidate isolates’ improvement for higher extremozyme production, their extraction, purification, characterization, and application trials are required for a detailed evaluation of their practical applications. Furthermore, the isolates also exhibited an appreciable potential for producing ectoine-compatible solute, as validated by the presence of its biosynthetic gene. Besides, the isolates also showed the presence of glycine betaine-compatible solute biosynthetic gene. Thus, these isolates may serve as a promising source of osmoprotectant-responsive genes for developing osmotolerant transgenic plants against salinity, heat, and drought stresses, as already reported by many scientists. The study also suggests that there may be a correlation between compatible solutes and extremozyme production of the isolates under saline conditions. The compatible solutes could be playing a vital role in aiding the maintenance of normal extremozyme production by protecting them from salt-induced denaturation effects, potentially enhancing their stability and activity. However, this hypothesis is purely our assumption, and further investigation is required to confirm it.
Supplementary Figure 1. Supplementary Figure 2. Supplementary Figure 3. Supplementary Figure 4. Supplementary Figure 5. Supplementary Figure 6. Supplementary Tables.
|
A practical guide to public involvement with children and young people in dental research | 33f2d3a0-a0a6-4c13-b916-a5a981d11d77 | 11761068 | Dentistry[mh] | Public involvement (PI) in health research is an umbrella term which describes the process by which research is undertaken ‘with' or ‘by' people rather than ‘to', ‘about' or ‘for' them. The terms patient and public involvement (PPI) and PI are often used interchangeably but have subtle differences in their definition. PPI provides separate definitions of patients and the public; patients are seen as current or former users of health and social care services, with the public seen as anybody else, such as potential users of healthcare services. PI encompasses both, including current, former or potential patients and those who represent patients, carers and family members. Public involvement differs to public engagement. Engagement focuses on the dissemination of research information and knowledge to the public, for example raising awareness of research or disseminating research findings. PI involves a partnership between the researchers and the public, empowering the public to influence decision-making at all stages of the research process. This may include a range of activities, including prioritising research themes, working as part of a project advisory group, informing the development of research materials, or carrying out user-led research. Regardless of the activity or stage, PI should be meaningful, empowering the public to inform research development, and not simply a tick-box exercise. It is also important to consider that children and young people (CYP) may also be current, former or potential service users, carers or family members and should be involved in PI. Health research should have the overarching aim of meeting the needs of the public, including where those groups are CYP. To meet this aim, it is important to work with those who have relevant lived experience or knowledge, including CYP, facilitating their voice to produce research which is relevant to their needs. In recent years, the expectation from funders for PI to be part of the research has increased, with many funders stipulating that applicants demonstrate how the public will actively be involved in the design and delivery of their research. , In 2022, several funding bodies, including UK Research and Innovation and the National Institute for Health and Social Care Research (NIHR) signed a shared commitment to improve public involvement in research, stating that ‘public involvement is important, expected and possible in all types of health and social care research'. In addition to this, there has been a drive for PI in wider fields, such as NHS service delivery, clinical guideline development and the UK parliamentary system. , , PI encompasses all activities which aim to include the public, including CYP, in the research process; however, the extent to which people are involved in the research process may differ. Involvement with CYP may take place at different phases of the research process and at different levels. highlights an adapted participation matrix, originally developed by Shier, which describes three levels of involvement - consultation, collaboration and user-led - across the different phases of the research process. , Consultation describes a one-off involvement process, where CYP provide opinions on certain aspects of a proposal to inform the research but are not actively involved in decision-making on an ongoing basis. , Collaboration describes ongoing involvement with CYP, where they are actively involved in the research process. , In this case, CYP work alongside researchers providing input into areas such as research design and/or data collection or analysis and/or dissemination. User-led describes a research process which is led by CYP, rather than the researcher. , With support from researchers, CYP design and deliver the research project. This may be the sole study, or there may be PI-led elements within a larger research study. To undertake high-quality research with CYP, it is important that they are involved in PI and also as participants in the research. It is important to include CYP as active participants in research where they are allowed to provide their experiences and opinions, rather than using proxies, such as parents. In a systematic review published in 2015, only 17.4% of dental research was undertaken with CYP where CYP were participants in the study. Additionally, 18.1% of studies used proxies for CYP and 64.2% undertook research on children, where they were subjects and not involved in the research. While this is an improvement from 2007, where only 7.3% of dental research was with CYP, there is still a need for significant improvement in the extent to which CYP are involved in research.
The United Nations Convention of the Rights of the Child (UNCRC) provides CYP with a comprehensive set of human rights. Article 12 of the UNCRC states that ‘every child has the right to express their views, feelings, and wishes in all matters affecting them, and to have their views considered and taken seriously'. CYP should have the opportunity to contribute directly and this input can have many benefits to both CYP and the research. CYP can be involved from the start, aiding in identification and prioritisation of research questions. Through the design, CYP can inform recruitment strategies, techniques for data collection and dissemination of results. This can have great benefits for research, including wider involvement of CYP and improved recruitment and retention. Involvement in research supports CYP to develop a wide range of research skills, such as writing and public speaking. This has been associated with a self-perceived improvement in confidence, self-esteem and employment opportunities. CYP report positive experiences of involvement in research, such as feeling part of a team, feeling listened too, empowerment and a greater understanding of their rights.
Early involvement of CYP may identify a research question relevant to your local community or context, which may not otherwise have been identified. While advocating for early involvement in the research design process, we note this may not be possible, for example, where funding is provided from a pre-defined research question. Despite this, engagement with CYP as early as possible is beneficial for the research and CYP. It is important to consider the level of involvement you anticipate using: consultation, collaboration, or user-led. There are many factors which may influence this decision, such as time availability, funding availability, type of research being undertaken and previous PI experience. Time and funding availability are some of the biggest limiting factors when considering public involvement, which can impact a researcher's ability to undertake meaningful PI. Where possible, appropriate time and funding should be incorporated into research design to facilitate ongoing PI. Where this is not possible, a pragmatic approach is needed to consider how CYP can be involved in the research. In these cases, consultation approaches are often used to gain feedback from CYP; however, it vital that this is appropriately planned and the feedback actioned to avoid this moving to a tick-box approach to PI. The nature of research can influence the type of PI planned and, in some studies, it may not be appropriate to have CYP involved in all aspects of the research. However, in such studies, CYP can have a vital role in developing techniques for disseminating research. Undertaking PI for the first time can be a daunting but it doesn't mean that you can't undertake meaningful PI. While the levels of PI are often described in isolation, there may be a natural development from consultation to collaboration or user-led research. We note that collaboration and user-led research can be easier once the research has developed a relationship with a community of CYP with an interest in this area. Meaningful consultation can have great benefit to research and can help foster partnerships with CYP, opening the door for further involvement where CYP have greater autonomy.
PI is a fluid and ever-evolving process and is highly dependent on the research area and the CYP involved. Considering the needs of CYP, it is almost impossible to create a one-size-fits-all approach to PI. However, there is guidance available from a range of sources: UNCRC Article 12: the right of the child to be heard UK Standards for Public Involvement in research NIHR: briefing notes for public involvement in the NHS, health and social research Top tips for involving CYP in research from CYP's point of view Royal College of Paediatrics and Child Health: engaging children and young people While ethical approval is often not needed, the underlying ethical principles should still apply to PI processes. These include areas such as informed consent, safeguarding, ensuring confidentiality, minimising risk of harm and training for researchers and PI members (where required). Ethical approval may be required, for example, for user-led research; although, this remains a point of discussion as outlined by Nollett et al . If unsure, it is important to discuss this with your local institution.
The CYP involved in PI will depend on the type of research being undertaken and the nature of the input required. It is important to be flexible in your approach in identifying those to be involved in your PI, as this may evolve as your research progresses. Firstly, consider the population you plan to be involved in the research. This may be associated with characteristics such as age, location or a certain health condition or lived experience. Secondly, consider the level of involvement you are looking for, for example, a one-off consultation or a long-term, user-led research project, as this may alter the initial approach. Once the target group has been identified, methods to advertise the involvement opportunities should be considered. It is useful to consider whether your organisation, such as an NHS trust or university, already has an established link to existing groups which can be used. These may include: Existing local/regional/national young person's advisory group (YPAG) - identify whether there is a local YPAG in your region. There may be a GenerationR YPAG near you, which is an alliance of YPAGs across the UK funded by NIHR and/or NHS organisations through various channels. Contact the co-ordinator to discuss involvement of the YPAG Existing PI groups relevant to your research theme - there may be a regional or national PI group relevant to your research area. They may be able to be involved in your work or may be able to provide input as to the best place to advertise for the CYP you are looking to involve Charities or support groups - a wide range of charity groups, support groups or patient networks exist locally, regionally, nationally and internationally. These groups may be aimed at certain populations, such as those with specific conditions or of a certain age, so it is important to identify if there is a group relevant to your research. These may be of benefit for research regarding rare diseases, as it can help identify those who may be current patients, carers or family members Relevant settings - there may be settings which may be best suited to the CYP you are looking to involve. Examples include healthcare settings, activity groups or clubs. It can be useful to contact these areas and discuss the possibility of advertising through these networks. When considering healthcare settings, there may be wider ethical considerations associated with these settings Social media - social media may be useful to disseminate this information through wider networks. The characteristics of those involved also needs to be considered, with a desire to maximise the diversity of the group. There may be groups of CYP who are less likely to be involved in research, and while there is a wide range of terminology used to describe these groups, they are often defined as under-served groups. This definition best reflects that research should better serve these groups and facilitate their involvement. Intention must be made to plan ways to actively offer these groups an opportunity to be involved in research. Researchers may consider contacting those who may have an existing relationship with these groups, who can be described as ‘gatekeepers'. These gatekeepers can be a range of people who work in different settings, such as healthcare professionals in the community, those in community groups, such as children's centres and clubs, or religious groups. Contacting such gatekeepers and explaining the rationale behind the research and the expectations of the PI will be useful. Gatekeepers may suggest adaptions to the planned PI to support CYP involvement and can suggest the best way to advertise to increase involvement. Additionally, advertisement through these gatekeepers, who are often a trusted person within the community, can facilitate rapport building and subsequent involvement, rather than advertising coming from an unknown researcher. Building trust and rapport with gatekeepers and communities takes time and this should be considered in the research planning.
Adapting the setting of PI can be useful to help people get CYP involved. Holding events in a location which is familiar to the public can be useful to aid involvement. For example, rather than inviting people to attend a meeting with you at a different location, try to hold a session in a convenient location, or attend a scheduled meeting with an existing group. This can ease the process of involvement and help reduce the burden for CYP. There are many methods you can use to involve CYP in research. Common examples include questionnaires, interview or focus group discussions, and interactive workshops. The methods used are flexible depending on the CYP involved but they should be encouraging and easy for CYP to give their opinions. Reasonable adjustments should be made to facilitate involvement of CYP who may need additional support for communication. While the input is coming from CYP, some CYP may prefer to have their parent or guardian present for support, while others may not. Discuss this with the CYP involved and make adaptions so that all CYP are comfortable. If parents are present, ensure they have information regarding their role in support to ensure that the CYP's voice leads the discussion. There may need to be several events where CYP of a similar age range are together so that discussions can be pitched at the appropriate level of understanding. When involving CYP, timing is also of particular importance. CYP often have busy lives, including extra-curricular activities, crucial timings, such as GCSE and A-level exams, and other personal responsibilities. It is important to be flexible, offering after-school times, weekends or school holidays, depending on their preference. Virtual events may be more convenient and can be useful for PI which is required over a large geographical area. Face-to-face events can allow CYP to discuss with each other, but it is important to hold these in a place convenient to the CYP. It is important to cost sufficient funding for PI. Guidance for remuneration is available from NIHR. Consider the length of time and level of commitment required for the involvement and be transparent with CYP regarding the commitment. Remuneration should be reflective of the level of commitment and any associated costs, such as travel expenses. Shopping vouchers are a popular method of remuneration for CYP. Direct monetary payment can be considered, such as for travel expenses; although, this can have complexities and local guidance should be followed. Some CYP will want to be involved in research and prefer not to be remunerated for this and it is important to respect these wishes.
Evaluation is a key component of PI. It is important to consider the impact involvement has on both the public and the research. Firstly, it is important that CYP are provided with feedback regarding the input they have provided and how this influenced the research. It is important that this feedback is transparent and timely. Failure to do so can leave those involved dissatisfied and with feelings of it being a tick-box exercise. There are many ways that impacts of PI can be shared, such as a newsletter or a website. It is beneficial to discuss the preferred ways of receiving updates with those involved to ensure it is timely and relevant to their needs. The research team should gain feedback from those involved regarding their experience of the PI process and how it met their expectations. Key areas to evaluate include setting, timing, activities, feeling heard, meeting expectations and areas for development. This can take many formats but questionnaires or open discussions with those involved are commonly used. Reflections from the researcher using a diary can also be helpful to note key discussions or developments from PI.
Sharing learning about CYP involvement in practice is aided through systematic evaluation and reporting of what works best for CYP and describes what impact their involvement has on the actual research and on those who get involved. To aid the reporting process, the Guidance for Reporting Involvement of Patients and the Public (GRIPP2) checklist has been developed to improve the quality and consistency of reporting. An example GRIPP2 short form can be seen in , outlining the key areas for reporting. In addition to reporting PI within scientific publications, it is important to consider dissemination of the overall research outputs and PI contributions to the public, relating public involvement to public engagement. There are many methods of public engagement, such as open days or community events, social media and websites, and this will likely vary depending on the nature of the research. This can be beneficial to both demonstrate the impact of PI and research and may encourage others to be involved in research in the future. This forms part of public engagement, which focuses on the dissemination of research to the public.
As demonstrated, there is great scope for PI in dental research, with benefits for the researcher, CYP and the research output. At all levels of PI, there is opportunity for meaningful relationships to be built with CYP to create research which is both achievable and relevant to their needs and desires. While PI can have great benefits to dental research, it is important to acknowledge challenges that research teams face during the PI process. Limitations in time, funding or ability to engage sufficient numbers of CYP are often reported by researchers. However, this does not mean that the PI produced will not be meaningful. Proactive engagement of CYP through a range of methods, transparent reporting and research reflection are key in preventing PI becoming a tick-box exercise. Using these principles, there is opportunity for involvement of CYP in a range of settings and the authors actively encourage readers to involve CYP in decision-making for what is, after all, their research.
|
Arcopilins:
A New
Family of | b4ef72ff-e629-404f-bac9-41e197601e5a | 11403616 | Microbiology[mh] | Sordarialean fungi, renowned for their
pivotal ecological roles
across various natural habitats, have emerged as instrumental contributors
in different fields of economic relevance. − This diverse taxonomic group is also a prolific source
of biologically active secondary metabolites, from which taxa belonging
to the Chaetomiaceae are particularly known to harbor a wealth of
unique and chemically diverse entities. , Despite the
extensive research on fungi within this family, the exploration of
untapped genera continues to offer opportunities for the discovery
of novel natural products with diverse biological activities. Over the past three decades, biofilms have been a relevant topic
due to their complex nature and impact on human health. Biofilms are
structured microbial communities, which adhere to any suitable living
or abiotic surface through a self-produced matrix of extracellular
polymeric substances (EPSs). The three-dimensional
EPS matrix provides several functions within biofilms, such as the
transportation of signals and nutrients between cells and the environment. − In addition, biofilms confer protection against environmental factors,
including high salt concentrations, ultraviolet radiation, extreme
temperatures, pH variations, high pressure, and chemicals. − As a result, biofilms also significantly enhance the tolerance and
resistance of pathogens to antibiotics when compared to planktonic
cells. According to a report by the National
Institutes of Health (NIH), bacterial pathogens forming biofilms are
responsible for 80% of the chronic infections in clinical trials. Among these pathogens, Staphylococcus
aureus , recognized as an ESKAPE pathogen, is one of
the most dangerous opportunistic organisms, causing a range of human
infections. Numerous diseases, including
osteomyelitis, cystic fibrosis, and otitis media, are therefore related
to the biofilm infection of S. aureus . − Microbial infections threaten the development of society,
as their
treatment remains a global challenge with the rapid increase and spread
of resistance. − To address this substantial challenge, combination
therapy has been increasingly accepted in recent years, building on
approaches established for anticancer treatment. This approach involves targeting multiple pathways within
important pathogen biological processes, circumventing their defense
mechanisms. For instance, the combination
of sublethal concentrations of bacteriophages with the antibiotic
vancomycin, or using biofilm-targeting antigens as a vaccine in conjunction
with vancomycin, has significantly reduced S. aureus biofilm formation. , These strategies are effective
through the disruption of the biofilm structure or cell membrane,
offering new avenues for therapeutic intervention. During an
ongoing project focused on the discovery of bioactive
compounds from taxa belonging to the Sordariales, six previously undescribed
tetramic acids ( 1 – 6 ) and a related
2-pyridone congener ( 7 ) were isolated from the soil-born
fungus Arcopilus navicularis CCF 3252 T . Herein, we report the isolation, structure elucidation,
antimicrobial activities, and biofilm disruption properties against S. aureus of arcopilins A–G ( 1 – 7 ). Due to the remarkable efficacy of arcopilin
A ( 1 ) to disrupt S. aureus biofilms at subtoxic concentrations, we decided to systematically
examine its synergistic effect in combination with the known antibiotics,
gentamicin (GM) and vancomycin (Vac), ineffective against the preformed
biofilms of this pathogen.
Isolation and Structure
Elucidation of Arcopilins The
strain CCF 3252 T was obtained from the Culture Collection
of Fungi (CCF) in Prague. This strain represents the type strain of A. navicularis . Morphologically,
this species is characterized by ascomata bearing arcuate hairs with
incurved to coiled apexes and navicular ascospores with two apical
germ pores . The production of secondary metabolites by the
chaetomiaceous fungus A. navicularis CCF 3252 T was evaluated
under its cultivation in three different liquid media (YM 6.3, ZM
1/2, Q6 1/2) and one solid medium (BRFT) ( Figure S1 ). Metabolomic analysis of the obtained crude extracts by
high-resolution electrospray ionization mass spectrometry (HR-ESI-MS)
discerned the production of nitrogen-containing molecules with unprecedented
molecular formulas and a distinctive UV/vis absorption at λ max 226, 288, and 346 nm in the Q6 1/2 medium. After the scaled-up
fermentation of A. navicularis CCF
3252 T in Q6 1/2 medium (8 L), targeted isolation by preparative
HPLC afforded compounds 1 – 6 as brown
to orange oils and 7 as an orange to white powder . Their planar structures
were elucidated by 1D and 2D NMR spectroscopy in combination with
tandem mass spectrometry analyses ( ; Figures S4–S45 ). The molecular formula of compound 1 was
determined
as C 22 H 27 NO 4 according to the quasimolecular
ion peak cluster at m / z 370.2015
[M + H] + in the HR-ESI-MS spectrum, indicating ten degrees
of unsaturation. 1 H and HSQC spectra revealed the presence
of four methyl, two methylene, and four olefinic/aromatic signals,
two of the aromatics with dual intensities. Since the 13 C NMR spectrum only contained signals for an additional ketone and
a further quaternary carbon without bound protons, signals of five
carbon atoms were missing according to the molecular formula. HMBC
correlations connected a styryl and an oxotrimethyleptyl moiety as
two isolated parts of the molecule ( Figure S8 ). Based on the coupling of 6–H to N–1 in the 1 H, 15 N HMBC spectrum, four unassigned degrees of
unsaturation, and chemical shifts, we deduce the tetramic acid backbone
for 1 . Tetramic acids are known for their tautomeric
exchange, explaining the missing signals in the NMR spectra. The rather
small shift difference of the germinal methylene protons of Δδ H = 0.20 and 0.10 ppm for 9–H 2 and 11–H 2 , respectively, is indicative of a trans / trans configuration of the methyl groups. The molecular formula of 2 was established
as C 22 H 29 NO 4 according to the quasimolecular
ion peak cluster at m / z 372.2168
[M + H] + in the HR-ESI-MS spectrum, corresponding to the
loss of one degree of unsaturation compared to 1 . NMR
data were highly similar to those of 1 , with the replacement
of the C–13 keto moiety by a hydroxyl. A J-resolved analysis
connected the stereochemistry of C–12 and C–13, while the patterns of the Δδ SR shift with a negative value for 14–H 3 (−0.09)
and positive ones for 12–H (+0.04) and 17–H 3 (+0.08) were indicative for an 8 S ,10 R ,12 R ,13 S absolute configuration. Compounds 3 and 4 were found to be the
18–formyl and 18–dehydroxy derivatives of 2 , respectively. Indicative for the structures were the molecular
formulas C 23 H 29 NO 5 and C 22 H 29 NO 3 , respectively, in addition to the additional
formyl group connected to C–13 by HMBC coupling in 3 as well as the lack of signals for the hydroxyl function at C–13
in 4 . HR-ESI-MS data revealed C 22 H 27 NO 6 as the molecular formula of compound 5 , meaning two additional oxygen atoms compared to 1 .
These were located at C–12 and C–14, as demonstrated
by the replacement of the methyl group CH 3 –14 as
well as methane CH–12 by an oxymethylene as well as a carbon
devoid of bound protons. Since ROESY correlations and coupling constants
remained largely unchanged, we ascribe 3 as the 8 S ,10 S ,12 S configuration.
HR-ESI-MS data disclosed the molecular formula C 22 H 27 NO 7 for 6 . In the structure of 6 , methyl C-14 and methine C–12 of 1 were
replaced by a carboxylic acid and an oxygenated carbon devoid of bound
protons, respectively. Coupling constants and ROESY correlations are
similar to those of 3 , and thus, we assign a common 8 S ,10 S ,12 S ,13 R configuration. Compound 7 had the same molecular
formula C 22 H 29 NO 4 as 2 . However, NMR data
showed apparent differences. The methane CH–6 was significantly
deshielded (δ H 7.86/ δ C 133.6) compared
to compounds 1 – 6 , and all expected
carbons were observed in the 13 C NMR spectrum, indicating
a strongly lesser degree of tautomerism. The same styryl and 6-keto-1,3,5-trimethyleptyl
moieties were assembled by COSY and HMBC data, but HMBC correlations
from 4–OH to C-3, C-4, and C-5 and from 6–H to C-2,
C-4, and C-5 connected the α-pyridone. Strong ROESY correlations
between 7–H and 11–H as well as 8–H and 16–H 3 established the 7 S ,8 S ,10 S ,11 S stereochemistry. The crystal
structure of compound 7 was determined
via a continuous rotation 3D electron diffraction (3D ED) experiment
collected on a XtaLAB Synergy-ED diffractometer. The structure was solved with direct methods, and the absolute configuration was determined
in the course of dynamical refinement in JANA. , The absolute configuration of the stereocenters, as well as the
molecular conformation within the crystal structure, is shown in . The experimental
and refinement details as well as the CSD deposition number of the
structure are given in the Supporting Information . Arcopilin G ( 7 ) is nearly the enantiomer of septoriamycin
A, which has been isolated from a culture medium of the ascomycete
fungus Septoria pistaciarum . A total synthesis of septoriamycin A has been
completed by Fotiadou and Zogrofos. Extensive knowledge of
the biosynthesis of tetramic acids and their
related 2-pyridones reveals a common progression catalyzed by polyketide
synthase-nonribosomal peptide synthetase (PKS-NRPS) hybrid machineries.
The diversity and evolution of these biosynthetic pathways are illustrated
in several natural products, including tenellin, aspyridone A, fusarin
C, leporin B, fischerin, PFF1140, sambutoxin, equisetin, etc. , In the late stages of tenellin biosynthesis, two cytochrome P450
oxidases are responsible for catalyzing the oxidative expansion and
N-hydroxylation of pretenellin A. Furthermore,
certain metabolites might undergo cyclization of their side chains
through processes such as inverse-electron demand Diels–Alder
reactions, as seen in the antifungal ilicicolin H, or through a Michael
addition, as observed in the biosynthesis of the mycotoxin -sambutoxin. , Since compounds 1 – 7 share the same
carbon skeleton except for 3 , it is likely that 7 is biosynthesized in a similar fashion as -sambutoxin,
a related PKS-NRPS hybrid product with a longer polyketide chain. Antimicrobial and Cytotoxic Activities of Arcopilins The
antimicrobial activities of compounds 1 – 7 (Acp A–G) were assessed against different bacterial
and fungal strains in addition to their cytotoxic effects on two mammalian
cell lines. The tested microorganisms comprised a diverse array of
clinically relevant pathogens, encompassing sensitive indicator strains.
Among the Gram-positive bacteria were Bacillus subtilis , Staphylococcus aureus , and Mycolicibacterium smegmatis . Gram-negative bacteria
included Acinetobacter baumannii , Chromobacterium violaceum , Escherichia
coli , and Pseudomonas aeruginosa . Additionally, filamentous fungi such as Mucor hiemalis and yeasts including Candida albicans , Wickerhamomyces anomalus , Rhodotorula glutinis , and Schizosaccharomyces
pombe were included. Generally, all compounds presented
similar biological properties, summarized in weak or no activity against
fungal pathogens as well as weak to moderate inhibition of Gram-positive
bacteria .
Acp E and F did not exhibit any antimicrobial activity in our assays. The above suggests that the hydroxylation at C-12
and C-14 in Acp
E has a negative effect on antibacterial activity. Similarly, the
hydroxylation at C-12 and C-13, in addition to the presence of carboxylic
acid at C-14 in Acp F, results in the loss of antibacterial activity.
In terms of their cytotoxic properties, note that 2-pyridone Acp G
was the most cytotoxic metabolite, while its tetramic acid congeners
presented rather weak or no cytotoxic effects as for compounds Acp
B, Acp E, and Acp F. The fact that hydrophilic arcopilins are less
cytotoxic suggests a possible correlation between the hydrophobicity
and the cytotoxicity of these molecules. While PKS-NRPS hybrid
products within the tetramic acid and pyridone
secondary metabolite families exert a wide range of biological activities
and are widespread in ascomycetes, only a limited number of examples
from the Sordariales order have been reported. Notably, the most notorious
examples are the decalin-containing tetramic acids, myceliothermophins,
originally discovered in Thermothelomyces thermophilus (syn. Myceliophthora thermophila ). The potent antitumor activity exhibited by myceliothermophins
C, D, and E against a number of human cancer cell lines has prompted
numerous total synthesis endeavors. , Similarly,
the chaetolivacines A–C, isolated from Chaetomium
olivaceum (Chaetomiaceae), represent another example
of decalin-containing tetramic acids. Only chaetolivacine B exerts moderate antibacterial properties against S. aureus and methicillin-resistant S. aureus (MRSA). Additionally, rare decalin-containing
tetramic acids such as zopfiellamide A and B, as well as zopfielliamides
A–D, have been isolated from Zopfiella latipes and Zopfiella sp., taxa with uncertain taxonomic
placement within this order. , , Arcopilins Are Able to Disrupt the Preformed Biofilms of Staphylococcus aureus After identifying
that arcopilins exhibit rather weak activities against the tested
organisms and cell lines, we decided to evaluate their efficacy toward
the disruption of preformed biofilms of the bacterial pathogen S. aureus . Therefore, Acp A–G were evaluated
against preformed biofilms of S. aureus using crystal violet staining. The
2-pyridone, 15-hydroxytenellin (15-Ht), produced by the entomopathogenic
fungus Beauveria neobassiana was also
used for comparison, as the tenellins are model compounds for the
study of fungal secondary metabolite biosynthesis and have displayed
inhibitory properties against the formation of biofilms by S. aureus . , Among the tested metabolites,
Acp A and C showed the most promising disrupting effects toward preformed
biofilms of S. aureus , whereas weak
to moderate effects were observed for Acp B, F, and G . Furthermore, Acp G and 15-Ht (data not shown), both belonging
to the
class
of 2-pyridones, were not active against preformed biofilms of S. aureus . Acp A displayed approximately 50–60%
efficacy toward preformed biofilms within the concentration range
of 15.6 to 250 μg/mL. Similarly, Acp C demonstrated ca. 50%
effectiveness in the dispersal of preformed biofilms between 7.8 μg/mL
and 250 μg/mL. Notably, both compounds exhibited a pronounced
efficacy of 35–45% even at a concentration as low as 3.9 μg/mL.
These results align with the growth curve of Acp A shown in Figure S47 , demonstrating that the growth of S. aureus was inhibited by Acp A treatment even at
concentrations as low as 3.9 and 7.8 μg/mL. Consequently, the
disruption of existing biofilms may be due to the downregulation of
cell growth. However, the precise mechanism behind this effect remains
unclear and it is out of the scope of the present study. In
the case of Acp A and C, a carbonyl group is present on the
side chain of these metabolites. However, the presence of this moiety
is not exclusively necessary for the observed activity, as demonstrated
by Acp D, which lacks a carbonyl group but still exhibits significant
dispersal effects at concentrations as low as 31.3 μg/mL. During
our examination of different tetramic acids and related 2-pyridones,
no discernible link between cytotoxicity and the dispersion of S. aureus preformed biofilms was found. For instance,
Acp G, the most cytotoxic metabolite within the tested congeners,
exhibited only weak disruptive effects on the biofilms. A link between
cytotoxicity and biofilm eradication could affect the applicability
of the metabolites, as increased cytotoxicity might also damage host
cells. Synergistic Effects of Arcopilin A in Combination with Gentamicin
and Vancomycin Interestingly, both Acp A and Acp C demonstrated
remarkable effectiveness in disrupting preformed biofilms of S. aureus . Given its promising activity and relatively
low cytotoxicity, we selected Acp A for further experiments. We investigated
its in-depth effects alone and in combination with the antibiotics
gentamicin (GM) and vancomycin (Vac) on planktonic cells and S. aureus biofilms. To evaluate the influence of
Acp A on both biofilm metabolic activity and planktonic cell growth,
XTT and growth curve analyses were conducted, respectively. The results
from XTT assay as depicted in Figure S46 revealed a significant reduction in metabolic activity even at low
concentrations of 3.9 μg/mL. These findings were consistent
with the outcomes of the antibiofilm assay, indicating that effective
concentrations of Acp A in dispersing S. aureus biofilms coincide with an alteration in the metabolic activity of
preformed biofilms. In line with this, inhibitory effects were observed
at concentrations between 7.8 μg/mL and 2 μg/mL according
to the growth curve analysis ( Figure S47 ). After assessing the effects of Acp A on the pathogen S. aureus , we delved deeper into the interaction
of Acp A with established antibiotics (GM and Vac). This exploration
focused on both planktonic cells and preformed biofilms of S. aureus . Consequently, we used a checkerboard assay
to determine the fractional inhibitory concentration index (FICI)
for combinations involving Acp A, GM, or Vac based on both their MIC
values in combination. Antibiotics
commonly used to fight bacterial infections often act
through diverse mechanisms to hinder the growth of these pathogens.
For instance, the well-known antibiotic GM functions as a protein
synthesis inhibitor, while Vac exerts inhibitory effects on this pathogen
by interfering with cell wall biosynthesis. Our findings revealed
that when Acp A (3.9 μg/mL) was used in combination with GM
or Vac, the MIC values of the established antibiotics were significantly
decreased from 15.6 to 0.13 μg/mL and from 2 to 0.065 μg/mL,
respectively. The combination treatment substantially increased the
potency of GM and Vac up to 115-fold and 31-fold, respectively, and
calculation of the FICI showed synergistic effects (FICI < 0.5)
for both combinations ( b). Similarly, the MIC value of Acp A decreased almost 10-fold
when combined with each antibiotic. Additionally, combined effects
were also assessed on the preformed biofilms. Consequently, a colony-forming
unit (CFU) count analysis treated with Acp A (7.8–3.9 μg/mL),
GM (7.8–2 μg/mL), or Vac (15.6–3.9 μg/mL)
alone, as well as their combinations, was carried out for preformed
biofilms. For both GM and Vac, roughly a 3-fold improvement in the
inhibitory effects was observed when used in combination with Acp
A (7.8–3.9 μg/mL) ( c). According to previous studies, tetramic acids with
long polyketide
side chains, such as the reutericyclins, have been shown to act against
bacteria by disrupting their proton gradient and membrane potential. , The cellular membrane potential is dynamic, and it is linked to
signal transmission between cells within biofilms and the overall
level of biofilm formation. In addition, tetramic acids are likely
to act as metal chelators, but the biological implications of this
phenomenon are poorly understood. Similarly,
it has been demonstrated that human-targeted drugs, when used at sublethal
concentrations, can be repurposed as new antimicrobials in combination
therapy. However, the specific mode of
action by which arcopilins disrupt S. aureus biofilms remains unclear and will require future investigation.
The
strain CCF 3252 T was obtained from the Culture Collection
of Fungi (CCF) in Prague. This strain represents the type strain of A. navicularis . Morphologically,
this species is characterized by ascomata bearing arcuate hairs with
incurved to coiled apexes and navicular ascospores with two apical
germ pores . The production of secondary metabolites by the
chaetomiaceous fungus A. navicularis CCF 3252 T was evaluated
under its cultivation in three different liquid media (YM 6.3, ZM
1/2, Q6 1/2) and one solid medium (BRFT) ( Figure S1 ). Metabolomic analysis of the obtained crude extracts by
high-resolution electrospray ionization mass spectrometry (HR-ESI-MS)
discerned the production of nitrogen-containing molecules with unprecedented
molecular formulas and a distinctive UV/vis absorption at λ max 226, 288, and 346 nm in the Q6 1/2 medium. After the scaled-up
fermentation of A. navicularis CCF
3252 T in Q6 1/2 medium (8 L), targeted isolation by preparative
HPLC afforded compounds 1 – 6 as brown
to orange oils and 7 as an orange to white powder . Their planar structures
were elucidated by 1D and 2D NMR spectroscopy in combination with
tandem mass spectrometry analyses ( ; Figures S4–S45 ). The molecular formula of compound 1 was
determined
as C 22 H 27 NO 4 according to the quasimolecular
ion peak cluster at m / z 370.2015
[M + H] + in the HR-ESI-MS spectrum, indicating ten degrees
of unsaturation. 1 H and HSQC spectra revealed the presence
of four methyl, two methylene, and four olefinic/aromatic signals,
two of the aromatics with dual intensities. Since the 13 C NMR spectrum only contained signals for an additional ketone and
a further quaternary carbon without bound protons, signals of five
carbon atoms were missing according to the molecular formula. HMBC
correlations connected a styryl and an oxotrimethyleptyl moiety as
two isolated parts of the molecule ( Figure S8 ). Based on the coupling of 6–H to N–1 in the 1 H, 15 N HMBC spectrum, four unassigned degrees of
unsaturation, and chemical shifts, we deduce the tetramic acid backbone
for 1 . Tetramic acids are known for their tautomeric
exchange, explaining the missing signals in the NMR spectra. The rather
small shift difference of the germinal methylene protons of Δδ H = 0.20 and 0.10 ppm for 9–H 2 and 11–H 2 , respectively, is indicative of a trans / trans configuration of the methyl groups. The molecular formula of 2 was established
as C 22 H 29 NO 4 according to the quasimolecular
ion peak cluster at m / z 372.2168
[M + H] + in the HR-ESI-MS spectrum, corresponding to the
loss of one degree of unsaturation compared to 1 . NMR
data were highly similar to those of 1 , with the replacement
of the C–13 keto moiety by a hydroxyl. A J-resolved analysis
connected the stereochemistry of C–12 and C–13, while the patterns of the Δδ SR shift with a negative value for 14–H 3 (−0.09)
and positive ones for 12–H (+0.04) and 17–H 3 (+0.08) were indicative for an 8 S ,10 R ,12 R ,13 S absolute configuration. Compounds 3 and 4 were found to be the
18–formyl and 18–dehydroxy derivatives of 2 , respectively. Indicative for the structures were the molecular
formulas C 23 H 29 NO 5 and C 22 H 29 NO 3 , respectively, in addition to the additional
formyl group connected to C–13 by HMBC coupling in 3 as well as the lack of signals for the hydroxyl function at C–13
in 4 . HR-ESI-MS data revealed C 22 H 27 NO 6 as the molecular formula of compound 5 , meaning two additional oxygen atoms compared to 1 .
These were located at C–12 and C–14, as demonstrated
by the replacement of the methyl group CH 3 –14 as
well as methane CH–12 by an oxymethylene as well as a carbon
devoid of bound protons. Since ROESY correlations and coupling constants
remained largely unchanged, we ascribe 3 as the 8 S ,10 S ,12 S configuration.
HR-ESI-MS data disclosed the molecular formula C 22 H 27 NO 7 for 6 . In the structure of 6 , methyl C-14 and methine C–12 of 1 were
replaced by a carboxylic acid and an oxygenated carbon devoid of bound
protons, respectively. Coupling constants and ROESY correlations are
similar to those of 3 , and thus, we assign a common 8 S ,10 S ,12 S ,13 R configuration. Compound 7 had the same molecular
formula C 22 H 29 NO 4 as 2 . However, NMR data
showed apparent differences. The methane CH–6 was significantly
deshielded (δ H 7.86/ δ C 133.6) compared
to compounds 1 – 6 , and all expected
carbons were observed in the 13 C NMR spectrum, indicating
a strongly lesser degree of tautomerism. The same styryl and 6-keto-1,3,5-trimethyleptyl
moieties were assembled by COSY and HMBC data, but HMBC correlations
from 4–OH to C-3, C-4, and C-5 and from 6–H to C-2,
C-4, and C-5 connected the α-pyridone. Strong ROESY correlations
between 7–H and 11–H as well as 8–H and 16–H 3 established the 7 S ,8 S ,10 S ,11 S stereochemistry. The crystal
structure of compound 7 was determined
via a continuous rotation 3D electron diffraction (3D ED) experiment
collected on a XtaLAB Synergy-ED diffractometer. The structure was solved with direct methods, and the absolute configuration was determined
in the course of dynamical refinement in JANA. , The absolute configuration of the stereocenters, as well as the
molecular conformation within the crystal structure, is shown in . The experimental
and refinement details as well as the CSD deposition number of the
structure are given in the Supporting Information . Arcopilin G ( 7 ) is nearly the enantiomer of septoriamycin
A, which has been isolated from a culture medium of the ascomycete
fungus Septoria pistaciarum . A total synthesis of septoriamycin A has been
completed by Fotiadou and Zogrofos. Extensive knowledge of
the biosynthesis of tetramic acids and their
related 2-pyridones reveals a common progression catalyzed by polyketide
synthase-nonribosomal peptide synthetase (PKS-NRPS) hybrid machineries.
The diversity and evolution of these biosynthetic pathways are illustrated
in several natural products, including tenellin, aspyridone A, fusarin
C, leporin B, fischerin, PFF1140, sambutoxin, equisetin, etc. , In the late stages of tenellin biosynthesis, two cytochrome P450
oxidases are responsible for catalyzing the oxidative expansion and
N-hydroxylation of pretenellin A. Furthermore,
certain metabolites might undergo cyclization of their side chains
through processes such as inverse-electron demand Diels–Alder
reactions, as seen in the antifungal ilicicolin H, or through a Michael
addition, as observed in the biosynthesis of the mycotoxin -sambutoxin. , Since compounds 1 – 7 share the same
carbon skeleton except for 3 , it is likely that 7 is biosynthesized in a similar fashion as -sambutoxin,
a related PKS-NRPS hybrid product with a longer polyketide chain.
The
antimicrobial activities of compounds 1 – 7 (Acp A–G) were assessed against different bacterial
and fungal strains in addition to their cytotoxic effects on two mammalian
cell lines. The tested microorganisms comprised a diverse array of
clinically relevant pathogens, encompassing sensitive indicator strains.
Among the Gram-positive bacteria were Bacillus subtilis , Staphylococcus aureus , and Mycolicibacterium smegmatis . Gram-negative bacteria
included Acinetobacter baumannii , Chromobacterium violaceum , Escherichia
coli , and Pseudomonas aeruginosa . Additionally, filamentous fungi such as Mucor hiemalis and yeasts including Candida albicans , Wickerhamomyces anomalus , Rhodotorula glutinis , and Schizosaccharomyces
pombe were included. Generally, all compounds presented
similar biological properties, summarized in weak or no activity against
fungal pathogens as well as weak to moderate inhibition of Gram-positive
bacteria .
Acp E and F did not exhibit any antimicrobial activity in our assays. The above suggests that the hydroxylation at C-12
and C-14 in Acp
E has a negative effect on antibacterial activity. Similarly, the
hydroxylation at C-12 and C-13, in addition to the presence of carboxylic
acid at C-14 in Acp F, results in the loss of antibacterial activity.
In terms of their cytotoxic properties, note that 2-pyridone Acp G
was the most cytotoxic metabolite, while its tetramic acid congeners
presented rather weak or no cytotoxic effects as for compounds Acp
B, Acp E, and Acp F. The fact that hydrophilic arcopilins are less
cytotoxic suggests a possible correlation between the hydrophobicity
and the cytotoxicity of these molecules. While PKS-NRPS hybrid
products within the tetramic acid and pyridone
secondary metabolite families exert a wide range of biological activities
and are widespread in ascomycetes, only a limited number of examples
from the Sordariales order have been reported. Notably, the most notorious
examples are the decalin-containing tetramic acids, myceliothermophins,
originally discovered in Thermothelomyces thermophilus (syn. Myceliophthora thermophila ). The potent antitumor activity exhibited by myceliothermophins
C, D, and E against a number of human cancer cell lines has prompted
numerous total synthesis endeavors. , Similarly,
the chaetolivacines A–C, isolated from Chaetomium
olivaceum (Chaetomiaceae), represent another example
of decalin-containing tetramic acids. Only chaetolivacine B exerts moderate antibacterial properties against S. aureus and methicillin-resistant S. aureus (MRSA). Additionally, rare decalin-containing
tetramic acids such as zopfiellamide A and B, as well as zopfielliamides
A–D, have been isolated from Zopfiella latipes and Zopfiella sp., taxa with uncertain taxonomic
placement within this order. , ,
Staphylococcus aureus After identifying
that arcopilins exhibit rather weak activities against the tested
organisms and cell lines, we decided to evaluate their efficacy toward
the disruption of preformed biofilms of the bacterial pathogen S. aureus . Therefore, Acp A–G were evaluated
against preformed biofilms of S. aureus using crystal violet staining. The
2-pyridone, 15-hydroxytenellin (15-Ht), produced by the entomopathogenic
fungus Beauveria neobassiana was also
used for comparison, as the tenellins are model compounds for the
study of fungal secondary metabolite biosynthesis and have displayed
inhibitory properties against the formation of biofilms by S. aureus . , Among the tested metabolites,
Acp A and C showed the most promising disrupting effects toward preformed
biofilms of S. aureus , whereas weak
to moderate effects were observed for Acp B, F, and G . Furthermore, Acp G and 15-Ht (data not shown), both belonging
to the
class
of 2-pyridones, were not active against preformed biofilms of S. aureus . Acp A displayed approximately 50–60%
efficacy toward preformed biofilms within the concentration range
of 15.6 to 250 μg/mL. Similarly, Acp C demonstrated ca. 50%
effectiveness in the dispersal of preformed biofilms between 7.8 μg/mL
and 250 μg/mL. Notably, both compounds exhibited a pronounced
efficacy of 35–45% even at a concentration as low as 3.9 μg/mL.
These results align with the growth curve of Acp A shown in Figure S47 , demonstrating that the growth of S. aureus was inhibited by Acp A treatment even at
concentrations as low as 3.9 and 7.8 μg/mL. Consequently, the
disruption of existing biofilms may be due to the downregulation of
cell growth. However, the precise mechanism behind this effect remains
unclear and it is out of the scope of the present study. In
the case of Acp A and C, a carbonyl group is present on the
side chain of these metabolites. However, the presence of this moiety
is not exclusively necessary for the observed activity, as demonstrated
by Acp D, which lacks a carbonyl group but still exhibits significant
dispersal effects at concentrations as low as 31.3 μg/mL. During
our examination of different tetramic acids and related 2-pyridones,
no discernible link between cytotoxicity and the dispersion of S. aureus preformed biofilms was found. For instance,
Acp G, the most cytotoxic metabolite within the tested congeners,
exhibited only weak disruptive effects on the biofilms. A link between
cytotoxicity and biofilm eradication could affect the applicability
of the metabolites, as increased cytotoxicity might also damage host
cells.
Interestingly, both Acp A and Acp C demonstrated
remarkable effectiveness in disrupting preformed biofilms of S. aureus . Given its promising activity and relatively
low cytotoxicity, we selected Acp A for further experiments. We investigated
its in-depth effects alone and in combination with the antibiotics
gentamicin (GM) and vancomycin (Vac) on planktonic cells and S. aureus biofilms. To evaluate the influence of
Acp A on both biofilm metabolic activity and planktonic cell growth,
XTT and growth curve analyses were conducted, respectively. The results
from XTT assay as depicted in Figure S46 revealed a significant reduction in metabolic activity even at low
concentrations of 3.9 μg/mL. These findings were consistent
with the outcomes of the antibiofilm assay, indicating that effective
concentrations of Acp A in dispersing S. aureus biofilms coincide with an alteration in the metabolic activity of
preformed biofilms. In line with this, inhibitory effects were observed
at concentrations between 7.8 μg/mL and 2 μg/mL according
to the growth curve analysis ( Figure S47 ). After assessing the effects of Acp A on the pathogen S. aureus , we delved deeper into the interaction
of Acp A with established antibiotics (GM and Vac). This exploration
focused on both planktonic cells and preformed biofilms of S. aureus . Consequently, we used a checkerboard assay
to determine the fractional inhibitory concentration index (FICI)
for combinations involving Acp A, GM, or Vac based on both their MIC
values in combination. Antibiotics
commonly used to fight bacterial infections often act
through diverse mechanisms to hinder the growth of these pathogens.
For instance, the well-known antibiotic GM functions as a protein
synthesis inhibitor, while Vac exerts inhibitory effects on this pathogen
by interfering with cell wall biosynthesis. Our findings revealed
that when Acp A (3.9 μg/mL) was used in combination with GM
or Vac, the MIC values of the established antibiotics were significantly
decreased from 15.6 to 0.13 μg/mL and from 2 to 0.065 μg/mL,
respectively. The combination treatment substantially increased the
potency of GM and Vac up to 115-fold and 31-fold, respectively, and
calculation of the FICI showed synergistic effects (FICI < 0.5)
for both combinations ( b). Similarly, the MIC value of Acp A decreased almost 10-fold
when combined with each antibiotic. Additionally, combined effects
were also assessed on the preformed biofilms. Consequently, a colony-forming
unit (CFU) count analysis treated with Acp A (7.8–3.9 μg/mL),
GM (7.8–2 μg/mL), or Vac (15.6–3.9 μg/mL)
alone, as well as their combinations, was carried out for preformed
biofilms. For both GM and Vac, roughly a 3-fold improvement in the
inhibitory effects was observed when used in combination with Acp
A (7.8–3.9 μg/mL) ( c). According to previous studies, tetramic acids with
long polyketide
side chains, such as the reutericyclins, have been shown to act against
bacteria by disrupting their proton gradient and membrane potential. , The cellular membrane potential is dynamic, and it is linked to
signal transmission between cells within biofilms and the overall
level of biofilm formation. In addition, tetramic acids are likely
to act as metal chelators, but the biological implications of this
phenomenon are poorly understood. Similarly,
it has been demonstrated that human-targeted drugs, when used at sublethal
concentrations, can be repurposed as new antimicrobials in combination
therapy. However, the specific mode of
action by which arcopilins disrupt S. aureus biofilms remains unclear and will require future investigation.
In summary, we discovered a new family
of tetramic
acids and related
2-pyridones named arcopilins, adding to the diversity of this class
of natural products. While their antimicrobial properties against
the tested microorganisms were relatively weak, these compounds exerted
varying effectiveness at disrupting preformed biofilms of S. aureus . Among them, arcopilin A ( 1 ) emerged as a particularly promising candidate for an in-depth investigation
of its effects on this bacterial pathogen solely and in combination
with established antibiotics like gentamicin and vancomycin. Notably,
arcopilin A exhibited synergistic effects on both planktonic cells
and preformed biofilms of S. aureus when paired with two antibiotics that operate through different
modes of action. These findings suggest the potential for arcopilin
A to be further developed for potent preclinical applications in combination
therapy.
Fermentation, Extraction,
and Isolation For the evaluation
of the production of secondary metabolites by Arcopilus
navicularis CCF 3252 T , three different
liquid media (YM 6.3: malt extract 10 g/L, yeast extract 4 g/L, d -glucose 4 g/L, pH 6.3 before autoclaving; ZM 1/2: molasses
5 g/L, oatmeal 5 g/L, sucrose 4 g/L, mannitol 4 g/L, d -glucose
1.5 g/L, CaCO 3 1.5 g/L, edamine 0.5 g/L, (NH 4 ) 2 SO 4 0.5 g/L, pH 7.2 before autoclaving; Q6
1/2: d -glucose 2.5 g/L, glycerin 10 g/L, cotton seed flour
5 g/L, pH 7.2 before autoclaving) and one solid medium (BRFT: brown
rice 28 g) as well as 0.1 L of base liquid (yeast extract 1 g/L, disodium
tartrate dihydrate 0.5 g/L, KH 2 PO 4 0.5 g/L)
were used. The fungus was grown in yeast malt agar (YM agar: malt
extract 10 g/L, yeast extract 4 g/L, d -glucose 4 g/L, agar
20 g/L, pH 6.3 before autoclaving) at 23 °C. Later, the colonies
were cut into small pieces using a cork borer (1 cm × 1 cm) and
eight pieces were placed into 500 mL Erlenmeyer flasks containing
200 mL of each liquid medium, which were incubated at 23 °C under
shaking conditions (140 rpm) in the darkness until 3 days after glucose
depletion. For the solid culture, an additional 500 Erlenmeyer flask
containing 200 mL of YM broth was incubated at 23 °C under shaking
conditions (140 rpm) in the darkness. After 7 days, 6 mL of this seed
culture was transferred to an Erlenmeyer flask of 500 mL containing
the BRFT medium. This solid culture was incubated for 15 days at 23
°C in the darkness without agitation. To extract the secondary
metabolites from the liquid cultures, the mycelia were initially separated
from the supernatant through filtration. The supernatant was extracted
with an equal volume of ethyl acetate in a separatory funnel. The
resulting ethyl acetate fraction was evaporated to dryness under vacuum
at 40 °C. Simultaneously, the mycelia, covered in acetone, were
sonicated in an ultrasonic bath for 30 min at 40 °C. The acetone
fraction was separated from the mycelia by filtration throughout a
cellulose filter paper (MN 615 1/4 Ø 185 mm, Macherey Macherey-Nagel,
Düren, Germany). The remaining mycelia underwent another round
of sonication and extraction. Both extracts were combined, and acetone
was evaporated to yield an aqueous residue in vacuo at 40 °C.
This aqueous phase was extracted similarly to the supernatant. For
solid cultures, the mycelia followed the same extraction process as
for the mycelia obtained from liquid cultures until the evaporation
of the ethyl acetate fraction. Subsequently, the ethyl acetate extract
was dissolved in methanol and partitioned with an equal volume of
heptane in a separatory funnel. This step was repeated with the obtained
methanol phase, which was then evaporated to dryness under vacuum
at 40 °C. Both methanol fractions were finally combined and dried
under vacuum at 40 °C. For the scaled-up cultivation, the
fungus was grown in YM agar
at 23 °C. Later, the colonies were cut into small pieces using
a cork borer (1 cm × 1 cm), and eight pieces were placed into
two 500 mL Erlenmeyer flasks each containing 200 mL of YM broth, which
were incubated at 23 °C under shaking conditions (140 rpm) in
the darkness for 7 days. Afterward, 6 mL of this seed culture was
transferred to each of the 40 Erlenmeyer flasks (500 mL) containing
200 mL of Q6 1/2 broth (8 L in total) and incubated at 23 °C
under shaking conditions (140 rpm) in the darkness until 3 days after
glucose depletion. Consequently, the cultures followed the extraction
procedure described above to afford 1845 and 558 mg of supernatant
and mycelial extract, respectively. The supernatant extract
(450 mg × 4) was preseparated using
reverse-phase HPLC (Büchi, Pure C-850, 2020, Switzerland) with
a Gemini C18 (250 mm × 50 mm, 10 μm, Phenomenex, Torrance,
CA) as the stationary phase and the following conditions as the mobile
phase: solvent A: deionized water (H 2 O) + 0.1% formic acid;
solvent B: acetonitrile (MeCN) + 0.1% formic acid; flow: 45 mL/min;
and collected fraction volume: 15 mL. The following gradient elution
was applied: holding in 5% B for 5 min, increasing from 5% B to 60%
B in 60 min and then from 60% B to 100% B in 10 min, and holding in
100% B for 15 min. Five fractions (SF1–SF5) were collected,
from which fraction SF5 was further purified (160 mg × 2) using
reverse-phase HPLC (Büchi, Pure C-850, 2020, Switzerland) with
a Gemini C18 (250 mm × 50 mm, 10 μm, Phenomenex, Torrance,
CA) as the stationary phase and the following conditions as the mobile
phase: solvent A: deionized water (H 2 O) + 0.1% formic acid;
solvent B: acetonitrile (MeCN) + 0.1% formic acid; flow: 40 mL/min;
and collected fraction volume: 15 mL. The following gradient elution
was applied: holding in 5% B for 5 min, increasing from 5% B to 65%
B in 15 min and then from 65% B to 100% B in 60 min, and holding in
100% B for 10 min. This resulted in the isolation of five pure compounds: 6 (3.2 mg, t R = 15 min), 5 (1.46 mg, t R = 32 min), 2 (4.45 mg, t R = 43 min), 1 (4.74 mg, t R = 45 min), and 3 (2.5 mg, t R = 54 min). The mycelial extract (225 mg × 2) was separated using reverse-phase
HPLC (Büchi, Pure C-850, 2020, Switzerland) with a Gemini C18
(250 × 50 mm, 10 μm, Phenomenex, Torrance, CA) as the stationary
phase and the following conditions as the mobile phase: solvent A,
deionized water (H 2 O) + 0.1% formic acid; solvent B, acetonitrile
(MeCN) + 0.1% formic acid; flow: 40 mL/min; and collected fraction
volume: 15 mL. The following gradient elution was applied: holding
in 5% B for 5 min, increasing from 5% B to 80% B in 15 min, and then
from 80% B to 100% B in 40 min. Twelve fractions (MF1−MF12)
were collected, from which MF12 corresponded to compound 4 (2.89 mg, t R = 55 min). The fraction
MF8 (25 mg) was further purified using reverse-phase HPLC (Büchi,
Pure C-850, 2020, Switzerland) with an X-Bridge C18 column (250 mm
× 19 mm, 5 μm, Waters, Milford, MA) as the stationary phase
and the following conditions as the mobile phase: solvent A: deionized
water (H 2 O) + 0.1% formic acid; solvent B: acetonitrile
(MeCN) + 0.1% formic acid; flow: 20 mL/min; and collected fraction
volume: 5 mL. The following gradient elution was applied: increasing
from 5% B to 45% B in 5 min and then from 45% B to 70% B in 40 min
and finally increasing from 70% B to 100% B in 10 min. This afforded
compound 7 (1.41 mg, t R =
29 min). Single-Crystal Structure Determination via 3D Electron Diffraction
of Arcopilin G ( 7 ) Electrons feature very strong
interactions with the electrostatic potential of the atoms. Subsequently,
electron diffraction allows for the performing of experiments with
crystallites in the nanometer range. However, it needs to be considered
that the absorption of the samples is much stronger, and the data
are affected by dynamical diffraction as well as ionic scattering
factors compared to X-ray diffraction. This can lead to seemingly
bad R-values for refinement in the simplistic kinematic approximation. Microcrystalline powder of 7 was spread on a standard
holey carbon-coated copper TEM grid. Colorless plate-like crystallites
with a few 100 nm thickness were selected for 3D ED/microED measurements.
Cryotransfer, i.e., freezing of samples prior to introduction to vacuum,
at −173.15 °C using a Gatan ELSA (Model 698) specimen
holder was applied here. As electron diffraction requires samples
to be studied under a high vacuum, the cryotransfer technique is essential
for many sensitive compounds, such as solvent-containing MOFs or proteins.
Next to stabilization in vacuo, other benefits are improving the resolution,
reducing disorder, and reducing beam damage. Crystallites of 7 suffered from the latter one when measured at ambient temperature,
resulting in no diffraction after some collected frames. The combination
of cryotransfer and measurement under cryogenic conditions prolonged
the lifetime of the grains. Electron diffraction measurements
for 7 were collected
using the Rigaku XtaLAB Synergy-ED, equipped with a Rigaku HyPix-ED
detector optimized for operation in the continuous rotation 3D-ED
experimental setup. Data acquisition
was performed at −173.15 °C under high vacuum with an
electron wavelength of 0.0251 Å (200 kV). The instrument was
operated, and the diffraction data were processed in the program CrysAlisPro. A multiscan absorption correction was performed
using spherical harmonics implemented in the SCALE3 ABSPACK scaling
algorithm in CrysAlisPro. The structure was solved using ShelXT and subsequently refined with kinematical approximation
using ShelXL in the crystallographic
program suite Olex2. , Since we wanted to conduct dynamical
refinement to determine the absolute configuration of 7 , a single dataset with as much completeness as possible was collected
(grain 1, 60.3%) and thus used for refinement, instead of collecting
several datasets followed by data merging for full data completeness.
For initial kinematical refinement, non-hydrogen atoms were assigned
isotropic displacement parameters. The hydrogen atoms bonded to the
oxygen atoms were located from Fourier difference maps. Other hydrogen
atoms were placed in idealized positions and included as riding. Isotropic
displacement parameters for all H atoms were constrained to multiples
of the equivalent displacement parameters of their parent atoms with
U iso (H) = 1.2 U eq (parent atom). The experimental
and refinement details are given below. CSD 2311560 contains the supplementary
crystallographic data for this publication. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif or by emailing [email protected] or by
contacting The Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, U.K.; fax: + 44 1223 336033. Spectral Data Optical rotations were recorded employing
an MCP 150 circular polarimeter (Anton Paar, Seelze, Germany) at 20
°C. UV/vis spectra were recorded with a UV-2450 spectrophotometer
(Shimadzu, Kyoto, Japan). Spectral data were measured in MeOH (Uvasol,
Merck, Darmstadt, Germany) for all compounds. All compounds used in
this study for the in vitro experiments were >95%
pure as confirmed by NMR analysis, which are included in the Supporting Information of the manuscript. The
respective 1D and 2D NMR spectra were recorded with an Avance III
700 spectrometer with a 5 mm TCI cryoprobe ( 1 H NMR: 700
MHz, 13 C: 175 MHz, Bruker, Billerica, MA) and an Avance
III 500 spectrometer ( 1 H NMR: 500 MHz, 13 C:
125 MHz, Bruker, Billerica, MA). The chemical shifts δ were
referenced to the solvent DMSO- d 6 ( 1 H, δ = 2.50; 13 C, δ = 39.51). Arcopilin
A ( 1 ) Brown to orange oil; [α] d 20 −12 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.175), 237.5 (0.158), 313 (0.275);
ESI-MS: m / z 368.20 [M – H] − , 370.20 [M + H] + , and 392.20 [M + Na] + ; HRESI-MS: m / z 370.2025
[M + H] + (calculated for C 22 H 28 NO 4 + : 370.2013 Da). Arcopilin B ( 2 ) Brown to orange oil; [α] d 20 −8 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.178), 238 (0.153), 313 (0.265);
ESI-MS: m / z 370.20 [M – H] − , 372.20 [M + H] + , and 394.20 [M + Na] + ; HRESI-MS: m / z 372.21512
[M + H] + (calculated for C 22 H 30 NO 4 + : 372.2169 Da). Arcopilin C ( 3 ) Brown to orange oil; [α] d 20 −3.5 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201.5 (0.688), 230.5 (0.426), 290
(0.727); ESI-MS: m / z 398.06 [M –
H] − , 400.25 [M + H] + , and 422.24 [M +
Na] + ; HRESI-MS: m / z 400.48531
[M + H] + (calculated for C 23 H 30 NO 5 + : 400.4868 Da). Arcopilin D ( 4 ) Brown to orange oil; [α] d 20 −20 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201 (0.524), 231.5 (0.340), 313.5
(0.650); ESI-MS: m / z 354.23 [M –
H] − , 356.26 [M + H] + , and 378.22 [M +
Na] + ; HRESI-MS: m / z 356.2237
[M + H] + (calculated for C 22 H 30 NO 3 + : 356.2220 Da). Arcopilin E ( 5 ) Brown to orange oil; [α] d 20 −3.4 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.297), 314 (0.357); ESI-MS: m / z 400.05 [M – H] − , 402.22 [M + H] + , and 424.21 [M + Na] + ; HRESI-MS: m / z 402.1921 [M + H] + (calculated
for C 22 H 28 NO 6 + : 402.1911
Da). Arcopilin F ( 6 ) Brown to orange oil; [α] d 20 −1.8 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201 (0.813), 232.5 (0.503), 312 (0.840);
ESI-MS: m / z 416.05 [M – H] − , 418.22 [M + H] + , and 440.20 [M + Na] + ; HRESI-MS: m / z 418.1861
[M + H] + (calculated for C 22 H 28 NO 7 + : 417.1860 Da). Arcopilin G ( 7 ) Orange to white powder;
[α] d 20 −15 (c 0.001, MeOH);
UV (MeOH) λmax (log ε) 207.5 (0.733), 241 (0.702),
304.5 (0.229); ESI-MS: m / z 370.21
[M – H] − , 372.23 [M + H] + , and
394.20 [M + Na] + ; HRESI-MS: m / z 372.2156 [M + H] + (calculated for C 22 H 30 NO 4 + : 372.2169 Da). Crystallographic
Data ( 7 ) Grain 1 only:
CSD 2311560, colorless plate, C 22 H 29 NO 4 , M r = 371.46 gmol –1 , monoclinic, space group I 2 (No. 5), a = 16.7(3) Å, b = 7.07(18) Å, c = 16.83(14) Å, α = 90°, β = 90.47(10)°,
γ = 90°, V = 1989(62) Å3, Z = 4, Z′ = 1, T = −173.15
°C, m (transmission electron microscope) = 0.000,
3761 total reflections, 1070 with I 0 > 2σ(I 0 ), resolution = 0.837 Å, completeness = 60.3%, redundancy
=
3.2, R int = 0.1079, R pim = 0.082, CC 1/2 = 0.990, 2127 data,
101 parameters, 15 restraints, GOF = 1.749, R 1 = 0.2075
and w R 2 = 0.4588 [I 0 > 2σ(I 0 )], R 1 = 0.2762 and w R 2 = 0.4988 (all reflections), 0.152 < dΔρ < −0.121.
Merged grain 1 and 2: C 22 H 29 NO 4 , M r = 371.46 gmol –1 , monoclinic,
space group I 2 (No. 5), a = 16.7(3)
Å, b = 7.07(18) Å, c =
16.83(14) Å, α = 90°, β = 90.47(10)°, γ
= 90°, V = 1989(62) Å3, Z = 4, Z′ = 1, T = −173.15 °C,
m(transmission electron microscope) = 0.000, 8321 total reflections,
1598 with I 0 > 2σ(I 0 ), resolution =
0.837
Å, completeness = 99.9%, redundancy = 4.3, R int =
0.1794, R pim = 0.101, CC 1/2 = 0.989, 3557
data, 236 parameters, 5 restraints, GOF = 1.605, R 1 = 0.2171
and w R 2 = 0.4702 [I 0 > 2σ(I 0 )], R 1 = 0.2910 and w R 2 = 0.5027 (all reflections), 0.226 < dΔρ < −0.188. Derivatization of Arcopilin B ( 2 ) with MTPA Arcopilin B ( 2 ) was dissolved in pyridine- d 5 (50 μL) and transferred into a 250 μL glass
vial, and ( R )-(−)-α-methoxy-α-(trifluoromethyl)
phenylacetyl chloride (4 μL) was added. The mixture was incubated
for 2 h at room temperature before being transferred to an NMR tube
(600 μL) and diluted to a final volume of 350 μL for the
measurement of 1 H, TOCSY, and HSQC NMR spectra. 1 H NMR data (700 MHz, pyridine- d 5 ): similar
to 2 , but δ H 5.17 (m, 13–H),
1.87 (m, 12–H), 1.17 (d, J = 6.3 Hz, 14–H 3 ) and 0.89 (d, J = 6.9 Hz, 17–H 3 ). The ( R )-MTPA ester derivative was obtained analogously with ( S )--α-methoxy-α-(trifluoromethyl) phenylacetyl
chloride (4 μL). 1 H NMR data (700 MHz, pyridine- d 5): similar to 2 , but δ H 5.15
(m, 13–H), 1.83 (m, 12–H), 1.26 (d, J = 6.3 Hz, 14–H 3 ) and 0.81 (d, J = 6.9 Hz, 17–H 3 ). Antimicrobial
and Cytotoxic Assays The antimicrobial
and cytotoxic assays were performed according to the methods reported
previously. Biofilm Assays Cultures of S. aureus DSM 1104 were
prepared by inoculating 1 mL aliquots from a frozen
stock (−20 °C) into 25 mL of CASO medium and incubating
them overnight at 37 °C with shaking at 130 rpm. Preformed Biofilms The crystal violet assay was performed
according to a previously reported procedure. Arcopilins A–G were tested in serial dilutions (250–2
μg/mL), with methanol and microporenic acid A (MAA) as negative
and positive controls, respectively. Statistical differences between
samples and the controls were determined using a two-tailed Student’s t test, with statistical significance defined as p < 0.01. Statistical analysis was carried out using
GraphPad Prism 9 (GraphPad Software, San Diego, CA). XTT Assay The seed culture of S. aureus DSM 1104 was prepared as previously described,
and its OD600 was
adjusted to match the turbidity of a 0.001 McFarland standard. Next,
150 μL of this bacterial solution in CASO with 4% glucose broth
was incubated in 96-well tissue plates (TPP tissue culture ref no.
92196, Switzerland) for 24 h at 150 rpm. After incubation, the supernatant
was discarded and 150 μL of the fresh media was added to the
wells, along with serially diluted arcopilin A (31.3–0.5 μg/mL).
The plate was further incubated for 24 h. Afterward, XTT (Cell profile
XTT kit, Roche, Switzerland) was prepared in phosphate-buffered saline
(PBS) at a final concentration of 0.3 mg/mL. The plate was washed
three times with PBS buffer, and then, 150 μL of the prepared
XTT solution was added to each well. Plates were further incubated
for an additional 4 h at 37 °C while shaking (150 rpm), and absorbance
was measured at 490 nm using a plate reader (Synergy 2, BioTek, Santa
Clara). Methanol (2.5%) was used as the solvent control. Error bars
indicate the standard deviation (SD) of duplicate with two repeats. Growth Curve of S. aureus The seed culture of S. aureus DSM
1104 was adjusted to match the turbidity of a 0.1 McFarland standard
and then cultured at 37 °C and 150 rpm in CASO with 4% glucose
broth. Subsequently, it was added together with arcopilin A to a 96-well
nontissue microtiter plate (TPP nontissue culture refno 92197, Switzerland)
and serially diluted (31.3–2 μg/mL). Absorbance was measured using a plate reader (Synergy 2,
BioTek, Santa Clara) at 530 nm every 90 min. Methanol (2.5%) was used
as the solvent control. Error bars indicate the SD of duplicate with
two repeats. Fractional Inhibitory Concentration Indices
(FICIs) The interaction between arcopilin A, vancomycin,
and gentamicin against S. aureus DSM
1104 was evaluated using a checkerboard
broth dilution method to determine the fractional inhibitory concentration
indices (FICIs), calculated as FIC = MIC of drug A in combination/MIC
of drug A alone + MIC of drug B in combination/MIC of drug B alone. The FICIs were interpreted as synergistic (FICI
≤ 0.5). For this assay, a seed culture was prepared as previously
described to inoculate fresh CASO with 4% glucose broth to match the
turbidity of a 0.1 McFarland standard suspension. Then, 100 μL
of bacterial suspension was distributed in 96-well nontissue microtiter
plates (TPP nontissue culture ref no. 92197, Switzerland). Arcopilin
A (7.8–2 μg/mL) and antibiotics (vancomycin and gentamicin:
31.3–0.016 μg/mL) were added in increasing concentrations
in columns and rows, respectively. The experiments were conducted
in duplicate. Synergistic Effects on Preformed Biofilms The seed
culture of S. aureus DSM 1104 was prepared
as previously described, and the OD600 was adjusted to match the turbidity
of a 0.001 McFarland standard. Then, 150 μL of bacterial solution
in CASO with 4% glucose broth was incubated in 96-well tissue plates
(TPP tissue culture ref no. 92196, Switzerland) for 24 h at 150 rpm.
Afterward, the supernatant was discarded, and 150 μL of the
fresh media was added to the wells, together with serially diluted
arcopilin A (15.6–3.9 μg/mL), vancomycin (15.6–2
μg/mL, Sigma Aldrich), and gentamicin (7.8–2 μg/mL,
Sigma Aldrich) as well as their combinations. Methanol (2.5%) was
used as the solvent control. The plates were incubated for a further
24 h at 37 °C. Colony-forming unit (CFU) count analysis of arcopilin
A, antibiotics (vancomycin and gentamicin), and their combinations
was performed as previously described. Cells were suspended in the well 50 times. Dilution series in 1
to 10 steps (20 μL in 200 μL) were prepared down to a
final dilution level of 10 –6 , and 100 μL of
this last dilution was platted on LB agar plates using 3 mm small
glass beads (5 to 10, Omnilab, Germany) to homogeneously spread the
liquid. Individual colonies on agar plates were counted after incubation
at 30 °C for 24 and 48 h. Afterward,
CFUs were calculated by considering the dilution factors. Error bars
indicate the SD of duplicate with two repeats.
For the evaluation
of the production of secondary metabolites by Arcopilus
navicularis CCF 3252 T , three different
liquid media (YM 6.3: malt extract 10 g/L, yeast extract 4 g/L, d -glucose 4 g/L, pH 6.3 before autoclaving; ZM 1/2: molasses
5 g/L, oatmeal 5 g/L, sucrose 4 g/L, mannitol 4 g/L, d -glucose
1.5 g/L, CaCO 3 1.5 g/L, edamine 0.5 g/L, (NH 4 ) 2 SO 4 0.5 g/L, pH 7.2 before autoclaving; Q6
1/2: d -glucose 2.5 g/L, glycerin 10 g/L, cotton seed flour
5 g/L, pH 7.2 before autoclaving) and one solid medium (BRFT: brown
rice 28 g) as well as 0.1 L of base liquid (yeast extract 1 g/L, disodium
tartrate dihydrate 0.5 g/L, KH 2 PO 4 0.5 g/L)
were used. The fungus was grown in yeast malt agar (YM agar: malt
extract 10 g/L, yeast extract 4 g/L, d -glucose 4 g/L, agar
20 g/L, pH 6.3 before autoclaving) at 23 °C. Later, the colonies
were cut into small pieces using a cork borer (1 cm × 1 cm) and
eight pieces were placed into 500 mL Erlenmeyer flasks containing
200 mL of each liquid medium, which were incubated at 23 °C under
shaking conditions (140 rpm) in the darkness until 3 days after glucose
depletion. For the solid culture, an additional 500 Erlenmeyer flask
containing 200 mL of YM broth was incubated at 23 °C under shaking
conditions (140 rpm) in the darkness. After 7 days, 6 mL of this seed
culture was transferred to an Erlenmeyer flask of 500 mL containing
the BRFT medium. This solid culture was incubated for 15 days at 23
°C in the darkness without agitation. To extract the secondary
metabolites from the liquid cultures, the mycelia were initially separated
from the supernatant through filtration. The supernatant was extracted
with an equal volume of ethyl acetate in a separatory funnel. The
resulting ethyl acetate fraction was evaporated to dryness under vacuum
at 40 °C. Simultaneously, the mycelia, covered in acetone, were
sonicated in an ultrasonic bath for 30 min at 40 °C. The acetone
fraction was separated from the mycelia by filtration throughout a
cellulose filter paper (MN 615 1/4 Ø 185 mm, Macherey Macherey-Nagel,
Düren, Germany). The remaining mycelia underwent another round
of sonication and extraction. Both extracts were combined, and acetone
was evaporated to yield an aqueous residue in vacuo at 40 °C.
This aqueous phase was extracted similarly to the supernatant. For
solid cultures, the mycelia followed the same extraction process as
for the mycelia obtained from liquid cultures until the evaporation
of the ethyl acetate fraction. Subsequently, the ethyl acetate extract
was dissolved in methanol and partitioned with an equal volume of
heptane in a separatory funnel. This step was repeated with the obtained
methanol phase, which was then evaporated to dryness under vacuum
at 40 °C. Both methanol fractions were finally combined and dried
under vacuum at 40 °C. For the scaled-up cultivation, the
fungus was grown in YM agar
at 23 °C. Later, the colonies were cut into small pieces using
a cork borer (1 cm × 1 cm), and eight pieces were placed into
two 500 mL Erlenmeyer flasks each containing 200 mL of YM broth, which
were incubated at 23 °C under shaking conditions (140 rpm) in
the darkness for 7 days. Afterward, 6 mL of this seed culture was
transferred to each of the 40 Erlenmeyer flasks (500 mL) containing
200 mL of Q6 1/2 broth (8 L in total) and incubated at 23 °C
under shaking conditions (140 rpm) in the darkness until 3 days after
glucose depletion. Consequently, the cultures followed the extraction
procedure described above to afford 1845 and 558 mg of supernatant
and mycelial extract, respectively. The supernatant extract
(450 mg × 4) was preseparated using
reverse-phase HPLC (Büchi, Pure C-850, 2020, Switzerland) with
a Gemini C18 (250 mm × 50 mm, 10 μm, Phenomenex, Torrance,
CA) as the stationary phase and the following conditions as the mobile
phase: solvent A: deionized water (H 2 O) + 0.1% formic acid;
solvent B: acetonitrile (MeCN) + 0.1% formic acid; flow: 45 mL/min;
and collected fraction volume: 15 mL. The following gradient elution
was applied: holding in 5% B for 5 min, increasing from 5% B to 60%
B in 60 min and then from 60% B to 100% B in 10 min, and holding in
100% B for 15 min. Five fractions (SF1–SF5) were collected,
from which fraction SF5 was further purified (160 mg × 2) using
reverse-phase HPLC (Büchi, Pure C-850, 2020, Switzerland) with
a Gemini C18 (250 mm × 50 mm, 10 μm, Phenomenex, Torrance,
CA) as the stationary phase and the following conditions as the mobile
phase: solvent A: deionized water (H 2 O) + 0.1% formic acid;
solvent B: acetonitrile (MeCN) + 0.1% formic acid; flow: 40 mL/min;
and collected fraction volume: 15 mL. The following gradient elution
was applied: holding in 5% B for 5 min, increasing from 5% B to 65%
B in 15 min and then from 65% B to 100% B in 60 min, and holding in
100% B for 10 min. This resulted in the isolation of five pure compounds: 6 (3.2 mg, t R = 15 min), 5 (1.46 mg, t R = 32 min), 2 (4.45 mg, t R = 43 min), 1 (4.74 mg, t R = 45 min), and 3 (2.5 mg, t R = 54 min). The mycelial extract (225 mg × 2) was separated using reverse-phase
HPLC (Büchi, Pure C-850, 2020, Switzerland) with a Gemini C18
(250 × 50 mm, 10 μm, Phenomenex, Torrance, CA) as the stationary
phase and the following conditions as the mobile phase: solvent A,
deionized water (H 2 O) + 0.1% formic acid; solvent B, acetonitrile
(MeCN) + 0.1% formic acid; flow: 40 mL/min; and collected fraction
volume: 15 mL. The following gradient elution was applied: holding
in 5% B for 5 min, increasing from 5% B to 80% B in 15 min, and then
from 80% B to 100% B in 40 min. Twelve fractions (MF1−MF12)
were collected, from which MF12 corresponded to compound 4 (2.89 mg, t R = 55 min). The fraction
MF8 (25 mg) was further purified using reverse-phase HPLC (Büchi,
Pure C-850, 2020, Switzerland) with an X-Bridge C18 column (250 mm
× 19 mm, 5 μm, Waters, Milford, MA) as the stationary phase
and the following conditions as the mobile phase: solvent A: deionized
water (H 2 O) + 0.1% formic acid; solvent B: acetonitrile
(MeCN) + 0.1% formic acid; flow: 20 mL/min; and collected fraction
volume: 5 mL. The following gradient elution was applied: increasing
from 5% B to 45% B in 5 min and then from 45% B to 70% B in 40 min
and finally increasing from 70% B to 100% B in 10 min. This afforded
compound 7 (1.41 mg, t R =
29 min).
7 ) Electrons feature very strong
interactions with the electrostatic potential of the atoms. Subsequently,
electron diffraction allows for the performing of experiments with
crystallites in the nanometer range. However, it needs to be considered
that the absorption of the samples is much stronger, and the data
are affected by dynamical diffraction as well as ionic scattering
factors compared to X-ray diffraction. This can lead to seemingly
bad R-values for refinement in the simplistic kinematic approximation. Microcrystalline powder of 7 was spread on a standard
holey carbon-coated copper TEM grid. Colorless plate-like crystallites
with a few 100 nm thickness were selected for 3D ED/microED measurements.
Cryotransfer, i.e., freezing of samples prior to introduction to vacuum,
at −173.15 °C using a Gatan ELSA (Model 698) specimen
holder was applied here. As electron diffraction requires samples
to be studied under a high vacuum, the cryotransfer technique is essential
for many sensitive compounds, such as solvent-containing MOFs or proteins.
Next to stabilization in vacuo, other benefits are improving the resolution,
reducing disorder, and reducing beam damage. Crystallites of 7 suffered from the latter one when measured at ambient temperature,
resulting in no diffraction after some collected frames. The combination
of cryotransfer and measurement under cryogenic conditions prolonged
the lifetime of the grains. Electron diffraction measurements
for 7 were collected
using the Rigaku XtaLAB Synergy-ED, equipped with a Rigaku HyPix-ED
detector optimized for operation in the continuous rotation 3D-ED
experimental setup. Data acquisition
was performed at −173.15 °C under high vacuum with an
electron wavelength of 0.0251 Å (200 kV). The instrument was
operated, and the diffraction data were processed in the program CrysAlisPro. A multiscan absorption correction was performed
using spherical harmonics implemented in the SCALE3 ABSPACK scaling
algorithm in CrysAlisPro. The structure was solved using ShelXT and subsequently refined with kinematical approximation
using ShelXL in the crystallographic
program suite Olex2. , Since we wanted to conduct dynamical
refinement to determine the absolute configuration of 7 , a single dataset with as much completeness as possible was collected
(grain 1, 60.3%) and thus used for refinement, instead of collecting
several datasets followed by data merging for full data completeness.
For initial kinematical refinement, non-hydrogen atoms were assigned
isotropic displacement parameters. The hydrogen atoms bonded to the
oxygen atoms were located from Fourier difference maps. Other hydrogen
atoms were placed in idealized positions and included as riding. Isotropic
displacement parameters for all H atoms were constrained to multiples
of the equivalent displacement parameters of their parent atoms with
U iso (H) = 1.2 U eq (parent atom). The experimental
and refinement details are given below. CSD 2311560 contains the supplementary
crystallographic data for this publication. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/data_request/cif or by emailing [email protected] or by
contacting The Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, U.K.; fax: + 44 1223 336033.
Optical rotations were recorded employing
an MCP 150 circular polarimeter (Anton Paar, Seelze, Germany) at 20
°C. UV/vis spectra were recorded with a UV-2450 spectrophotometer
(Shimadzu, Kyoto, Japan). Spectral data were measured in MeOH (Uvasol,
Merck, Darmstadt, Germany) for all compounds. All compounds used in
this study for the in vitro experiments were >95%
pure as confirmed by NMR analysis, which are included in the Supporting Information of the manuscript. The
respective 1D and 2D NMR spectra were recorded with an Avance III
700 spectrometer with a 5 mm TCI cryoprobe ( 1 H NMR: 700
MHz, 13 C: 175 MHz, Bruker, Billerica, MA) and an Avance
III 500 spectrometer ( 1 H NMR: 500 MHz, 13 C:
125 MHz, Bruker, Billerica, MA). The chemical shifts δ were
referenced to the solvent DMSO- d 6 ( 1 H, δ = 2.50; 13 C, δ = 39.51). Arcopilin
A ( 1 ) Brown to orange oil; [α] d 20 −12 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.175), 237.5 (0.158), 313 (0.275);
ESI-MS: m / z 368.20 [M – H] − , 370.20 [M + H] + , and 392.20 [M + Na] + ; HRESI-MS: m / z 370.2025
[M + H] + (calculated for C 22 H 28 NO 4 + : 370.2013 Da). Arcopilin B ( 2 ) Brown to orange oil; [α] d 20 −8 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.178), 238 (0.153), 313 (0.265);
ESI-MS: m / z 370.20 [M – H] − , 372.20 [M + H] + , and 394.20 [M + Na] + ; HRESI-MS: m / z 372.21512
[M + H] + (calculated for C 22 H 30 NO 4 + : 372.2169 Da). Arcopilin C ( 3 ) Brown to orange oil; [α] d 20 −3.5 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201.5 (0.688), 230.5 (0.426), 290
(0.727); ESI-MS: m / z 398.06 [M –
H] − , 400.25 [M + H] + , and 422.24 [M +
Na] + ; HRESI-MS: m / z 400.48531
[M + H] + (calculated for C 23 H 30 NO 5 + : 400.4868 Da). Arcopilin D ( 4 ) Brown to orange oil; [α] d 20 −20 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201 (0.524), 231.5 (0.340), 313.5
(0.650); ESI-MS: m / z 354.23 [M –
H] − , 356.26 [M + H] + , and 378.22 [M +
Na] + ; HRESI-MS: m / z 356.2237
[M + H] + (calculated for C 22 H 30 NO 3 + : 356.2220 Da). Arcopilin E ( 5 ) Brown to orange oil; [α] d 20 −3.4 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.297), 314 (0.357); ESI-MS: m / z 400.05 [M – H] − , 402.22 [M + H] + , and 424.21 [M + Na] + ; HRESI-MS: m / z 402.1921 [M + H] + (calculated
for C 22 H 28 NO 6 + : 402.1911
Da). Arcopilin F ( 6 ) Brown to orange oil; [α] d 20 −1.8 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201 (0.813), 232.5 (0.503), 312 (0.840);
ESI-MS: m / z 416.05 [M – H] − , 418.22 [M + H] + , and 440.20 [M + Na] + ; HRESI-MS: m / z 418.1861
[M + H] + (calculated for C 22 H 28 NO 7 + : 417.1860 Da). Arcopilin G ( 7 ) Orange to white powder;
[α] d 20 −15 (c 0.001, MeOH);
UV (MeOH) λmax (log ε) 207.5 (0.733), 241 (0.702),
304.5 (0.229); ESI-MS: m / z 370.21
[M – H] − , 372.23 [M + H] + , and
394.20 [M + Na] + ; HRESI-MS: m / z 372.2156 [M + H] + (calculated for C 22 H 30 NO 4 + : 372.2169 Da). Crystallographic
Data ( 7 ) Grain 1 only:
CSD 2311560, colorless plate, C 22 H 29 NO 4 , M r = 371.46 gmol –1 , monoclinic, space group I 2 (No. 5), a = 16.7(3) Å, b = 7.07(18) Å, c = 16.83(14) Å, α = 90°, β = 90.47(10)°,
γ = 90°, V = 1989(62) Å3, Z = 4, Z′ = 1, T = −173.15
°C, m (transmission electron microscope) = 0.000,
3761 total reflections, 1070 with I 0 > 2σ(I 0 ), resolution = 0.837 Å, completeness = 60.3%, redundancy
=
3.2, R int = 0.1079, R pim = 0.082, CC 1/2 = 0.990, 2127 data,
101 parameters, 15 restraints, GOF = 1.749, R 1 = 0.2075
and w R 2 = 0.4588 [I 0 > 2σ(I 0 )], R 1 = 0.2762 and w R 2 = 0.4988 (all reflections), 0.152 < dΔρ < −0.121.
Merged grain 1 and 2: C 22 H 29 NO 4 , M r = 371.46 gmol –1 , monoclinic,
space group I 2 (No. 5), a = 16.7(3)
Å, b = 7.07(18) Å, c =
16.83(14) Å, α = 90°, β = 90.47(10)°, γ
= 90°, V = 1989(62) Å3, Z = 4, Z′ = 1, T = −173.15 °C,
m(transmission electron microscope) = 0.000, 8321 total reflections,
1598 with I 0 > 2σ(I 0 ), resolution =
0.837
Å, completeness = 99.9%, redundancy = 4.3, R int =
0.1794, R pim = 0.101, CC 1/2 = 0.989, 3557
data, 236 parameters, 5 restraints, GOF = 1.605, R 1 = 0.2171
and w R 2 = 0.4702 [I 0 > 2σ(I 0 )], R 1 = 0.2910 and w R 2 = 0.5027 (all reflections), 0.226 < dΔρ < −0.188.
1 ) Brown to orange oil; [α] d 20 −12 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.175), 237.5 (0.158), 313 (0.275);
ESI-MS: m / z 368.20 [M – H] − , 370.20 [M + H] + , and 392.20 [M + Na] + ; HRESI-MS: m / z 370.2025
[M + H] + (calculated for C 22 H 28 NO 4 + : 370.2013 Da).
2 ) Brown to orange oil; [α] d 20 −8 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.178), 238 (0.153), 313 (0.265);
ESI-MS: m / z 370.20 [M – H] − , 372.20 [M + H] + , and 394.20 [M + Na] + ; HRESI-MS: m / z 372.21512
[M + H] + (calculated for C 22 H 30 NO 4 + : 372.2169 Da).
3 ) Brown to orange oil; [α] d 20 −3.5 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201.5 (0.688), 230.5 (0.426), 290
(0.727); ESI-MS: m / z 398.06 [M –
H] − , 400.25 [M + H] + , and 422.24 [M +
Na] + ; HRESI-MS: m / z 400.48531
[M + H] + (calculated for C 23 H 30 NO 5 + : 400.4868 Da).
4 ) Brown to orange oil; [α] d 20 −20 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201 (0.524), 231.5 (0.340), 313.5
(0.650); ESI-MS: m / z 354.23 [M –
H] − , 356.26 [M + H] + , and 378.22 [M +
Na] + ; HRESI-MS: m / z 356.2237
[M + H] + (calculated for C 22 H 30 NO 3 + : 356.2220 Da).
5 ) Brown to orange oil; [α] d 20 −3.4 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 203 (0.297), 314 (0.357); ESI-MS: m / z 400.05 [M – H] − , 402.22 [M + H] + , and 424.21 [M + Na] + ; HRESI-MS: m / z 402.1921 [M + H] + (calculated
for C 22 H 28 NO 6 + : 402.1911
Da).
6 ) Brown to orange oil; [α] d 20 −1.8 (c 0.001, MeOH); UV (MeOH)
λmax (log ε) 201 (0.813), 232.5 (0.503), 312 (0.840);
ESI-MS: m / z 416.05 [M – H] − , 418.22 [M + H] + , and 440.20 [M + Na] + ; HRESI-MS: m / z 418.1861
[M + H] + (calculated for C 22 H 28 NO 7 + : 417.1860 Da).
7 ) Orange to white powder;
[α] d 20 −15 (c 0.001, MeOH);
UV (MeOH) λmax (log ε) 207.5 (0.733), 241 (0.702),
304.5 (0.229); ESI-MS: m / z 370.21
[M – H] − , 372.23 [M + H] + , and
394.20 [M + Na] + ; HRESI-MS: m / z 372.2156 [M + H] + (calculated for C 22 H 30 NO 4 + : 372.2169 Da).
7 ) Grain 1 only:
CSD 2311560, colorless plate, C 22 H 29 NO 4 , M r = 371.46 gmol –1 , monoclinic, space group I 2 (No. 5), a = 16.7(3) Å, b = 7.07(18) Å, c = 16.83(14) Å, α = 90°, β = 90.47(10)°,
γ = 90°, V = 1989(62) Å3, Z = 4, Z′ = 1, T = −173.15
°C, m (transmission electron microscope) = 0.000,
3761 total reflections, 1070 with I 0 > 2σ(I 0 ), resolution = 0.837 Å, completeness = 60.3%, redundancy
=
3.2, R int = 0.1079, R pim = 0.082, CC 1/2 = 0.990, 2127 data,
101 parameters, 15 restraints, GOF = 1.749, R 1 = 0.2075
and w R 2 = 0.4588 [I 0 > 2σ(I 0 )], R 1 = 0.2762 and w R 2 = 0.4988 (all reflections), 0.152 < dΔρ < −0.121.
Merged grain 1 and 2: C 22 H 29 NO 4 , M r = 371.46 gmol –1 , monoclinic,
space group I 2 (No. 5), a = 16.7(3)
Å, b = 7.07(18) Å, c =
16.83(14) Å, α = 90°, β = 90.47(10)°, γ
= 90°, V = 1989(62) Å3, Z = 4, Z′ = 1, T = −173.15 °C,
m(transmission electron microscope) = 0.000, 8321 total reflections,
1598 with I 0 > 2σ(I 0 ), resolution =
0.837
Å, completeness = 99.9%, redundancy = 4.3, R int =
0.1794, R pim = 0.101, CC 1/2 = 0.989, 3557
data, 236 parameters, 5 restraints, GOF = 1.605, R 1 = 0.2171
and w R 2 = 0.4702 [I 0 > 2σ(I 0 )], R 1 = 0.2910 and w R 2 = 0.5027 (all reflections), 0.226 < dΔρ < −0.188.
2 ) with MTPA Arcopilin B ( 2 ) was dissolved in pyridine- d 5 (50 μL) and transferred into a 250 μL glass
vial, and ( R )-(−)-α-methoxy-α-(trifluoromethyl)
phenylacetyl chloride (4 μL) was added. The mixture was incubated
for 2 h at room temperature before being transferred to an NMR tube
(600 μL) and diluted to a final volume of 350 μL for the
measurement of 1 H, TOCSY, and HSQC NMR spectra. 1 H NMR data (700 MHz, pyridine- d 5 ): similar
to 2 , but δ H 5.17 (m, 13–H),
1.87 (m, 12–H), 1.17 (d, J = 6.3 Hz, 14–H 3 ) and 0.89 (d, J = 6.9 Hz, 17–H 3 ). The ( R )-MTPA ester derivative was obtained analogously with ( S )--α-methoxy-α-(trifluoromethyl) phenylacetyl
chloride (4 μL). 1 H NMR data (700 MHz, pyridine- d 5): similar to 2 , but δ H 5.15
(m, 13–H), 1.83 (m, 12–H), 1.26 (d, J = 6.3 Hz, 14–H 3 ) and 0.81 (d, J = 6.9 Hz, 17–H 3 ).
The antimicrobial
and cytotoxic assays were performed according to the methods reported
previously.
Cultures of S. aureus DSM 1104 were
prepared by inoculating 1 mL aliquots from a frozen
stock (−20 °C) into 25 mL of CASO medium and incubating
them overnight at 37 °C with shaking at 130 rpm.
The crystal violet assay was performed
according to a previously reported procedure. Arcopilins A–G were tested in serial dilutions (250–2
μg/mL), with methanol and microporenic acid A (MAA) as negative
and positive controls, respectively. Statistical differences between
samples and the controls were determined using a two-tailed Student’s t test, with statistical significance defined as p < 0.01. Statistical analysis was carried out using
GraphPad Prism 9 (GraphPad Software, San Diego, CA).
The seed culture of S. aureus DSM 1104 was prepared as previously described,
and its OD600 was
adjusted to match the turbidity of a 0.001 McFarland standard. Next,
150 μL of this bacterial solution in CASO with 4% glucose broth
was incubated in 96-well tissue plates (TPP tissue culture ref no.
92196, Switzerland) for 24 h at 150 rpm. After incubation, the supernatant
was discarded and 150 μL of the fresh media was added to the
wells, along with serially diluted arcopilin A (31.3–0.5 μg/mL).
The plate was further incubated for 24 h. Afterward, XTT (Cell profile
XTT kit, Roche, Switzerland) was prepared in phosphate-buffered saline
(PBS) at a final concentration of 0.3 mg/mL. The plate was washed
three times with PBS buffer, and then, 150 μL of the prepared
XTT solution was added to each well. Plates were further incubated
for an additional 4 h at 37 °C while shaking (150 rpm), and absorbance
was measured at 490 nm using a plate reader (Synergy 2, BioTek, Santa
Clara). Methanol (2.5%) was used as the solvent control. Error bars
indicate the standard deviation (SD) of duplicate with two repeats.
S. aureus The seed culture of S. aureus DSM
1104 was adjusted to match the turbidity of a 0.1 McFarland standard
and then cultured at 37 °C and 150 rpm in CASO with 4% glucose
broth. Subsequently, it was added together with arcopilin A to a 96-well
nontissue microtiter plate (TPP nontissue culture refno 92197, Switzerland)
and serially diluted (31.3–2 μg/mL). Absorbance was measured using a plate reader (Synergy 2,
BioTek, Santa Clara) at 530 nm every 90 min. Methanol (2.5%) was used
as the solvent control. Error bars indicate the SD of duplicate with
two repeats.
The interaction between arcopilin A, vancomycin,
and gentamicin against S. aureus DSM
1104 was evaluated using a checkerboard
broth dilution method to determine the fractional inhibitory concentration
indices (FICIs), calculated as FIC = MIC of drug A in combination/MIC
of drug A alone + MIC of drug B in combination/MIC of drug B alone. The FICIs were interpreted as synergistic (FICI
≤ 0.5). For this assay, a seed culture was prepared as previously
described to inoculate fresh CASO with 4% glucose broth to match the
turbidity of a 0.1 McFarland standard suspension. Then, 100 μL
of bacterial suspension was distributed in 96-well nontissue microtiter
plates (TPP nontissue culture ref no. 92197, Switzerland). Arcopilin
A (7.8–2 μg/mL) and antibiotics (vancomycin and gentamicin:
31.3–0.016 μg/mL) were added in increasing concentrations
in columns and rows, respectively. The experiments were conducted
in duplicate.
The seed
culture of S. aureus DSM 1104 was prepared
as previously described, and the OD600 was adjusted to match the turbidity
of a 0.001 McFarland standard. Then, 150 μL of bacterial solution
in CASO with 4% glucose broth was incubated in 96-well tissue plates
(TPP tissue culture ref no. 92196, Switzerland) for 24 h at 150 rpm.
Afterward, the supernatant was discarded, and 150 μL of the
fresh media was added to the wells, together with serially diluted
arcopilin A (15.6–3.9 μg/mL), vancomycin (15.6–2
μg/mL, Sigma Aldrich), and gentamicin (7.8–2 μg/mL,
Sigma Aldrich) as well as their combinations. Methanol (2.5%) was
used as the solvent control. The plates were incubated for a further
24 h at 37 °C. Colony-forming unit (CFU) count analysis of arcopilin
A, antibiotics (vancomycin and gentamicin), and their combinations
was performed as previously described. Cells were suspended in the well 50 times. Dilution series in 1
to 10 steps (20 μL in 200 μL) were prepared down to a
final dilution level of 10 –6 , and 100 μL of
this last dilution was platted on LB agar plates using 3 mm small
glass beads (5 to 10, Omnilab, Germany) to homogeneously spread the
liquid. Individual colonies on agar plates were counted after incubation
at 30 °C for 24 and 48 h. Afterward,
CFUs were calculated by considering the dilution factors. Error bars
indicate the SD of duplicate with two repeats.
|
Cross-cultural adaptation of the Quebecois Patient-Centered Coordination by a Care Team Questionnaire for use in France | 5fe3bcc3-b38f-4a8c-9258-d1b87aba472f | 11471038 | Patient-Centered Care[mh] | Prevalence of chronic diseases is increasing due to the aging population, medical advances, and certain socio-environmental factors . Multimorbidity, defined as the co-existence of several diseases in the same individual, affects nearly 65% of people over 65 years of age . The burden for those living with chronic diseases is multidimensional and involves physical, mental, occupational, financial, social, and familial impacts. It is particularly heavy for people with multimorbidity and those close to them , and this is worsened by the burden of treatment . Multidisciplinary care is required to meet the health needs of people living with chronic diseases and multimorbidity . This care is more effective if coordination between the different healthcare professionals is optimal . Good care coordination is associated with improved symptoms, treatment adherence and patient satisfaction, reduced numbers of hospitalizations, better control of chronic diseases (such as glycated hemoglobin), and reduced mortality . In addition, patient-centered care is particularly relevant in improving the health of people living with chronic diseases . This approach requires a relationship of trust, collaboration between the patient and clinician(s) and shared decision-making to ensure their health needs are met . The benefits of patient-centered care include increased patient satisfaction, greater autonomy to manage their health, improved quality of care, reduced risk of complications from chronic diseases, reduced rehospitalization and improved quality of life and well-being . Positive effects have also been observed for those close to the patient and healthcare professionals . Several countries have initiated reforms aimed at strengthening and structuring primary care to ensure that it is more coordinated and patient-centered [ – ]. This most often involves supporting the creation of formalized multidisciplinary teams , for example with multi-professional health centers in France, or family medicine groups and patient medical homes in Canada [ – ]. Until recently, there was no valid tool to measure the quality of coordinated patient-centered care provided by a multidisciplinary team from the patient's perspective. As part of the Canadian research program , PACEinMM (Patient-Centered Innovations for Persons with Multimorbidity Study), aimed at evaluating clinical innovations for multimorbid people, the 2-part Patient-Centered Coordination by a Care Team (PCCCT) questionnaire, was developed. The first part is a checklist of healthcare professionals involved in the patient’s care. The second part contains 14 items assessing the patient’s opinion about how patient-centered and coordinated their care is. Each item is rated from zero to three points, which generates a total score ranging from 0 (imperceptible coordination) to 42 (optimal coordination). The PCCCT questionnaire was initially developed in English language (Fig. A) and in French language for Quebec (Fig. B). The English version has not been evaluated in terms of psychometric properties to date. Evaluations of the French version for Quebec revealed that the questionnaire has satisfactory validity and reliability . The original version, designed and evaluated in Quebec, is not directly transferable to France due to linguistic, cultural and health system organizational differences between the two countries. These differences can make some items irrelevant or change their meaning and affect how people answer a given question. This study aims to perform cross-cultural adaptation of the Quebecois PCCCT questionnaire to obtain a new version adapted for use in France ensuring item and semantic equivalences between the two versions.
The cross cultural method used to adapt the PCCCT questionnaire in this study was based on guidelines for the adaption of self-reported measures . During cross-cultural adaptation, equivalence is based on five dimensions: conceptual, item, semantic, operational and measurement equivalence . Due to linguistic, cultural and health system organizational differences, item and semantic equivalences are unlikely. This study therefore focuses on these two equivalences. Conceptual and operational equivalence already exist since coordinated patient-centered care has the same relevance in Quebec and France and the similar literacy levels mean the questionnaire can be used in the same way in both settings. Measurement equivalence will be addressed in a future study. The cross-cultural adaption process took place between September 2020 and February 2021. It consisted of two stages: the Delphi consensus, carried out between September 2020 and October 2020, and the cognitive interviews, between November 2020 and February 2021. Both stages were supervised by a single scientific committee made up of five healthcare professionals with academic activity, two of whom had a dual French-Quebecois perspective (Table ). Stage 1: Delphi consensus The first stage was a Delphi consensus involving a multidisciplinary health professional panel to evaluate and harmonize the clarity and appropriateness of the questionnaire for patients using the French health care system. Purposive sampling was used to recruit healthcare professionals with experience in coordinated care from diverse professions, practice structures, and cultural backgrounds. They were invited by e-mail to participate and a detailed information letter on the project was sent. All the professionals contacted accepted and participated in the Delphi rounds. The expert panel consisted of ten healthcare professionals, one of whom was from Quebec and was involved in creating the original questionnaire (Table ). Firstly, the scientific committee proposed an initial adaptation of the 14 items from the original Quebecois version (V0), called version V1. The list of the healthcare professionals visited by patients, which precedes the 14 items, was also modified by the scientific committee: some occupational categories, such as “nurse practitioner” or “kinesiologist” which are not present in France, were removed. All the questionnaire adaptations made by the scientific committee (related to items or list of professionals) were carried out by oral consensus, via live synchronous videoconferences. During the first Delphi round, versions V0 and V1 were submitted to the group of experts using an electronic LimeSurvey® form. Each expert was contacted individually by e-mail asking for a response within one week. For each item, the expert was asked: “Do you think this new formulation is sufficiently clear and adapted to a patient using the French health system?”. The expert was reminded of the need to maintain the original meaning of the item. Each expert indicated their degree of agreement with the reformulation by assigning a score between 1 (completely unsuitable) and 9 (completely suitable). Each score below 7 had to be justified with a comment. At the end of the first round, the scientific committee met to examine all the items with an average score below 8 and/or with less than 70% of scores greater than or equal to 7. No specific tools or programs have been used to analyze scores, except for the Excel tool. For each item not reaching consensus (as defined just above), the scientific committee used the experts’ comments to reformulate the item creating version V2. Only those items that had not yet achieved consensus, were submitted, in versions V0 and V2, to the group of experts in the second Delphi round. In order to not influence their views and decisions, experts were not provided with other experts' comments from the first Delphi round. Additional rounds would be performed until consensus was reached for all items. Stage 2: Cognitive interviews During the second stage, patients assessed the comprehensibility and conformity of the adapted version of the questionnaire resulting from stage 1 and improved it if necessary. The checklist of health professionals was also submitted to the patients. This was achieved using cognitive interviews. Two members of the scientific committee (SL and AP) recruited patients, first from a rural general practice, and then from an urban community pharmacy, as complementary support. Adult patients who could speak French, had one or more chronic illnesses (physical and/or mental) and benefited from regular monitoring were eligible for inclusion. For each patient, sociodemographic data were collected (gender, age, profession and socio-professional category, health condition). Purposive maximum variation sampling was used to obtain a diverse patient panel including gender, age, socio-professional category (proxy for literacy level) and chronic diseases. Eligible patients were invited to participate in the study at the end of a medical consultation or a visit to the pharmacy. Potential participants were provided with an information letter explaining the nature of the study. Those who accepted provided written consent. Two patients refused to participate. Two accepted but one was excluded before data collection due to cognitive disorders preventing them from being able to read and sadly, the other died unexpectedly before the interview. One female general practitioner (SL) conducted the cognitive interviews. She had no previous interview experience and received training from the senior researcher who supervised her (ARR). Each patient was interviewed individually. The objective was to systematically identify any discrepancies between the researchers’ and the patient’s understanding of the item and to improve the formulation if necessary . The patients were provided with the adapted version of the questionnaire resulting from stage 1 and the checklist of healthcare professionals involved in their care at recruitment. They were instructed to only read it during the interview to ensure the patient's spontaneous reactions were collected. Patients were asked to read each item aloud, rephrase it in their own words, and point out any ambiguity or misunderstanding. They then answered the item, thinking aloud and giving reasons for their response. If discrepancy or ambiguity was identified, the patient was invited to suggest a possible reformulation following an explanation from the researcher. The interviews took place by telephone due to the contact restrictions imposed in France during the COVID-19 epidemic. Each interview was audio recorded, transcribed item by item, then compiled interview by interview. This enabled discrepancies, ambiguities, and suggested reformulations for each item to be extracted and reconciled. Any items that patients felt were not applicable to them were also listed. Audio files were archived only on the password-protected computer of the interviewer (SL), and deleted once transcription and de-identification of the excerpts processes were completed. One researcher (SL) analyzed the cognitive interviews under the supervision of a senior researcher (ARR). The analysis process focused on the patients' reaction to the question, their ability to rephrase it, their answer to the question, and the justification they provided for their response. Based on these points, the researcher assessed whether the item was completely understood and if the understanding was adequate in regard to its original meaning. The outcome of her interpretation was noted, as well as patient’s comments, in a de-identified report, before submitting it to the scientific committee. The scientific committee met every 3 to 4 interviews to validate the results and modify the questionnaire, if necessary, before the updated version was tested during subsequent interviews with new patients. Data interpretation was carried out blindly by the scientific committee, since patients’ identities were not disclosed. The minimum number of cognitive interviews was set at eight, and interviews ended when data sufficiency was reached (no new relevant data after two additional interviews ). The Ethics Committee of Angers University Hospital approved the study on September 1, 2020 (number 2020/106). Neither the health professionals of the panel nor the patients received financial compensation for participating in this project.
The first stage was a Delphi consensus involving a multidisciplinary health professional panel to evaluate and harmonize the clarity and appropriateness of the questionnaire for patients using the French health care system. Purposive sampling was used to recruit healthcare professionals with experience in coordinated care from diverse professions, practice structures, and cultural backgrounds. They were invited by e-mail to participate and a detailed information letter on the project was sent. All the professionals contacted accepted and participated in the Delphi rounds. The expert panel consisted of ten healthcare professionals, one of whom was from Quebec and was involved in creating the original questionnaire (Table ). Firstly, the scientific committee proposed an initial adaptation of the 14 items from the original Quebecois version (V0), called version V1. The list of the healthcare professionals visited by patients, which precedes the 14 items, was also modified by the scientific committee: some occupational categories, such as “nurse practitioner” or “kinesiologist” which are not present in France, were removed. All the questionnaire adaptations made by the scientific committee (related to items or list of professionals) were carried out by oral consensus, via live synchronous videoconferences. During the first Delphi round, versions V0 and V1 were submitted to the group of experts using an electronic LimeSurvey® form. Each expert was contacted individually by e-mail asking for a response within one week. For each item, the expert was asked: “Do you think this new formulation is sufficiently clear and adapted to a patient using the French health system?”. The expert was reminded of the need to maintain the original meaning of the item. Each expert indicated their degree of agreement with the reformulation by assigning a score between 1 (completely unsuitable) and 9 (completely suitable). Each score below 7 had to be justified with a comment. At the end of the first round, the scientific committee met to examine all the items with an average score below 8 and/or with less than 70% of scores greater than or equal to 7. No specific tools or programs have been used to analyze scores, except for the Excel tool. For each item not reaching consensus (as defined just above), the scientific committee used the experts’ comments to reformulate the item creating version V2. Only those items that had not yet achieved consensus, were submitted, in versions V0 and V2, to the group of experts in the second Delphi round. In order to not influence their views and decisions, experts were not provided with other experts' comments from the first Delphi round. Additional rounds would be performed until consensus was reached for all items.
During the second stage, patients assessed the comprehensibility and conformity of the adapted version of the questionnaire resulting from stage 1 and improved it if necessary. The checklist of health professionals was also submitted to the patients. This was achieved using cognitive interviews. Two members of the scientific committee (SL and AP) recruited patients, first from a rural general practice, and then from an urban community pharmacy, as complementary support. Adult patients who could speak French, had one or more chronic illnesses (physical and/or mental) and benefited from regular monitoring were eligible for inclusion. For each patient, sociodemographic data were collected (gender, age, profession and socio-professional category, health condition). Purposive maximum variation sampling was used to obtain a diverse patient panel including gender, age, socio-professional category (proxy for literacy level) and chronic diseases. Eligible patients were invited to participate in the study at the end of a medical consultation or a visit to the pharmacy. Potential participants were provided with an information letter explaining the nature of the study. Those who accepted provided written consent. Two patients refused to participate. Two accepted but one was excluded before data collection due to cognitive disorders preventing them from being able to read and sadly, the other died unexpectedly before the interview. One female general practitioner (SL) conducted the cognitive interviews. She had no previous interview experience and received training from the senior researcher who supervised her (ARR). Each patient was interviewed individually. The objective was to systematically identify any discrepancies between the researchers’ and the patient’s understanding of the item and to improve the formulation if necessary . The patients were provided with the adapted version of the questionnaire resulting from stage 1 and the checklist of healthcare professionals involved in their care at recruitment. They were instructed to only read it during the interview to ensure the patient's spontaneous reactions were collected. Patients were asked to read each item aloud, rephrase it in their own words, and point out any ambiguity or misunderstanding. They then answered the item, thinking aloud and giving reasons for their response. If discrepancy or ambiguity was identified, the patient was invited to suggest a possible reformulation following an explanation from the researcher. The interviews took place by telephone due to the contact restrictions imposed in France during the COVID-19 epidemic. Each interview was audio recorded, transcribed item by item, then compiled interview by interview. This enabled discrepancies, ambiguities, and suggested reformulations for each item to be extracted and reconciled. Any items that patients felt were not applicable to them were also listed. Audio files were archived only on the password-protected computer of the interviewer (SL), and deleted once transcription and de-identification of the excerpts processes were completed. One researcher (SL) analyzed the cognitive interviews under the supervision of a senior researcher (ARR). The analysis process focused on the patients' reaction to the question, their ability to rephrase it, their answer to the question, and the justification they provided for their response. Based on these points, the researcher assessed whether the item was completely understood and if the understanding was adequate in regard to its original meaning. The outcome of her interpretation was noted, as well as patient’s comments, in a de-identified report, before submitting it to the scientific committee. The scientific committee met every 3 to 4 interviews to validate the results and modify the questionnaire, if necessary, before the updated version was tested during subsequent interviews with new patients. Data interpretation was carried out blindly by the scientific committee, since patients’ identities were not disclosed. The minimum number of cognitive interviews was set at eight, and interviews ended when data sufficiency was reached (no new relevant data after two additional interviews ). The Ethics Committee of Angers University Hospital approved the study on September 1, 2020 (number 2020/106). Neither the health professionals of the panel nor the patients received financial compensation for participating in this project.
Stage 1 The ten included experts completed all the Delphi rounds (100% participation rate). Most responded within the allotted time of seven days per round (maximum three days late). First Delphi round The experts accepted seven of the 14 questionnaire items proposed in version V1 (nos. 2, 4, 8, 9, 10, 11 and 12) (Table ). The scientific committee reformulated five of the seven remaining items, which did not reach consensus (nos. 1, 3, 5, 6, and 13) using the experts’ comments (Supplementary Material 1A). However, they were unable to identify a satisfactory alternative for the last two items (nos. 7 and 14) which were resubmitted to the panel during the next round without reformulation. Second Delphi round The second round began three weeks after the end of the first. On version V2, three reformulated items (nos. 1, 5 and 6) fulfilled the validation criteria and were accepted (Table ). The scientific committee could not agree on a reformulation, based on experts’ comments and suggestions, for the four remaining items which did not reach consensus (nos. 3, 7, 13 and 14). This was either because the expert comments diverged significantly from each other, or because they diverged from the initial intent of the Quebecois questionnaire (Supplementary Material 1B). The scientific committee, therefore, considered it irrelevant to conduct a third Delphi round. In addition, items 3 and 7 were considered essential to evaluate the person-centred care while items 13 and 14 were considered essential to evaluate the care coordination during transitions. Therefore, items 3, 7, 13, and 14 were not deleted but submitted unchanged to the patients in version V3. Stage 2 Patient cognitive interviews A total of 14 cognitive interviews were conducted lasting an average of 30 min. Patient characteristics are summarized in Table . Seven interviews took place using version V3 of the questionnaire and were analyzed before the scientific committee met (Supplementary Material 2A). Several suggestions concerned the checklist of healthcare professionals involved in the patient’s care leading to the addition of “ostéopathe” (equivalent to an osteopath) and pédicure-podologue (allied health professional specialized in foot care). No patients highlighted any misunderstanding or ambiguity with items 1, 2, 3, 7, 8, 9, 10, 11, 12 so they remained unchanged. The term “parcours de soins” ( “care plan” ) in items 5, 6 and 13 was interpreted differently among patients (as had already been the case among the experts during stage 1) without a satisfactory alternative being found. The scientific committee therefore decided to test the addition of a “parcours de soins” definition, that was inspired by existing literature . Items 4 and 14 were a source of ambiguity or misunderstanding for several patients so were reformulated. In item 4, “mes médicaments étaient connus par ces professionnels de santé” (“ My list of medications was known by these healthcare professionals ”) was changed to “ces professionnels de santé savaient quels médicaments je prenais” (“ These healthcare professionals knew what medications I was taking ”). In item 14 “une conduite à tenir était proposée pour la prochaine étape” (“ a course of action was proposed for the next step ”) was replaced by “celui-ci avait déjà réfléchi aux soins à me proposer pour ma santé” (“ they had already thought about the care plan to offer me next, for my health” ). Finally, several patients wondered about the expression “ces professionnels de santé” (“these healthcare professionals” ), without making the expected link with the healthcare professionals designated in the checklist preceding the 14 items. The scientific committee proposed changing it to “mes professionnels de santé” (“my healthcare professionals”). Four interviews were conducted using this new V4 version (Supplementary Material 2B). Several patients mentioned healthcare professionals they had seen more than four months ago so the instructions relating to the checklist of healthcare professionals were revised to better highlight the required 4-month period. Overall, the items seemed to be better understood. The reformulated item 4 seemed clearer, and the “parcours de soins” definition made the meaning of items 5, 6 and 13 more evident. However, for item 14, the understanding seemed improved but moved away from its initial meaning. The scientific committee therefore reformulated this in version V5 (Table ). Version V5 was tested with three additional patients (Supplementary Material 2C). They did not identify any residual difficulty. The scientific committee therefore considered this version to be the final version resulting from the cross-cultural adaptation process (Fig. C).
The ten included experts completed all the Delphi rounds (100% participation rate). Most responded within the allotted time of seven days per round (maximum three days late). First Delphi round The experts accepted seven of the 14 questionnaire items proposed in version V1 (nos. 2, 4, 8, 9, 10, 11 and 12) (Table ). The scientific committee reformulated five of the seven remaining items, which did not reach consensus (nos. 1, 3, 5, 6, and 13) using the experts’ comments (Supplementary Material 1A). However, they were unable to identify a satisfactory alternative for the last two items (nos. 7 and 14) which were resubmitted to the panel during the next round without reformulation. Second Delphi round The second round began three weeks after the end of the first. On version V2, three reformulated items (nos. 1, 5 and 6) fulfilled the validation criteria and were accepted (Table ). The scientific committee could not agree on a reformulation, based on experts’ comments and suggestions, for the four remaining items which did not reach consensus (nos. 3, 7, 13 and 14). This was either because the expert comments diverged significantly from each other, or because they diverged from the initial intent of the Quebecois questionnaire (Supplementary Material 1B). The scientific committee, therefore, considered it irrelevant to conduct a third Delphi round. In addition, items 3 and 7 were considered essential to evaluate the person-centred care while items 13 and 14 were considered essential to evaluate the care coordination during transitions. Therefore, items 3, 7, 13, and 14 were not deleted but submitted unchanged to the patients in version V3.
The experts accepted seven of the 14 questionnaire items proposed in version V1 (nos. 2, 4, 8, 9, 10, 11 and 12) (Table ). The scientific committee reformulated five of the seven remaining items, which did not reach consensus (nos. 1, 3, 5, 6, and 13) using the experts’ comments (Supplementary Material 1A). However, they were unable to identify a satisfactory alternative for the last two items (nos. 7 and 14) which were resubmitted to the panel during the next round without reformulation.
The second round began three weeks after the end of the first. On version V2, three reformulated items (nos. 1, 5 and 6) fulfilled the validation criteria and were accepted (Table ). The scientific committee could not agree on a reformulation, based on experts’ comments and suggestions, for the four remaining items which did not reach consensus (nos. 3, 7, 13 and 14). This was either because the expert comments diverged significantly from each other, or because they diverged from the initial intent of the Quebecois questionnaire (Supplementary Material 1B). The scientific committee, therefore, considered it irrelevant to conduct a third Delphi round. In addition, items 3 and 7 were considered essential to evaluate the person-centred care while items 13 and 14 were considered essential to evaluate the care coordination during transitions. Therefore, items 3, 7, 13, and 14 were not deleted but submitted unchanged to the patients in version V3.
Patient cognitive interviews A total of 14 cognitive interviews were conducted lasting an average of 30 min. Patient characteristics are summarized in Table . Seven interviews took place using version V3 of the questionnaire and were analyzed before the scientific committee met (Supplementary Material 2A). Several suggestions concerned the checklist of healthcare professionals involved in the patient’s care leading to the addition of “ostéopathe” (equivalent to an osteopath) and pédicure-podologue (allied health professional specialized in foot care). No patients highlighted any misunderstanding or ambiguity with items 1, 2, 3, 7, 8, 9, 10, 11, 12 so they remained unchanged. The term “parcours de soins” ( “care plan” ) in items 5, 6 and 13 was interpreted differently among patients (as had already been the case among the experts during stage 1) without a satisfactory alternative being found. The scientific committee therefore decided to test the addition of a “parcours de soins” definition, that was inspired by existing literature . Items 4 and 14 were a source of ambiguity or misunderstanding for several patients so were reformulated. In item 4, “mes médicaments étaient connus par ces professionnels de santé” (“ My list of medications was known by these healthcare professionals ”) was changed to “ces professionnels de santé savaient quels médicaments je prenais” (“ These healthcare professionals knew what medications I was taking ”). In item 14 “une conduite à tenir était proposée pour la prochaine étape” (“ a course of action was proposed for the next step ”) was replaced by “celui-ci avait déjà réfléchi aux soins à me proposer pour ma santé” (“ they had already thought about the care plan to offer me next, for my health” ). Finally, several patients wondered about the expression “ces professionnels de santé” (“these healthcare professionals” ), without making the expected link with the healthcare professionals designated in the checklist preceding the 14 items. The scientific committee proposed changing it to “mes professionnels de santé” (“my healthcare professionals”). Four interviews were conducted using this new V4 version (Supplementary Material 2B). Several patients mentioned healthcare professionals they had seen more than four months ago so the instructions relating to the checklist of healthcare professionals were revised to better highlight the required 4-month period. Overall, the items seemed to be better understood. The reformulated item 4 seemed clearer, and the “parcours de soins” definition made the meaning of items 5, 6 and 13 more evident. However, for item 14, the understanding seemed improved but moved away from its initial meaning. The scientific committee therefore reformulated this in version V5 (Table ). Version V5 was tested with three additional patients (Supplementary Material 2C). They did not identify any residual difficulty. The scientific committee therefore considered this version to be the final version resulting from the cross-cultural adaptation process (Fig. C).
A total of 14 cognitive interviews were conducted lasting an average of 30 min. Patient characteristics are summarized in Table . Seven interviews took place using version V3 of the questionnaire and were analyzed before the scientific committee met (Supplementary Material 2A). Several suggestions concerned the checklist of healthcare professionals involved in the patient’s care leading to the addition of “ostéopathe” (equivalent to an osteopath) and pédicure-podologue (allied health professional specialized in foot care). No patients highlighted any misunderstanding or ambiguity with items 1, 2, 3, 7, 8, 9, 10, 11, 12 so they remained unchanged. The term “parcours de soins” ( “care plan” ) in items 5, 6 and 13 was interpreted differently among patients (as had already been the case among the experts during stage 1) without a satisfactory alternative being found. The scientific committee therefore decided to test the addition of a “parcours de soins” definition, that was inspired by existing literature . Items 4 and 14 were a source of ambiguity or misunderstanding for several patients so were reformulated. In item 4, “mes médicaments étaient connus par ces professionnels de santé” (“ My list of medications was known by these healthcare professionals ”) was changed to “ces professionnels de santé savaient quels médicaments je prenais” (“ These healthcare professionals knew what medications I was taking ”). In item 14 “une conduite à tenir était proposée pour la prochaine étape” (“ a course of action was proposed for the next step ”) was replaced by “celui-ci avait déjà réfléchi aux soins à me proposer pour ma santé” (“ they had already thought about the care plan to offer me next, for my health” ). Finally, several patients wondered about the expression “ces professionnels de santé” (“these healthcare professionals” ), without making the expected link with the healthcare professionals designated in the checklist preceding the 14 items. The scientific committee proposed changing it to “mes professionnels de santé” (“my healthcare professionals”). Four interviews were conducted using this new V4 version (Supplementary Material 2B). Several patients mentioned healthcare professionals they had seen more than four months ago so the instructions relating to the checklist of healthcare professionals were revised to better highlight the required 4-month period. Overall, the items seemed to be better understood. The reformulated item 4 seemed clearer, and the “parcours de soins” definition made the meaning of items 5, 6 and 13 more evident. However, for item 14, the understanding seemed improved but moved away from its initial meaning. The scientific committee therefore reformulated this in version V5 (Table ). Version V5 was tested with three additional patients (Supplementary Material 2C). They did not identify any residual difficulty. The scientific committee therefore considered this version to be the final version resulting from the cross-cultural adaptation process (Fig. C).
The objective of this study was to perform cross-cultural adaptation of the validated Quebecois PCCCT questionnaire for use in France. The two-stage method involving both healthcare professional and patient expertise produced a version of the questionnaire aimed at being used in France and which has item and semantic equivalence to the original Quebecois version. To our knowledge, until now there has not been a simple and valid tool to measure the patient’s perspective on patient-centered care coordination by a care team in France. Some of the existing questionnaires only consider the healthcare professionals’ perspective or the health system perspective, which do not necessarily reflect the patient's subjective experience of good patient-centered care coordination . Other existing questionnaires consider the patient's perspective but focus on one specific condition such as diabetes or cancer [ – ]. Others evaluate continuity of care which is a concept close to but distinct from that of patient-centered coordination of care . Of the existing non-condition specific questionnaires focusing on the patient’s perspective in primary care [ , – ], the PCCCT was best suited to our needs since it was concise and simple. It also included three essential elements: 1) the checklist of healthcare professionals involved in the patient’s care; 2) the degree of coordination between them; 3) the patient's own problems and objectives. Having such a tool is necessary for documenting and evaluating the health system reforms and public policies aimed explicitly at improving patient-centered care coordination. When implementing these reforms and policies, the patient’s perception of their care experience is an essential element that must be documented and considered. Our study is in line with the work carried out by the Organization for Economic Cooperation and Development (OECD) as part of the Patient-Reported Indicator Surveys initiative (PaRIS) . In this initiative, several countries, including France and Canada, are working together to develop, standardize and implement a new generation of indicators to measure the healthcare experiences and outcomes that matter most to patients. Ultimately, the PaRIS project will provide access to patient-reported data and leverage for better clinical, managerial, and political decision-making. Therefore, the adapted PCCCT can be used to enrich this patient-reported data about healthcare experiences which is an essential dimension for improving the quality of care according to the five health goals principles These are enhancing population health, improving care experience, reducing costs, limiting burnout among members of the healthcare workforce and advancing health equity. The psychometric qualities of the French version of the PCCCT questionnaire still need to be evaluated to verify measurement equivalence compared to the original Quebecois version. This step will complete the cross-cultural adaptation process. We chose to involve both professionals and patients because the two panels have complementary expertise for the adaptation process. The participating professionals reflected various professional skills, practice environments, geographic and linguistic origins, while the patients reflected different care experiences related to their gender, age, socio-professional category and somatic and/or mental health problems. The study was implemented with few deviations from the pre-established protocol and with excellent involvement of all participants. Especially, patients’ participation in cognitive interviews faced few barriers, they appear to appreciate feeling concerned and sharing their care experience during this study that was not very binding and time-consuming. However, there were no patients older than 80 years since the two patients in this age group who agreed to participate were not able to do so, as discussed previously. It would have been beneficial to have more elderly participants since multimorbidity is particularly prevalent in this age group meaning patient-centered care coordination is particularly relevant to them. As concerns the Delphi stage, the integration of patients in this step could have sped up the adaptation process since the patient’s opinions would have been considered earlier, while the participation of a language expert could have made a complementary contribution and provided a relevant point of view on the interpretation and reformulation of the questionnaire. In addition, we made a conscious choice not to share in round 2 the experts’ comments from round 1. We considered that sharing this information would have impacted the experts' responses, while our study relied heavily on the personal understanding of the items and their subjective interpretation. However, this choice may have contributed to not reaching consensus on all items at the end of the Delphi stage. The study took place in a single region of France, taking care to recruit patients from two different care settings and two distinct departments to ensure regional diversity. Conducting the interviews by telephone limited the collection of certain data, particularly non-verbal communication such as facial expressions indicating incomprehension or hesitation. Using a video call may have helped with this but it was not used as some patients did not have or were reluctant to use the required technology, especially the elderly and those living in rural areas. It was decided to use the same media for all the participants. Furthermore, the instruction not to look at the questionnaire before the interview had not always been respected. Patients who had not confronted certain situations mentioned in the items were unable to respond based on their own experience. In this situation, it was suggested that the patients used the experiences of relatives to assess the level of understanding and response mechanisms.
The adaptation process used in this study means the PCCCT questionnaire may soon be available for use in France, after appropriate studies on equivalence measures will be carried out. This short and easily administered questionnaire can be employed for people with one or more chronic conditions to assesses their perspectives about patient-centered care coordination, particularly in primary care. It is a useful resource for the French health system reforms aimed at promoting more integrated and patient-centered care pathways.
Supplementary Material 1. Supplementary Material 2.
|
Mechanistic insights into Guizhi Fuzi decoction for lumbar disc herniation: Integrating network pharmacology and bioinformatics approach | edf658cd-1546-46c4-88fd-309a28125497 | 11936605 | Musculoskeletal System[mh] | Lumbar disc herniation (LDH) refers to the displacement of the nucleus pulposus (NP) through the annulus fibrosus beyond the normal intervertebral disc space, resulting in irritation or compression of adjacent spinal nerve roots. Clinically, patients may present with lumbar pain, which can be accompanied by radiating pain and sensory decrement in the lower limbs. In severe cases, they may develop cauda equina syndrome or become unable to walk. The treatment of LDH can be categorized into nonsurgical and surgical approaches. Nonsurgical treatments, including bed rest, exercise, physiotherapy (traction, acupuncture, massage, etc), and pharmacological therapy (nonsteroidal anti-inflammatory drugs, muscle relaxants, etc). It is noteworthy that most patients experience spontaneous regression of the herniated disc tissue, and approximately 60% to 90% of LDH cases can be managed with conservative treatment strategies. Therefore, in-depth research on conservative treatment methods is essential. In TCM theory, LDH is referred to as “Bi syndrome,” typically caused by Qi stagnation, blood stasis, dampness-cold, or deficiency in liver and kidney functions, and oral administration of traditional Chinese medicine is considered one of the important ways to treat LDH. The traditional Chinese medicine formula Guizhi Fuzi Decoction (GZFZT), recorded in ancient literature for the treatment of “bi syndrome,” consists of “Guizhi” (Cassia Twig), “Fuzi” (Monkshood Root), “Shengjiang” (Fresh Ginger), “Dazao” (Jujube), and “Gancao” (Licorice Root). Studies have indicated that Guizhi and Fuzi can alleviate the symptoms of lumbar pain and reduce inflammation in the lumbar spine. Moreover, pharmacological research on other components of GZFZT suggests that they are also associated with the pathogenesis of LDH. [ – ] In modern medicine, compared with nonsteroidal anti-inflammatory and analgesic drugs and muscle relaxants, researchers have found that GZFZT can equally improve symptoms in LDH patients without adverse reactions, and its long-term effects are more significant and stable. Furthermore, for LDH patients undergoing surgical treatment, GZFZT can facilitate their postoperative recovery and enhance the therapeutic efficacy. Importantly, due to the complexity of patients’ conditions and the diversity of disease mechanisms, a single drug is often insufficient to address the diverse manifestations of LDH. However, GZFZT utilizes the 4 properties (cold, hot, warm, and cool) and 5 tastes (sweet, sour, bitter, pungent, and salty) of 5 herbal medicines to provide targeted symptomatic treatment for the complex conditions of LDH patients. Currently, despite the clinical validation of the overall therapeutic efficacy of GZFZT, there is a lack of in-depth research into the mechanism by which the individual herbal components of GZFZT synergistically treat LDH. Therefore, our team has employed a comprehensive approach using network pharmacology, bioinformatics, and molecular docking techniques to explore the active ingredients and target proteins of GZFZT, and to elaborate on its mechanism of action in treating LDH. We hope that our research can provide a foundation for future studies on LDH treatment, offer more therapeutic insights to clinicians, and ultimately benefit patients. The research process is illustrated in Fig .
2.1. Identification of potential therapeutic targets of GZFZT Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, https://tcmsp-e.com/ ) is a database of systems pharmacology for Traditional Chinese Medicine, used to screen the active components and potential therapeutic targets of GZFZT. This database offers valuable connections between herbal treatments, disease targets, and diseases, making it an essential resource for TCM research. The screening criteria for GZFZT herbs included an oral bioavailability ≥ 30% and drug-likeness ≥ 0.18%. These parameters are essential for evaluating the ADME characteristics and are used to screen active compounds for drug selection. The effective active components and protein targets of the 5 herbs in GZFZT were retrieved using the keywords “Guizhi,” “Fuzi,” “Shengjiang,” “Dazao,” and “Gancao.” The targets were standardized using the UniProt protein database ( https://www.uniprot.org ). 2.2. GEO differential gene expression analysis Search the GEO database ( www.ncbi.nlm.nih.gov/geo/ ) using the keyword “lumbar disc herniation.” Select an appropriate GEO data chip and download the corresponding GSE files for subsequent analysis using R. Identify differentially expressed genes using the Limma package with criteria of logFC > 1 and P < .05. Use the ggplot2 package to create a volcano plot of the differentially expressed genes and the pheatmap package to generate heatmaps for the top 10 upregulated and downregulated differentially expressed genes. 2.3. Collecting therapeutic targets for LDH The targets associated with LDH were obtained by searching the DisGeNET, GeneCards, and OMIM databases using the keyword “lumbar disc herniation.” These targets were integrated with the differentially expressed genes from GEO. The UniProt database was used to convert target names to gene symbols and eliminate duplicates. 2.4. Active ingredient-disease target network construction of Chinese medicine The potential targets of active ingredients in GZFZT intersect with potential targets for LDH. Networks and type files were constructed and imported into Cytoscape 3.8.0 to build the GZFZT active ingredient-disease target network diagram. 2.5. Analysis of PPI network, GO and KEGG The intersecting target genes were carefully selected and imported into the STRING database with a minimum interaction score > 0.9, specifying the species as ``Homo sapiens’’. Using Cytoscape 3.8.0 software, a protein–protein interaction (PPI) network was constructed, taking into account hidden nodes. Gene ontology (GO) function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were conducted through online bioinformatics platforms ( https://www.bioinformatics.com.cn/ , https://www.omicshare.com/ ) to explore biological data of core target genes. This analysis provides insights into the complex regulatory mechanisms and interactions among these genes. 2.6. Molecular docking Molecular docking is a critical technique in network pharmacology used to validate compound-target interactions and determine binding affinity. Based on the degree values in the PPI network, 6 potential target proteins were selected for molecular docking studies. The PDB files of the main targets were obtained from the RCSB database ( https://www.rcsb.org ), while the 3D structures of the primary active compounds in GZFZT were sourced from the PubChem database. Pymol software was used to prepare the core target proteins by removing solvent molecules, and AutoDock Tools 1.5.7 was employed for further hydrogenation and charging. The core target proteins and active compounds were saved as “pdbqt” format files, and AutoDock Vina was utilized to set the position and size of the docking box for the docking process.
Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP, https://tcmsp-e.com/ ) is a database of systems pharmacology for Traditional Chinese Medicine, used to screen the active components and potential therapeutic targets of GZFZT. This database offers valuable connections between herbal treatments, disease targets, and diseases, making it an essential resource for TCM research. The screening criteria for GZFZT herbs included an oral bioavailability ≥ 30% and drug-likeness ≥ 0.18%. These parameters are essential for evaluating the ADME characteristics and are used to screen active compounds for drug selection. The effective active components and protein targets of the 5 herbs in GZFZT were retrieved using the keywords “Guizhi,” “Fuzi,” “Shengjiang,” “Dazao,” and “Gancao.” The targets were standardized using the UniProt protein database ( https://www.uniprot.org ).
Search the GEO database ( www.ncbi.nlm.nih.gov/geo/ ) using the keyword “lumbar disc herniation.” Select an appropriate GEO data chip and download the corresponding GSE files for subsequent analysis using R. Identify differentially expressed genes using the Limma package with criteria of logFC > 1 and P < .05. Use the ggplot2 package to create a volcano plot of the differentially expressed genes and the pheatmap package to generate heatmaps for the top 10 upregulated and downregulated differentially expressed genes.
The targets associated with LDH were obtained by searching the DisGeNET, GeneCards, and OMIM databases using the keyword “lumbar disc herniation.” These targets were integrated with the differentially expressed genes from GEO. The UniProt database was used to convert target names to gene symbols and eliminate duplicates.
The potential targets of active ingredients in GZFZT intersect with potential targets for LDH. Networks and type files were constructed and imported into Cytoscape 3.8.0 to build the GZFZT active ingredient-disease target network diagram.
The intersecting target genes were carefully selected and imported into the STRING database with a minimum interaction score > 0.9, specifying the species as ``Homo sapiens’’. Using Cytoscape 3.8.0 software, a protein–protein interaction (PPI) network was constructed, taking into account hidden nodes. Gene ontology (GO) function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were conducted through online bioinformatics platforms ( https://www.bioinformatics.com.cn/ , https://www.omicshare.com/ ) to explore biological data of core target genes. This analysis provides insights into the complex regulatory mechanisms and interactions among these genes.
Molecular docking is a critical technique in network pharmacology used to validate compound-target interactions and determine binding affinity. Based on the degree values in the PPI network, 6 potential target proteins were selected for molecular docking studies. The PDB files of the main targets were obtained from the RCSB database ( https://www.rcsb.org ), while the 3D structures of the primary active compounds in GZFZT were sourced from the PubChem database. Pymol software was used to prepare the core target proteins by removing solvent molecules, and AutoDock Tools 1.5.7 was employed for further hydrogenation and charging. The core target proteins and active compounds were saved as “pdbqt” format files, and AutoDock Vina was utilized to set the position and size of the docking box for the docking process.
3.1. Potential therapeutic targets for GZFZT Using the TCMSP database, 154 active compounds were screened, including 7 from Guizhi, 21 from Fuzi, 29 from Dazao, 5 from Shengjiang, and 92 from Gancao. Detailed data can be found in Table S1, Supplemental Digital Content, http://links.lww.com/MD/O583 . This analysis identified 230 drug targets. The data were visualized using Cytoscape 3.8.0, generating a network diagram of the “GZFZT Active Ingredients-Targets” relationship (Fig. ). Different shapes and colors represent herbal components and target genes respectively. Diamonds indicate drug targets, and circles represent active ingredients. Guizhi, Fuzi, Shengjiang, Dazao, and Gancao are represented in green, purple, blue, orange, and cyan, respectively. 3.2. GEO differential gene analysis Download the gene expression data for LDH from the GEO database (GSE124272), which includes 8 samples from normal individuals and 8 samples from LDH patients. A total of 417 differentially expressed genes were identified, of which 251 were upregulated and 166 were downregulated, as shown in Fig. B. Detailed data can be found in Table S2, Supplemental Digital Content, http://links.lww.com/MD/O584 . The heatmap of the top 10 upregulated and downregulated differentially expressed genes is shown in Fig. A, with information on these genes provided in Table . 3.3. Collecting therapeutic targets for LDH Integrating the GEO database with Disgenet, Genecards, and OMIM databases, 1492 LDH targets were identified, as shown in Fig. A. Detailed information is available in Table S3, Supplemental Digital Content, http://links.lww.com/MD/O585 . After removing duplicate targets, the intersection of LDH active compound targets and disease targets resulted in the identification of 90 common targets. These 90 targets are considered the primary therapeutic targets for LDH, as shown in Fig. B. Detailed information is available in Table S4, Supplemental Digital Content, http://links.lww.com/MD/O586 . 3.4. Active ingredient-disease target network construction of Chinese medicine Using Cytoscape 3.8.0 software, we constructed a network of therapeutic targets for GZFZT active components, comprising 239 nodes and 984 edges. The network was subjected to visual analysis, with results shown in Fig . In the network diagram, squares represent disease targets, while circles represent medicinal herb components. Gui Zhi, Fu Zi, Sheng Jiang, Da Zao, and Gan Cao are represented in purple, blue, pink, orange, and green, respectively. The main active components identified for GZFZT include quercetin, kaempferol, 7-methoxy-2-methyl isoflavone, naringenin, formononetin, licochalcone A, isorhamnetin, medicarpin and stigmasterol, as shown in Table . 3.5. Analysis of PPI network, GO and KEGG To obtain the PPI network, we analyzed 90 common targets using the String database ( https://cn.string-db.org/ ), resulting in a network with 78 nodes and 281 edges as shown in Fig. . This network reveals relationships between proteins. We conducted topological analysis of potential targets using the Cyto-CAN plugin and selected the top 6 targets based on their degree values, as shown in Table . Using R language, we performed GO and KEGG enrichment analyses on these targets, revealing their involvement in various biological processes, cellular components, and molecular functions, indicating their therapeutic potential, as illustrated in Fig. A. KEGG enrichment analysis emphasized their significance in multiple signaling pathways, suggesting their potential for tailored treatments against different diseases, as depicted in Fig. B. 3.6. Molecular docking verification Molecular docking was employed to verify interactions between screened active compounds and targets in the context of LDH using bioinformatics. High-ranking crucial targets including IL6, MAPK3, STAT3, MAPK1, TNF, and AKT1 were selected from the PPI network results for docking analysis with 10 compounds identified from the GZFZT active ingredient-disease target network. Two-dimensional structures of the main active ingredients were obtained from the TCMSP and PubChem databases, followed by energy minimization and preparation of small molecule ligands using Chem3D software. Protein receptors were prepared by removing water molecules and ligands using PYMOL software. The binding energies (kcal/mol) between the main targets and active compounds were determined, demonstrating their binding affinities as shown in Fig. and Table . Finally, molecular docking of ligands with receptors was conducted using AutoDock software. Compounds such as naringenin, β-sitosterol, medioresinol, and stigmasterol exhibited strong binding with core targets, while the MAPK3 active compound showed good binding with disease targets. Figure displays 4 functional groups with favorable docking results, indicating a higher likelihood of molecular interactions. These results attribute enhanced stability of binding molecules due to lower binding energies between small molecule ligands and protein receptors.
Using the TCMSP database, 154 active compounds were screened, including 7 from Guizhi, 21 from Fuzi, 29 from Dazao, 5 from Shengjiang, and 92 from Gancao. Detailed data can be found in Table S1, Supplemental Digital Content, http://links.lww.com/MD/O583 . This analysis identified 230 drug targets. The data were visualized using Cytoscape 3.8.0, generating a network diagram of the “GZFZT Active Ingredients-Targets” relationship (Fig. ). Different shapes and colors represent herbal components and target genes respectively. Diamonds indicate drug targets, and circles represent active ingredients. Guizhi, Fuzi, Shengjiang, Dazao, and Gancao are represented in green, purple, blue, orange, and cyan, respectively.
Download the gene expression data for LDH from the GEO database (GSE124272), which includes 8 samples from normal individuals and 8 samples from LDH patients. A total of 417 differentially expressed genes were identified, of which 251 were upregulated and 166 were downregulated, as shown in Fig. B. Detailed data can be found in Table S2, Supplemental Digital Content, http://links.lww.com/MD/O584 . The heatmap of the top 10 upregulated and downregulated differentially expressed genes is shown in Fig. A, with information on these genes provided in Table .
Integrating the GEO database with Disgenet, Genecards, and OMIM databases, 1492 LDH targets were identified, as shown in Fig. A. Detailed information is available in Table S3, Supplemental Digital Content, http://links.lww.com/MD/O585 . After removing duplicate targets, the intersection of LDH active compound targets and disease targets resulted in the identification of 90 common targets. These 90 targets are considered the primary therapeutic targets for LDH, as shown in Fig. B. Detailed information is available in Table S4, Supplemental Digital Content, http://links.lww.com/MD/O586 .
Using Cytoscape 3.8.0 software, we constructed a network of therapeutic targets for GZFZT active components, comprising 239 nodes and 984 edges. The network was subjected to visual analysis, with results shown in Fig . In the network diagram, squares represent disease targets, while circles represent medicinal herb components. Gui Zhi, Fu Zi, Sheng Jiang, Da Zao, and Gan Cao are represented in purple, blue, pink, orange, and green, respectively. The main active components identified for GZFZT include quercetin, kaempferol, 7-methoxy-2-methyl isoflavone, naringenin, formononetin, licochalcone A, isorhamnetin, medicarpin and stigmasterol, as shown in Table .
To obtain the PPI network, we analyzed 90 common targets using the String database ( https://cn.string-db.org/ ), resulting in a network with 78 nodes and 281 edges as shown in Fig. . This network reveals relationships between proteins. We conducted topological analysis of potential targets using the Cyto-CAN plugin and selected the top 6 targets based on their degree values, as shown in Table . Using R language, we performed GO and KEGG enrichment analyses on these targets, revealing their involvement in various biological processes, cellular components, and molecular functions, indicating their therapeutic potential, as illustrated in Fig. A. KEGG enrichment analysis emphasized their significance in multiple signaling pathways, suggesting their potential for tailored treatments against different diseases, as depicted in Fig. B.
Molecular docking was employed to verify interactions between screened active compounds and targets in the context of LDH using bioinformatics. High-ranking crucial targets including IL6, MAPK3, STAT3, MAPK1, TNF, and AKT1 were selected from the PPI network results for docking analysis with 10 compounds identified from the GZFZT active ingredient-disease target network. Two-dimensional structures of the main active ingredients were obtained from the TCMSP and PubChem databases, followed by energy minimization and preparation of small molecule ligands using Chem3D software. Protein receptors were prepared by removing water molecules and ligands using PYMOL software. The binding energies (kcal/mol) between the main targets and active compounds were determined, demonstrating their binding affinities as shown in Fig. and Table . Finally, molecular docking of ligands with receptors was conducted using AutoDock software. Compounds such as naringenin, β-sitosterol, medioresinol, and stigmasterol exhibited strong binding with core targets, while the MAPK3 active compound showed good binding with disease targets. Figure displays 4 functional groups with favorable docking results, indicating a higher likelihood of molecular interactions. These results attribute enhanced stability of binding molecules due to lower binding energies between small molecule ligands and protein receptors.
In clinical practice, several traditional Chinese medicine formulas, such as Bu Yang Huan Wu Tang and Bu Shen Huo Xue Fang, have been effectively used for the treatment of LDH. GZFZT, rooted in ancient medical literature, has demonstrated efficacy in the treatment of LDH, akin to traditional Chinese medicine prescriptions previously mentioned. According to the theory of traditional Chinese medicine, the combination of Fuzi with Guizhi serves to warm yang and disperse cold, while the synergy of Shengjiang, Dazao, and Gancao harmonizes qi and blood, enhances the righteous qi, thereby alleviating patients’ symptoms and strengthening their constitution. To further explore the molecular mechanisms underlying GZFZT’s therapeutic effects on LDH, we have integrated bioinformatics, network pharmacology approaches, and molecular docking techniques, aiming to provide a scientific basis for optimizing clinical treatment strategies. The network diagram of GZFZT active ingredient-disease targets in Fig. shows a total of 105 components. Studies have shown that compounds such as quercetin, kaempferol, 7-methoxy-2-methylisoflavone, naringenin, and β-sitosterol exert effects through various mechanisms. These mechanisms include anti-inflammatory, antioxidative stress, anti-osteoporosis, neuroprotection, and promoting extracellular matrix (ECM) homeostasis. Intervertebral disc degeneration (IDD) is a major cause of disc herniation and can lead to lower back pain. Research indicates that the inflammatory factor IL-1β is a critical factor in IDD, and quercetin, a major component of GZFZT, can inhibit IL-1β-induced senescence-associated secretory phenotype in NP cells while promoting ECM homeostasis, thereby improving IDD progression. Additionally, long-term use of senolytic drugs Dasatinib and Quercetin can improve disc degeneration, reduce systemic inflammation, and ameliorate adverse physical conditions associated with aging. Kaempferol, one of the main components of GZFZT, has various beneficial effects and is used as a nutritional supplement for its anti-inflammatory, anti-osteoporosis, antiaging, and antioxidative properties. Studies have shown that kaempferol can alleviate IDD by reducing lipopolysaccharide-induced pro-inflammatory cytokines in bone marrow-derived mesenchymal stem cells, including IL-1β, iNOS, TNF-α, and IL-6, while inhibiting NF-κB activation. Kaempferol, naringenin, β-sitosterol, formononetin, medicarpin, and licochalcone A can prevent osteoporosis; formononetin, medicarpin, and licochalcone A do so by inhibiting osteoblast apoptosis, while kaempferol, naringenin, and β-sitosterol additionally promote osteoblast formation, providing bone protection. [ – ] Most studies suggest a connection between osteoporosis and disc degeneration, but the specific mechanisms remain highly controversial and warrant further investigation. Moreover, naringenin exhibits anti-inflammatory and analgesic effects, making it effective for pain caused by nerve damage and inflammation. Recent research shows that naringenin has neuroprotective effects and, when used in combination with hydroxypropyl-β-cyclodextrin, can improve sciatic nerve pain in mice. Since the mouse model is created by directly damaging the sciatic nerve, further verification is needed to determine if naringenin can alleviate pain caused by sciatic nerve compression due to disc herniation. Stigmasterol, a plant-derived sterol, has antioxidant and anti-inflammatory properties and is effective in treating osteoarthritis. In vivo studies indicate that stigmasterol can upregulate the expression of anti-inflammatory cytokine IL-10 through the NF-κB and MAPK pathways, improving osteoarthritis. Local inflammation is a significant factor in disc degeneration and herniation, thus stigmasterol has great potential for treating LDH. In summary, the main active components of GZFZT exhibit multiple mechanisms of action related to LDH. Additionally, other active ingredients, such as 7-methoxy-2-methylisoflavone and isorhamnetin, may have potential therapeutic effects in treating LDH. Exploring these components can enhance future research on LDH treatment, potentially bringing relief to patients with LDH. Among the top 10 upregulated and downregulated differentially expressed genes, we identified a connection between SIRPB2 and LDH. SIRPB2 is a member of the SIRP receptor family. Analyzing whole blood RNA-seq transcriptome data from 25 LDH patients and healthy volunteers using WGCNA, researchers found that SIRPB2 expression was significantly upregulated and identified as one of the hub genes. SIRPB2 itself does not transmit activating or inhibitory signals, but when it binds to CD47 on antigen-presenting cells, it mediates cell–cell adhesion, promoting antigen-specific T cell proliferation and cytokine secretion. Therefore, SIRPB2 may play an important role in the immune response within the inflamed disc environment. SCARA5 is a class A scavenger receptor encoded on chromosome 8. Studies have found that SCARA5 expression levels in LDH patients are negatively correlated with neutrophil expression and activated mast cell expression. Researchers have speculated that the downregulation of SCARA5 in LDH patients may be related to promoting the proliferation of the NP and the occurrence of inflammation. S100B, an extracellular factor, was found to be generally upregulated in LDH patients in this study. This may be associated with S100B causing myocyte damage and inhibiting osteoblasts. OLFM4 is an antiapoptotic protein that has been reported to play a role in cell growth and inflammation, potentially promoting fibroblast-like synoviocytes proliferation or antiapoptosis, exacerbating synovial inflammation. OLFM4 is significantly upregulated in LDH patients, suggesting that OLFM4 could be a potential target for LDH treatment. Through bioinformatics analysis, we identified several differentially expressed genes. However, some of these genes have not been extensively studied or validated in related research. This may be due to their recent discovery, limited research, or insufficient exploration. Despite the lack of supporting literature, we believe these genes merit further investigation. They may play crucial roles in specific biological processes, disease progression, or therapeutic interventions, yet remain inadequately understood. Therefore, we recommend additional experiments and studies to gain a comprehensive understanding of the functions and significance of these under-researched genes, providing valuable insights. PPI network analysis identified core targets, including TNF, STAT3, MAPK1, IL-6, MAPK3, and AKT1. IDD is a common degenerative disease, leading to NP collapse or herniation, causing abnormal nerve root conditions in patients. TNF-α is considered a key component exacerbating inflammation during disc degeneration and affecting ECM homeostasis. Studies have shown a significant increase in TNF-α expression in the intervertebral disc and adjacent muscles of discogenic low back pain group rats compared to controls. IL-6 is a cytokine signaling through the type I cytokine receptor complex, exhibiting both anti-inflammatory and pro-inflammatory properties. In addition to T cells and macrophages, IL-6 is also secreted by intervertebral disc cells. Research indicates that protectin DX reduces IL-6 and IL-1β mRNA levels in non-compressive lumbar disc herniation rats, promoting transforming growth factor β mRNA transcription and improving lumbar nerve root pain. Furthermore, IL-6 not only impacts the catabolic metabolism of NP cells but also induces TNF expression in the dorsal root ganglion and apoptosis of neuronal cells, potentially leading to abnormal pain perception. [ – ] STAT is a DNA-binding protein family, and the JAK/STAT signaling pathway is activated in LDH and plays a role in inflammatory responses. Studies indicate that under negative regulation by SOCS3, the JAK/STAT signaling pathway can be activated by IL-6 and transmit information from the cell surface to the nucleus. Inhibiting this pathway may attenuate cell apoptosis and ECM degradation, thereby regulating the degeneration process of lumbar intervertebral disc chondrocytes. MAPK1 is considered a central regulator of various biochemical signals that modulate cellular functions such as proliferation, differentiation, and apoptosis. Previous research has implicated MAPK1 in the regulation of inflammation and catabolic metabolism in intervertebral disc, with MAPK1 regarded as a pivotal gene in Lumbar disc degeneration. Upregulated expression of MAPK1 inhibits NP cell proliferation, promotes NP cell senescence, ultimately leading to the development of disc degeneration and even LDH. Akt has 3 subtypes (AKT1, AKT2, and AKT3), all sharing a conserved structure domain with 80% sequence similarity. The Akt/PKB signaling pathway is considered a major mediator of cell survival, controlling cell growth, proliferation, and regulating cell movement and migration. Research concludes that Akt1/PKB constitutes approximately 75% of Akt/PKB total activity, and activation of only the Akt1/PKBa subtype correlates with the severity of intervertebral disc herniation. [ – ] Interestingly, AKT1 and AKT3 mRNA show significant positive correlation only in intervertebral disc herniation, suggesting potential synergistic effects of these subtypes in this condition. In summary, GZFZT may exert its therapeutic effects on LDH by targeting relevant targets such as TNF, STAT3, MAPK1, IL-6, and AKT1. GO enrichment analysis indicated that GZFZT is involved in multiple biological processes related to LDH, including regulation of apoptotic signaling pathway, response to oxidative stress, reactive oxygen species metabolic process, and muscle cell proliferation. Apoptosis is a programmed cell death process involved in normal development, tissue renewal, and immune system function, and is associated with many physiological and pathological conditions. Currently, 3 common apoptosis-inducing factors—Fas Ligand, TNF-α, and tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)—can trigger apoptosis in intervertebral disc cells through different signaling pathways. [ – ] Studies have shown that Fas can be detected in herniated lumbar disc tissue, with higher expression levels in discs with annulus fibrosus disruption compared to intact discs, and demonstrating that disc cells undergo apoptosis via the FAS and FasL system. The Type II (mitochondrial) pathway, one of the main pathways for Fas-mediated apoptosis, has been shown to occur in disc cell apoptosis. miR-495-3p is a novel miRNA whose overexpression can significantly inhibit TNF-α-induced apoptosis in NP cells. DR4 (death receptor 4) and DR5 (death receptor 5) are TRAIL’s agonistic receptors, and TRAIL binding to either can induce apoptosis. Research has demonstrated that the TRAIL/DR4 pathway is involved in the development of LDH, with DR4 expression positively correlating with disease severity. Additionally, the A20 protein encoded by the tumor necrosis factor α-induced protein 3 (TNFAIP3) gene has been shown to alleviate pain in LDH patients by inhibiting the NF-κB pathway. The KEGG pathway enrichment analysis indicates that GZFZT might be involved in pathways in cancer, lipid and atherosclerosis, AGE-RAGE signaling pathway in diabetic complications, and IL-17 signaling pathway. Notably, the AGE-RAGE signaling pathway, a crucial pathway involved in various biological processes, is found to be dysregulated in LDH as shown in Fig. . Carboxymethyl-lysine, one of the main components of advanced glycation end products (AGEs), is derived from glucose and oxidized lipids, and is both stable and irreversible. The receptor for AGEs act as inducers of intracellular pathways, and when carboxymethyl-lysine binds to the receptor for AGEs, NF-κB is released and translocates from the cytoplasm to the nucleus, generating inflammatory signals and oxidative stress responses, inducing the upregulation of MMPs, apoptosis, and other processes. This causes the release of proteolytic enzymes and an imbalance in the metabolism of matrix components, affecting the collagen and elastic connective tissues of the intervertebral disc. Recent studies suggest that the excessive accumulation of AGEs in the human body leads to the loss of cell adhesion and increased membrane permeability, resulting in muscle fiber damage. However, it is yet to be determined whether this impacts muscle function and subsequently the intervertebral disc tissue. Studies have found that high levels of IL-17A are associated with lumbar disc degeneration and LDH, and IL-17A is considered a key factor in disc pathology as shown in Fig. . IL-17 (also known as IL-17A) is produced by T helper 17 cells, a subset of CD4+ T cells. It has been reported that IL-17A, in conjunction with other cytokines such as TNFα, stimulates the NF-κB pathway, leading to the release of inflammatory cytokines IL-2, IL-6, and TNFα, as well as catabolic enzymes MMP-3, MMP-9, and MMP-13. These factors contribute to the further progression of LDH pathology. [ – ] Compared to healthy controls, LDH patients with NP protrusion combined with AF rupture show increased levels of IL-17A, which correlate with the intensity of sciatic nerve pain. Notably, the frequency of T helper 17 cells and the concentration of IL-17A positively correlate with the visual analog scale scores for low back pain. Although Pathways in cancer and Lipid and atherosclerosis play significant roles in the development of LDH, the current research has limitations, and the mechanisms of LDH remain unclear, necessitating further investigation. To explore the molecular mechanisms of GZFZT in treating LDH, we employed molecular docking techniques to predict interactions between 6 key target proteins—IL6, MAPK3, STAT3, MAPK1, TNF, and AKT1—and active compounds such as quercetin, kaempferol, 7-methoxy-2-methyl isoflavone, naringenin, formononetin, licochalcone A, isorhamnetin, medicarpin and stigmasterol. Molecular docking results indicate strong binding affinity ranging from −6.9 to −11.0 (kcal/mol) between these proteins and ligands. Although modern bioinformatics approaches have provided valuable insights into the effectiveness of GZFZT in treating LDH, this study still has certain limitations. Firstly, while network pharmacology and molecular docking techniques have proven to be of great significance, the limitations and delays inherent in network information technology and database data may lead to incomplete or biased research results, thereby affecting the accuracy and reliability of the studies. Therefore, further validation through pharmacokinetic and molecular biology experiments is necessary. Meanwhile, given that network pharmacology research inherently involves highly complex interdisciplinary collaboration, such collaboration inevitably leads to increased research time and costs. Therefore, in future research practices, it is imperative to actively explore and promote opportunities for interdisciplinary cooperation, synergistically accelerating the development of network pharmacology research. Nonetheless, the development of GZFZT holds significant importance and potential, demonstrating substantial therapeutic efficacy and relative safety in the treatment of LDH. Furthermore, the research progress on GZFZT provides a certain decision-making basis for the research directions of scientists and offers inspiration for clinical treatment management.
In this study, we systematically elucidated the relationship between GZFZT and LDH at the molecular level, indicating the potential of GZFZT as a novel oral drug for the treatment of LDH. Our study identified quercetin, kaempferol, 7-methoxy-2-methylisoflavone, naringenin, β-sitosterol, formononetin, licochalcone A, isorhamnetin, medicarpin, and stigmasterol as the key active components in GZFZT for the treatment of LDH. The mechanism of action involves core targets such as IL6, MAPK3, STAT3, MAPK1, TNF, and AKT1. Notably, naringenin, β-sitosterol, and stigmasterol exhibited prominent targeting effects on MAPK3 in molecular docking studies. Further research is recommended to validate the findings of this study.
The authors acknowledge the Kunming Institute of Botany’s Bioactivity Screening Center for its assistance in evaluating the compounds’ bioactivity.
Conceptualization: Jiafeng Peng. Formal analysis: Jiafeng Peng, Ran Xu. Funding acquisition: Danyang Li. Investigation: Jiafeng Peng, Hongxing Zhang. Methodology: Jiafeng Peng, Li Xia. Project administration: Yingchun Li. Resources: Qianqian Meng. Software: Huaize Wang. Supervision: Minglei Gao, Junchen Zhu. Validation: Xingfu Ma. Visualization: Junchen Zhu. Writing – review & editing: Junchen Zhu.
SUPPLEMENTARY MATERIAL
|
An educational intervention for improving knowledge, attitude, and practice of dietary salt intake among individuals with hypertension in public sector secondary care facilities, Agra, India, 2021 | e6d23d88-5565-429b-959d-d425e4ed5a16 | 11180682 | Patient Education as Topic[mh] | INTRODUCTION Excess dietary salt intake significantly contributes to high blood pressure, which, in turn, increases the risk of cardiovascular and progressive renal diseases. A meta‐analysis of 34 trials showed that reducing dietary salt by 4.4 g/day could reduce the mean systolic blood pressure by 4 mm Hg and diastolic by 2 mm Hg. Countries worldwide have adopted strategies, including decreasing dietary salt consumption, salt and food reformulation, consumer education, front‐of‐pack labeling, and salt taxation. In 2017−2018, 29% of adult Indians aged 18–69 years had hypertension. An average Indian consumed 11 g of salt daily, double the recommended limit. Despite the global target of achieving a 30% reduction in population salt intake by 2025, no national‐level strategies have been put in place to reduce salt usage in the Indian population to date. The primary source of salt intake in India comes from added salt during home cooking. Thus, consumer education should focus on the importance of limiting salt or choosing substitutes while cooking. This study examined the effect of a nurse‐led counseling intervention on salt‐related knowledge, attitude, and practices (KAP) among hypertensive patients at public‐sector secondary care facilities in Agra, Uttar Pradesh, India.
METHODS We conducted a prepost intervention study without a comparison group at noncommunicable diseases (NCDs) clinics of three purposively selected public‐sector secondary care facilitites, known as Community Health Centers, in Agra, India, from April to June 2021. The study included all hypertensive patients aged 18–60 under treatment for ≥ 6 months. We excluded individuals with severe comorbidities and those on a prescribed diet. We assumed the intervention would bring a 10% reduction in adding salt to the dough. We calculated the sample size with 95% confidence and 80% power as 77, but adjusted to 185 after factoring in a design effect of 2 and a 20% nonresponse adjustment. Nonetheless, we included all eligible patients receiving care in the facility for ethical reasons. Trained assistants (TAs) used a semistructured questionnaire adapted from the Euro‐WHO Salt tool to collect salt‐related KAP at baseline, immediately, and 1‐month postcounseling session. TAs contacted nonreporting individuals by phone for follow‐up. We developed counseling materials in local language following the WHO salt reduction guidelines with tips for reducing salt consumption during shopping, cooking, and dining. We trained one nurse per facility to deliver a one‐time salt reduction counseling session, lasting 20−25 min, to the participants individually after their baseline assessment. Nurses educated the participants on instructions towards salt reduction in the kitchen, the dining table, and also for shopping. We excluded individuals with incomplete follow‐up or missing data. We used Pearson's Chi‐square or Fisher exact test to compare the characteristics of the final study sample with those of loss‐to‐follow‐up (LTFU) individuals. We summarized the KAP indicators as proportions and compared the baseline with immediate and 1‐month postintervention indicators using the McNemar test. We used STATA SE (version 17.0) software (StataCorp LLC, College Station, Texas, USA) for statistical analysis and set the significance level at p < .05. The Human Ethics Committee of ICMR National Institute of Epidemiology approved the study; TAs obtained written consent from all participants.
RESULTS Of 263 patients assessed at baseline, we included 153 (58%) in the final analysis. We excluded 48 individuals due to incomplete data at the baseline and immediate post‐counseling session. We excluded another 62 due to loss‐to‐follow up at the 1‐month post counselling. Nearly half ( n = 78, 51%) were males, and 54% ( n = 83) were in the 41−60 age group. All sociodemographic characteristics of the included participants were comparable to those of LTFU participants (Table ). At baseline, 27% ( n = 41) knew the health risks of a high salt intake; this increased to 61% ( n = 93, p < .001) post‐intervention and remained unchanged ( n = 93 [61%], p < .001) at the 1‐month follow‐up. Knowledge regarding the recommended daily salt limit increased from 2% ( n = 3) at baseline to 36% ( n = 55, p < .001) at post‐intervention and then to 38% ( n = 58, p < .001) at the 1‐month follow‐up. Knowledge about low‐sodium alternatives showed the highest increase (4% at baseline to 42% at 1‐month post‐intervention, p < .001). At baseline, 16% ( n = 25) of participants agreed that reducing salt is beneficial for health, which increased to 36% ( n = 55, p < .001) post‐intervention and remained at 34% ( n = 52, p = .22) after 1 month. Individuals who considered prioritizing salt reduction increased from 34% ( n = 52) to 53% ( n = 81, p < .001) immediately after the intervention and remained almost similar ( n = 79 [52%], p < .001) at the 1‐month post‐intervention. None of the participants used low‐sodium salt for cooking at baseline or 1‐month post‐intervention. At baseline, 30% ( n = 46) of the participants reported consuming high‐salt and instant food items daily; there was no significant change in the daily use of high‐salt and instant food items at both follow‐up assessments ( n = 46 [30%] vs. n = 42 [27%], p = .10). The practice of adding salt regularly to dough reduced from 48% ( n = 74) at baseline to 41% ( n = 63, p = .002) at the 1‐month follow‐up. Adding salt to the cooking rice also decreased from 44% ( n = 68) at baseline to 37% ( n = 57, p < .001) at the 1‐month follow‐up (Table ).
DISCUSSION We documented that nurse‐led one‐on‐one counseling improved KAP toward dietary salt intake among individuals with hypertension in public sector health facilities in Agra, India. The post‐intervention knowledge and attitude changes remained consistent for one month, with the highest improvement in knowledge, followed by attitudes and practices at the 1‐month follow‐up. Our findings are consistent with the literature that consumer education could significantly change KAPs toward dietary salt. , The improvement of knowledge and attitude could be due to socio‐culturally relevant counseling material delivered in the regional language. However, our study also suggested that educational intervention alone may not improve the uptake of low‐salt alternatives. Changing food behaviors is a challenging process that requires time and sustained effort. Furthermore, low‐sodium alternatives' availability and affordability may influence adopting such practices (12). The availability of low‐sodium alternatives in the study site is limited. Globally, government‐led consumer education programs are common salt reduction strategies. Implementing small‐scale counseling sessions, as demonstrated in the present study, could serve as the initial steps toward a more extensive planned intervention in the country. The high burden of NCDs, and human resource staffing at public sector facilities should be considered when implementing counseling on a larger scale, with respect for nurse's workload and schedules in busy clinics. One of the main strengths of this study is the use of a realistic and representative sample that enhances the validity and applicability of the results. Although there was a high LTFU, similar characteristics of LTFU participants to those of the included suggest no or low selection bias. Other limitations included the absence of a comparison group and the lack of measures for salt intake. The social desirability bias likely affected the baseline and follow‐ups, uniformly dividing its potential impact. Nevertheless, the changes in KAP were comparable to other studies with a similar setting. In conclusion, nurse‐led one‐on‐one counseling on dietary salt reduction improved knowledge, attitude, and a few practices toward salt use in a Agra, Uttar Pradesh, India. Interventional research is required to understand the effect of counseling on reducing blood pressure, which can later be included in the hypertension management program. We plan to explore the prolonged impact of our counseling on dietary salt intake and blood pressure in the future.
The authors do not have any conflict of interest to declare.
|
Influence of minimally invasive cavities on color stability of dental crowns with different filling sealers | 012a1f82-8db8-4fb1-80aa-23db0145335a | 11552461 | Dentistry[mh] | Minimally invasive endodontic cavities preserving part of the pulp chamber roof and pericervical dentin have been proposed to conserve the remaining tooth structure and increase its fracture resistance after performing definitive restorative procedures on teeth. – This approach involves creating smaller access cavities that avoid extensive removal of the tooth's structural components. However, preserving parts of the pulp chamber roof may prevent the complete removal of pulpal tissue and facilitate the retention of staining substances from antibiotic pastes, intracanal medications, root canal sealers, and irrigating solutions. , Color changes in endodontically treated teeth can be associated with the presence of endodontic sealers containing silver and bismuth oxides, as radiopacifying agents, in the pulp chamber. , The color of bismuth oxide changes from yellow to dark brown in the presence of irrigating solutions ( e.g. , sodium hypochlorite), collagen from remaining dentin, blood, and lack of free oxygen. – Therefore, radiopacifiers based on silver oxide and bismuth oxide have been replaced by zirconium oxide and iron ( e.g. , AH Plus), , and bioceramic sealers were developed. , Epoxy resin-based sealers may lead to, however, progressive coronal discoloration over time. On the other hand, only imperceptible discoloration has been reported with bioceramic sealers. , Color evaluations with spectrophotometers are commonly employed to reduce possible biases related to subjective analysis. The Commission Internationale l'Eclariage (CIE) has defined systems to quantify color using coordinates indicating lightness (L* or Y) and chromaticity. The CIELAB system quantifies the chromaticity through coordinates a* (red-green axis) and b* (yellow-blue axis). Color differences are usually measured using either the CIE76 (∆E ab ) or CIEDE2000 (∆E 00 ) formulas. Although ∆E values are commonly used in dental studies, conclusions based only on these data are problematic. For instance, they do not indicate whether the specimens became yellower or whiter. Thus, using the yellowness index can help to better understand the color changes observed after endodontic treatment. , Tooth discoloration can also be due to voids around the restoration and remnants of filling material ( i.e. sealer and gutta-percha). Micro-computed tomography (micro-CT) is a useful tool to evaluate the volume of remaining sealer and voids in the restoration. – Therefore, this study evaluated the impact of the type of access surgery on the color stability of endodontically treated teeth obturated with either an epoxy resin-based sealer or a bioceramic sealer. Moreover, the remaining filling material in the pulp chamber and voids in the restoration were also assessed. The null hypothesis of this study is that neither: a) the endodontic access approach or b) the sealer composition affects the tooth color.
This in vitro study was approved by the Research Ethics Committee (CAAE: 40959720.0.000.5419) of the Faculty of Dentistry of Ribeirão Preto, University of São Paulo. Fifty sound human maxillary central incisors were scanned in a high-resolution micro-CT device (SkyScan 1174, Bruker, Kontich, Belgium), and 32 were selected based on the similarity of enamel (0.7–1.0 mm) and dentin (2.0–2.5 mm) thicknesses. The sample size was calculated for the continuous outcome ∆E 00 , defined as primary. A pilot study was carried out and showed a standard deviation of 1.13 (mean 2.41) in ∆E 00 values due to root canal obturation. Therefore, considering the threshold of 1.8 as clinically acceptable, the Cohen's D effect size calculated was 1.59. The sample size calculation was performed for a two-sided T-test assuming a type-I error of 5% and a test power of 80%. The G*Power software (version 3.1.9.6; Duesseldorf, Germany) was used for the calculation, and at least eight specimens per experimental condition were defined as necessary to meet these parameters. All selected teeth presented complete rhizogenesis, a single root canal, mild root curvature (°r > 8 mm), and no calcifications or resorptions. The initial tooth color was recorded on the buccal surface with a spectrophotometer (Vita Easyshade V, VITA Zahnfabrik, Bad Säckinge, Germany) based on the CIELAB system. An individualized silicone index was used to standardize the color reading area for the entire experiment, and readings were performed in triplicate. A tool available on the website Nix™ Color Sensor ( https://www.nixsensor.com/free-color-converter ) was used to convert the color coordinates from the CIELAB system to the CIEXYZ. The conversion considered a 2° observer and D65 illuminant, the defaults of the spectrophotometer used. YI of specimens was calculated using the ASTM Method E313: Y I = 100 × ( C x × X – C z × Z ) Y Where C x (1.2985) and C z (1.1335) are the coefficients for the X and Z coordinates, respectively, for the observer angle and illuminant used in the present study. Specimens were also scanned with the high-resolution micro-CT scanner. Geometric and flat-field corrections were obtained by positioning a 0.5-mm aluminum filter between the specimens and the X-ray source, altering the sensitivity to polychromatic radiation and reducing the possible hardening effect. A polystyrene tube was used to keep specimens perpendicular to the radiation source and reduce possible image distortions. Specimens were scanned under 50 kV, 800 mA, isotropic resolution of 26.7 mm, and 360° rotation. Images were reconstructed using the NRecon v.1.6.3 software (Bruker, Kontich, Belgium) and CTAN v.1.15.4 software (Bruker, Kontich, Belgium). A single experienced endodontist treated and restored all the canals using an operating microscope. The access to the pulp chamber for the conservative endodontic cavity (CEC) was carried out with high-speed rotating spheric diamond burs (#1013 and 1014, KG Sorensen, São Paulo, Brazil), followed by EndoZ burs (KG Sorensen, São Paulo, Brazil). Only the spheric diamond burs 1011HL (KG Sorensen, São Paulo, Brazil) were used for the minimally invasive endodontic cavity (MEC) (n = 16). After that, the root canal was explored with a #15 K-type hand file (Dentsply Maillefer, Baillagues, Switzerland) until its tip was visible at the apical foramen and the working length (WL) was set 0.5 mm shorter than this measurement. Then, following the manufacturer's instructions, the patency of the root canal was performed with the WaveOne Gold Glider (#15) instrument (Dentsply Maillefer, Baillagues, Switzerland), and the anatomical diameter was determined with a #20 K-type hand file (Dentsply Maillefer, Baillagues, Switzerland) at the WL. The biomechanical instrumentation was performed with a single WaveOne Gold Large (#45) instrument (Dentsply Maillefer, Baillagues, Switzerland), under a reciprocating movement, following the manufacturer's directions. After three pecking motions, the instrument was removed, and the root canals were irrigated with 5 mL of 2.5% sodium hypochlorite . Root canals were irrigated with 2 mL of 17% EDTA for 5 minutes, dried with absorbent paper points, and filled with a single gutta-percha cone corresponding to the final diameter #45 (Conform fit, Dentsply Maillefer, Baillagues, Switzerland). The obturation was performed using the single cone technique. Before its insertion, the cone was covered with either (AH Plus Dentsply, De Trey, Konstanz, Germany) or Bio-C Sealer (Angelus, Londrina, PR, Brazil) (n = 8). The excess obturating material was removed with a heated instrument 2 mm below the cementoenamel junction. Vertical condensation of plasticized gutta-percha was performed using Paiva condensers. Finally, the endodontic cavity was cleaned with sterile cotton soaked in 70% alcohol and an ultrasonic tip without refrigeration (E1 – Irrisonic Tip, Helse, Santa Rosa do Viterbo, SP, Brazil). The tooth color was measured again, and YI was calculated using previously described procedures. Data from CIELab were converted into CIELCh, and the overall color change caused by the obturation was calculated using the following formula (CIEDE2000): ∆ E 00 = ( ∆ L ' K L S L ) 2 + ( ∆ C ' K C S C ) 2 + ( ∆ H ' K H S H ) 2 + R T ∆ C ' K C S C ∆ H ' K H S H ' Changes in lightness (∆L’), chroma (∆C’), and hue (∆H’) were calculated by subtracting the value measured from the values of the prior step. S L , S C , and S H are the weighted functions for each component. K L , K C , and K H are the weighted factors for lightness, chroma, and hue, respectively (it was used K L = K C = K H = 1). R T is the interactive term between chroma and hue differences. Immediately after the obturation, the enamel surrounding the cavity was etched with 37% phosphoric acid (Dentsply Maillefer, Ballaigues, Switzerland) for 30s, rinsed with water and dried. A universal adhesive (Single Bond Universal, 3M ESPE, St. Paul, USA) was applied on the cavity with vigorous rubbing for 15 s and light-cured for 20 s. Cavities were incrementally restored with a nanofiller resin composite (Filtek Z-350 XT, shade A2, 3M ESPE, St. Paul, USA). Each increment was cured for 20 s with a LED-based light-curing unit (Radii Plus, SDI, Limited, Bayswater, Victoria, Australia) with an irradiance of 1,500 mW/cm 2 . Specimens were stored at 37°C under absolute humidity for 24h. The tooth color was measured again, and YI and ∆E 00 resulting from the restorative procedure were calculated. Specimens were also scanned, and images were reconstructed with the same parameters as the initial scan. The occurrence of voids around the restoration and remnants of filling material (sealer and gutta-percha) in the tooth crown was analyzed. The bottom of the section was considered the cementoenamel junction and the top, the incisal edge. Then, the voids were identified using the threshold scale in the CTAn software (Bruker, Kontich, Belgium). The "3D analysis tool" automatically calculated the volume of voids, considering the selected threshold . The same procedure was carried out to calculate the remnants of filling material in the pulp chamber. Three-dimensional images were used for qualitative analysis using the Data Viewer v.1.5.1 64-bit (Bruker, Kontich, Belgium) and CTVol v.2.3.1 (Bruker, Kontich, Belgium) software. The specimens were stored in distilled water (replaced every day) for one year in an oven at 37°C. Data of each outcome were individually tested regarding the normal distribution (Shapiro-Wilk test) and sphericity (Mauchly's W, Greenhouse-Geisser, and Huynh-Feldt tests). Two-way ANOVA ("sealer" vs. "endodontic access") was used to analyze data for voids in restorations and remnants of filling material in the pulpal chamber. For ∆E 00 and YI, statistical analysis was carried out using repeated-measured (RM) ANOVA to evaluate the independent variables "sealer", "endodontic access", and "assessment time". The last variable was defined as a factor for repeated measures. Pair-wise comparisons were performed using Tukey's test. A confidence level of 95% was pre-set for all analyses, which were carried out using the open statistical platform Jamovi 1.6.15 ( www.jamovi.org ).
Micro-CT analysis Neither "sealer" (p = 0.218) nor "endodontic access" (p = 0.522) affected the presence of filling material in the pulpal chamber, and the interaction between these factors was also not significant (p = 0.399) . Two-way ANOVA showed that both independent variables (sealer, p = 0.077; endodontic access, p = 0.848) had no significant effect on the occurrence of voids in restorations. However, the interaction between the variables was significant (p = 0.022) . Differences between sealers were observed only with MEC. Root canals accessed using MEC and filled with Bio-C Sealer resulted in more voids around the crown restoration compared to using AH Plus. The endodontic access did not affect the presence of voids when AH Plus was used but using Bio-C Sealer resulted in more voids for MEC than CEC. shows representative images reconstructed using micro-CT scans. It is possible to observe the presence of voids in green for Bio-C Sealer and MEC. Yellowness index Results for YI are presented in . RM ANOVA showed that only the variable "assessment time" (p < .001) affected the values of YI. Neither "sealer" (p = 0.307) nor "endodontic access" (p = 0.227) had a significant effect on this outcome. Only the double interaction "sealer vs. assessment time" (p = 0.021) was significant. The p-values for other interactions were: "sealer vs. endodontic access", p = 0.887; "endodontic access vs. assessment time", p = 0.666), and the triple interaction (p = 0.560). presents the results of Tukey's test for the interaction between the variables "assessment time" and "sealer". Regardless of the endodontic sealer, the specimens were yellower after the obturation than at baseline. The restoration slightly reduced the yellowness but without difference to the values observed after obturation. The difference between restored specimens and baseline was maintained only for the AH Plus. For both sealers, the highest values of YI were observed after the 1-year storage in the oven. No difference between the sealers was observed for all assessment times. Color changes after each step Results for ∆E 00 are displayed in . RM ANOVA showed that only the variable "assessment time" (p < 0.001) affected the values of ∆E 00 . Neither "sealer" (p = 0.537) nor "endodontic access" (p = 0.239) had a significant effect on this outcome. Moreover, all double interactions ("sealer vs. endodontic access", p = 0.908; "sealer vs. assessment time", p = 0.830; and "endodontic access vs. assessment time", p = 0.609), and the triple interaction (p = 0.391) were not significant. Pooled averages for the variable "assessment time" are presented in . Similar color changes were caused by both obturation and restoration, and the highest values of ∆E00 were observed after 1-year storage in the oven. Color coordinates were used to draw , which illustrates the color changes caused by each step according to the experimental conditions.
Neither "sealer" (p = 0.218) nor "endodontic access" (p = 0.522) affected the presence of filling material in the pulpal chamber, and the interaction between these factors was also not significant (p = 0.399) . Two-way ANOVA showed that both independent variables (sealer, p = 0.077; endodontic access, p = 0.848) had no significant effect on the occurrence of voids in restorations. However, the interaction between the variables was significant (p = 0.022) . Differences between sealers were observed only with MEC. Root canals accessed using MEC and filled with Bio-C Sealer resulted in more voids around the crown restoration compared to using AH Plus. The endodontic access did not affect the presence of voids when AH Plus was used but using Bio-C Sealer resulted in more voids for MEC than CEC. shows representative images reconstructed using micro-CT scans. It is possible to observe the presence of voids in green for Bio-C Sealer and MEC.
Results for YI are presented in . RM ANOVA showed that only the variable "assessment time" (p < .001) affected the values of YI. Neither "sealer" (p = 0.307) nor "endodontic access" (p = 0.227) had a significant effect on this outcome. Only the double interaction "sealer vs. assessment time" (p = 0.021) was significant. The p-values for other interactions were: "sealer vs. endodontic access", p = 0.887; "endodontic access vs. assessment time", p = 0.666), and the triple interaction (p = 0.560). presents the results of Tukey's test for the interaction between the variables "assessment time" and "sealer". Regardless of the endodontic sealer, the specimens were yellower after the obturation than at baseline. The restoration slightly reduced the yellowness but without difference to the values observed after obturation. The difference between restored specimens and baseline was maintained only for the AH Plus. For both sealers, the highest values of YI were observed after the 1-year storage in the oven. No difference between the sealers was observed for all assessment times.
Results for ∆E 00 are displayed in . RM ANOVA showed that only the variable "assessment time" (p < 0.001) affected the values of ∆E 00 . Neither "sealer" (p = 0.537) nor "endodontic access" (p = 0.239) had a significant effect on this outcome. Moreover, all double interactions ("sealer vs. endodontic access", p = 0.908; "sealer vs. assessment time", p = 0.830; and "endodontic access vs. assessment time", p = 0.609), and the triple interaction (p = 0.391) were not significant. Pooled averages for the variable "assessment time" are presented in . Similar color changes were caused by both obturation and restoration, and the highest values of ∆E00 were observed after 1-year storage in the oven. Color coordinates were used to draw , which illustrates the color changes caused by each step according to the experimental conditions.
This study's results confirmed the hypotheses that neither endodontic access approach nor sealer composition affect tooth color. Tooth discoloration (darkening and yellowing) after endodontic treatment can result in patient dissatisfaction with the treatment in esthetic areas, negatively impacting quality of life. , Therefore, establishing clinical protocols and materials that prevent significant and undesirable tooth color changes might be as important to treatment success as eliminating the infection in root canals. A factor associated with tooth discoloration is residual filling material ( i.e. , sealers and gutta-percha) in the pulpal chamber. Hence, in the present study, to evaluate the influence of type of cavity and filling material on the color stability of the crown, other variables, such as saliva and food residue, were considered constant. Some studies have suggested that using MEC can result in more filling material remnants in the pulp chamber, including areas below the pulpal chamber roof. , However, the micro-CT analysis did not demonstrate any significant effect of the endodontic access protocols evaluated in the present study on the volume of remaining filling material in the pulpal chamber. Moreover, the sealer used had no effect. The use of assistive technologies, such as magnification by operating microscope and ultrasound, along with the operator's experience in the present study, facilitates the complete removal of the filling material and can explain the results. , It is noteworthy that the operating microscopy improves visualization and allows the operator to access difficult areas with the ultrasonic insert, which breaks and extracts the filling material by acoustic transmission. , On the other hand, the highest occurrence of voids surrounding the restoration was observed for the obturation using Bio-C Sealer in MEC. Bio-C Sealer contains high free CaO content, low concentrations of C 3 S (tricalcium silicate) and C 2 S (dicalcium silicate), and a long-chain polymer. This composition favors crystal formation by hydration reactions, which impairs crystal interlacing. , This disrupted crystal structure in Bio-C Sealer could explain the higher number of voids observed, which might weaken the bond between the composite filling and the dentin walls of the cavity. This weakening effect could be more pronounced with higher shrinkage stress in the composite. With MEC, the ratio between bonded and non-bonded areas of the restoration (the so-called C-factor) tends to be higher than with CEC. Consequently, higher shrinkage stress of the composite is expected, leading to a more pronounced effect of the presence of Bio-C Sealer. While this explanation is a possibility, additional studies are needed to definitively confirm the link between Bio-C Sealer and increased voids, particularly in the context of MEC. Regarding tooth discoloration, the present findings showed that the obturation of the root canal caused the specimen's zx yellowing. For specimens filled with Bio-C Sealer, however, the tooth color after restoring the endodontic access cavities was similar to that measured at baseline. Besides, the 1-year storage resulted in the most severe yellowing for both sealers. For the epoxy resin-based sealer AH Plus, tooth yellowing may be related to its chemical interaction with the dentin collagen matrix. The high flowability of this sealer improves its penetration into the dentinal tubules favoring some color change of the dentin. It is also worth noting that thermoset polymers, including epoxy resins, are susceptible to physical aging over time. Physical aging relies on changes in polymer structure towards a thermodynamic equilibrium, which is ultimately followed by a polymer color change. Moreover, the spatial geometric configuration of the epoxy molecule is prone to water absorption due to its open structure and the presence of cross-linking agents. Therefore, a gradual color change can be expected due to the degradation of curing initiators over time. Regarding the Bio-C Sealer, the discoloration mechanism is associated with sealer oxidation. During its setting, the calcium ion released by the hydration reaction reacts with phosphate ions and plasma proteins from the dentinal fluid, resulting in pigmented subproducts, such as calcium phosphate. In addition, the iron oxide from the radiopacifying agent is reduced to aluminoferrite during the hydration process of bioceramic sealers, modifying the color of the sealer from white to yellow-grey. Finally, as the sealer sets, hydrated compounds such as hydrated calcium silicates (an amorphous structure) and calcium hydroxide are formed, which crystallize in superposed hexagonal plates. As the sealer becomes more porous and chemically reactive, it can absorb higher content of organic and inorganic components (e.g., necrotic tissue, smear layer, and blood cells), favoring color changes over time. It is important to emphasize that both sealers contain zirconium oxide as a radiopacifying agent that degrades in the presence of sodium hypochlorite and EDTA remaining in the dentinal tubules. , – The degradation of this agent also contributes to tooth discoloration. The findings of this study highlight the clinical relevance of minimally invasive endodontic access, demonstrating that it is not directly associated with tooth discoloration when using bioceramic or epoxy resin-based root canal sealers. This suggests that dentists can adopt minimally invasive techniques without increased risk of esthetic compromise due to discoloration. However, it is important to acknowledge the limitations of this study, as it was conducted in vitro. The controlled environment of an in vitro study cannot fully replicate the complex conditions encountered in clinical practice, such as variations in oral flora, patient-specific factors, and long-term outcomes. Future research should aim to validate these results in clinical trials to better understand the implications of minimally invasive endodontic access in real-life scenarios. Residues of filling materials on the dentin surface may lead to tooth crown darkening over time. This occurrence was not dependent on the root canal sealer composition (epoxy resin-based or bioceramic-based) and the type of endodontic access cavity (minimally invasive or conservative). The physicochemical changes in sealers are responsible for tooth discoloration, impacting the esthetics of the patient's smile. Thus, protocols to effectively remove the endodontic sealer remnants from the pulpal chamber should be proposed to minimize this clinical effect.
A minimally invasive endodontic access does not affect the tooth color in the presence of either bioceramic or epoxy resin-based sealers. Irrespective of the sealer, the yellowing of specimens was observed after the obturation of the root canal. The restorative procedure only compensated this color change for specimens obturated with Bio-C Sealer. The more pronounced color changes occurred after a 1-year storage of the specimens in distilled water.
|
Polyomaviruses After Allogeneic Hematopoietic Stem Cell Transplantation | 1f7fa8a7-c877-456c-b683-e28ef2f14083 | 11946477 | Surgery[mh] | Polyomaviruses (PyVs) are ubiquitous and can infect human and other animal hosts. The best described PyVs are BK polyomavirus (BKPyV) and JC polyomavirus (JCPyV), which were named after patients with BKPyV-associated ureteric obstruction after kidney transplantation and progressive multifocal leukoencephalopathy (PML) after treatment of lymphoma, respectively . After allogeneic hematopoietic stem cell transplant (allo-HSCT), PyVs can replicate asymptomatically or cause morbid disease. BKPyV can cause hemorrhagic cystitis (BKPyV-HC) in allo-HSCT recipients, and although JCPyV has been commonly associated with PML in other patient populations (classically among those with HIV/AIDS or patients receiving specific immunomodulators), PML has also been reported after HSCT. In this review, we will primarily focus on BKPyV-associated diseases as BKPyV-HC is uniquely associated with allogeneic HSCT, and to a lesser extent JCPyV-associated disease; we will also briefly discuss reports of other PyVs causing disease following allo-HSCT. Due to a paucity of data, some information is extrapolated from other immunocompromised hosts such as kidney transplant recipients. We will refer to detection of PyVs in the blood and urine as “DNAemia” and “DNAuria”, respectively. Human polyomaviruses (HPyVs) have been linked with transformation to malignancy , including the association between BKPyV with bladder and urothelial carcinoma [ , , , ], Merkel cell PyV (MCPyV) as a cause of Merkel cell carcinoma , and JCPyV with a several solid tumors . For the purposes of this review, we will not discuss the association between HPyVs and malignancy, but this topic has been reviewed elsewhere [ , , ]. Virology of Polyomaviruses PyVs are non-enveloped double stranded DNA (dsDNA) viruses within the family Polyomaviridae that infect humans and animals . BKPyV and JCPyV share 70–75% identity across their genomes and are closely related to SV-40. The genome of all three viruses contains two transcriptional units separated by a non-coding control region (NCCR) that contains the origin of replication and promoters/enhancers for early and late gene expression. One transcriptional unit encodes the early genes: large T antigen, small t antigen, and T antigen splice variants (depending on which PyV). The other transcriptional unit encodes the late genes: three structural proteins (VP1, VP2, and VP3), and agnoprotein, and two micro RNAs that downregulate large T antigen . Rearrangements in NCCR sequences occur in vivo and are thought to be responsible for or associated with progression to some PyV-diseases (e.g., PML and BKPyV-Associated Nephropathy [BKPyVAN] after kidney transplant) but findings are mixed in others (e.g., BKPyV-HC) [ , , , ]. Specific mechanisms of cell entry and viral replication within the cell have been reviewed extensively elsewhere . In renal or other tissues, a cross-reacting antibody to SV40-LTag can identify either BKPyV or JCPyV and virus-specific nucleic acid amplification testing (NAAT) is required to distinguish between them . In HSCT recipients, the diagnosis of BKPyV and JCPyV replication relies on detection of replication using NAAT assays obtained from cerebrospinal fluid (CSF), serum, or urine. Importantly, since there is inter-assay variability (due to differences in amplicon size, sample type, target, detection of unencapsidated DNA, etc.) and challenges with the WHO international standard, absolute levels of DNAuria (viruria) or DNAemia (viremia) cannot be directly compared across assays, which has limited the ability to establish diagnostic criteria that are based on absolute DNA load thresholds [ , , , ]. Methods to increase precision and compare results between assays are a major focus of ongoing work, including the ability to distinguish replicating virus from DNA shedding [ , , , ].
PyVs are non-enveloped double stranded DNA (dsDNA) viruses within the family Polyomaviridae that infect humans and animals . BKPyV and JCPyV share 70–75% identity across their genomes and are closely related to SV-40. The genome of all three viruses contains two transcriptional units separated by a non-coding control region (NCCR) that contains the origin of replication and promoters/enhancers for early and late gene expression. One transcriptional unit encodes the early genes: large T antigen, small t antigen, and T antigen splice variants (depending on which PyV). The other transcriptional unit encodes the late genes: three structural proteins (VP1, VP2, and VP3), and agnoprotein, and two micro RNAs that downregulate large T antigen . Rearrangements in NCCR sequences occur in vivo and are thought to be responsible for or associated with progression to some PyV-diseases (e.g., PML and BKPyV-Associated Nephropathy [BKPyVAN] after kidney transplant) but findings are mixed in others (e.g., BKPyV-HC) [ , , , ]. Specific mechanisms of cell entry and viral replication within the cell have been reviewed extensively elsewhere . In renal or other tissues, a cross-reacting antibody to SV40-LTag can identify either BKPyV or JCPyV and virus-specific nucleic acid amplification testing (NAAT) is required to distinguish between them . In HSCT recipients, the diagnosis of BKPyV and JCPyV replication relies on detection of replication using NAAT assays obtained from cerebrospinal fluid (CSF), serum, or urine. Importantly, since there is inter-assay variability (due to differences in amplicon size, sample type, target, detection of unencapsidated DNA, etc.) and challenges with the WHO international standard, absolute levels of DNAuria (viruria) or DNAemia (viremia) cannot be directly compared across assays, which has limited the ability to establish diagnostic criteria that are based on absolute DNA load thresholds [ , , , ]. Methods to increase precision and compare results between assays are a major focus of ongoing work, including the ability to distinguish replicating virus from DNA shedding [ , , , ].
BKPyV-associated clinical syndromes are more common than PML following allo-HSCT. Among BKPyV-associated syndromes, hemorrhagic cystitis (BKPyV-HC) is best characterized, however, analogous to JCPyV, BKPyV DNAuria and DNAemia are common. 2.1. Clinical Presentation of BKPyV Syndromes and Incidence After HSCT a. Asymptomatic DNAuria BKPyV DNAuria after HSCT is common . In studies of HSCT recipients who undergo regular monitoring, BKPyV DNAuria increases from <10% before HSCT to 50–100% after transplant, typically identified between 1–8 weeks post-HSCT [ , , , ]. Progression to BKPyV-HC occurs in a minority of patients with DNAuria but is associated with high urinary BKPyV loads [ , , , ]. b. Asymptomatic DNAemia In studies that prospectively monitored patients after allo-HSCT, BKPyV DNAemia was reported in 50–80% of pediatric patients and 20–60% of adults, with a median onset of 30–40 days post-HSCT [ , , , ]. Reactivation and detection of dsDNA viruses, including BKPyV, were independently associated with increased risk for both early and late mortality after accounting for immune reconstitution, acute graft versus host disease (GVHD) severity, and steroid use ; the majority of virus reactivations occurred without clinical symptoms. c. BKPyV-associated hemorrhagic cystitis: The incidence of BKPyV-HC in patients after allo-HSCT patients varies across studies; incidence after allogeneic HSCT in adults is 16% (range, 7–54%) [ , , , , ] and 18% (range, 7–25%) in children [ , , ]. The variation in observed incidence may be due to differences in study populations, conditioning regimens, and diagnostic criteria followed by each study. BKPyV-HC is generally characterized by hematuria, symptoms of UTI, and detection of BKPyV in urine. Other potential etiologies of cystitis should be ruled out, which can make diagnosis difficult in the presence of co-pathogens, which are common . Hematuria is graded according to the Bedi et al. classification : Grade 1: Microscopic hematuria Grade 2: Macroscopic hematuria Grade 3: Macroscopic hematuria with clots Grade 4: Macroscopic hematuria requiring instrumentation. Stricter consensus diagnostic criteria were developed by ECIL that do not include Bedi grade 1 (microscopic) hematuria in the case definition: Clinical symptoms/signs of cystitis, such as dysuria and lower abdominal pain; Hematuria of grade 2 or higher; Demonstration of BKPyV DNAuria, with viral loads exceeding 7 log10 copies/mL . Irritative bladder symptoms most commonly include dysuria, frequency, or pelvic discomfort [ , , ]. BKPyV-HC typically develops between 2 and 8 weeks post-hematopoietic cell transplantation, with a range from 1 week to 6 months . The time to symptom resolution varies among studies. One retrospective analysis reported that the median time for macroscopic hematuria resolution was 17 days (range: 10–30 days), while the resolution of all symptoms, including cystitis, occurred at a median of 24 days (range: 15–44 days). In multivariable models, a high plasma viral load (≥10,000 copies/mL) and cytopenias at the onset of BKPyV-HC were significantly associated with prolonged macroscopic hematuria and cystitis symptoms . Severe BKPyV-HC is characterized by hematuria and blood clots that can obstruct bladder or ureteral flow, causing obstructive kidney injury or requiring surgical intervention; the incidence of severe BKPyV-HC varies across studies, between 32–59% for HC grade 3 or higher . The diagnosis of BKPyV-HC is associated with longer hospitalizations, more transfusions, and higher healthcare cost . Importantly, some data indicate that the duration and severity of HC symptoms have no significant impact on progression-free survival , but results are mixed . BKPyV loads in urine are significantly higher than in blood among patients with BKPyV-HC and not all cases of BKPyV-HC are accompanied by BKPyV DNAemia . Relative to patients with asymptomatic DNAuria or DNAemia, patients with BKPyV-HC of any grade of hematuria have significantly higher BKPyV loads in urine or blood [ , , , , ]; limited data suggest that BKPyV plasma loads are similar between different grades of hematuria . BKPyV plasma loads of >10,000 copies/mL have been associated with higher BKPyV-HC severity and longer duration [ , , , , , ]. Since BKPyV replication can be detected a median of 8 days before the onset of BKPyV-HC , prospective monitoring has been proposed as a strategy to identify high-risk patients. If a future effective treatment is identified, monitoring could identify a high-risk group of patients who could benefit from pre-emptive intervention. Cesaro et al. conducted a prospective study of 107 pediatric HSCT recipients and demonstrated that plasma BKPyV-DNAemia monitoring outperformed urine BKPyV load in predicting BKPyV-HC. A plasma BKPyV DNA load of 10 3 copies/mL exhibited 100% sensitivity, 86% specificity, a negative predictive value (NPV) of 100%, and a positive predictive value (PPV) of 39% for subsequent HC. In contrast, a urine BKPyV-DNA load > 10⁷ copies/mL showed 86% sensitivity, 60% specificity, an NPV of 98%, and a PPV of 14% for HC46. Similar findings have been reported in other pediatric studies [ , , , ]. Small studies in adults have also suggested that plasma monitoring of BKPyV-DNA may identify patients at risk of HC46,47,49. In one study, the presence of BKPyV-DNA in serum on day 21 post-HSCT (>0.75 × 10 3 BKPyV copies/mL) was a statistically significant risk factor for HC and survival outcomes50. Another study demonstrated that plasma BKPyV PCR titters on days 0, 30, and 60 post-transplant were sensitive tools for predicting clinically significant HC51. A recent retrospective study demonstrated that high BKPyV viremia is associated with an early decline in renal function and worse survival . Current guidelines from the European Conference on Infections in Leukemia (ECIL) do not recommend routine screening for BKPyV virus in the urine or blood in asymptomatic patients. BKPyV DNAuria is common in this population and even though the presence and degree of BKPyV DNAemia is linked with BKPyV-HC, detection of DNAemia is not sensitive for BKPyV-HC . Furthermore, the lack of effective prevention or management strategies limits the actionability of routine surveillance. d. Non-hemorrhagic cystitis syndromes associated with BKPyV Cases of BKPyV non-hemorrhagic cystitis have been reported in non-allo-HSCT recipients as a diagnosis of exclusion when no other urinary pathogens were identified. BKPyV cystitis without macroscopic hematuria, either with microscopic (grade 1) hematuria or without documented hematuria at all, are included in observational studies of BKPyV disease after allo-HSCT. For example, in a prospective study of 99 HSCT patients diagnosed with BKPyV disease, only 38% reported hematuria, while cystitis symptoms such as increased urinary frequency and dysuria were observed in 88% and 63% of cases, respectively . Similarly, another study on BKPyV-associated cystitis identified dysuria as the most common symptom in 88.5%, followed by hematuria in 79% . Microscopic (grade 1) hematuria has been reported in 5.7% to 30.2% of cases across various studies [ , , ]. Along the same lines, a subset of retrospective study of 128 allo-HSCT recipients who ultimately developed BKPyV-HC at grade 2 or above found no statistically significant difference in viral load at time points of cystitis symptoms alone, cystitis with macroscopic hematuria, and cystitis with hematuria and clots . In many patients, cystitis symptoms preceded hematuria . However, the overall incidence of BKPyV-associated cystitis without micro or macroscopic hematuria following allo-HSCT remains unknown. In immunocompetent patients with interstitial cystitis, the potential contribution of PyV has been proposed based on higher PyV detection compared to matched controls, however PyV detection is inconsistent in biopsy specimens and this association requires confirmation in larger studies [ , , , , ]. e. BKPyV-associated nephropathy BKPyVAN is the major complication of BKPyV replication after kidney transplant recipients (KTR) and is best characterized in this setting. In KTRs, BKPyV DNAemia is thought to be a result of BKPyV replication and cell lysis within the allograft . Definitive diagnosis relies on allograft biopsy with demonstration of SV-40 staining, but high-level BKPyV DNAemia is clinically used as a tool for presumptive diagnosis [ , , ]. BKPyVAN after KTR is a major cause of allograft dysfunction and loss . Unlike after KTR, renal dysfunction among patients with BKPyV-HC is commonly caused by obstructive renal injury . However, BKPyVAN after allo-HSCT has been reported [ , , , ], and renal dysfunction has been observed among HSCT recipients with BKPyV DNAemia even without known BKPyV-HC-related obstruction [ , , , , ]. Several studies have demonstrated long-term renal dysfunction associated with BKPyV DNAemia > 10,000 copies/mL, a threshold that is associated with BKPyVAN after KT [ , , , , ] Since renal biopsy is rarely feasible or safe in the post-HSCT period, the true incidence of BKPyVAN after HSCT, or whether or not it occurs concomitantly with BKPyV-HC, is unknown. f. Cases of BKPyV replication linked to clinical disease Other potential cases of BKPyV replication have occurred in patients with encephalitis, reported in allo-HSCT recipients with high BKPyV loads in CSF [ , , , , , ], and pneumonia diagnosed based on demonstration of viral inclusion bodies and SV-40 staining from lung tissue . Some of these reported cases have occurred in patients with other concurrent manifestations of BKPyV such as HC or PyVAN. In vitro studies have demonstrated the potential for BKPyV to infect brain endothelial cells, ependymal cells, and astrocytes; however, the incidence of these syndromes, or whether BKPyV replication is a cause or a bystander in all of these cases is unknown [ , , , ]. The presence of BKPyV-HC has been statistically significant risk factor associated with complement-associated thrombotic microangiopathy, with a hazard ratio (HR) of 2.55 demonstrated in adults . 2.2. Pathogenesis of BKPyV Syndromes After HSCT Analogous to JCPyV, primary infection with BKPyV occurs early in life and persists asymptomatically within the renourinary tract, including renal tubular epithelial cells, renal glomerular cells, and potentially bladder microvascular endothelial cells . BKPyV replication in urine or plasma occurs intermittently in immunocompetent patients but increases following HSCT-associated immunosuppression, resulting in BKPyV-associated disease [ , , , , , ]. a. Transmission and persistence Approximately 90% of BKPyV transmission is thought to occur during childhood through respiratory or oral transmission [ , , ]. This hypothesis was supported by BKPyV seroconversion among 7 children at the time of an upper respiratory infection and detection of BKPyV DNA in respiratory tract or tonsillar tissues in other studies . From the respiratory tract, the virus enters the bloodstream, infects peripheral blood leukocytes, and subsequently disseminates to various tissues . Once acquired, BKPyV persists predominantly in urothelial cells although it has also been identified in lymphocytes and monocytes . Although most BKPyV disease after allo-HSCT is thought to be a result of reactivation of previously acquired disease, there are reports of nosocomial BKPyV transmission [ , , ]. By adulthood, more than 80% of the population is seropositive . b. BKPyV-HC Determining the precise relationship between BKPyV and clinical disease or identifying the pathophysiology is hampered by the lack of histological data and lack of a precise definition of clinical disease. There are several theories of how BKPyV replication after HSCT results in BKPyV-HC. The first is that immunosuppression causes unchecked BKPyV replication, which has direct cytopathic effect on urothelial cells and mucosa ( ). Viral replication is particularly prominent in urothelial cells already damaged by prior therapies, such as cyclophosphamide. This replication, combined with the reduced activity of cytotoxic T cells, increases the risk of uncontrolled viral replication . This theory is supported by studies demonstrating a relationship between viral load and HC severity [ , , ] and studies demonstrating a relationship between lack of immune reconstitution and more severe or longer duration of BKPyV- HC . Another theory suggests that immune reconstitution exacerbates mucosal damage by reacting to replicating virus or viral-induced changes to urothelial epithelium, potentiating further damage to bladder and urothelial tissues . Leung et al. proposed a three-phase model for the development of BKPyV-associated HC. In the first phase, the conditioning regimen damages the bladder mucosa, creating a favorable environment for BKPYV replication. In the second phase, viral replication becomes unchecked due to impaired immunity. In the third phase, immune reconstitution and the return of anti-BKPyV immunity cause additional damage to the bladder mucosa ( ). c. Future directions of study Several questions remain regarding BKPyV-HC pathogenesis. It is unknown whether BKPyV replication detected in urine or blood represents replication in urothelial tissue, kidney tissue, or a combination of both , whether BKPyV causes nephropathy in the native kidneys of HSCT recipients even in the absence of HC [ , , ], or how the type of immunosuppressive procedure (allo-HSCT vs KT) so strongly impacts the clinical manifestations of BKPyV replication (BKPyV-HC vs BKPyVAN). In BKPyV-HC the relationship between BKPyV coming from renal epithelium vs bladder epithelium remains unclear and, in contrast to BKPyVAN after KT, some studies have observed a nonlinear relationship between urine and plasma BKPyV load . Similarly, the relationship between immune reconstitution and the occurrence and severity of BKPyV-HC has not been clearly established. Lastly, some definitions of BKPyV-HC include cystitis with microscopic hematuria and some definitions require Bedi grade ≥ 2 hematuria. It is unknown whether BKPyV DNAuria, cystitis symptoms, and microscopic hematuria represent a more mild form of BKPyV-HC or separate syndrome entirely, particularly as the degree of BKPyV DNAemia may be similar across groups of patients with cystitis symptoms regardless of the presence of macroscopic hematuria or clots . 2.3. Risk Factors for BKPyV-HC Several factors associated with BKPyV-HC have been identified, which are often markers of higher severity of immunosuppression or lack of immune reconstitution. These risk factors include: male sex, haploidentical, matched unrelated donors [ , , , ], myeloablative conditioning [ , , ], acute/chronic GVHD , the presence of cytomegalovirus (CMV) reactivation or initiation of CMV treatment , CMV DNAuria , older age, use of lymphocyte depletion such as ATG, alemtuzumab, or T cell depleted grafts [ , , , , ]. The subtype of BKPyV does not appear to be a risk factor for developing BKPyV-HC . In the pediatric population, risk factors are similar to the adult patients including matched unrelated donors, a prior diagnosis of acute myeloblastic leukemia, and GVHD . Post-Transplant Cyclophosphamide Post-transplant cyclophosphamide (PTCy) has been increasingly adopted as GVHD prophylaxis following allo-HSCT and has been identified as a risk for BKPyV-HC . A study presented at the American Society of Hematology in 2024 reported an incidence of 72% at one hundred days post-HSCT, with a median onset of 30 days post-transplant. Notably, reducing the PTCy dose did not decrease the risk of BKPyV-HC but delayed its onset and reduced its duration . Another retrospective study of patients receiving PTCy reported a median time to BKPyV-HC diagnosis of 29 days post-HSCT, with JCPyV coinfection detected in 24% of patients and cytomegalovirus DNAuria in 17%. Additionally, a higher prevalence of grade 3–4 HC was observed in younger patients and those who had a haploidentical donor . The exact mechanisms by which PTCy elevates the risk are not fully understood. One hypothesis suggests that PTCy may cause a delay in immune reconstitution, affecting T cell recovery , leading to prolonged immunosuppression and creating an environment conducive to viral reactivation. Additionally, PTCy may induce direct urothelial toxicity, secondary to acrolein, a toxic metabolite of Cy , which could facilitate viral replication in the urinary tract . 2.4. Relationship of BKPyV DNAemia and HC with Immune System Like other post-allo-HSCT viruses (e.g., Cytomegalovirus, Adenovirus), cellular immunity is thought to play a major role in the control of BKPyV. Viral variations in VP1 and LTag may escape control by neutralizing antibodies and T cells . Sustained BK virus (BKPyV) DNAemia has been associated with impaired recovery of CD4+ and CD8+ T-cell subsets, possibly due to control by regulatory T cells [ , , , ]. Similarly, reconstitution of CD4+ and CD8+ cells is associated with BKPyV clearance in HSCT and resolution of BKPyV-HC symptoms is associated with BKPyV-specific T cell responses . At day 100, higher ALC, CD3, and CD8 counts were associated with lower hazard of BKPyV DNAuria in multivariable models . Espada et al. characterized the recovery of BK virus–specific T-cell immunity in 77 adult patients with urinary symptoms after HSCT, comparing those with and without BKPyV DNAuria. Patients with BKPyV DNAuria had delayed CD4 T-cell recovery post-transplant but had faster recovery of BK virus–specific Th1 CD4 T cells, which were more frequent than cytolytic CD8 T cells. Among those with BKPyV DNAuria, patients with early reconstitution of BKPyV-specific interferon-γ+ and cytolytic CD4 T cells was linked to lower hematuria rates, highlighting a potential role in prevention of symptoms . Humoral immunity likely also contributes to control of BKPyV replication; however, studies have not found a clear relationship between the presence of BKPyV-antibodies (which are not routinely obtained) and BKPyV-HC . Studies in seropositive patients have demonstrated an association between high BKPyV IgG titers with subsequent BKPyV DNAuria, although this was not compared to seronegative patients . Limited information exists regarding BKPyV syndromes in patients who were seronegative patients prior to transplantation. Among six children with BKPyV-HC after allo-HSCT, BKPyV viremia decreased as IgM levels initially increased, followed by IgG, suggesting the role of the humoral response in recovery . 2.5. Management of BKPyV The two major approaches to BKPyV disease include symptom management and antiviral strategies intended to prevent or reduce BKPyV replication. There are no high-quality or randomized data to support any specific prophylaxis or management strategy for BKPyV-HC or other BKPyV-associated diseases. Among the therapies studied, intravenous (IV) and intravesical cidofovir were the most reported strategy, but no clear benefit was demonstrated, and the level of evidence has been graded low . a. Supportive measures Without significant data to support an antiviral strategy, supportive care is recommended for BKPyV-HC (grade AIII) . Supportive measures include hydration, platelet transfusions, and analgesics . HBO therapy has been explored in case series123 and case reports122,124,125. However, there is no good quality data to support its use. b. Antiviral strategies While the cornerstone of BKPyVAN management after kidney transplantation is reduction in immunosuppression (RIS), this is frequently not feasible after allo-HSCT due to the threat of donor alloreactivity and GVHD . Several antivirals have been proposed or studied, including fluoroquinolones, leflunomide, cidofovir (IV or intravesical), brincidofovir, and virus-specific T cell (VST) therapy. 2.5.1. Fluoroquinolones Although quinolone antibiotics were initially thought to have modest anti-BKPyV activity in vitro and potential benefit in observational studies of BKPyV-HC after HSCT [ , , ], three RCTs in KTRs have not shown antiviral effect either as prevention of BKPyV DNAuria or DNAemia. In one RCT of 3 months of levofloxacin vs placebo, BKPyV DNAuria was observed in 22/76 patients (29%) in the levofloxacin group and 26/78 patients (33.3%) in the comparison group . In a second RCT, BKPyV viremia occurred in 25/133 patients (18.8%) in the ciprofloxacin group compared to 5/67 patients (7.5%) in the placebo group ( p = 0.03) . Lastly, among KTRs with BKPyV DNAemia the percentage of patients with BKPyV load reduction were 70.3% and 69.1% in the levofloxacin group (n = 22) and the placebo group (n = 21), respectively ( p = 0.93) . Several studies have also identified a higher risk of quinolone-resistant bacterial infections [ , , ]. As a result, quinolones are not recommended for BKPyV treatment or prevention either after KT or HSCT [ , , ]. 2.5.2. Leflunomide Leflunomide is an antimetabolite drug with immunomodulatory and antiviral activity; its use has been supported by case reports and small retrospective studies [ , , ]. However, a phase II study of a leflunomide derivative did not show substantial benefit in the treatment of BKPyVAN after KT as treatment in 30 patients led to a slightly greater reduction in urine BKPyV DNA load (3.1 log10 copies/μL in treatment group vs. 2.8 log10 copies/μL with placebo) and a small statistically significant reduction in viremia (1.9 log10 copies/μL in treatment group vs. 1.3 log10 copies/μL in placebo group, p = 0.049) ; leflunomide is not currently recommended . 2.5.3. Cidofovir Cidofovir is a nucleotide analog that inhibits a broad range of DNA viruses in vitro, although BKPyV does not possess its own polymerase. Its active metabolite has a long half-life of 15–65 h, allowing for administration at weekly intervals . Intravenous (IV) or intravesical cidofovir have been commonly used for BKPyV-HC, but there is no consensus on the optimal dose, route, or frequency of administration . When given IV, doses ranging from 3–5 mg/kg/week with probenecid, or lower doses of 0.5–1.5 mg/kg/week without probenecid are typically given . To mitigate the risk of ocular and nephrotoxicity, cidofovir may be administered intravesically, although the potential for systemic exposure remains . Small retrospective studies, case series, and case reports (without a comparator group) have suggested varying clinical response rates both intravenous (IV) and intravesical cidofovir in the treatment of BK polyomavirus-associated hemorrhagic cystitis (BKPyV-HC). Clinical response was generally defined as complete symptom resolution, while partial response (PR) referred to significant symptom improvement with persistent hematuria, though definitions varied across studies. Study sizes ranged from 4 to 57 patients in both pediatric and adult populations, with complete response (CR) rates between 60–85% for IV cidofovir and 59–92% for intravesical administration. A systematic review reported a CR in 117 of 172 patients (68%) treated with IV cidofovir and 15 of 17 patients (88.2%) treated with intravesical cidofovir [ , , , , , , , , , , , , , , , ]. However, there was inconsistency in the reporting and definitions of virologic responses, and other studies have demonstrated no reductions in BKPyV DNAemia associated with cidofovir administration [ , , ]. Furthermore, a small randomized, placebo-controlled phase I/II trial of low-dose IV cidofovir given for treatment of BKPyVAN after KT did not significantly reduce BKPyV DNAemia . 2.5.4. Brincidofovir Brincidofovir (CMX001) is a lipid-conjugated prodrug of cidofovir with in vitro activity against a variety of double-stranded DNA viruses, including BKPyV. Importantly, brincidofovir is associated with lower rates of renal dysfunction compared to cidofovir . Several case reports suggest potential benefits of this medication in managing BKPyV-associated conditions [ , , ]. However, in a phase III study of brincidofovir for CMV prophylaxis after allo-HSCT, the percentage of patients with BKPyV infection was not different in the treatment vs placebo arms. Brincidofovir is not currently available . 2.5.5. Virus-Specific T-Cells (VSTs) VSTs are a targeted therapy used to treat viral infections in immunocompromised patients. Various approaches have been developed, including ex-vivo expansion of donor-derived VSTs or banking VSTs from healthy donors with diverse haplotypes for immediate infusion . To address the multiple viral infections that affect immunocompromised patients, multi-virus VSTs have been developed and studied. Although adoptive transfer of VSTs is generally safe, the complexity of preparing these cells and their limited antiviral range have posed significant hurdles. Case reports and series have reported favorable clinical outcomes including reduction in viral load and symptomatic improvement [ , , , , ]. Two single arm, uncontrolled studies evaluated safety and efficacy of VSTs targeted against BKPyV. One study included 38 HSCT recipients and 3 SOT recipients with BKPyV DNAemia or BKPyV-HC; among patients with cystitis, 31/38 achieved complete response (CR); when assessing both BKPyV PCR levels and cystitis response, 22/38 patients achieved a CR; CR was defined as undetectable plasma BKPyV PCR within four weeks after the final VST infusion or resolution of hemorrhagic cystitis to below grade 2 without the need for medical therapy. No infusion-related toxicity, de novo GVHD, or organ rejection were found . Another study of third-party BKPyV-specific cytotoxic T lymphocytes for patients with BKPyV-HC after allo HSCT demonstrated that by day 21, 33/56 patients achieved CR (defined as full resolution of symptoms and gross hematuria or a reduction in hemorrhagic cystitis (HC) grade from 2, 3, or 4 to 0 or 1), while by day 45, 34 of 49 patients achieved CR . Post-infusion T cell detections have been observed and sustained for several months [ , , ]. However, these promising results need to be confirmed in RCTs. Posoleucel, an investigational multivirus-specific T-cell therapy, was found to be safe, well tolerated, and associated with greater reductions in BK viremia compared to placebo in a phase two study of 61 kidney transplant recipients . However, a phase three study evaluating posoleucel for preventing viral infections after allo-HSCT, treating virus-associated hemorrhagic cystitis, and adenovirus were terminated early due to futility .
a. Asymptomatic DNAuria BKPyV DNAuria after HSCT is common . In studies of HSCT recipients who undergo regular monitoring, BKPyV DNAuria increases from <10% before HSCT to 50–100% after transplant, typically identified between 1–8 weeks post-HSCT [ , , , ]. Progression to BKPyV-HC occurs in a minority of patients with DNAuria but is associated with high urinary BKPyV loads [ , , , ]. b. Asymptomatic DNAemia In studies that prospectively monitored patients after allo-HSCT, BKPyV DNAemia was reported in 50–80% of pediatric patients and 20–60% of adults, with a median onset of 30–40 days post-HSCT [ , , , ]. Reactivation and detection of dsDNA viruses, including BKPyV, were independently associated with increased risk for both early and late mortality after accounting for immune reconstitution, acute graft versus host disease (GVHD) severity, and steroid use ; the majority of virus reactivations occurred without clinical symptoms. c. BKPyV-associated hemorrhagic cystitis: The incidence of BKPyV-HC in patients after allo-HSCT patients varies across studies; incidence after allogeneic HSCT in adults is 16% (range, 7–54%) [ , , , , ] and 18% (range, 7–25%) in children [ , , ]. The variation in observed incidence may be due to differences in study populations, conditioning regimens, and diagnostic criteria followed by each study. BKPyV-HC is generally characterized by hematuria, symptoms of UTI, and detection of BKPyV in urine. Other potential etiologies of cystitis should be ruled out, which can make diagnosis difficult in the presence of co-pathogens, which are common . Hematuria is graded according to the Bedi et al. classification : Grade 1: Microscopic hematuria Grade 2: Macroscopic hematuria Grade 3: Macroscopic hematuria with clots Grade 4: Macroscopic hematuria requiring instrumentation. Stricter consensus diagnostic criteria were developed by ECIL that do not include Bedi grade 1 (microscopic) hematuria in the case definition: Clinical symptoms/signs of cystitis, such as dysuria and lower abdominal pain; Hematuria of grade 2 or higher; Demonstration of BKPyV DNAuria, with viral loads exceeding 7 log10 copies/mL . Irritative bladder symptoms most commonly include dysuria, frequency, or pelvic discomfort [ , , ]. BKPyV-HC typically develops between 2 and 8 weeks post-hematopoietic cell transplantation, with a range from 1 week to 6 months . The time to symptom resolution varies among studies. One retrospective analysis reported that the median time for macroscopic hematuria resolution was 17 days (range: 10–30 days), while the resolution of all symptoms, including cystitis, occurred at a median of 24 days (range: 15–44 days). In multivariable models, a high plasma viral load (≥10,000 copies/mL) and cytopenias at the onset of BKPyV-HC were significantly associated with prolonged macroscopic hematuria and cystitis symptoms . Severe BKPyV-HC is characterized by hematuria and blood clots that can obstruct bladder or ureteral flow, causing obstructive kidney injury or requiring surgical intervention; the incidence of severe BKPyV-HC varies across studies, between 32–59% for HC grade 3 or higher . The diagnosis of BKPyV-HC is associated with longer hospitalizations, more transfusions, and higher healthcare cost . Importantly, some data indicate that the duration and severity of HC symptoms have no significant impact on progression-free survival , but results are mixed . BKPyV loads in urine are significantly higher than in blood among patients with BKPyV-HC and not all cases of BKPyV-HC are accompanied by BKPyV DNAemia . Relative to patients with asymptomatic DNAuria or DNAemia, patients with BKPyV-HC of any grade of hematuria have significantly higher BKPyV loads in urine or blood [ , , , , ]; limited data suggest that BKPyV plasma loads are similar between different grades of hematuria . BKPyV plasma loads of >10,000 copies/mL have been associated with higher BKPyV-HC severity and longer duration [ , , , , , ]. Since BKPyV replication can be detected a median of 8 days before the onset of BKPyV-HC , prospective monitoring has been proposed as a strategy to identify high-risk patients. If a future effective treatment is identified, monitoring could identify a high-risk group of patients who could benefit from pre-emptive intervention. Cesaro et al. conducted a prospective study of 107 pediatric HSCT recipients and demonstrated that plasma BKPyV-DNAemia monitoring outperformed urine BKPyV load in predicting BKPyV-HC. A plasma BKPyV DNA load of 10 3 copies/mL exhibited 100% sensitivity, 86% specificity, a negative predictive value (NPV) of 100%, and a positive predictive value (PPV) of 39% for subsequent HC. In contrast, a urine BKPyV-DNA load > 10⁷ copies/mL showed 86% sensitivity, 60% specificity, an NPV of 98%, and a PPV of 14% for HC46. Similar findings have been reported in other pediatric studies [ , , , ]. Small studies in adults have also suggested that plasma monitoring of BKPyV-DNA may identify patients at risk of HC46,47,49. In one study, the presence of BKPyV-DNA in serum on day 21 post-HSCT (>0.75 × 10 3 BKPyV copies/mL) was a statistically significant risk factor for HC and survival outcomes50. Another study demonstrated that plasma BKPyV PCR titters on days 0, 30, and 60 post-transplant were sensitive tools for predicting clinically significant HC51. A recent retrospective study demonstrated that high BKPyV viremia is associated with an early decline in renal function and worse survival . Current guidelines from the European Conference on Infections in Leukemia (ECIL) do not recommend routine screening for BKPyV virus in the urine or blood in asymptomatic patients. BKPyV DNAuria is common in this population and even though the presence and degree of BKPyV DNAemia is linked with BKPyV-HC, detection of DNAemia is not sensitive for BKPyV-HC . Furthermore, the lack of effective prevention or management strategies limits the actionability of routine surveillance. d. Non-hemorrhagic cystitis syndromes associated with BKPyV Cases of BKPyV non-hemorrhagic cystitis have been reported in non-allo-HSCT recipients as a diagnosis of exclusion when no other urinary pathogens were identified. BKPyV cystitis without macroscopic hematuria, either with microscopic (grade 1) hematuria or without documented hematuria at all, are included in observational studies of BKPyV disease after allo-HSCT. For example, in a prospective study of 99 HSCT patients diagnosed with BKPyV disease, only 38% reported hematuria, while cystitis symptoms such as increased urinary frequency and dysuria were observed in 88% and 63% of cases, respectively . Similarly, another study on BKPyV-associated cystitis identified dysuria as the most common symptom in 88.5%, followed by hematuria in 79% . Microscopic (grade 1) hematuria has been reported in 5.7% to 30.2% of cases across various studies [ , , ]. Along the same lines, a subset of retrospective study of 128 allo-HSCT recipients who ultimately developed BKPyV-HC at grade 2 or above found no statistically significant difference in viral load at time points of cystitis symptoms alone, cystitis with macroscopic hematuria, and cystitis with hematuria and clots . In many patients, cystitis symptoms preceded hematuria . However, the overall incidence of BKPyV-associated cystitis without micro or macroscopic hematuria following allo-HSCT remains unknown. In immunocompetent patients with interstitial cystitis, the potential contribution of PyV has been proposed based on higher PyV detection compared to matched controls, however PyV detection is inconsistent in biopsy specimens and this association requires confirmation in larger studies [ , , , , ]. e. BKPyV-associated nephropathy BKPyVAN is the major complication of BKPyV replication after kidney transplant recipients (KTR) and is best characterized in this setting. In KTRs, BKPyV DNAemia is thought to be a result of BKPyV replication and cell lysis within the allograft . Definitive diagnosis relies on allograft biopsy with demonstration of SV-40 staining, but high-level BKPyV DNAemia is clinically used as a tool for presumptive diagnosis [ , , ]. BKPyVAN after KTR is a major cause of allograft dysfunction and loss . Unlike after KTR, renal dysfunction among patients with BKPyV-HC is commonly caused by obstructive renal injury . However, BKPyVAN after allo-HSCT has been reported [ , , , ], and renal dysfunction has been observed among HSCT recipients with BKPyV DNAemia even without known BKPyV-HC-related obstruction [ , , , , ]. Several studies have demonstrated long-term renal dysfunction associated with BKPyV DNAemia > 10,000 copies/mL, a threshold that is associated with BKPyVAN after KT [ , , , , ] Since renal biopsy is rarely feasible or safe in the post-HSCT period, the true incidence of BKPyVAN after HSCT, or whether or not it occurs concomitantly with BKPyV-HC, is unknown. f. Cases of BKPyV replication linked to clinical disease Other potential cases of BKPyV replication have occurred in patients with encephalitis, reported in allo-HSCT recipients with high BKPyV loads in CSF [ , , , , , ], and pneumonia diagnosed based on demonstration of viral inclusion bodies and SV-40 staining from lung tissue . Some of these reported cases have occurred in patients with other concurrent manifestations of BKPyV such as HC or PyVAN. In vitro studies have demonstrated the potential for BKPyV to infect brain endothelial cells, ependymal cells, and astrocytes; however, the incidence of these syndromes, or whether BKPyV replication is a cause or a bystander in all of these cases is unknown [ , , , ]. The presence of BKPyV-HC has been statistically significant risk factor associated with complement-associated thrombotic microangiopathy, with a hazard ratio (HR) of 2.55 demonstrated in adults .
Analogous to JCPyV, primary infection with BKPyV occurs early in life and persists asymptomatically within the renourinary tract, including renal tubular epithelial cells, renal glomerular cells, and potentially bladder microvascular endothelial cells . BKPyV replication in urine or plasma occurs intermittently in immunocompetent patients but increases following HSCT-associated immunosuppression, resulting in BKPyV-associated disease [ , , , , , ]. a. Transmission and persistence Approximately 90% of BKPyV transmission is thought to occur during childhood through respiratory or oral transmission [ , , ]. This hypothesis was supported by BKPyV seroconversion among 7 children at the time of an upper respiratory infection and detection of BKPyV DNA in respiratory tract or tonsillar tissues in other studies . From the respiratory tract, the virus enters the bloodstream, infects peripheral blood leukocytes, and subsequently disseminates to various tissues . Once acquired, BKPyV persists predominantly in urothelial cells although it has also been identified in lymphocytes and monocytes . Although most BKPyV disease after allo-HSCT is thought to be a result of reactivation of previously acquired disease, there are reports of nosocomial BKPyV transmission [ , , ]. By adulthood, more than 80% of the population is seropositive . b. BKPyV-HC Determining the precise relationship between BKPyV and clinical disease or identifying the pathophysiology is hampered by the lack of histological data and lack of a precise definition of clinical disease. There are several theories of how BKPyV replication after HSCT results in BKPyV-HC. The first is that immunosuppression causes unchecked BKPyV replication, which has direct cytopathic effect on urothelial cells and mucosa ( ). Viral replication is particularly prominent in urothelial cells already damaged by prior therapies, such as cyclophosphamide. This replication, combined with the reduced activity of cytotoxic T cells, increases the risk of uncontrolled viral replication . This theory is supported by studies demonstrating a relationship between viral load and HC severity [ , , ] and studies demonstrating a relationship between lack of immune reconstitution and more severe or longer duration of BKPyV- HC . Another theory suggests that immune reconstitution exacerbates mucosal damage by reacting to replicating virus or viral-induced changes to urothelial epithelium, potentiating further damage to bladder and urothelial tissues . Leung et al. proposed a three-phase model for the development of BKPyV-associated HC. In the first phase, the conditioning regimen damages the bladder mucosa, creating a favorable environment for BKPYV replication. In the second phase, viral replication becomes unchecked due to impaired immunity. In the third phase, immune reconstitution and the return of anti-BKPyV immunity cause additional damage to the bladder mucosa ( ). c. Future directions of study Several questions remain regarding BKPyV-HC pathogenesis. It is unknown whether BKPyV replication detected in urine or blood represents replication in urothelial tissue, kidney tissue, or a combination of both , whether BKPyV causes nephropathy in the native kidneys of HSCT recipients even in the absence of HC [ , , ], or how the type of immunosuppressive procedure (allo-HSCT vs KT) so strongly impacts the clinical manifestations of BKPyV replication (BKPyV-HC vs BKPyVAN). In BKPyV-HC the relationship between BKPyV coming from renal epithelium vs bladder epithelium remains unclear and, in contrast to BKPyVAN after KT, some studies have observed a nonlinear relationship between urine and plasma BKPyV load . Similarly, the relationship between immune reconstitution and the occurrence and severity of BKPyV-HC has not been clearly established. Lastly, some definitions of BKPyV-HC include cystitis with microscopic hematuria and some definitions require Bedi grade ≥ 2 hematuria. It is unknown whether BKPyV DNAuria, cystitis symptoms, and microscopic hematuria represent a more mild form of BKPyV-HC or separate syndrome entirely, particularly as the degree of BKPyV DNAemia may be similar across groups of patients with cystitis symptoms regardless of the presence of macroscopic hematuria or clots .
Several factors associated with BKPyV-HC have been identified, which are often markers of higher severity of immunosuppression or lack of immune reconstitution. These risk factors include: male sex, haploidentical, matched unrelated donors [ , , , ], myeloablative conditioning [ , , ], acute/chronic GVHD , the presence of cytomegalovirus (CMV) reactivation or initiation of CMV treatment , CMV DNAuria , older age, use of lymphocyte depletion such as ATG, alemtuzumab, or T cell depleted grafts [ , , , , ]. The subtype of BKPyV does not appear to be a risk factor for developing BKPyV-HC . In the pediatric population, risk factors are similar to the adult patients including matched unrelated donors, a prior diagnosis of acute myeloblastic leukemia, and GVHD . Post-Transplant Cyclophosphamide Post-transplant cyclophosphamide (PTCy) has been increasingly adopted as GVHD prophylaxis following allo-HSCT and has been identified as a risk for BKPyV-HC . A study presented at the American Society of Hematology in 2024 reported an incidence of 72% at one hundred days post-HSCT, with a median onset of 30 days post-transplant. Notably, reducing the PTCy dose did not decrease the risk of BKPyV-HC but delayed its onset and reduced its duration . Another retrospective study of patients receiving PTCy reported a median time to BKPyV-HC diagnosis of 29 days post-HSCT, with JCPyV coinfection detected in 24% of patients and cytomegalovirus DNAuria in 17%. Additionally, a higher prevalence of grade 3–4 HC was observed in younger patients and those who had a haploidentical donor . The exact mechanisms by which PTCy elevates the risk are not fully understood. One hypothesis suggests that PTCy may cause a delay in immune reconstitution, affecting T cell recovery , leading to prolonged immunosuppression and creating an environment conducive to viral reactivation. Additionally, PTCy may induce direct urothelial toxicity, secondary to acrolein, a toxic metabolite of Cy , which could facilitate viral replication in the urinary tract .
Post-transplant cyclophosphamide (PTCy) has been increasingly adopted as GVHD prophylaxis following allo-HSCT and has been identified as a risk for BKPyV-HC . A study presented at the American Society of Hematology in 2024 reported an incidence of 72% at one hundred days post-HSCT, with a median onset of 30 days post-transplant. Notably, reducing the PTCy dose did not decrease the risk of BKPyV-HC but delayed its onset and reduced its duration . Another retrospective study of patients receiving PTCy reported a median time to BKPyV-HC diagnosis of 29 days post-HSCT, with JCPyV coinfection detected in 24% of patients and cytomegalovirus DNAuria in 17%. Additionally, a higher prevalence of grade 3–4 HC was observed in younger patients and those who had a haploidentical donor . The exact mechanisms by which PTCy elevates the risk are not fully understood. One hypothesis suggests that PTCy may cause a delay in immune reconstitution, affecting T cell recovery , leading to prolonged immunosuppression and creating an environment conducive to viral reactivation. Additionally, PTCy may induce direct urothelial toxicity, secondary to acrolein, a toxic metabolite of Cy , which could facilitate viral replication in the urinary tract .
Like other post-allo-HSCT viruses (e.g., Cytomegalovirus, Adenovirus), cellular immunity is thought to play a major role in the control of BKPyV. Viral variations in VP1 and LTag may escape control by neutralizing antibodies and T cells . Sustained BK virus (BKPyV) DNAemia has been associated with impaired recovery of CD4+ and CD8+ T-cell subsets, possibly due to control by regulatory T cells [ , , , ]. Similarly, reconstitution of CD4+ and CD8+ cells is associated with BKPyV clearance in HSCT and resolution of BKPyV-HC symptoms is associated with BKPyV-specific T cell responses . At day 100, higher ALC, CD3, and CD8 counts were associated with lower hazard of BKPyV DNAuria in multivariable models . Espada et al. characterized the recovery of BK virus–specific T-cell immunity in 77 adult patients with urinary symptoms after HSCT, comparing those with and without BKPyV DNAuria. Patients with BKPyV DNAuria had delayed CD4 T-cell recovery post-transplant but had faster recovery of BK virus–specific Th1 CD4 T cells, which were more frequent than cytolytic CD8 T cells. Among those with BKPyV DNAuria, patients with early reconstitution of BKPyV-specific interferon-γ+ and cytolytic CD4 T cells was linked to lower hematuria rates, highlighting a potential role in prevention of symptoms . Humoral immunity likely also contributes to control of BKPyV replication; however, studies have not found a clear relationship between the presence of BKPyV-antibodies (which are not routinely obtained) and BKPyV-HC . Studies in seropositive patients have demonstrated an association between high BKPyV IgG titers with subsequent BKPyV DNAuria, although this was not compared to seronegative patients . Limited information exists regarding BKPyV syndromes in patients who were seronegative patients prior to transplantation. Among six children with BKPyV-HC after allo-HSCT, BKPyV viremia decreased as IgM levels initially increased, followed by IgG, suggesting the role of the humoral response in recovery .
The two major approaches to BKPyV disease include symptom management and antiviral strategies intended to prevent or reduce BKPyV replication. There are no high-quality or randomized data to support any specific prophylaxis or management strategy for BKPyV-HC or other BKPyV-associated diseases. Among the therapies studied, intravenous (IV) and intravesical cidofovir were the most reported strategy, but no clear benefit was demonstrated, and the level of evidence has been graded low . a. Supportive measures Without significant data to support an antiviral strategy, supportive care is recommended for BKPyV-HC (grade AIII) . Supportive measures include hydration, platelet transfusions, and analgesics . HBO therapy has been explored in case series123 and case reports122,124,125. However, there is no good quality data to support its use. b. Antiviral strategies While the cornerstone of BKPyVAN management after kidney transplantation is reduction in immunosuppression (RIS), this is frequently not feasible after allo-HSCT due to the threat of donor alloreactivity and GVHD . Several antivirals have been proposed or studied, including fluoroquinolones, leflunomide, cidofovir (IV or intravesical), brincidofovir, and virus-specific T cell (VST) therapy. 2.5.1. Fluoroquinolones Although quinolone antibiotics were initially thought to have modest anti-BKPyV activity in vitro and potential benefit in observational studies of BKPyV-HC after HSCT [ , , ], three RCTs in KTRs have not shown antiviral effect either as prevention of BKPyV DNAuria or DNAemia. In one RCT of 3 months of levofloxacin vs placebo, BKPyV DNAuria was observed in 22/76 patients (29%) in the levofloxacin group and 26/78 patients (33.3%) in the comparison group . In a second RCT, BKPyV viremia occurred in 25/133 patients (18.8%) in the ciprofloxacin group compared to 5/67 patients (7.5%) in the placebo group ( p = 0.03) . Lastly, among KTRs with BKPyV DNAemia the percentage of patients with BKPyV load reduction were 70.3% and 69.1% in the levofloxacin group (n = 22) and the placebo group (n = 21), respectively ( p = 0.93) . Several studies have also identified a higher risk of quinolone-resistant bacterial infections [ , , ]. As a result, quinolones are not recommended for BKPyV treatment or prevention either after KT or HSCT [ , , ]. 2.5.2. Leflunomide Leflunomide is an antimetabolite drug with immunomodulatory and antiviral activity; its use has been supported by case reports and small retrospective studies [ , , ]. However, a phase II study of a leflunomide derivative did not show substantial benefit in the treatment of BKPyVAN after KT as treatment in 30 patients led to a slightly greater reduction in urine BKPyV DNA load (3.1 log10 copies/μL in treatment group vs. 2.8 log10 copies/μL with placebo) and a small statistically significant reduction in viremia (1.9 log10 copies/μL in treatment group vs. 1.3 log10 copies/μL in placebo group, p = 0.049) ; leflunomide is not currently recommended . 2.5.3. Cidofovir Cidofovir is a nucleotide analog that inhibits a broad range of DNA viruses in vitro, although BKPyV does not possess its own polymerase. Its active metabolite has a long half-life of 15–65 h, allowing for administration at weekly intervals . Intravenous (IV) or intravesical cidofovir have been commonly used for BKPyV-HC, but there is no consensus on the optimal dose, route, or frequency of administration . When given IV, doses ranging from 3–5 mg/kg/week with probenecid, or lower doses of 0.5–1.5 mg/kg/week without probenecid are typically given . To mitigate the risk of ocular and nephrotoxicity, cidofovir may be administered intravesically, although the potential for systemic exposure remains . Small retrospective studies, case series, and case reports (without a comparator group) have suggested varying clinical response rates both intravenous (IV) and intravesical cidofovir in the treatment of BK polyomavirus-associated hemorrhagic cystitis (BKPyV-HC). Clinical response was generally defined as complete symptom resolution, while partial response (PR) referred to significant symptom improvement with persistent hematuria, though definitions varied across studies. Study sizes ranged from 4 to 57 patients in both pediatric and adult populations, with complete response (CR) rates between 60–85% for IV cidofovir and 59–92% for intravesical administration. A systematic review reported a CR in 117 of 172 patients (68%) treated with IV cidofovir and 15 of 17 patients (88.2%) treated with intravesical cidofovir [ , , , , , , , , , , , , , , , ]. However, there was inconsistency in the reporting and definitions of virologic responses, and other studies have demonstrated no reductions in BKPyV DNAemia associated with cidofovir administration [ , , ]. Furthermore, a small randomized, placebo-controlled phase I/II trial of low-dose IV cidofovir given for treatment of BKPyVAN after KT did not significantly reduce BKPyV DNAemia . 2.5.4. Brincidofovir Brincidofovir (CMX001) is a lipid-conjugated prodrug of cidofovir with in vitro activity against a variety of double-stranded DNA viruses, including BKPyV. Importantly, brincidofovir is associated with lower rates of renal dysfunction compared to cidofovir . Several case reports suggest potential benefits of this medication in managing BKPyV-associated conditions [ , , ]. However, in a phase III study of brincidofovir for CMV prophylaxis after allo-HSCT, the percentage of patients with BKPyV infection was not different in the treatment vs placebo arms. Brincidofovir is not currently available . 2.5.5. Virus-Specific T-Cells (VSTs) VSTs are a targeted therapy used to treat viral infections in immunocompromised patients. Various approaches have been developed, including ex-vivo expansion of donor-derived VSTs or banking VSTs from healthy donors with diverse haplotypes for immediate infusion . To address the multiple viral infections that affect immunocompromised patients, multi-virus VSTs have been developed and studied. Although adoptive transfer of VSTs is generally safe, the complexity of preparing these cells and their limited antiviral range have posed significant hurdles. Case reports and series have reported favorable clinical outcomes including reduction in viral load and symptomatic improvement [ , , , , ]. Two single arm, uncontrolled studies evaluated safety and efficacy of VSTs targeted against BKPyV. One study included 38 HSCT recipients and 3 SOT recipients with BKPyV DNAemia or BKPyV-HC; among patients with cystitis, 31/38 achieved complete response (CR); when assessing both BKPyV PCR levels and cystitis response, 22/38 patients achieved a CR; CR was defined as undetectable plasma BKPyV PCR within four weeks after the final VST infusion or resolution of hemorrhagic cystitis to below grade 2 without the need for medical therapy. No infusion-related toxicity, de novo GVHD, or organ rejection were found . Another study of third-party BKPyV-specific cytotoxic T lymphocytes for patients with BKPyV-HC after allo HSCT demonstrated that by day 21, 33/56 patients achieved CR (defined as full resolution of symptoms and gross hematuria or a reduction in hemorrhagic cystitis (HC) grade from 2, 3, or 4 to 0 or 1), while by day 45, 34 of 49 patients achieved CR . Post-infusion T cell detections have been observed and sustained for several months [ , , ]. However, these promising results need to be confirmed in RCTs. Posoleucel, an investigational multivirus-specific T-cell therapy, was found to be safe, well tolerated, and associated with greater reductions in BK viremia compared to placebo in a phase two study of 61 kidney transplant recipients . However, a phase three study evaluating posoleucel for preventing viral infections after allo-HSCT, treating virus-associated hemorrhagic cystitis, and adenovirus were terminated early due to futility .
Although quinolone antibiotics were initially thought to have modest anti-BKPyV activity in vitro and potential benefit in observational studies of BKPyV-HC after HSCT [ , , ], three RCTs in KTRs have not shown antiviral effect either as prevention of BKPyV DNAuria or DNAemia. In one RCT of 3 months of levofloxacin vs placebo, BKPyV DNAuria was observed in 22/76 patients (29%) in the levofloxacin group and 26/78 patients (33.3%) in the comparison group . In a second RCT, BKPyV viremia occurred in 25/133 patients (18.8%) in the ciprofloxacin group compared to 5/67 patients (7.5%) in the placebo group ( p = 0.03) . Lastly, among KTRs with BKPyV DNAemia the percentage of patients with BKPyV load reduction were 70.3% and 69.1% in the levofloxacin group (n = 22) and the placebo group (n = 21), respectively ( p = 0.93) . Several studies have also identified a higher risk of quinolone-resistant bacterial infections [ , , ]. As a result, quinolones are not recommended for BKPyV treatment or prevention either after KT or HSCT [ , , ].
Leflunomide is an antimetabolite drug with immunomodulatory and antiviral activity; its use has been supported by case reports and small retrospective studies [ , , ]. However, a phase II study of a leflunomide derivative did not show substantial benefit in the treatment of BKPyVAN after KT as treatment in 30 patients led to a slightly greater reduction in urine BKPyV DNA load (3.1 log10 copies/μL in treatment group vs. 2.8 log10 copies/μL with placebo) and a small statistically significant reduction in viremia (1.9 log10 copies/μL in treatment group vs. 1.3 log10 copies/μL in placebo group, p = 0.049) ; leflunomide is not currently recommended .
Cidofovir is a nucleotide analog that inhibits a broad range of DNA viruses in vitro, although BKPyV does not possess its own polymerase. Its active metabolite has a long half-life of 15–65 h, allowing for administration at weekly intervals . Intravenous (IV) or intravesical cidofovir have been commonly used for BKPyV-HC, but there is no consensus on the optimal dose, route, or frequency of administration . When given IV, doses ranging from 3–5 mg/kg/week with probenecid, or lower doses of 0.5–1.5 mg/kg/week without probenecid are typically given . To mitigate the risk of ocular and nephrotoxicity, cidofovir may be administered intravesically, although the potential for systemic exposure remains . Small retrospective studies, case series, and case reports (without a comparator group) have suggested varying clinical response rates both intravenous (IV) and intravesical cidofovir in the treatment of BK polyomavirus-associated hemorrhagic cystitis (BKPyV-HC). Clinical response was generally defined as complete symptom resolution, while partial response (PR) referred to significant symptom improvement with persistent hematuria, though definitions varied across studies. Study sizes ranged from 4 to 57 patients in both pediatric and adult populations, with complete response (CR) rates between 60–85% for IV cidofovir and 59–92% for intravesical administration. A systematic review reported a CR in 117 of 172 patients (68%) treated with IV cidofovir and 15 of 17 patients (88.2%) treated with intravesical cidofovir [ , , , , , , , , , , , , , , , ]. However, there was inconsistency in the reporting and definitions of virologic responses, and other studies have demonstrated no reductions in BKPyV DNAemia associated with cidofovir administration [ , , ]. Furthermore, a small randomized, placebo-controlled phase I/II trial of low-dose IV cidofovir given for treatment of BKPyVAN after KT did not significantly reduce BKPyV DNAemia .
Brincidofovir (CMX001) is a lipid-conjugated prodrug of cidofovir with in vitro activity against a variety of double-stranded DNA viruses, including BKPyV. Importantly, brincidofovir is associated with lower rates of renal dysfunction compared to cidofovir . Several case reports suggest potential benefits of this medication in managing BKPyV-associated conditions [ , , ]. However, in a phase III study of brincidofovir for CMV prophylaxis after allo-HSCT, the percentage of patients with BKPyV infection was not different in the treatment vs placebo arms. Brincidofovir is not currently available .
VSTs are a targeted therapy used to treat viral infections in immunocompromised patients. Various approaches have been developed, including ex-vivo expansion of donor-derived VSTs or banking VSTs from healthy donors with diverse haplotypes for immediate infusion . To address the multiple viral infections that affect immunocompromised patients, multi-virus VSTs have been developed and studied. Although adoptive transfer of VSTs is generally safe, the complexity of preparing these cells and their limited antiviral range have posed significant hurdles. Case reports and series have reported favorable clinical outcomes including reduction in viral load and symptomatic improvement [ , , , , ]. Two single arm, uncontrolled studies evaluated safety and efficacy of VSTs targeted against BKPyV. One study included 38 HSCT recipients and 3 SOT recipients with BKPyV DNAemia or BKPyV-HC; among patients with cystitis, 31/38 achieved complete response (CR); when assessing both BKPyV PCR levels and cystitis response, 22/38 patients achieved a CR; CR was defined as undetectable plasma BKPyV PCR within four weeks after the final VST infusion or resolution of hemorrhagic cystitis to below grade 2 without the need for medical therapy. No infusion-related toxicity, de novo GVHD, or organ rejection were found . Another study of third-party BKPyV-specific cytotoxic T lymphocytes for patients with BKPyV-HC after allo HSCT demonstrated that by day 21, 33/56 patients achieved CR (defined as full resolution of symptoms and gross hematuria or a reduction in hemorrhagic cystitis (HC) grade from 2, 3, or 4 to 0 or 1), while by day 45, 34 of 49 patients achieved CR . Post-infusion T cell detections have been observed and sustained for several months [ , , ]. However, these promising results need to be confirmed in RCTs. Posoleucel, an investigational multivirus-specific T-cell therapy, was found to be safe, well tolerated, and associated with greater reductions in BK viremia compared to placebo in a phase two study of 61 kidney transplant recipients . However, a phase three study evaluating posoleucel for preventing viral infections after allo-HSCT, treating virus-associated hemorrhagic cystitis, and adenovirus were terminated early due to futility .
JCPyV is best known as a cause of PML due to different types of immunosuppression and has been reported as an uncommon complication after allo-HSCT . JCPyV DNAuria and DNAemia are common after all-HSCT but less commonly associated with clinically significant disease. We review the clinical presentation, epidemiology, pathophysiology, and management below. 3.1. Clinical Presentation of JCPyV Syndromes After HSCT a. Progressive multifocal leukoencephalopathy (PML) Classically, the development of PML has been associated with specific types of immunosuppression, such as natalizumab or HIV/AIDS , however PML is seen among patients with immunosuppression of a variety of causes and occasionally among immunocompetent patients. Around 8% of cases of PML are made up of patients with hematological malignancies including HSCT recipients . PML typically presents months to years after HSCT and presenting symptoms are similar to those in other immunocompromised, non-allogeneic HSCT hosts: weakness, cerebellar symptoms, visual changes, aphasia, and hemiparesis [ , , , , ]. The clinical diagnosis of PML relies on the presence of compatible symptoms, imaging features, and the presence of JCPyV by PCR from CSF; diagnosis may also be made based on histopathology with detection of JCPyV by electron microscopy, immunohistochemistry, or tissue PCR . Detection of JCPyV in CSF alone without symptoms or radiographic signs of infection has poor predictive value for PML . Cytological features of PML include hypertrophy of astrocytes, oligodendrocytes with enlarged, round nuclei, and disseminated foci of demyelination . b. Viremia Asymptomatic JCPyV DNAemia is uncommonly detected in blood samples of healthy adults , but is common after HSCT. Among 164 allo HSCT recipients who were retrospectively tested for JCPyV from whole blood samples originally sent for CMV, any DNAemia was detected in 40 (24%) and DNAemia in at least two samples was detected in 20 (12%); two patients were ultimately diagnosed with PML . Other studies show a lower incidence (0–20%) of post-allo-HSCT viremia . Among patients ultimately diagnosed with PML, JCPyV replication in peripheral blood preceded PML but JCPyV detection in peripheral blood is an insensitive predictor as most patients with JCPyV viremia will not go on to develop PML. c. JCPyV-associated renourinary syndromes Asymptomatic JCPyV DNAuria is common both pre and post HSCT , and there is a positive quantitative relationship between DNA load in the urine and IgG index . JCPyV is a recognized but uncommon cause of PyVAN after kidney transplant . The true incidence of PyVAN after HSCT is unknown, and JCPyV DNAuria is less common after HSCT than BKPyV DNAuria , but it is possible that JCPyVAN could also affect HSCT recipients. BKPyV-HC is a well-described syndrome after HSCT; JCPyV-HC has been reported, but the incidence is unknown . 3.2. Pathogenesis of JCPyV Syndromes After HSCT JCPyV is thought to spread ubiquitously throughout the population early in life through the respiratory or oral route; by the time of adulthood, 60–90% of the population is seropositive for JCPyV [ , , , ]. Following transmission and primary viremia, JCPyV persists primarily within the kidneys, and 20% of asymptomatic healthy individuals can intermittently shed JCPyV in their urine . JCPyV can also be detected in tonsillar tissue, lymphocytes, and in brain tissue of healthy patients . JCPyV is best known as a cause of PML but has also been reported as a cause of encephalopathy, meningitis, granule cell neuronopathy, PyV associated nephropathy (PyVAN), or HC [ , , ]. PML is thought to arise when immunosuppression allows for JCPyV replication within the kidney, hematogenous spread via infected lymphocytes to the central nervous system, and replication within glial cells in the CNS . The pathophysiology of PML after allo-HSCT is thought to be similar to PML associated with other forms of immunosuppression. Genotypic differences in the NCCR between JCPyV associated with PML (“prototypical” JCPyV) and JCPyV detected in urine and sewage (“archetypal” JCPyV) have been observed, giving rise to multiple hypotheses regarding pathogenesis. Archetypal JCPyV is thought to be the strain transmitted from person to person, which then persists in tonsils, renal cells, and other tissues. It is unknown whether genetic rearrangements in the NCCR occur in peripheral tissues, in the CNS, or in both to produce the prototypical strain of JCPyV. However, immunosuppression is thought to permit prototypical JCPyV to spread to or replication within the CNS, potentially via generation of VP1 escape mutant strains that confer resistance to circulating VP1 antibodies [ , , , , ]. The degree of mutations (number of repeats) within the NCCR has been correlated with poorer PML prognosis [ , , ]. The control of JCPyV replication and PML has been strongly associated with cellular immunity, although humoral immunity likely also plays a role . In a small cohort of HSCT recipients prospectively monitored pre- and post-transplant, the presence of JCPyV DNAemia inversely correlated with detection of cellular immune responses; JCPyV-specific CD4+ and CD8+ T cell responses increased 12–18 months post-HSCT . In another small case series of patients with PML (none post-HSCT), however, plasma neutralizing antibody levels were highest among PML survivors . 3.3. Management of PML As below with BKPyV-associated diseases, there is no demonstrated effective treatment for JCPyV disease, including PML, and the cornerstone of management is immune restoration . Numerous medications have been investigated for their antiviral activity including via large-scale drug screens , some with mechanisms that could disrupt known components of the JCPyV life cycle, including trifluoperazine (calcium signaling-related inhibition), topotecan (topoisomerase inhibitor), cytarabine (polymerase inhibition), mefloquine (antagonize endosomal acidification), mirtazapine (interrupt viral attachment/entry and trafficking), pembrolizumab or nivolumab (boost cellular immune activity), and cidofovir/brincidofovir (polymerase inhibition) ; however many medications have had mixed outcomes or lack of benefit in larger observational studies [ , , , , , , , , ]. JCPyV- or BKPyV-specific T cells (leveraging cross reactivity due to virus similarity) have been used successfully in patients with PML, including after HSCT, and a few studies have examined the impact of VSTs in multiple patients with PML [ , , ]. In the first, nine patients with definite PML received JCPyV-specific T cells (autologous or allogeneic HLA-matched); 6/9 patients achieved control of their JCPyV, evidenced by symptomatic stability or improvement and a lower JC load in CSF . In a second study of 4 HSCT (2 autologous, 2 allogeneic) recipients with PML, no cases of graft versus host disease or infusion reactions were associated with transfer of third party BKPyV-directed VSTs. One patient’s disease clinically stabilized, with clearance of JCPyV from blood and CSF, but the other three patients had progressive neurologic decline despite some improvement JCPyV load . Lastly, 28 patients received BKPyV-directed VSTs isolated from healthy donors; 22/28 (79%) demonstrated clinical stabilization/improvement and a reduced JCPyV load over >1 year of follow up. Three of 4 enrolled allo-HSCT recipients responded. Survival was significantly higher (hazard ratio, 0.42, p = 0.02) in comparison to a historical control group (n = 113) . These results are promising and should be further examined in larger, well-controlled studies.
a. Progressive multifocal leukoencephalopathy (PML) Classically, the development of PML has been associated with specific types of immunosuppression, such as natalizumab or HIV/AIDS , however PML is seen among patients with immunosuppression of a variety of causes and occasionally among immunocompetent patients. Around 8% of cases of PML are made up of patients with hematological malignancies including HSCT recipients . PML typically presents months to years after HSCT and presenting symptoms are similar to those in other immunocompromised, non-allogeneic HSCT hosts: weakness, cerebellar symptoms, visual changes, aphasia, and hemiparesis [ , , , , ]. The clinical diagnosis of PML relies on the presence of compatible symptoms, imaging features, and the presence of JCPyV by PCR from CSF; diagnosis may also be made based on histopathology with detection of JCPyV by electron microscopy, immunohistochemistry, or tissue PCR . Detection of JCPyV in CSF alone without symptoms or radiographic signs of infection has poor predictive value for PML . Cytological features of PML include hypertrophy of astrocytes, oligodendrocytes with enlarged, round nuclei, and disseminated foci of demyelination . b. Viremia Asymptomatic JCPyV DNAemia is uncommonly detected in blood samples of healthy adults , but is common after HSCT. Among 164 allo HSCT recipients who were retrospectively tested for JCPyV from whole blood samples originally sent for CMV, any DNAemia was detected in 40 (24%) and DNAemia in at least two samples was detected in 20 (12%); two patients were ultimately diagnosed with PML . Other studies show a lower incidence (0–20%) of post-allo-HSCT viremia . Among patients ultimately diagnosed with PML, JCPyV replication in peripheral blood preceded PML but JCPyV detection in peripheral blood is an insensitive predictor as most patients with JCPyV viremia will not go on to develop PML. c. JCPyV-associated renourinary syndromes Asymptomatic JCPyV DNAuria is common both pre and post HSCT , and there is a positive quantitative relationship between DNA load in the urine and IgG index . JCPyV is a recognized but uncommon cause of PyVAN after kidney transplant . The true incidence of PyVAN after HSCT is unknown, and JCPyV DNAuria is less common after HSCT than BKPyV DNAuria , but it is possible that JCPyVAN could also affect HSCT recipients. BKPyV-HC is a well-described syndrome after HSCT; JCPyV-HC has been reported, but the incidence is unknown .
JCPyV is thought to spread ubiquitously throughout the population early in life through the respiratory or oral route; by the time of adulthood, 60–90% of the population is seropositive for JCPyV [ , , , ]. Following transmission and primary viremia, JCPyV persists primarily within the kidneys, and 20% of asymptomatic healthy individuals can intermittently shed JCPyV in their urine . JCPyV can also be detected in tonsillar tissue, lymphocytes, and in brain tissue of healthy patients . JCPyV is best known as a cause of PML but has also been reported as a cause of encephalopathy, meningitis, granule cell neuronopathy, PyV associated nephropathy (PyVAN), or HC [ , , ]. PML is thought to arise when immunosuppression allows for JCPyV replication within the kidney, hematogenous spread via infected lymphocytes to the central nervous system, and replication within glial cells in the CNS . The pathophysiology of PML after allo-HSCT is thought to be similar to PML associated with other forms of immunosuppression. Genotypic differences in the NCCR between JCPyV associated with PML (“prototypical” JCPyV) and JCPyV detected in urine and sewage (“archetypal” JCPyV) have been observed, giving rise to multiple hypotheses regarding pathogenesis. Archetypal JCPyV is thought to be the strain transmitted from person to person, which then persists in tonsils, renal cells, and other tissues. It is unknown whether genetic rearrangements in the NCCR occur in peripheral tissues, in the CNS, or in both to produce the prototypical strain of JCPyV. However, immunosuppression is thought to permit prototypical JCPyV to spread to or replication within the CNS, potentially via generation of VP1 escape mutant strains that confer resistance to circulating VP1 antibodies [ , , , , ]. The degree of mutations (number of repeats) within the NCCR has been correlated with poorer PML prognosis [ , , ]. The control of JCPyV replication and PML has been strongly associated with cellular immunity, although humoral immunity likely also plays a role . In a small cohort of HSCT recipients prospectively monitored pre- and post-transplant, the presence of JCPyV DNAemia inversely correlated with detection of cellular immune responses; JCPyV-specific CD4+ and CD8+ T cell responses increased 12–18 months post-HSCT . In another small case series of patients with PML (none post-HSCT), however, plasma neutralizing antibody levels were highest among PML survivors .
As below with BKPyV-associated diseases, there is no demonstrated effective treatment for JCPyV disease, including PML, and the cornerstone of management is immune restoration . Numerous medications have been investigated for their antiviral activity including via large-scale drug screens , some with mechanisms that could disrupt known components of the JCPyV life cycle, including trifluoperazine (calcium signaling-related inhibition), topotecan (topoisomerase inhibitor), cytarabine (polymerase inhibition), mefloquine (antagonize endosomal acidification), mirtazapine (interrupt viral attachment/entry and trafficking), pembrolizumab or nivolumab (boost cellular immune activity), and cidofovir/brincidofovir (polymerase inhibition) ; however many medications have had mixed outcomes or lack of benefit in larger observational studies [ , , , , , , , , ]. JCPyV- or BKPyV-specific T cells (leveraging cross reactivity due to virus similarity) have been used successfully in patients with PML, including after HSCT, and a few studies have examined the impact of VSTs in multiple patients with PML [ , , ]. In the first, nine patients with definite PML received JCPyV-specific T cells (autologous or allogeneic HLA-matched); 6/9 patients achieved control of their JCPyV, evidenced by symptomatic stability or improvement and a lower JC load in CSF . In a second study of 4 HSCT (2 autologous, 2 allogeneic) recipients with PML, no cases of graft versus host disease or infusion reactions were associated with transfer of third party BKPyV-directed VSTs. One patient’s disease clinically stabilized, with clearance of JCPyV from blood and CSF, but the other three patients had progressive neurologic decline despite some improvement JCPyV load . Lastly, 28 patients received BKPyV-directed VSTs isolated from healthy donors; 22/28 (79%) demonstrated clinical stabilization/improvement and a reduced JCPyV load over >1 year of follow up. Three of 4 enrolled allo-HSCT recipients responded. Survival was significantly higher (hazard ratio, 0.42, p = 0.02) in comparison to a historical control group (n = 113) . These results are promising and should be further examined in larger, well-controlled studies.
Fifteen PyVs have been isolated from humans—around half have been linked with clinical disease; Karolinska Institute PyV (KIPyV), Washington University PyV (WUPyV), Merkel cell PyV (MCPyV), HPyV6, HPyV7, Trichodysplasia Spinulosa PyV (TSPyV) . A similar disease paradigm to JCPyV and BKPyV exists for the remainder of the common HPyVs: viral acquisition is thought to occur early in life, detection of HPyV can occur in healthy, asymptomatic individuals, and immunosuppression is thought to permit increased replication and the potential for symptomatic disease. TSPyV may be the exception as some data suggest that primary infection is associated with symptoms in at least some cases . Several HPyVs are hypothesized to be a cause of clinical syndromes in immunosuppressed individuals due to reports of detection, but causality has yet to be definitively established: HPyV 6 and HPyV 7 (identified in cases of pruritic dermatoses) and KIPyV and WUPyV (identified in cases of respiratory infection). Other HPyVs have been causally linked with clinical manifestations; these include MCPyV (cause of Merkel cell carcinoma; reported after HSCT) and TSPyV (trichodysplasia spinulosa, characterized by spiny follicular papules and projections; cases reported after organ transplant but not allo-HSCT) . Subclinical detection from patients undergoing allo-HSCT is common from plasma or other samples. In a study that performed regular plasma sampling for 109 consecutive allo-HSCT recipients, several HPyVs were detected at least once at or after HSCT, including HPyV6 (24% of patients), HPyV7 (14% of patients), and MCPyV (21%); HPyV9 and TSPyV were not detected in any samples . A similar study in pediatric patients also had a low rate of detection of TSPyV (1.9%) . HPyV7 has been identified in urine samples of one patient following allo-HSCT along with BKPyV . Although KIPyV and WUPyV are not commonly detected in plasma sampling , they have each been detected in 5–20% of allo-HSCT recipients with regular upper respiratory sampling [ , , ]. Detection of WU or KI (relative to no detection) was associated with some respiratory symptoms (e.g., wheezing, sputum production) . WUPyV is suspected to have caused respiratory disease and death in one reported patient after allo-HSCT . The diagnosis was based on molecular detection of WUPyV, WU-VP1-specific immunohistochemical detection, and electron microscopy visualization of viral capsids of compatible size within lung tissue from autopsy specimens. There are insufficient data to guide treatment in cases where clinical disease is suspected, though cases of treatment with valganciclovir, cidofovir, and reduction in immunosuppression are reported .
Polyomavirus-associated diseases, particularly BKPyV-HC and JCPyV-PML, remain significant complications in allo-HSCT recipients, with limited effective therapies and high morbidity. Future research should focus on establishing the precise pathophysiology of PyV reactivation, including the role of immune reconstitution and host-virus interactions. Development of targeted antiviral agents and virus-specific T-cell therapies shows promise but requires validation in controlled trials. Improved diagnostic strategies, including standardized thresholds for DNAemia and DNAuria, are needed to enhance early detection and risk stratification. Finally, understanding the interplay between immunosuppression, viral pathogenesis, and host immunity may pave the way for innovative prophylactic and therapeutic approaches.
|
Protective Effects of PACAP in a Rat Model of Diabetic Neuropathy | fb91a813-db1a-41b3-8435-607fcdd15d01 | 8509403 | Anatomy[mh] | Pituitary adenylate cyclase-activating peptide (PACAP) is a member of the vasoactive intestinal peptide (VIP)/secretin/glucagon peptide family and has two biologically active isoforms—PACAP-27 and PACAP-38, with the latter being predominant in mammals . PACAP acts through G-protein-coupled receptors: its specific receptor is PAC1, while VPAC1 and VPAC2 receptors bind PACAP and VIP with the same affinity. PACAP has a widespread distribution in the body, with the highest expression levels in the nervous system and endocrine glands, where it acts as a neurotransmitter, neuromodulator, and neurohormone [ , , , , ]. In addition, PACAP and its receptors are widely expressed in peripheral organs [ , , , , ], and the peptide plays different roles in numerous physiological processes in the cardiovascular, respiratory, urogenital, musculoskeletal, and digestive systems [ , , , , , ]. One of the established effects of the neuropeptide is its neurotrophic/neuroprotective action [ , , , , , ]. This has been proven in numerous neuronal insults and models of neurodegenerative diseases, such as spinal atrophy , amyotrophic lateral sclerosis , Alzheimer’s disease [ , , ], stroke [ , , ], Parkinson’s disease [ , , , ], Huntington chorea , and several types of retinal injuries [ , , , ]. Several reports have proven that PACAP is also protective in diabetes-related pathological conditions, such as diabetic nephropathy and retinopathy [ , , , , , , ]. It has been demonstrated that PACAP is protective in the inner, neuronal retinal layers in diabetic retinopathy and also acts on pigment epithelial cells in hyperglycemic conditions [ , , ]. Vasculopathy is in the background of several diabetic complications. PACAP has been shown to ameliorate hyperglycemia-induced vascular dysfunction in isolated vessels . Diabetic neuropathy is a common microvascular complication of diabetes, affecting around 50–70% of diabetic patients . PACAP is involved in glucose metabolism and insulin secretion , in addition to protective effects exerted on the insulin-producing pancreatic beta cells . More importantly, from a neuropathy point of view, PACAP influences Schwann cell functions , stimulates axonal growth , and promotes regeneration of peripheral nerves [ , , , ]. However, it is not known whether it has protective effects in diabetic neuropathy. Therefore, in the present study, we aimed at investigating the neuroprotective effects of PACAP in an experimental model of diabetic neuropathy in rats. The two major predictors of developing neuropathy are the duration of diabetes and the degree of metabolic instability . Hyperglycemia results in excessive production of reactive oxygen species (ROS), overproduction of advanced glycation end products and glutamate, and decreased production of neuroprotective factors and hyperglycemia-activated signaling pathways, such as the polyol, hexosamine, and DAG-PKC pathways [ , , , , ]. Signs and symptoms of diabetic neuropathy include loss of reflexes, dysesthesia, and paresthesia, along with neuropathic pain—hyperalgesia and allodynia . Diabetic neuropathy is associated with alterations in the structure of peripheral nerves as well as in central structures of the pain processing pathway. However, the mechanism of neuropathic pain is still not understood, and it is an unmet medical need due to the ineffectiveness of the currently available therapy in a great proportion of cases. Previously, it has been shown that streptozotocin (STZ)-induced diabetes leads to alterations in the dorsal horn of the spinal cord and the mesencephalic periaqueductal grey matter (PAG) . Among others, altered c-Fos and FosB expressions in these centers have been reported [ , , , , ]. Therefore, in the present study, we also investigated the expression of FosB, a marker of chronic neuronal activation , in the above pain processing centers, in addition to the detailed morphological analysis of peripheral nerve fibers and functional assessment of pain sensation.
2.1. Blood Glucose Levels and Body Weight Vehicle (saline) treated diabetic and PACAP-treated diabetic groups had a significant rise in blood sugar levels after the 7th day of the experiment. Blood glucose level was 6.6 ± 0.34 mmol/L in the vehicle-treated control and 6.7 ± 0.24 mmol/L in the PACAP-treated control group on week 8. These values were 27.1 ± 2.02 mmol/L in the vehicle-treated diabetic and 26.4 ± 1.65 mmol/L in the PACAP-treated diabetic group ( a). These data show that blood glucose levels were not significantly affected by PACAP treatment. Vehicle-treated diabetic and PACAP-treated diabetic groups showed a significant decrease in their body weight from the second week. On week 8, we measured 477 ± 15 g in the vehicle-treated control, 473 ± 11 g in the PACAP-treated control groups, while only 302 ± 23 g in the diabetic and 292 ± 6 g in the PACAP-treated diabetic groups ( b). PACAP-38 treatment did not influence the body weight loss. 2.2. Functional Tests 2.2.1. Randall–Selitto Test The Randal–Selitto test revealed an increased pressure sensitivity in the vehicle-treated diabetic and PACAP-treated diabetic groups after the 6th week of the experiment, compared to the vehicle-treated control and PACAP-treated control groups. Pressure sensitivity increased to a lower extent in the PACAP-treated diabetic group on the 7th and 8th weeks, compared to the vehicle-treated diabetic group ( a). 2.2.2. Dynamic Plantar Aesthesiometer (DPA) The DPA test showed an increased touch sensitivity in the vehicle-treated diabetic and PACAP-treated diabetic groups after the 5th week of the experiment, compared to the vehicle-treated control and PACAP-treated control groups. It was less prominent in the PACAP-treated diabetic group from the 5th week, compared to the vehicle-treated diabetic group, but the difference was not statistically significant ( b). 2.3. Immunohistochemistry Our study showed that the number of FosB immunoreactive nuclei in laminae I-III of the spinal cord dorsal horn of segments L4–L5 was higher in the vehicle-treated diabetic and PACAP-treated diabetic groups, compared to the vehicle-treated control and PACAP-treated control groups. PACAP-treated diabetic animals had, however, significantly fewer FosB positive nuclei than vehicle-treated diabetic rats ( ). The number of FosB-positive nuclei in the lateral part of PAG was also investigated. We found that both vehicle- (saline) and PACAP-treated diabetic animals showed an elevated number of FosB immunoreactive nuclei in the lateral PAG, compared to the vehicle-treated control and PACAP-treated control groups. When comparing the PACAP-treated diabetic group to the vehicle-treated diabetic group, we found that PACAP was effective in significantly reducing the neuronal activity in the lateral PAG ( ). 2.4. Electron Microscopy of the Sciatic Nerve In vehicle-treated control and PACAP-treated control animals, normal peripheral nerve structure was found without any signs of myelin or axonal injury. PACAP treatment did not cause any changes under control (no diabetes) situations. However, the sciatic nerve of diabetic animals showed signs of neuropathy: axon–myelin separation, elevated average mitochondrial number in the axons, unmyelinated fiber atrophy, and basement membrane thickening. The percentage of the axon–myelin separation was significantly higher in the diabetic groups, compared to the control groups. However, it was significantly less prominent in the PACAP-treated diabetic group ( ). A marked elevation in the mitochondrial number in the myelinated axons was found in the vehicle-treated diabetic group. This could not be observed in the PACAP-treated diabetic group; thus, PACAP successfully prevented the rise in mitochondrial number ( ). We also found unmyelinated fiber atrophy, characterized by a decrease in the unmyelinated fiber area in the vehicle-treated diabetic and PACAP-treated diabetic groups, compared to the vehicle-treated control and PACAP-treated control groups. This decrease was significantly less severe in the PACAP-treated diabetic group. No difference was observed between the vehicle-treated control and PACAP-treated control groups ( ). Electron microscopy also revealed thickening of the basement membrane in the endoneurial capillaries in the vehicle-treated diabetic group. PACAP-treated diabetic animals did not show any basement membrane thickening; there was no difference between the vehicle-treated control, PACAP-treated control, and PACAP-treated diabetic groups ( ).
Vehicle (saline) treated diabetic and PACAP-treated diabetic groups had a significant rise in blood sugar levels after the 7th day of the experiment. Blood glucose level was 6.6 ± 0.34 mmol/L in the vehicle-treated control and 6.7 ± 0.24 mmol/L in the PACAP-treated control group on week 8. These values were 27.1 ± 2.02 mmol/L in the vehicle-treated diabetic and 26.4 ± 1.65 mmol/L in the PACAP-treated diabetic group ( a). These data show that blood glucose levels were not significantly affected by PACAP treatment. Vehicle-treated diabetic and PACAP-treated diabetic groups showed a significant decrease in their body weight from the second week. On week 8, we measured 477 ± 15 g in the vehicle-treated control, 473 ± 11 g in the PACAP-treated control groups, while only 302 ± 23 g in the diabetic and 292 ± 6 g in the PACAP-treated diabetic groups ( b). PACAP-38 treatment did not influence the body weight loss.
2.2.1. Randall–Selitto Test The Randal–Selitto test revealed an increased pressure sensitivity in the vehicle-treated diabetic and PACAP-treated diabetic groups after the 6th week of the experiment, compared to the vehicle-treated control and PACAP-treated control groups. Pressure sensitivity increased to a lower extent in the PACAP-treated diabetic group on the 7th and 8th weeks, compared to the vehicle-treated diabetic group ( a). 2.2.2. Dynamic Plantar Aesthesiometer (DPA) The DPA test showed an increased touch sensitivity in the vehicle-treated diabetic and PACAP-treated diabetic groups after the 5th week of the experiment, compared to the vehicle-treated control and PACAP-treated control groups. It was less prominent in the PACAP-treated diabetic group from the 5th week, compared to the vehicle-treated diabetic group, but the difference was not statistically significant ( b).
The Randal–Selitto test revealed an increased pressure sensitivity in the vehicle-treated diabetic and PACAP-treated diabetic groups after the 6th week of the experiment, compared to the vehicle-treated control and PACAP-treated control groups. Pressure sensitivity increased to a lower extent in the PACAP-treated diabetic group on the 7th and 8th weeks, compared to the vehicle-treated diabetic group ( a).
The DPA test showed an increased touch sensitivity in the vehicle-treated diabetic and PACAP-treated diabetic groups after the 5th week of the experiment, compared to the vehicle-treated control and PACAP-treated control groups. It was less prominent in the PACAP-treated diabetic group from the 5th week, compared to the vehicle-treated diabetic group, but the difference was not statistically significant ( b).
Our study showed that the number of FosB immunoreactive nuclei in laminae I-III of the spinal cord dorsal horn of segments L4–L5 was higher in the vehicle-treated diabetic and PACAP-treated diabetic groups, compared to the vehicle-treated control and PACAP-treated control groups. PACAP-treated diabetic animals had, however, significantly fewer FosB positive nuclei than vehicle-treated diabetic rats ( ). The number of FosB-positive nuclei in the lateral part of PAG was also investigated. We found that both vehicle- (saline) and PACAP-treated diabetic animals showed an elevated number of FosB immunoreactive nuclei in the lateral PAG, compared to the vehicle-treated control and PACAP-treated control groups. When comparing the PACAP-treated diabetic group to the vehicle-treated diabetic group, we found that PACAP was effective in significantly reducing the neuronal activity in the lateral PAG ( ).
In vehicle-treated control and PACAP-treated control animals, normal peripheral nerve structure was found without any signs of myelin or axonal injury. PACAP treatment did not cause any changes under control (no diabetes) situations. However, the sciatic nerve of diabetic animals showed signs of neuropathy: axon–myelin separation, elevated average mitochondrial number in the axons, unmyelinated fiber atrophy, and basement membrane thickening. The percentage of the axon–myelin separation was significantly higher in the diabetic groups, compared to the control groups. However, it was significantly less prominent in the PACAP-treated diabetic group ( ). A marked elevation in the mitochondrial number in the myelinated axons was found in the vehicle-treated diabetic group. This could not be observed in the PACAP-treated diabetic group; thus, PACAP successfully prevented the rise in mitochondrial number ( ). We also found unmyelinated fiber atrophy, characterized by a decrease in the unmyelinated fiber area in the vehicle-treated diabetic and PACAP-treated diabetic groups, compared to the vehicle-treated control and PACAP-treated control groups. This decrease was significantly less severe in the PACAP-treated diabetic group. No difference was observed between the vehicle-treated control and PACAP-treated control groups ( ). Electron microscopy also revealed thickening of the basement membrane in the endoneurial capillaries in the vehicle-treated diabetic group. PACAP-treated diabetic animals did not show any basement membrane thickening; there was no difference between the vehicle-treated control, PACAP-treated control, and PACAP-treated diabetic groups ( ).
In the present study, we demonstrated that in vivo PACAP treatment is protective in diabetic neuropathy. An 8-week PACAP-38 treatment effectively counteracted the functional and morphological changes observed in diabetic rats without altering the blood glucose levels or body weight. This observation is in accordance with our previous results studying the effects of PACAP treatment in diabetic nephropathy . These data suggest that PACAP, in spite of its effects on glucose homeostasis and insulin secretion , is protective in our diabetic neuropathy model, not by acting directly on glucose levels, but most probably due to its neuro- and general cytoprotective effects . Although, as outlined above, PACAP is known to stimulate insulin secretion, it did not alter blood glucose levels in this model of type I diabetes. Whether it affects insulin levels in this model or in a model of type II diabetes awaits further investigations. In the present study, we found that STZ-treated control diabetic rats displayed the morphological signs of diabetic neuropathy, i.e., axon–myelin separation, an increase in axonal mitochondria number, unmyelinated fiber atrophy, and basement membrane thickening of perineurial vessels. All these signs were attenuated by in vivo PACAP treatment. Previous studies have reported that axon–myelin separation is due to hyperglycemia , Na + channel dysfunction , and glycogen accumulation in Schwann cells , which led to a hyperosmolar perineurial environment, causing axon–myelin separation and demyelination . Our study showed that PACAP treatment markedly attenuated this axon–myelin separation. PACAP is known to be involved in myelin maturation and synthesis by inducing the expression of myelin markers , and it has a trophic and antiapoptotic effect on Schwann cells . PACAP receptors are upregulated in peripheral nerve injury in the Schwann cells, and the peptide promotes myelin gene expression, inhibits the release of pro-inflammatory cytokines, and stimulates anti-inflammatory cytokines . In experimental diabetic neuropathy, the increase in mitochondrial number has been reported in myelinated axons [ , , ]. Presumably, hyperglycemia-induced oxidative stress stimulates mitochondrial fission, which leads to the overproduction of mitochondrial ROS resulting in small aberrant, more electron-dense mitochondria with a reduced respiratory capacity . It has been suggested that the inhibition of mitochondrial fission would relieve the ROS-induced oxidative stress . The attenuated response in PACAP-treated animals might be due to the ability of PACAP to decrease oxidative metabolite levels, increase antioxidant potential , and stimulate antioxidant enzymes, such as peroxiredoxin , heme oxygenase-1 , superoxide-dismutase , and glutathione . Unmyelinated fiber atrophy was found in STZ-induced diabetic rats, with a reduced cross-sectional area of the axons, similar to other reports . This could also be attenuated by PACAP treatment. The protective effects of PACAP in nerve degeneration have been confirmed by dozens of studies . Among others, PACAP promotes cell survival and neurite outgrowth , enhances axonal sprouting , and stimulates neuronal differentiation during development and regeneration [ , , , , ]. In peripheral nerve injuries, PACAP has been shown to be upregulated and to promote regeneration partly through stimulation of other growth factors, such as glial cell line-derived neurotrophic factor [ , , , ]. Given the importance of endogenous PACAP in nerve regeneration, not surprisingly, mice lacking endogenous PACAP show a slower axonal regeneration with an increased pro-inflammatory environment . These authors suggested that endogenous PACAP is involved in the controlled immune response that is necessary for proper nerve regeneration after injury . This action of PACAP has been recently supported by human data: transcriptional profiling of the skin from patients with carpal tunnel syndrome revealed that the gene encoding PACAP was the most strongly upregulated gene and its expression was associated with recovery of intraepidermal nerve fibers . Diabetes is also associated with the thickening of the basement membrane of the vasa nervosum as a consequence of the increased expression and decreased breakdown of collagen IV [ , , ]. We found that diabetes resulted in the thickening of the basement membrane, attenuated by PACAP treatment. Similar to our present results, PACAP was found to attenuate basement membrane thickening in diabetic nephropathy . Similar vascular protective effects have been observed in murine endothelial cells exposed to glucose: PACAP elicited an antiproliferative effect under chronic hyperglycemic conditions . PACAP has also been demonstrated to protect endothelial cells against oxidative stress . The protective effects on vessels are reflected in morphological signs, but PACAP has also been shown to reduce hyperglycemia-induced vascular dysfunction . In that study, PACAP restored the disturbed relaxation of the vessels to an extent comparable to superoxide dismutase without direct scavenging of ROS. The elevated levels of fibroblast growth factor basic, matrix metalloproteinase 9, and nephroblastoma associated with endothelial dysfunction could be reduced by PACAP administration . The model used in our study mimics type I diabetes, as streptozotocin leads to toxic degeneration of the insulin-producing beta cells. Based on these observations, however, it is plausible that PACAP could also be protective in neuropathies observed in type II diabetes, as the main mechanism of neuropathic induction is not directly related to the model itself but more to the increased glucose levels. However, in order to prove this point, further experiments are required. The observed morphological signs of the protective effects of PACAP were also reflected in functional improvement in our study. Mechanical hyperalgesia is present in early diabetic neuropathy . In our study, the intraperitoneally administered PACAP significantly attenuated the enhancement of pressure sensitivity by measuring mechanical hyperalgesia. In addition to the protective effects on nerve fibers, the anti-nociceptive and anti-hyperalgesic effects of PACAP might also be due to the decreased release of the pro-nociceptive neuropeptides . In our present study, we also investigated the activation of pain-processing central structures. The expression of the acute neuronal activity marker c-Fos has been shown to be increased in the STZ diabetic model in the PAG and dorsal horn of the spinal cord . Previous studies have found that FosB expression is significantly elevated in rats in the case of chronic pain and stress but not acute pain . The dorsal horn of segment L4 of the spinal cord corresponds to the primary nociceptive afferent regions of the rat’s hind paw , while the lateral PAG is an important center of the descending anti-nociceptive system . Here, we described chronic neuronal activity (i.e., FosB expression) in the spinal dorsal horn of segments L4–L5 and in the lateral part of PAG in our STZ diabetes model and found that PACAP treatment effectively prevented FosB activation in these centers. Our earlier findings in mice lacking endogenous PACAP support these findings, as PACAP knockout mice showed increased basal neuronal activity (i.e., c-Fos) in the lateral PAG . The importance of endogenous PACAP in pain-processing centers has been highlighted also by several other studies [ , , , ]. In conclusion, here we show, for the first time, that PACAP treatment can attenuate or moderate the pathological changes of diabetic neuropathy, suggesting that PACAP could have therapeutic potential in diabetic neuropathy.
4.1. Animals The experiment was carried out using adult male Wistar rats ( n = 22) weighing 360–420 g. The experimental animals were housed under light/dark cycles of 12:12 h and received normal rat chow and drinking water ad libitum. Rats were randomly divided into four groups: (1) vehicle (saline)-treated control (non-diabetic) ( n = 5); (2) PACAP-treated control (non-diabetic) ( n = 5); (3) vehicle-treated diabetic ( n = 5), and (4) PACAP-treated diabetic ( n = 5) groups (in the figures, these groups are referred to as control, control + PACAP, diabetic, diabetic + PACAP groups, respectively). Diabetes was induced by a single dose of 65 mg/kg intravenous streptozotocin injection (Sigma, Budapest, Hungary). PACAP (20 µg PACAP1-38/100 μL saline solution) was injected intraperitoneally every second day for eight weeks, starting simultaneously with the streptozotocin injection to PACAP-treated control and PACAP-treated diabetic groups. The dose of PACAP was based on previous observations where this dose was effective in a rat model of diabetic nephropathy . Vehicle-treated control and vehicle-treated diabetic groups received 100 μL saline intraperitoneally. Body weight and blood glucose levels (Accu-Check Active, Roche, Budapest, Hungary) were measured weekly, rats with glucose levels higher than 11 mmol/L were considered diabetic. Experimental procedures were carried out in accordance with approved protocols (University of Pecs; BA02/2000-24/2011). We performed in vivo behavioral tests on all experimental animals of the four groups. Following 8 weeks of survival, animals were processed for histological analysis. 4.2. Functional Tests 4.2.1. Mechanical Nociceptive Threshold—Randall–Selitto Test The pressure sensitivity of the hind paw was measured by the Randall–Selitto test using Ugo Basile Analgesia Meter on a weekly basis. During the Randall–Selitto test, a continuously increasing pressure—at a maximum of 250 g—was applied to the hind paw of the rats. The increasing pressure caused the withdrawal of the paw, which was considered as the mechanical nociceptive threshold in grams . Three measurements were made on both left and right hind paws, and the average of the assessments was taken. A decreased mechanical nociceptive threshold in this test can be considered as hyperalgesia . 4.2.2. Mechanical Nociceptive Threshold—Dynamic Plantar Aesthesiometer Test (DPA) Touch sensitivity on the plantar surface of the hind paws was determined by a dynamic plantar aesthesiometer (DPA) (Ugo Basile, Gemonio, Italy) on a weekly basis. During the DPA test, a continuously increasing force—at a maximum of 50 g in 10 s—was applied to the hind paw by the elevation of a blunt-end needle and the aesthesiometer automatically detected the latency time and force (in grams) at the time of paw withdrawal. The decreased mechanical nociceptive threshold in the DPA test is a sign of allodynia since this mechanical stimulus is not painful to the rats . 4.3. Histology 4.3.1. Tissue Collection and Preparation for Histology Animals were anesthetized with an overdose of isoflurane (Forane, Abbott Hungary, Budapest, Hungary) on week 8. Rats were transcardially perfused with 25 mL of phosphate-buffered saline (PBS), followed by 300 mL of 4% paraformaldehyde solution in Millonig buffer for 20 min. The brain and spinal cord were dissected and then placed into the same fixative solution for post-fixation for 72 h at 4 °C. The sciatic nerve was also removed from all animals and further processed for electron microscopy. A tissue block containing the midbrain was isolated from the brains by cutting at the frontal planes of the posterior border of the median eminence and the transverse fissure. A tissue block of the L4-L5 spinal cord segments was also dissected. Blocks were sectioned by a Leica VT S 1000 (Leica, Wetzlar, Germany) vibratome. Five series of 30 µm coronal sections, interspaced by 120 µm, were collected in anti-freeze solution and stored at −20 °C until further use. 4.3.2. Immunohistochemistry for FosB Free-floating labeling for the chronic neuronal activation marker FosB was performed on a series of the midbrain and spinal cord sections, as published earlier . Briefly, sections were permeabilized with 0.5% Triton X-100 solution. Normal goat serum (2%, NGS, Jackson Immunoresearch, Europe Ltd., Suffolk, UK) was used to reduce non-specific binding. Subsequently, sections were treated with rabbit anti-FosB antibodies (1:500, Santa Cruz, sc-48 Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) in PBS with 2% NGS overnight. After washes, sections were treated with biotinylated goat anti-rabbit serum and avidin–biotin complex (Vectastain ABC Elite Kit Vector Lbs., Burlingame, CA, USA) according to the supplier’s protocol. The labeling was developed in Tris buffer (pH 7.4) with 0.02% 3, 3′ diamino-benzidine (DAB) (Sigma) and 0.00003 v / v % H 2 O 2 . The reaction was carried out under visual control and was stopped after 7 min with PBS. Then, preparations were washed and mounted on gelatin-covered slides. After drying, sections were dehydrated with ethanol solutions (50%, 70%, 96%, absolute, absolute 5 min, respectively), cleared by xylene (2 × 20 min), and coverslipped using Depex (Fluka, Heidelberg, Germany). Specificity of the FosB antiserum (Santa Cruz, sc-48) was characterized earlier . Western blot studies also support the specificity ( http://datasheets.scbt.com/sc-48.pdf , 2017). Preabsorption experiment in the rat revealed that the blocking peptide (sc-48-P, Santa Cruz) prevented the immunolabeling. In line with this, omission and/or replacement of the primary or secondary antibodies by non-immune sera abolished the signal in all tests (images not shown). 4.3.3. Digital Imaging and Morphometry at Light Microscopic Level The DAB-labeled FosB immunohistochemistry was studied and digitalized by a Nikon Microphot FXA microscope with a Spot RT camera (Nikon, Tokyo, Japan). For each animal, five serial sections were photographed. The number of FosB positive nuclei was determined using non-edited digital images by manual cell counting. The whole cross-sectional surface areas of the lateral PAG, as well as the dorsal horn laminae I, II, and III, individually were measured, and summed (laminae I + II + II) at L4–L5 spinal segments were evaluated, as marked in and . Cell counting was carried out by a skilled neurophysiologist who was not informed about the identity of preparations. For documentation and publication purposes, the micro-photos were grey-scaled and contrasted using Adobe Photoshop 7.0.1 software. 4.3.4. Electron Microscopy Sciatic nerve samples were placed in fixing solution (2.5% glutaraldehyde + 2% formalin + 0.1 M PBS) immediately after dissection in +4 °C for 24 h. A post-fixation procedure was performed with 1% osmium tetroxide. After dehydration in ascending alcohol (50%–70%–90%–96%) and subsequent transfer to propylene oxide, samples were embedded in Araldite resin. Semithin sections (0.5 µm) were cut by ultramicrotome (Leica Ultracut R), stained with 1% toluidine blue (Sigma), and examined with a Nikon Eclipse 80i microscope. Ultrathin sections were prepared from the area of interest and were contrasted by 2.5% uranyl–acetic acid and lead citrate. Slides were examined using a JEM-1200 EX-II electron microscope. The following parameters were analyzed: percentage of the axon-myelin separation, mitochondrial number in myelinated axons, area of the unmyelinated fibers, and thickness of the basement membrane. 4.4. Statistical Analysis Statistical analysis was performed in GraphPad Prism 6.01 software. Two-way analysis of variance (ANOVA) was used to detect significant differences between groups. Multiple comparisons were performed by Tukey’s test. Data are presented as means ± S.E.M (standard error of the mean). Differences were considered statistically significant when p < 0.05.
The experiment was carried out using adult male Wistar rats ( n = 22) weighing 360–420 g. The experimental animals were housed under light/dark cycles of 12:12 h and received normal rat chow and drinking water ad libitum. Rats were randomly divided into four groups: (1) vehicle (saline)-treated control (non-diabetic) ( n = 5); (2) PACAP-treated control (non-diabetic) ( n = 5); (3) vehicle-treated diabetic ( n = 5), and (4) PACAP-treated diabetic ( n = 5) groups (in the figures, these groups are referred to as control, control + PACAP, diabetic, diabetic + PACAP groups, respectively). Diabetes was induced by a single dose of 65 mg/kg intravenous streptozotocin injection (Sigma, Budapest, Hungary). PACAP (20 µg PACAP1-38/100 μL saline solution) was injected intraperitoneally every second day for eight weeks, starting simultaneously with the streptozotocin injection to PACAP-treated control and PACAP-treated diabetic groups. The dose of PACAP was based on previous observations where this dose was effective in a rat model of diabetic nephropathy . Vehicle-treated control and vehicle-treated diabetic groups received 100 μL saline intraperitoneally. Body weight and blood glucose levels (Accu-Check Active, Roche, Budapest, Hungary) were measured weekly, rats with glucose levels higher than 11 mmol/L were considered diabetic. Experimental procedures were carried out in accordance with approved protocols (University of Pecs; BA02/2000-24/2011). We performed in vivo behavioral tests on all experimental animals of the four groups. Following 8 weeks of survival, animals were processed for histological analysis.
4.2.1. Mechanical Nociceptive Threshold—Randall–Selitto Test The pressure sensitivity of the hind paw was measured by the Randall–Selitto test using Ugo Basile Analgesia Meter on a weekly basis. During the Randall–Selitto test, a continuously increasing pressure—at a maximum of 250 g—was applied to the hind paw of the rats. The increasing pressure caused the withdrawal of the paw, which was considered as the mechanical nociceptive threshold in grams . Three measurements were made on both left and right hind paws, and the average of the assessments was taken. A decreased mechanical nociceptive threshold in this test can be considered as hyperalgesia . 4.2.2. Mechanical Nociceptive Threshold—Dynamic Plantar Aesthesiometer Test (DPA) Touch sensitivity on the plantar surface of the hind paws was determined by a dynamic plantar aesthesiometer (DPA) (Ugo Basile, Gemonio, Italy) on a weekly basis. During the DPA test, a continuously increasing force—at a maximum of 50 g in 10 s—was applied to the hind paw by the elevation of a blunt-end needle and the aesthesiometer automatically detected the latency time and force (in grams) at the time of paw withdrawal. The decreased mechanical nociceptive threshold in the DPA test is a sign of allodynia since this mechanical stimulus is not painful to the rats .
The pressure sensitivity of the hind paw was measured by the Randall–Selitto test using Ugo Basile Analgesia Meter on a weekly basis. During the Randall–Selitto test, a continuously increasing pressure—at a maximum of 250 g—was applied to the hind paw of the rats. The increasing pressure caused the withdrawal of the paw, which was considered as the mechanical nociceptive threshold in grams . Three measurements were made on both left and right hind paws, and the average of the assessments was taken. A decreased mechanical nociceptive threshold in this test can be considered as hyperalgesia .
Touch sensitivity on the plantar surface of the hind paws was determined by a dynamic plantar aesthesiometer (DPA) (Ugo Basile, Gemonio, Italy) on a weekly basis. During the DPA test, a continuously increasing force—at a maximum of 50 g in 10 s—was applied to the hind paw by the elevation of a blunt-end needle and the aesthesiometer automatically detected the latency time and force (in grams) at the time of paw withdrawal. The decreased mechanical nociceptive threshold in the DPA test is a sign of allodynia since this mechanical stimulus is not painful to the rats .
4.3.1. Tissue Collection and Preparation for Histology Animals were anesthetized with an overdose of isoflurane (Forane, Abbott Hungary, Budapest, Hungary) on week 8. Rats were transcardially perfused with 25 mL of phosphate-buffered saline (PBS), followed by 300 mL of 4% paraformaldehyde solution in Millonig buffer for 20 min. The brain and spinal cord were dissected and then placed into the same fixative solution for post-fixation for 72 h at 4 °C. The sciatic nerve was also removed from all animals and further processed for electron microscopy. A tissue block containing the midbrain was isolated from the brains by cutting at the frontal planes of the posterior border of the median eminence and the transverse fissure. A tissue block of the L4-L5 spinal cord segments was also dissected. Blocks were sectioned by a Leica VT S 1000 (Leica, Wetzlar, Germany) vibratome. Five series of 30 µm coronal sections, interspaced by 120 µm, were collected in anti-freeze solution and stored at −20 °C until further use. 4.3.2. Immunohistochemistry for FosB Free-floating labeling for the chronic neuronal activation marker FosB was performed on a series of the midbrain and spinal cord sections, as published earlier . Briefly, sections were permeabilized with 0.5% Triton X-100 solution. Normal goat serum (2%, NGS, Jackson Immunoresearch, Europe Ltd., Suffolk, UK) was used to reduce non-specific binding. Subsequently, sections were treated with rabbit anti-FosB antibodies (1:500, Santa Cruz, sc-48 Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) in PBS with 2% NGS overnight. After washes, sections were treated with biotinylated goat anti-rabbit serum and avidin–biotin complex (Vectastain ABC Elite Kit Vector Lbs., Burlingame, CA, USA) according to the supplier’s protocol. The labeling was developed in Tris buffer (pH 7.4) with 0.02% 3, 3′ diamino-benzidine (DAB) (Sigma) and 0.00003 v / v % H 2 O 2 . The reaction was carried out under visual control and was stopped after 7 min with PBS. Then, preparations were washed and mounted on gelatin-covered slides. After drying, sections were dehydrated with ethanol solutions (50%, 70%, 96%, absolute, absolute 5 min, respectively), cleared by xylene (2 × 20 min), and coverslipped using Depex (Fluka, Heidelberg, Germany). Specificity of the FosB antiserum (Santa Cruz, sc-48) was characterized earlier . Western blot studies also support the specificity ( http://datasheets.scbt.com/sc-48.pdf , 2017). Preabsorption experiment in the rat revealed that the blocking peptide (sc-48-P, Santa Cruz) prevented the immunolabeling. In line with this, omission and/or replacement of the primary or secondary antibodies by non-immune sera abolished the signal in all tests (images not shown). 4.3.3. Digital Imaging and Morphometry at Light Microscopic Level The DAB-labeled FosB immunohistochemistry was studied and digitalized by a Nikon Microphot FXA microscope with a Spot RT camera (Nikon, Tokyo, Japan). For each animal, five serial sections were photographed. The number of FosB positive nuclei was determined using non-edited digital images by manual cell counting. The whole cross-sectional surface areas of the lateral PAG, as well as the dorsal horn laminae I, II, and III, individually were measured, and summed (laminae I + II + II) at L4–L5 spinal segments were evaluated, as marked in and . Cell counting was carried out by a skilled neurophysiologist who was not informed about the identity of preparations. For documentation and publication purposes, the micro-photos were grey-scaled and contrasted using Adobe Photoshop 7.0.1 software. 4.3.4. Electron Microscopy Sciatic nerve samples were placed in fixing solution (2.5% glutaraldehyde + 2% formalin + 0.1 M PBS) immediately after dissection in +4 °C for 24 h. A post-fixation procedure was performed with 1% osmium tetroxide. After dehydration in ascending alcohol (50%–70%–90%–96%) and subsequent transfer to propylene oxide, samples were embedded in Araldite resin. Semithin sections (0.5 µm) were cut by ultramicrotome (Leica Ultracut R), stained with 1% toluidine blue (Sigma), and examined with a Nikon Eclipse 80i microscope. Ultrathin sections were prepared from the area of interest and were contrasted by 2.5% uranyl–acetic acid and lead citrate. Slides were examined using a JEM-1200 EX-II electron microscope. The following parameters were analyzed: percentage of the axon-myelin separation, mitochondrial number in myelinated axons, area of the unmyelinated fibers, and thickness of the basement membrane.
Animals were anesthetized with an overdose of isoflurane (Forane, Abbott Hungary, Budapest, Hungary) on week 8. Rats were transcardially perfused with 25 mL of phosphate-buffered saline (PBS), followed by 300 mL of 4% paraformaldehyde solution in Millonig buffer for 20 min. The brain and spinal cord were dissected and then placed into the same fixative solution for post-fixation for 72 h at 4 °C. The sciatic nerve was also removed from all animals and further processed for electron microscopy. A tissue block containing the midbrain was isolated from the brains by cutting at the frontal planes of the posterior border of the median eminence and the transverse fissure. A tissue block of the L4-L5 spinal cord segments was also dissected. Blocks were sectioned by a Leica VT S 1000 (Leica, Wetzlar, Germany) vibratome. Five series of 30 µm coronal sections, interspaced by 120 µm, were collected in anti-freeze solution and stored at −20 °C until further use.
Free-floating labeling for the chronic neuronal activation marker FosB was performed on a series of the midbrain and spinal cord sections, as published earlier . Briefly, sections were permeabilized with 0.5% Triton X-100 solution. Normal goat serum (2%, NGS, Jackson Immunoresearch, Europe Ltd., Suffolk, UK) was used to reduce non-specific binding. Subsequently, sections were treated with rabbit anti-FosB antibodies (1:500, Santa Cruz, sc-48 Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) in PBS with 2% NGS overnight. After washes, sections were treated with biotinylated goat anti-rabbit serum and avidin–biotin complex (Vectastain ABC Elite Kit Vector Lbs., Burlingame, CA, USA) according to the supplier’s protocol. The labeling was developed in Tris buffer (pH 7.4) with 0.02% 3, 3′ diamino-benzidine (DAB) (Sigma) and 0.00003 v / v % H 2 O 2 . The reaction was carried out under visual control and was stopped after 7 min with PBS. Then, preparations were washed and mounted on gelatin-covered slides. After drying, sections were dehydrated with ethanol solutions (50%, 70%, 96%, absolute, absolute 5 min, respectively), cleared by xylene (2 × 20 min), and coverslipped using Depex (Fluka, Heidelberg, Germany). Specificity of the FosB antiserum (Santa Cruz, sc-48) was characterized earlier . Western blot studies also support the specificity ( http://datasheets.scbt.com/sc-48.pdf , 2017). Preabsorption experiment in the rat revealed that the blocking peptide (sc-48-P, Santa Cruz) prevented the immunolabeling. In line with this, omission and/or replacement of the primary or secondary antibodies by non-immune sera abolished the signal in all tests (images not shown).
The DAB-labeled FosB immunohistochemistry was studied and digitalized by a Nikon Microphot FXA microscope with a Spot RT camera (Nikon, Tokyo, Japan). For each animal, five serial sections were photographed. The number of FosB positive nuclei was determined using non-edited digital images by manual cell counting. The whole cross-sectional surface areas of the lateral PAG, as well as the dorsal horn laminae I, II, and III, individually were measured, and summed (laminae I + II + II) at L4–L5 spinal segments were evaluated, as marked in and . Cell counting was carried out by a skilled neurophysiologist who was not informed about the identity of preparations. For documentation and publication purposes, the micro-photos were grey-scaled and contrasted using Adobe Photoshop 7.0.1 software.
Sciatic nerve samples were placed in fixing solution (2.5% glutaraldehyde + 2% formalin + 0.1 M PBS) immediately after dissection in +4 °C for 24 h. A post-fixation procedure was performed with 1% osmium tetroxide. After dehydration in ascending alcohol (50%–70%–90%–96%) and subsequent transfer to propylene oxide, samples were embedded in Araldite resin. Semithin sections (0.5 µm) were cut by ultramicrotome (Leica Ultracut R), stained with 1% toluidine blue (Sigma), and examined with a Nikon Eclipse 80i microscope. Ultrathin sections were prepared from the area of interest and were contrasted by 2.5% uranyl–acetic acid and lead citrate. Slides were examined using a JEM-1200 EX-II electron microscope. The following parameters were analyzed: percentage of the axon-myelin separation, mitochondrial number in myelinated axons, area of the unmyelinated fibers, and thickness of the basement membrane.
Statistical analysis was performed in GraphPad Prism 6.01 software. Two-way analysis of variance (ANOVA) was used to detect significant differences between groups. Multiple comparisons were performed by Tukey’s test. Data are presented as means ± S.E.M (standard error of the mean). Differences were considered statistically significant when p < 0.05.
|
Pharmacological inhibition of USP7 suppresses growth and metastasis of melanoma cells in vitro and in vivo | 6afd448a-238e-4e49-a9b6-dc326fcd579e | 8500953 | Anatomy[mh] | INTRODUCTION Melanoma, representing the most aggressive and the deadliest type of skin cancer, is a malignancy originating from the neural crest‐derived melanocytes in skin, uvea and mucosal tissues. The incidence rate of melanoma is lower than that of many common cancers worldwide, such as lung cancer, liver cancer and colorectal cancer, but the number of melanoma cases is increasing faster than any other types of cancer. Besides, melanoma has unusual age demography, which caused particular concern. Unlike other solid malignancies, where the majority of cases are diagnosed at the age older than 65, melanoma affects a higher proportion of younger patients, with a median age of diagnosis of 57 years. Although timely recognition, detection and appropriate surgery improved the outcomes, the prognosis of metastatic melanoma remains poor. More than 95% of melanoma patients with three or more sites of the metastatic disease die within 1 year. 75% of melanoma patients usually suffer from brain metastases that cause death in 95% of total cases. Over the past years, the inauguration of several novel systematic therapies such as BRAF‐targeted therapy and immunotherapy and advancement in local therapy have contributed to improve survival rate. Unfortunately, a large fraction of patients failed to benefit from these targeted therapies due to the skin and gastrointestinal toxicity and low efficiency. Only a minority of patients responded to the treatments. Therefore, there is an urgent need to figure out the disease pathogenesis to find new therapeutic targets and develop new drugs with low toxicity for melanoma treatment. Ubiquitination is an important type of protein post‐translational modification (PTM), which plays a crucial role in controlling substrate degradation to ensure protein homeostasis in the cell. Deubiquitinases (DUBs) can reverse the effect of E3 ligases to remove ubiquitin from ubiquitylated proteins to regulate the stability, subcellular localization or activity of modified proteins. The role of DUBs in cancer is multifaceted, which includes proliferation, cell cycle control, cell apoptosis and DNA damage response. USP7 (ubiquitin‐specific protease‐7) is one of the deubiquitinases that belong to the UBQ‐specific protease family and plays critical roles in cancers, neurological disorders, cell differentiation, immune dysfunction, etc.. , USP7 activity is highly context‐specific and exhibits its versatility in substrate selection. The multifaceted role of USP7 in various cancers is established, including lung cancer, colon cancer, breast cancer and leukaemia. Studies revealed numerous substrates of USP7 and demonstrated that the USP7 exhibits oncogenic properties. These findings make USP7 as an attractive target for pharmacological discoveries and specific treatment strategy designs. However, the role of USP7 and its therapeutic value for melanoma remain unclear. P22077 is a specific USP7 inhibitor, which has been identified by activity‐based chemical proteomics. P22077 inhibits neuroblastoma and colon carcinoma growth via inducing p53‐mediated apoptosis. Current studies have also shown that USP7 suppressed proliferation and the colony formation capacity of lung cancer cells. However, the antitumour effect of P22077 in melanoma has not yet been studied. We found that a high expression of USP7 in melanoma patients correlated with poor overall survival. Further, using P22077 to inhibit USP7 demonstrated a suppression in cell growth and an induction of cell cycle arrest and cell apoptosis via ROS accumulation–induced DNA damage. Moreover, P22077 also suppressed the metastasis of melanoma cells by inhibiting β‐catenin. These data indicate that USP7 inhibition could be a potential strategy for melanoma treatment.
MATERIALS AND METHODS 2.1 Human melanoma tissue microarray Human melanoma tissue microarray (malignant melanoma with skin tissue array, 80 cores) was purchased from www.alenabio.com . Immunohistochemistry (IHC) staining assays using anti‐human USP7 antibody (1:1000, ab4080, abcam) were performed following standard protocols (Wuhan Servicebio Technology Co., Ltd., Hubei, China) and slides scanned using an automatic digital slide scanner (Pannoramic MIDI II). The density was quantified by using Quant software (3D Histech), and the histochemistry score (H‐score) system was used to assess the protein levels. 2.2 Reagents and antibodies P22077 was purchased from Selleck (S713301). HDM2 (86934), ATM (2873), P‐ATM(13050), P‐ATR (30632), γ‐H2AX (9718), MMP9 (13667), vimentin (5741), β‐catenin (8480) and cyclin B (12231) antibodies were purchased from Cell Signaling (Cell Signaling Technology, Danvers, MA, USA). USP7(ab4080), caspase‐3 (ab13847) and cleaved‐caspase‐3 (ab214430) antibodies were purchased from abcam, and ATR (19787–1‐AP), cCdc‐2 (27334–1‐AP), P‐cCdc‐2 (21082–1‐AP) and GAPDH (60004–1‐Ig) antibodies were purchased from Protech. 2.3 Cell lines and cell culture The human malignant melanoma cell lines A375 and SK‐Mel‐28 were maintained in Dulbecco's modified Eagle's medium (BI, Israel) supplemented with 10% heat‐inactivated foetal bovine serum (FBS) (BI, Israel), 100 units/ml penicillin, 100 μg/ml streptomycin and 2 mM glutamine. The mouse cell line B16F10 was maintained in RPMI 1640 containing 10% FBS, 100 units/ml penicillin, 100 μg/ml streptomycin and 2 mM glutamine. All cells were grown in a humidified incubator containing 5% CO 2 at 37℃. 2.4 Cell viability assay Cell viability assays were assessed by the Cell Counting Kit‐8 reagent (Selleck) following the manufacturer's instructions. Cells were seeded into 96‐well plates (2.5 × 10 3 cells per well) and incubated overnight. Then, cells were cultured at various concentrations of P22077 or DMSO (control). 24, 48 and 72 h later, 10 μl of CCK‐8 was added into each well, and after 2 h of incubation, the absorbance was measured at 450 nm using the spectrophotometer (Beckman). Each experiment was performed in 6 replicates. 2.5 Apoptosis and cell cycle analysis Cells were seeded into 6‐well plates (3 × 10 5 cells per well) and incubated overnight at 37℃ in media containing 10% FBS at various concentrations of P22077 or DMSO (control). For cell cycle assays, the cells were harvested by trypsinization after 48 h. The cells were resuspended in 300 µl PBS, and subsequently, 700 µl 100% ethanol was added to the cell suspension. After incubation at 4℃. for at least overnight, the cells were washed with PBS three times. The next day, according to the manufacturer's instructions, the collected cells were incubated at room temperature in the dark with propidium iodide (PI) (Becton Dickinson and Company). The cell cycle profile was analysed using a flow cytometer (BD Biosciences). For apoptosis assays, cells were harvested after 48 h by trypsin digestion without EDTA. The collected cells were washed with cold PBS and incubated with an Annexin V/propidium iodide stain (Becton, Dickinson and Company), according to the manufacturer's instruction. Cell apoptosis was detected by flow cytometry and analysed using FlowJo software. 2.6 Wound healing 5 × 10 5 cells/well were seeded into six‐well plates and incubated under standard conditions overnight. The cells reached confluence using a 1000‐μL pipette tip to scrape the cell monolayer to create a wound. The cells were treated at various concentrations of P22077 or DMSO (control), and photomicrographs were taken at 0 h and 24 h. Representative images were captured using an inverted light microscope. 2.7 Transwell chamber invasion and migration assay The Transwell chamber was placed into the corresponding culture plate. The upper chamber contains the serum‐free medium and corresponding P22077 or DMSO (control), and the lower chamber contains media containing 30% FBS. The melanoma cells were seeded in the upper chamber. After 24 h, the cells were fixed with 4% PFA and stained by 0.1% crystal violet, and migrated cell number was counted by phase‐contrast microscope and statistically analysed. To perform the invasion assay, transwell chambers were precoated with ECM Matrix gel solution (Sigma‐Aldrich) for 24 h. Residual cells on the upper transwel L chambers were counted and statistically analysed in five random fields per chamber. 2.8 Immunoblotting The collected cells were treated with P22077 or DMSO for 48 h, and 200 ul of PMSF‐containing lysate was added to each sample and lysed on ice for 30 min. and The supernatant was collected by centrifugation of cells at 12000 rpm for 5 min at 4℃. Then, the BCA protein assay kit (Beyotime) was used to measure protein concentration. Proteins were detected by following the standard SDS‐PAGE and immunoblotting protocols and incubated with primary and secondary antibodies. 2.9 Detection of intracellular ROS generation Cells were seeded into 6‐well plates (3 × 10 5 cells per well) and incubated overnight at 37℃ in medium containing 10% FBS treated with P22077 or DMSO (control). The cells were processed with DCFH‐DA incubated for 20 min according to the manufacturer's instructions, and then, the intracellular fluorescence intensities were detected using a flow cytometer and analysed by FlowJo software. 2.10 Xenograft tumour model 1 × 10 6 A375 cells were surgically injected into the left renal capsule of 5‐week‐old female SD nude mice. The xenografts were allowed to grow for 2 weeks before randomizing the mice into a control group and a P22077 treatment group. Animals were treated with DMSO or P22077 by intraperitoneal (i.p.) injection every day for 12 days. At the end of the experiments, all mice were killed and weighed and photographed. 2.11 Syngeneic lung tumour metastasis models 5 × 10 5 B16F10 cells were injected into the tail veins of 6‐ to 8‐week‐old C57BL/6 mice maintained under specific pathogen–free conditions. Animals were treated with DMSO or P22077 by intraperitoneal (i.p.) injection every day for 16 days. At the end‐point, the mice were killed by CO 2 asphyxiation, the lung tumours were counted, and lung images were captured. 2.12 Statistical analysis Student's t tests and one‐ or two‐way ANOVA tests were conducted to analyse the data using the GraphPad Prism software (version 6.01). The quantified data are presented as the mean ±SEM. Differences were considered to be significant when p < 0.05.
Human melanoma tissue microarray Human melanoma tissue microarray (malignant melanoma with skin tissue array, 80 cores) was purchased from www.alenabio.com . Immunohistochemistry (IHC) staining assays using anti‐human USP7 antibody (1:1000, ab4080, abcam) were performed following standard protocols (Wuhan Servicebio Technology Co., Ltd., Hubei, China) and slides scanned using an automatic digital slide scanner (Pannoramic MIDI II). The density was quantified by using Quant software (3D Histech), and the histochemistry score (H‐score) system was used to assess the protein levels.
Reagents and antibodies P22077 was purchased from Selleck (S713301). HDM2 (86934), ATM (2873), P‐ATM(13050), P‐ATR (30632), γ‐H2AX (9718), MMP9 (13667), vimentin (5741), β‐catenin (8480) and cyclin B (12231) antibodies were purchased from Cell Signaling (Cell Signaling Technology, Danvers, MA, USA). USP7(ab4080), caspase‐3 (ab13847) and cleaved‐caspase‐3 (ab214430) antibodies were purchased from abcam, and ATR (19787–1‐AP), cCdc‐2 (27334–1‐AP), P‐cCdc‐2 (21082–1‐AP) and GAPDH (60004–1‐Ig) antibodies were purchased from Protech.
Cell lines and cell culture The human malignant melanoma cell lines A375 and SK‐Mel‐28 were maintained in Dulbecco's modified Eagle's medium (BI, Israel) supplemented with 10% heat‐inactivated foetal bovine serum (FBS) (BI, Israel), 100 units/ml penicillin, 100 μg/ml streptomycin and 2 mM glutamine. The mouse cell line B16F10 was maintained in RPMI 1640 containing 10% FBS, 100 units/ml penicillin, 100 μg/ml streptomycin and 2 mM glutamine. All cells were grown in a humidified incubator containing 5% CO 2 at 37℃.
Cell viability assay Cell viability assays were assessed by the Cell Counting Kit‐8 reagent (Selleck) following the manufacturer's instructions. Cells were seeded into 96‐well plates (2.5 × 10 3 cells per well) and incubated overnight. Then, cells were cultured at various concentrations of P22077 or DMSO (control). 24, 48 and 72 h later, 10 μl of CCK‐8 was added into each well, and after 2 h of incubation, the absorbance was measured at 450 nm using the spectrophotometer (Beckman). Each experiment was performed in 6 replicates.
Apoptosis and cell cycle analysis Cells were seeded into 6‐well plates (3 × 10 5 cells per well) and incubated overnight at 37℃ in media containing 10% FBS at various concentrations of P22077 or DMSO (control). For cell cycle assays, the cells were harvested by trypsinization after 48 h. The cells were resuspended in 300 µl PBS, and subsequently, 700 µl 100% ethanol was added to the cell suspension. After incubation at 4℃. for at least overnight, the cells were washed with PBS three times. The next day, according to the manufacturer's instructions, the collected cells were incubated at room temperature in the dark with propidium iodide (PI) (Becton Dickinson and Company). The cell cycle profile was analysed using a flow cytometer (BD Biosciences). For apoptosis assays, cells were harvested after 48 h by trypsin digestion without EDTA. The collected cells were washed with cold PBS and incubated with an Annexin V/propidium iodide stain (Becton, Dickinson and Company), according to the manufacturer's instruction. Cell apoptosis was detected by flow cytometry and analysed using FlowJo software.
Wound healing 5 × 10 5 cells/well were seeded into six‐well plates and incubated under standard conditions overnight. The cells reached confluence using a 1000‐μL pipette tip to scrape the cell monolayer to create a wound. The cells were treated at various concentrations of P22077 or DMSO (control), and photomicrographs were taken at 0 h and 24 h. Representative images were captured using an inverted light microscope.
Transwell chamber invasion and migration assay The Transwell chamber was placed into the corresponding culture plate. The upper chamber contains the serum‐free medium and corresponding P22077 or DMSO (control), and the lower chamber contains media containing 30% FBS. The melanoma cells were seeded in the upper chamber. After 24 h, the cells were fixed with 4% PFA and stained by 0.1% crystal violet, and migrated cell number was counted by phase‐contrast microscope and statistically analysed. To perform the invasion assay, transwell chambers were precoated with ECM Matrix gel solution (Sigma‐Aldrich) for 24 h. Residual cells on the upper transwel L chambers were counted and statistically analysed in five random fields per chamber.
Immunoblotting The collected cells were treated with P22077 or DMSO for 48 h, and 200 ul of PMSF‐containing lysate was added to each sample and lysed on ice for 30 min. and The supernatant was collected by centrifugation of cells at 12000 rpm for 5 min at 4℃. Then, the BCA protein assay kit (Beyotime) was used to measure protein concentration. Proteins were detected by following the standard SDS‐PAGE and immunoblotting protocols and incubated with primary and secondary antibodies.
Detection of intracellular ROS generation Cells were seeded into 6‐well plates (3 × 10 5 cells per well) and incubated overnight at 37℃ in medium containing 10% FBS treated with P22077 or DMSO (control). The cells were processed with DCFH‐DA incubated for 20 min according to the manufacturer's instructions, and then, the intracellular fluorescence intensities were detected using a flow cytometer and analysed by FlowJo software.
Xenograft tumour model 1 × 10 6 A375 cells were surgically injected into the left renal capsule of 5‐week‐old female SD nude mice. The xenografts were allowed to grow for 2 weeks before randomizing the mice into a control group and a P22077 treatment group. Animals were treated with DMSO or P22077 by intraperitoneal (i.p.) injection every day for 12 days. At the end of the experiments, all mice were killed and weighed and photographed.
Syngeneic lung tumour metastasis models 5 × 10 5 B16F10 cells were injected into the tail veins of 6‐ to 8‐week‐old C57BL/6 mice maintained under specific pathogen–free conditions. Animals were treated with DMSO or P22077 by intraperitoneal (i.p.) injection every day for 16 days. At the end‐point, the mice were killed by CO 2 asphyxiation, the lung tumours were counted, and lung images were captured.
Statistical analysis Student's t tests and one‐ or two‐way ANOVA tests were conducted to analyse the data using the GraphPad Prism software (version 6.01). The quantified data are presented as the mean ±SEM. Differences were considered to be significant when p < 0.05.
RESULTS 3.1 USP7 is overexpressed and correlates with poor prognosis in melanoma To explore the relationship between USP7 and malignancy of melanoma, first, we examined the expression level of USP7 in a human tissue microarray comprising 40 melanoma specimens from 40 patients (23 female and 17 male, age: 15–80; mean 55). The histochemical staining images were randomly selected, which showed the USP7 expression level (Figure ). The USP7 expression significantly upregulated in 77.5% of all melanomas (31/40), whereas USP7 expression was significantly downregulated in non‐malignant melanocytes (Figure ). Besides, compared with the normal epithelial cells, the USP7 protein level was also higher in melanoma cell lines (Figure ). Next, we analysed the USP7 expression in melanoma from TCGA data programmed online tool GEPIA and found that tumours with higher USP7 expression correlated with poor overall survival in melanoma (Figure ). Altogether, these results demonstrate that USP7 expression is upregulated in melanoma and is correlated with patients’ poor outcomes. 3.2 Pharmacological inhibition of USP7 by P22077 inhibits proliferation and induces cell cycle arrest and apoptosis in melanoma cells To evaluate the potential therapeutic role of USP7 inhibition in melanoma, we treated the TP53‐mutated melanoma cell line SK‐Mel‐28 and wild‐type cell line A375 with P22077. P22077 treatment significantly suppressed the proliferation of A375 and SK‐Mel‐28 cells in a dose‐dependent manner but caused no cytotoxic effect on normal human skin keratinocyte cell line (HaCat) (Figure ). Moreover, we found that USP7 knockdown also inhibited cell proliferation (Figure A,B). To investigate the effect of P22077‐induced cytotoxicity in melanoma cells, we analysed the apoptosis and cell cycle status in A375 and SK‐Mel‐28 cell lines. P22077 and USP7 knockdown also inhibited the proliferation of A375 and SK‐Mel‐28 cells by increasing the percentage of cells in G2/M and decreased the percentage of cells in G0/G1 and S phase (Figure , supplement Figure B, supplement Figure C). This was confirmed by specific cell cycle protein signalling with reduced phosphorylation of cell cycle regulatory proteins Cdc2 and cyclin B (Figure ). We also found that USP7 inhibition and knockdown increased significantly early (Annexin+/PI−) and late apoptosis (Annexin+/PI+) regardless of P53 status (Figure , supplement Figure A, supplement Figure D). Apoptosis was further confirmed by immunoblotting showing induction of the cleavage of caspase‐3. At the same time, the inhibition efficiency of P22077 was confirmed by HMD2, the known substrate of USP7. In addition, the expression of USP7 has not been affected by P22077 (Figure ). Together, the results suggest that USP7 inhibitor P22077 potentially induces cell cycle arrest and apoptosis in melanoma cell lines independent of the TP53 status. 3.3 P22077 induced DNA damage by increasing intracellular ROS level in melanoma cells At moderate levels, ROS contributes to tumour growth by acting as signalling molecules or promoting the mutation of genomic DNA, but overload ROS also prompts oxidative damage to biomacromolecules in the cell, which leads to cell dysfunction and death. Based on the previous results that P22077 treatment suppressed proliferation and induced apoptosis in melanoma cells, we hypothesized that P22077 increased ROS levels to induce melanoma cell cycle arrest and apoptosis. We applied flow cytometry to assess ROS level by intracellular DCF fluorescence and found a significant and dose‐dependent increase of intracellular levels of ROS in melanoma cell lines after an increasing concentration of P22077 treatment (Figure , supplement Figure A,B). Furthermore, the cell apoptosis induced by P22077 could be partially blocked by the antioxidant N‐acetylcysteine (NAC), a known scavenger of ROS (Figure , supplement Figure C). ROS generation is well known to lead to DNA damage and subsequently antitumour activity. ATM and ATR play critical roles in DNA damage responses, and these checkpoint pathways are activated by ROS‐induced DNA damage. In melanoma cell lines, P22077 treatment remarkably activated the ATM/ATR signalling pathway. The results showed that P22077 induced the upregulation of p‐ATM, p‐ATR and γH2AX expression levels (Figure ). These data indicate that the antimelanoma effect of P22077 is mediated through DNA damage induced by intracellular ROS. 3.4 P22077 significantly inhibits melanoma tumour growth in vivo Next, we investigated the in vivo antitumour effect of P22077. First, the melanoma cell line A375 was used to create a subcutaneous xenograft model in nude mice. The tumour‐bearing mice were treated with vehicle and P22077 (10 mg/kg) by intravenous (IV) injection for 12 days (Figure ). The results showed that P22077 treatment reduced the mouse tumour growth rate without bodyweight changes, indicating the safety of the P22077 treatment group (Figure B,C). Both tumour size and tumour weight were significantly decreased in the P22077‐treated group compared with the control group (Figure D,E). Further, the haematoxylin and eosin staining revealed a relatively much higher density of necrotic cells in P22077‐treated xenograft tumour sections compared with the control group (Figure ). The TdT‐mediated dUTP nick‐end labelling (TUNEL) assay showed that the percentage of apoptotic tumour cells was increased in the P22077 group compared with the vehicle group (Figure ). These results demonstrate the efficacy of P22077 in inhibiting tumour growth in vivo without toxicity at therapeutic doses. 3.5 P22077 inhibits metastasis and invasion in melanoma in vitro and in vivo The epithelial‐mesenchymal transition (EMT) is a key process for promoting tumour cell invasion and metastasis. Based on the aforementioned results, we evaluated the effect of P22077 on the capability of migration by performing a wound‐healing assay. The results revealed a decreased closure of the wound area compared with control cells after P22077 treatment and knockdown (Figure , supplement Figure ). Next, the effect of USP7 inhibition on the invasive ability of cells was accessed by transwell assays. P22077 and USP7 knockdown showed a dramatic attenuation of the invasive capacities of A375 and SK‐Mel‐28 cells compared with untreated cells (Figure , supplement Figure ). To further explore the underlying molecular mechanism of the P22077 effect on the EMT, the EMT‐associated protein markers were examined after P22077 treatment. The results demonstrated that the inhibition of USP7 significantly induced the expression of the epithelial marker snail and reduced the mesenchymal markers vimentin and MMP9 significantly (Figure ). The expression of β‐catenin, a key nuclear effector of canonical Wnt signalling, was also downregulated, which implied that P22077 could inactivate the Wnt/β‐catenin pathway to inhibit EMT (Figure ). To further demonstrate the effect of P22077 on melanoma metastasis, the B16F10 cells were injected through the lateral tail vein to induce melanoma lung metastasis in C57BL/6 mice. The results showed that after P22077 treatment, the lung metastases were decreased significantly and without toxicity (Figure A‐C). These findings suggest that P22077 exerts antimetastatic activity in vitro and in vivo.
USP7 is overexpressed and correlates with poor prognosis in melanoma To explore the relationship between USP7 and malignancy of melanoma, first, we examined the expression level of USP7 in a human tissue microarray comprising 40 melanoma specimens from 40 patients (23 female and 17 male, age: 15–80; mean 55). The histochemical staining images were randomly selected, which showed the USP7 expression level (Figure ). The USP7 expression significantly upregulated in 77.5% of all melanomas (31/40), whereas USP7 expression was significantly downregulated in non‐malignant melanocytes (Figure ). Besides, compared with the normal epithelial cells, the USP7 protein level was also higher in melanoma cell lines (Figure ). Next, we analysed the USP7 expression in melanoma from TCGA data programmed online tool GEPIA and found that tumours with higher USP7 expression correlated with poor overall survival in melanoma (Figure ). Altogether, these results demonstrate that USP7 expression is upregulated in melanoma and is correlated with patients’ poor outcomes.
Pharmacological inhibition of USP7 by P22077 inhibits proliferation and induces cell cycle arrest and apoptosis in melanoma cells To evaluate the potential therapeutic role of USP7 inhibition in melanoma, we treated the TP53‐mutated melanoma cell line SK‐Mel‐28 and wild‐type cell line A375 with P22077. P22077 treatment significantly suppressed the proliferation of A375 and SK‐Mel‐28 cells in a dose‐dependent manner but caused no cytotoxic effect on normal human skin keratinocyte cell line (HaCat) (Figure ). Moreover, we found that USP7 knockdown also inhibited cell proliferation (Figure A,B). To investigate the effect of P22077‐induced cytotoxicity in melanoma cells, we analysed the apoptosis and cell cycle status in A375 and SK‐Mel‐28 cell lines. P22077 and USP7 knockdown also inhibited the proliferation of A375 and SK‐Mel‐28 cells by increasing the percentage of cells in G2/M and decreased the percentage of cells in G0/G1 and S phase (Figure , supplement Figure B, supplement Figure C). This was confirmed by specific cell cycle protein signalling with reduced phosphorylation of cell cycle regulatory proteins Cdc2 and cyclin B (Figure ). We also found that USP7 inhibition and knockdown increased significantly early (Annexin+/PI−) and late apoptosis (Annexin+/PI+) regardless of P53 status (Figure , supplement Figure A, supplement Figure D). Apoptosis was further confirmed by immunoblotting showing induction of the cleavage of caspase‐3. At the same time, the inhibition efficiency of P22077 was confirmed by HMD2, the known substrate of USP7. In addition, the expression of USP7 has not been affected by P22077 (Figure ). Together, the results suggest that USP7 inhibitor P22077 potentially induces cell cycle arrest and apoptosis in melanoma cell lines independent of the TP53 status.
P22077 induced DNA damage by increasing intracellular ROS level in melanoma cells At moderate levels, ROS contributes to tumour growth by acting as signalling molecules or promoting the mutation of genomic DNA, but overload ROS also prompts oxidative damage to biomacromolecules in the cell, which leads to cell dysfunction and death. Based on the previous results that P22077 treatment suppressed proliferation and induced apoptosis in melanoma cells, we hypothesized that P22077 increased ROS levels to induce melanoma cell cycle arrest and apoptosis. We applied flow cytometry to assess ROS level by intracellular DCF fluorescence and found a significant and dose‐dependent increase of intracellular levels of ROS in melanoma cell lines after an increasing concentration of P22077 treatment (Figure , supplement Figure A,B). Furthermore, the cell apoptosis induced by P22077 could be partially blocked by the antioxidant N‐acetylcysteine (NAC), a known scavenger of ROS (Figure , supplement Figure C). ROS generation is well known to lead to DNA damage and subsequently antitumour activity. ATM and ATR play critical roles in DNA damage responses, and these checkpoint pathways are activated by ROS‐induced DNA damage. In melanoma cell lines, P22077 treatment remarkably activated the ATM/ATR signalling pathway. The results showed that P22077 induced the upregulation of p‐ATM, p‐ATR and γH2AX expression levels (Figure ). These data indicate that the antimelanoma effect of P22077 is mediated through DNA damage induced by intracellular ROS.
P22077 significantly inhibits melanoma tumour growth in vivo Next, we investigated the in vivo antitumour effect of P22077. First, the melanoma cell line A375 was used to create a subcutaneous xenograft model in nude mice. The tumour‐bearing mice were treated with vehicle and P22077 (10 mg/kg) by intravenous (IV) injection for 12 days (Figure ). The results showed that P22077 treatment reduced the mouse tumour growth rate without bodyweight changes, indicating the safety of the P22077 treatment group (Figure B,C). Both tumour size and tumour weight were significantly decreased in the P22077‐treated group compared with the control group (Figure D,E). Further, the haematoxylin and eosin staining revealed a relatively much higher density of necrotic cells in P22077‐treated xenograft tumour sections compared with the control group (Figure ). The TdT‐mediated dUTP nick‐end labelling (TUNEL) assay showed that the percentage of apoptotic tumour cells was increased in the P22077 group compared with the vehicle group (Figure ). These results demonstrate the efficacy of P22077 in inhibiting tumour growth in vivo without toxicity at therapeutic doses.
P22077 inhibits metastasis and invasion in melanoma in vitro and in vivo The epithelial‐mesenchymal transition (EMT) is a key process for promoting tumour cell invasion and metastasis. Based on the aforementioned results, we evaluated the effect of P22077 on the capability of migration by performing a wound‐healing assay. The results revealed a decreased closure of the wound area compared with control cells after P22077 treatment and knockdown (Figure , supplement Figure ). Next, the effect of USP7 inhibition on the invasive ability of cells was accessed by transwell assays. P22077 and USP7 knockdown showed a dramatic attenuation of the invasive capacities of A375 and SK‐Mel‐28 cells compared with untreated cells (Figure , supplement Figure ). To further explore the underlying molecular mechanism of the P22077 effect on the EMT, the EMT‐associated protein markers were examined after P22077 treatment. The results demonstrated that the inhibition of USP7 significantly induced the expression of the epithelial marker snail and reduced the mesenchymal markers vimentin and MMP9 significantly (Figure ). The expression of β‐catenin, a key nuclear effector of canonical Wnt signalling, was also downregulated, which implied that P22077 could inactivate the Wnt/β‐catenin pathway to inhibit EMT (Figure ). To further demonstrate the effect of P22077 on melanoma metastasis, the B16F10 cells were injected through the lateral tail vein to induce melanoma lung metastasis in C57BL/6 mice. The results showed that after P22077 treatment, the lung metastases were decreased significantly and without toxicity (Figure A‐C). These findings suggest that P22077 exerts antimetastatic activity in vitro and in vivo.
DISCUSSION Melanoma is one of the most malignant diseases worldwide, the most aggressive and deadliest type of skin cancer. Uncontrolled tumour growth and distant metastasis remain to be huge obstacles to effective treatments. Although the favourable results were obtained from immunotherapy, in general, only a small subset of patients showed a promising response. To identify novel targets that contribute to the growth and metastasis of melanoma to develop innovative strategies is still needed. In this study, we found that a notable number of melanoma patients contain high USP7 expression, which correlates with reduced overall survival. Pharmacological inhibition of USP7 by P22077 reduces melanoma cell proliferation and induces cell cycle arrest and apoptosis independent of the TP53 status. The antimelanoma activity of P22077 might be mediated through induced intracellular ROS leading to DNA damage and subsequent cell death. P22077 is also active without toxicity at a therapeutic dose against melanoma xenograft and inhibits invasion and metastasis of melanoma cells in vitro and in vivo. Our data suggest that targeting USP7 would provide a potential strategy for melanoma treatment. USP7 is the most widely studied deubiquitinating enzyme with numerous substrates including viral proteins, transcription factors and epigenetic modulators. In a wide variety of cancers, USP7 acts as an oncogene a good therapeutic target for cancer, and the high expression of USP7 leads to the progression of cancer. However, USP7 plays a variety of roles in tumours through the stabilization of different substrates. For example, the tumour suppressor protein p53 was the first identified deubiquitination substrate for USP7, this attributed to USP7 as a tumour suppressor, given its ability to increase the stability of p53, resulting in repression of cancer cell growth and activation of apoptotic pathways. A later study showed the p53 E3 ligase mouse double minute 2 homolog (MDM2) is also regulated by USP7‐mediated deubiquitination, leading to the degradation of p53 and reversal of the above‐mentioned cellular phenotype. Though, in normal conditions without DNA damage induction, the USP7 has a higher affinity with MDM2. Our data showed an increase in USP7 levels in melanoma cells independent of TP53 status, which indicated that USP7 may be active through other substrates to act as an oncogene in melanoma. Previous studies also have shown that USP7 regulates serval substrates to promote carcinogenesis independent of p53. These include controlling subcellular localization of PTEN to promote AML progression, stabilization of the histone demethylase PHF8 to promote breast carcinogenesis, and stabilization of N‐Myc to promote neuroblastoma progression, among others. , , Though we found an oncogenic role of USP7 with melanoma, nonetheless, the substrate requirement for USP7 to promote the progression in melanoma required further exploration. USP7 is a promising anticancer therapeutic target because of its aberrant expression and the oncogenic role in various cancers. There are various small‐molecular inhibitors such as HBX41108, P5091, FT671 and P22077 that have been reported to inhibit USP7 in cells. , , , In recent years, several groups reported the structures of USP7 in complex with small‐molecule inhibitors , and these structures give guidance to further develop new inhibitors. The P22077 covalently modifies the catalytic cysteine of USP7 and induces a conformational switch in the enzyme associated with active site rearrangement. This makes P22077 well‐characterized tool compounds for exploring the USP7 function. In melanoma, P22077 induced an excessive ROS, which further leads to DNA damage. The excessive amounts of ROS enhance cellular oxidative stress and directly activate ATM and ATR for the DNA damage response pathway. Previous studies have shown that USP7 acts as an important regulator for DNA damage response, and inhibition or knockdown of USP7 leads to increased DNA damage via destabilization of several regulators of DNA damage such as RAD18 (replication‐associated repair), CSB (nucleotide excision repair) and ALKBH2/3 (alkylation repair). , , Moreover, USP7 also induced unfolded protein accumulation causing ER stress in cancer cells, which leads to oxidative stress to induce DNA damage. On the other hand, we also found that P22077 led inhibition of invasion and migration in vitro and in vivo in melanoma. Several substrates contributed to invasion and migration regulated by USP7, such as EZH2 in prostate cancer, and LSD1 in glioblastoma. , On the contrary, the role of ROS in triggering signalling pathways, such as activation of MAPK, ERK, JNK and p‐38 MAPK, via growth factor‐mediated stimulation of receptor tyrosine kinases (RTKs) for cell migration and invasion has been well established. Those studies indicated USP7 as a multifaceted regulator of tumorigenesis that may act through different mechanisms at the same time. Collectively, this study demonstrated USP7 as a possible oncogenic molecule for melanoma progression, and inhibition of USP7 might be a potential strategy for the suppression of melanoma growth. Importantly, USP7 inhibitor P22077 possess antimelanoma activities in vitro and in vivo, induces apoptosis and cell cycle arrest by DNA damage and markedly impairs melanoma cell migration and invasion, indicating a prospective value of the application of P22077 as a promising novel effective strategy for melanoma therapy.
The authors declared that they have no competing interests.
Minmin Xiang: Conceptualization (lead); Data curation (lead); Formal analysis (supporting). Long Liang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal). Xinwei Kuang: Investigation (equal); Methodology (equal). Zuozhong Xie: Investigation (equal); Methodology (equal). Jing Liu: Supervision (equal). Shuang Zhao: Methodology (equal). Juan Su: Formal analysis (equal). Xiang Chen: Funding acquisition (equal); Investigation (equal). Hong Liu: Investigation (equal); Methodology (equal); Project administration (equal).
Figure S1 Click here for additional data file. Figure S2 Click here for additional data file. Figure S3 Click here for additional data file.
|
Lipid-based nano-formulation platform for eplerenone oral delivery as a potential treatment of chronic central serous chorioretinopathy: | 588d1f54-5e81-40b7-ad9a-0e5465b9df68 | 8023249 | Pharmacology[mh] | Introduction Chronic central serous chorioretinopathy (CSCR), a vision hostile disease, is a common cause of visual impairment in the working-age population and has been estimated as the fourth most frequently encountered retinopathy after age-related macular degeneration, diabetic retinopathy, and retinal vein occlusion (Gemenetzi et al., ; Salehi et al., ). CSCR is characterized by accumulation of the subretinal fluid causing a localized area of retinal detachment (Bouzas et al., ; Bousquet et al., ). Unfortunately, the causative factor behind CSCR is not completely understood (idiopathic). The use of mineralocorticoid receptor antagonist in CSCR treatment refers to the existence of an independent renin–angiotensin aldosterone system at the ocular level. This system has greater expression in retinal and uveal tissue and acts on the ocular vasculature, aqueous humor, and intraocular pressure control (Savaskan et al., ; Nakashima et al., ). Eplerenone (EPL) is a selective mineralocorticoid receptor antagonist. EPL is analogues to the commonly used diuretic spironolactone, with an increased mineralocorticoid receptor selectivity and higher affinity (Cook et al., ; Yang and Eliott, ). It has the ability to act at neuroretinal cell types (Zhao et al., ) modifying the above-mentioned physiopathological processes (Yang and Eliott, ; Campos Polo et al., ; Chatziralli et al., ). Multiple studies have investigated the success and safety of oral EPL in the treatment of patients with chronic CSCR (Breukink et al., ; Salz et al., ; Chin et al., ; Pichi et al., ; Daruich et al., ; Pichi et al., ), where a rapid subretinal fluid resolution and improvement in the visual acuity have been observed (Zhao et al., ; Haghighi et al., ). Unfortunately, EPL is class II drug according to biopharmaceutical classification system with limited aqueous solubility (less than 1 mg/mL) (Khames, ), having a maximum aqueous solubility of 0.00908 mg/ml only (Kendre & Chaudhari, ), in addition to being extensively metabolized into the liver to inactive metabolites (Almeida et al., ; Salehi et al., ). This leads to poor oral bioavailability and consequently low therapeutic response (Özdemir et al., ). Previous studies succeeded to improve its oral bioavailability and biodistribution through the formulation of EPL nano-emulsion and EPL solid dispersion (Almeida et al. ; Kendre & Chaudhari, ; Haghighi et al., ; Özdemir et al., ). EPL oral bioadhesive pellets succeeded to improve EPL solubility and hence the bioavailability (Kendre & Chaudhari, ). Nanostructured lipid carriers (NLCs) are modified solid lipid nanoparticles, compromising solid lipid and liquid lipid (oil) matrix, such like a regime enriched fat. After oral administration, these lipids have the ability to impel the bile secretion in the small intestine then the NLCs loaded with the drug become coupled with bile salts to form mixed micelles (Khan et al., ), allowing the intact NLCs to be transferred directly to the portal circulation through paracellular route or become gripped by microfold cells avoiding the first pass metabolism leading to a higher oral bioavailability (Khan et al., ; Ghasemiyeh & Mohammadi-Samani, ). In addition, NLCs with certain surfactants and lipids have been reported to efficiently avoid the efflux by P-glycoprotein and hence improvement in oral drug absorption (Bayón-Cordero et al., ). In addition, this unique composition allows to initiate a less organized structure diminishing the delocalization of the encapsulated drug and improving the loading ability (Radtke et al., ). Hence, NLCs are superior over other traditional carriers regarding enhancing the solubility of hydrophobic drugs, imparting controlled and sustained drug release characteristics and enhancing drug permeability. Moreover, NLCs could maintain the stability of drug during storage, by protecting it from chemical or enzymatic degradation (Radtke et al., ; Fang et al., ; Khan et al., ). In our study, EPL-loaded NLCs were formulated by emulsification solvent evaporation technique (Shahgaldian et al., ), using glyceryl monostearate (GMS) as solid lipid and Miglyol ® 812 as liquid lipid. A d -optimal statistical study design was employed to study the effects of different formulation, and process variables and their interaction with limited number of experimental runs. The particle size, polydispersity index, zeta potential, and drug entrapment within the prepared NLCs were evaluated. Furthermore, the physicochemical properties of the drug within the optimized NLCs system were studied using infrared spectroscopy and X-ray diffraction. The morphology of the prepared NLCs was evaluated using TEM. Finally, an ex-vivo permeation study through rabbit intestine was performed to prove the capability of the optimized system to cross the intestinal mucosa following oral delivery of EPL. To our knowledge, this is the first study to fabricate oral EPL-loaded NLCs as a potential treatment for CSCR.
Materials and methods 2.1. Materials Eplerenone was obtained from Shenzhen Oriental Pharmaceutical Co., Ltd., Guangdong, China. Glyceryl monostearate (Imwitor ® 900K) was purchased from Changwei Pharmaceutical Excipients Technology Co., Ltd. (Shanghai, China). Cremophor ® RH40 (Polyoxyl 40 Hydrogenated Castor Oil USP/NF), Pluronic ® F127 and Miglyol ® 81N2 were purchased from BASF chemical company (Ludwigshafen, Germany). Solutol ® HS15, Disodium hydrogen phosphate, Sodium chloride and Potassium dihydrogen phosphate were purchased from Sigma Aldrich VR (St. Louis, MO). Ethanol and acetone were obtained from El-Nasr Pharmaceutical Chemicals, Cairo, Egypt. All other chemicals were of analytical grade and were used as received. 2.2. Preparation of EPL-loaded NLCs EPL-loaded NLCs were prepared using the emulsification solvent evaporation technique using glyceryl monostearate (GMS) as solid lipid and Miglyol ® 812N as liquid lipid. In brief, 25 mg of EPL were dispersed in the determined amount of Miglyol ® 812N, then added to determined amount of molten GMS kept at 80°C using thermostatically controlled magnetic stirrer (WiseStir, Wisd Lab. Instruments, Tulsa, OK). Exactly, 10 mL mixture of ethanol and acetone (1:1 v/v) were added to the molten lipids maintained at 80 °C and stirred until complete dissolution in the organic phase. The organic phase was then added to 20 mL of an aqueous solution containing the chosen surface-active agent(s) to form a primary o/w emulsion under stirring at 1000 rpm for 1 min using magnetic stirrer. The obtained emulsion was subsequently subjected to 3 min of sonication using probe sonicator at room temperature (Probe Sonicator Ultrasonic Processor model VCX 750, Newtown, CT) adjusted at 40 W. Following this, the formed emulsion was stirred using magnetic stirrer at 500 rpm for 2 h at room temperature to allow evaporation of the organic solvent and formation of NLCs (Aburahma et al., ). Each formula was prepared three times, and the results are presented as mean ± SD ( n = 3). 2.3. Evaluation of the prepared EPL-loaded NLCs 2.3.1. Determination of particle size, polydispersity index, and zeta potential The mean particle size (PS), polydispersity index (PDI), and zeta potential (ZP) were determined using dynamic light scattering (DLS, Malvern Zetasizer, Malvern, UK) at 25 °C (Imam et al., ). Before measurement, 0.1 mL of the dispersion was properly diluted with distilled water (10 mL) in a glass tube and shaken to have an appropriate scattering intensity. Data presented as mean values ( n = 3± SD). 2.3.2. Entrapment efficiency The entrapment efficiency ( EE% w/w) of EPL in the prepared EPL-loaded NLCs was determined indirectly by measuring the concentration of free drug in the aqueous phase of the NLCs dispersion. A definite volume (1 mL) of the prepared NLCs dispersion was centrifuged using cooling centrifuge (Sigma 3-30 KS, Sigma Laborzentrifugen GmbH, Osterode am Harz, Germany) at 22,000 rpm for 1 h at 4 °C. The supernatant was separated and properly diluted with ethanol, then the un-entrapped drug concentration was estimated spectrophotometrically at λ max 241 nm. The EE% was calculated using the following equation: The E E % = W initial − W free W initial × 100 where W initial is the initial drug amount used in the preparation, and W free is the un-entrapped drug amount. 2.4. Statistical design and optimization of EPL-loaded NLCs The response surface methodology with polynomial equations is a beneficial numerical mean to investigate the effect of independent variables on the dependent variables (responses) based on a limited number of trails (Heurtault et al., ; Huang et al., ). Using d -optimal mixture design, maximum prediction power could be obtained in selected set of experimental runs, minimizing the discrepancy associated with estimates of coefficients in the model (Bodea & Leucuta, ). A d -optimal statistical design was employed to study the effect of different formulation variables on the prepared EPL-loaded NLCs. In this design, three liquid lipid to solid lipid ratios (1:1, 1:1.5, and 1:2 w/w), three surfactant types (Pluronic ® F127, Cremophor ® RH40 and Solutol ® HS15) and three concentrations of surfactant (0.2, 0.4, and 0.6%w/v) were evaluated. shows the different levels of the studied factors as well as the constraints applied in selection of the optimized EPL-loaded NLCs (minimize the PS, maximize the EE % and for the PDI to be less than 0.5). The composition of the prepared EPL-loaded NLCs is presented in along with the measured responses. For this study, 19 runs were considered. 2.5. In vitro characterization of the optimized EPL-loaded NLC system The optimized EPL-loaded NLCs system was prepared and evaluated for the PS, PDI, ZP, and EE%. The residual difference between the predicted and the measured responses was measured to ensure the validity of the design (Sayed et al., ). 2.6. In-vitro drug release study The cumulative % EPL released from the optimized EPL-loaded NLCs system was performed using dialysis bag technique (Üner et al., ). Briefly 2 mL of the NLCs system (equivalent to 2.5 mg EPL) were placed in a dialysis bag and the ends of dialysis bag were sealed with standard closures, and then immersed in 100 mL phosphate buffer solution (PBS pH 7.4) in a tightly closed flask kept in thermostatically controlled water bath shaker (Gesellschatt Laboratories, Berlin, Germany) at 37 ± 0.5 °C and shaken at 100 rpm. At different time intervals (0.25 h, 0.5 h, 1 h, 1.5 h, 2 h, 4 h, 6 h, 8 h, and 24 h), 3 mL sample were withdrawn from the release medium and replaced by equal volume of fresh medium. For comparison, EPL aqueous suspension (2.5 mg/2 mL) was prepared and EPL release was determined using the same procedure mentioned above and compared to that of the EPL-loaded NLCs optimized system. Samples were analyzed using UV-spectrophotometry at λ max 245 nm using UV/VIS spectrophotometer (Shimadzu, UV-1601 PC, Kyoto, Japan). 2.7. Morphological evaluation of the optimized EPL-loaded NLCs system Morphological parameters of the optimized EPL-loaded NLCs system were observed using transmission electron microscopy (TEM) at 70 kV (JEOL JEM1230, Tokyo, Japan). The optimized EPL-loaded NLCs system was first diluted with 10-fold its volume distilled water, then added to copper grid coated with collodion film after being stained by 2% (w/v) phosphotungestic acid solution (negative stain) and dried at room temperature, then observed under the TEM. 2.8. Fourier infrared spectroscopy and X-ray diffraction The FTIR spectroscopy study was employed using a Bruker FTIR spectrophotometer (Model 22; Bruker, Coventry, UK) to detect the chances of any chemical interactions between EPL and other ingredients used in the preparation of the optimized NLCs system. The IR spectra of plain EPL, physical mixture (composed of EPL, GMS and Pluronic ® F127 in equal amounts) and the freeze-dried optimized EPL-loaded NLCs system were recorded. Samples were mixed with KBr and compressed into disk, then scanned in the range 4000–400 cm −1 at ambient temperature. For freeze-drying, EPL-loaded NLCs system was first frozen at −22 °C for 24 h and then placed in the freeze-dryer (Novalyphe-NL 500, Halprook, NY) for 24 h. The freeze-dryer was operated under vacuum at pressure of 7 × 10 −2 mBAR. The condenser temperature was adjusted at − 45 °C. X-ray powder diffraction (X-RPD) experiments were carried out using X-ray diffractometer (XGEN-4000, Scintag Corp., Sunnyvale, CA), with Cu Ka radiation at 40 mA current and 45 kV Voltage. Diffraction patterns for plain EPL, physical mixture (composed of EPL, GMS, and Pluronic ® F127 in equal amounts) and the freeze-dried optimized EPL-loaded NLCs system were recorded as the X-ray intensity as a function of 2 θ angle covering from 2.0° to 50.0°. The scanning rate was 6°/minute. The X-ray diffraction patterns were displayed using Diffrac AT software. 2.9. Ex vivo permeation study 2.9.1. Animals Male albino rabbits (2.5–3 kg) were housed in a controlled temperature and humidity room (25 °C, 55% air humidity with free access to water ad libitum). Experiments were approved by the Research Ethics Committee (REC; PT 2682) at Faculty of Pharmacy, Cairo University (Cairo, Egypt). One day before the experiment, rabbits were fasted overnight with free access to water. Rabbits received an intramuscular injection of ketamine (35 mg/kg) as an anesthetizing agent and 5 mg/kg xylazine as a relaxing agent, then sacrificed and the small intestine was removed by a midline abdominal incision. The small intestine cleaned carefully using a syringe filled with warm (37 °C) saline (pH = 7.4), then the intestinal segments (∼7.8 cm 2 ) were separated. 2.9.2. Study design Accurately, 4 mL of EPL-loaded optimized NLCs system (equivalent to 5 mg of drug) were filled in an intestinal segment (of area 7.8 cm 2 ) via micropipette, then the two sides of the intestine were tangled with a thread. Each intestinal sac was placed in a bottle containing 500 mL of normal saline (pH = 7.4) maintained at 37 ± 0.5 °C. The bottle was placed in a thermostatically controlled shaking water bath (GFL, Gesellschatt Laboratories, Berlin, Germany) maintained at 37 ± 0.5 °C and 100 rpm. Samples (3 mL) were withdrawn at specific time intervals (0.5 h, 1 h, 2 h, 4 h, 6 h, 8 h, and 24 h) and replaced by equal volume of fresh medium. In order to determine the amount of EPL permeated through the intestine, samples were diluted with equal amount of acetonitrile (HPLC grade) followed by sonication for 10 min in order to detect the total drug permeated (as free and vesicles-encapsulated drug). Samples were analyzed according to a previously validated HPLC method (Rane et al., ). For comparison, EPL permeation from EPL aqueous suspension containing EPL 1.25 mg/mL was examined using the same procedure mentioned above. Chromatographic separation was achieved on RP C 18 column (250 mm × 4.6 mm, 5 μm). The system consisted of an Agilent 1260 infinity series connected to a DAD detector with a quaternary pump and an injector with a 20 μl loop. Manual injections were carried out using a 100 μl Hamilton syringe. The mobile phase consisted of a mixture acetonitrile and ammonium acetate buffer 50 mM adjusted at pH 7.0 (at ratio 55:45 v/v). The determination was performed using ultraviolet detector (Shimadzu, Tokyo, Japan) at 240 nm and injection volume was 5 μL.
Materials Eplerenone was obtained from Shenzhen Oriental Pharmaceutical Co., Ltd., Guangdong, China. Glyceryl monostearate (Imwitor ® 900K) was purchased from Changwei Pharmaceutical Excipients Technology Co., Ltd. (Shanghai, China). Cremophor ® RH40 (Polyoxyl 40 Hydrogenated Castor Oil USP/NF), Pluronic ® F127 and Miglyol ® 81N2 were purchased from BASF chemical company (Ludwigshafen, Germany). Solutol ® HS15, Disodium hydrogen phosphate, Sodium chloride and Potassium dihydrogen phosphate were purchased from Sigma Aldrich VR (St. Louis, MO). Ethanol and acetone were obtained from El-Nasr Pharmaceutical Chemicals, Cairo, Egypt. All other chemicals were of analytical grade and were used as received.
Preparation of EPL-loaded NLCs EPL-loaded NLCs were prepared using the emulsification solvent evaporation technique using glyceryl monostearate (GMS) as solid lipid and Miglyol ® 812N as liquid lipid. In brief, 25 mg of EPL were dispersed in the determined amount of Miglyol ® 812N, then added to determined amount of molten GMS kept at 80°C using thermostatically controlled magnetic stirrer (WiseStir, Wisd Lab. Instruments, Tulsa, OK). Exactly, 10 mL mixture of ethanol and acetone (1:1 v/v) were added to the molten lipids maintained at 80 °C and stirred until complete dissolution in the organic phase. The organic phase was then added to 20 mL of an aqueous solution containing the chosen surface-active agent(s) to form a primary o/w emulsion under stirring at 1000 rpm for 1 min using magnetic stirrer. The obtained emulsion was subsequently subjected to 3 min of sonication using probe sonicator at room temperature (Probe Sonicator Ultrasonic Processor model VCX 750, Newtown, CT) adjusted at 40 W. Following this, the formed emulsion was stirred using magnetic stirrer at 500 rpm for 2 h at room temperature to allow evaporation of the organic solvent and formation of NLCs (Aburahma et al., ). Each formula was prepared three times, and the results are presented as mean ± SD ( n = 3).
Evaluation of the prepared EPL-loaded NLCs 2.3.1. Determination of particle size, polydispersity index, and zeta potential The mean particle size (PS), polydispersity index (PDI), and zeta potential (ZP) were determined using dynamic light scattering (DLS, Malvern Zetasizer, Malvern, UK) at 25 °C (Imam et al., ). Before measurement, 0.1 mL of the dispersion was properly diluted with distilled water (10 mL) in a glass tube and shaken to have an appropriate scattering intensity. Data presented as mean values ( n = 3± SD). 2.3.2. Entrapment efficiency The entrapment efficiency ( EE% w/w) of EPL in the prepared EPL-loaded NLCs was determined indirectly by measuring the concentration of free drug in the aqueous phase of the NLCs dispersion. A definite volume (1 mL) of the prepared NLCs dispersion was centrifuged using cooling centrifuge (Sigma 3-30 KS, Sigma Laborzentrifugen GmbH, Osterode am Harz, Germany) at 22,000 rpm for 1 h at 4 °C. The supernatant was separated and properly diluted with ethanol, then the un-entrapped drug concentration was estimated spectrophotometrically at λ max 241 nm. The EE% was calculated using the following equation: The E E % = W initial − W free W initial × 100 where W initial is the initial drug amount used in the preparation, and W free is the un-entrapped drug amount.
Determination of particle size, polydispersity index, and zeta potential The mean particle size (PS), polydispersity index (PDI), and zeta potential (ZP) were determined using dynamic light scattering (DLS, Malvern Zetasizer, Malvern, UK) at 25 °C (Imam et al., ). Before measurement, 0.1 mL of the dispersion was properly diluted with distilled water (10 mL) in a glass tube and shaken to have an appropriate scattering intensity. Data presented as mean values ( n = 3± SD).
Entrapment efficiency The entrapment efficiency ( EE% w/w) of EPL in the prepared EPL-loaded NLCs was determined indirectly by measuring the concentration of free drug in the aqueous phase of the NLCs dispersion. A definite volume (1 mL) of the prepared NLCs dispersion was centrifuged using cooling centrifuge (Sigma 3-30 KS, Sigma Laborzentrifugen GmbH, Osterode am Harz, Germany) at 22,000 rpm for 1 h at 4 °C. The supernatant was separated and properly diluted with ethanol, then the un-entrapped drug concentration was estimated spectrophotometrically at λ max 241 nm. The EE% was calculated using the following equation: The E E % = W initial − W free W initial × 100 where W initial is the initial drug amount used in the preparation, and W free is the un-entrapped drug amount.
Statistical design and optimization of EPL-loaded NLCs The response surface methodology with polynomial equations is a beneficial numerical mean to investigate the effect of independent variables on the dependent variables (responses) based on a limited number of trails (Heurtault et al., ; Huang et al., ). Using d -optimal mixture design, maximum prediction power could be obtained in selected set of experimental runs, minimizing the discrepancy associated with estimates of coefficients in the model (Bodea & Leucuta, ). A d -optimal statistical design was employed to study the effect of different formulation variables on the prepared EPL-loaded NLCs. In this design, three liquid lipid to solid lipid ratios (1:1, 1:1.5, and 1:2 w/w), three surfactant types (Pluronic ® F127, Cremophor ® RH40 and Solutol ® HS15) and three concentrations of surfactant (0.2, 0.4, and 0.6%w/v) were evaluated. shows the different levels of the studied factors as well as the constraints applied in selection of the optimized EPL-loaded NLCs (minimize the PS, maximize the EE % and for the PDI to be less than 0.5). The composition of the prepared EPL-loaded NLCs is presented in along with the measured responses. For this study, 19 runs were considered.
In vitro characterization of the optimized EPL-loaded NLC system The optimized EPL-loaded NLCs system was prepared and evaluated for the PS, PDI, ZP, and EE%. The residual difference between the predicted and the measured responses was measured to ensure the validity of the design (Sayed et al., ).
In-vitro drug release study The cumulative % EPL released from the optimized EPL-loaded NLCs system was performed using dialysis bag technique (Üner et al., ). Briefly 2 mL of the NLCs system (equivalent to 2.5 mg EPL) were placed in a dialysis bag and the ends of dialysis bag were sealed with standard closures, and then immersed in 100 mL phosphate buffer solution (PBS pH 7.4) in a tightly closed flask kept in thermostatically controlled water bath shaker (Gesellschatt Laboratories, Berlin, Germany) at 37 ± 0.5 °C and shaken at 100 rpm. At different time intervals (0.25 h, 0.5 h, 1 h, 1.5 h, 2 h, 4 h, 6 h, 8 h, and 24 h), 3 mL sample were withdrawn from the release medium and replaced by equal volume of fresh medium. For comparison, EPL aqueous suspension (2.5 mg/2 mL) was prepared and EPL release was determined using the same procedure mentioned above and compared to that of the EPL-loaded NLCs optimized system. Samples were analyzed using UV-spectrophotometry at λ max 245 nm using UV/VIS spectrophotometer (Shimadzu, UV-1601 PC, Kyoto, Japan).
Morphological evaluation of the optimized EPL-loaded NLCs system Morphological parameters of the optimized EPL-loaded NLCs system were observed using transmission electron microscopy (TEM) at 70 kV (JEOL JEM1230, Tokyo, Japan). The optimized EPL-loaded NLCs system was first diluted with 10-fold its volume distilled water, then added to copper grid coated with collodion film after being stained by 2% (w/v) phosphotungestic acid solution (negative stain) and dried at room temperature, then observed under the TEM.
Fourier infrared spectroscopy and X-ray diffraction The FTIR spectroscopy study was employed using a Bruker FTIR spectrophotometer (Model 22; Bruker, Coventry, UK) to detect the chances of any chemical interactions between EPL and other ingredients used in the preparation of the optimized NLCs system. The IR spectra of plain EPL, physical mixture (composed of EPL, GMS and Pluronic ® F127 in equal amounts) and the freeze-dried optimized EPL-loaded NLCs system were recorded. Samples were mixed with KBr and compressed into disk, then scanned in the range 4000–400 cm −1 at ambient temperature. For freeze-drying, EPL-loaded NLCs system was first frozen at −22 °C for 24 h and then placed in the freeze-dryer (Novalyphe-NL 500, Halprook, NY) for 24 h. The freeze-dryer was operated under vacuum at pressure of 7 × 10 −2 mBAR. The condenser temperature was adjusted at − 45 °C. X-ray powder diffraction (X-RPD) experiments were carried out using X-ray diffractometer (XGEN-4000, Scintag Corp., Sunnyvale, CA), with Cu Ka radiation at 40 mA current and 45 kV Voltage. Diffraction patterns for plain EPL, physical mixture (composed of EPL, GMS, and Pluronic ® F127 in equal amounts) and the freeze-dried optimized EPL-loaded NLCs system were recorded as the X-ray intensity as a function of 2 θ angle covering from 2.0° to 50.0°. The scanning rate was 6°/minute. The X-ray diffraction patterns were displayed using Diffrac AT software.
Ex vivo permeation study 2.9.1. Animals Male albino rabbits (2.5–3 kg) were housed in a controlled temperature and humidity room (25 °C, 55% air humidity with free access to water ad libitum). Experiments were approved by the Research Ethics Committee (REC; PT 2682) at Faculty of Pharmacy, Cairo University (Cairo, Egypt). One day before the experiment, rabbits were fasted overnight with free access to water. Rabbits received an intramuscular injection of ketamine (35 mg/kg) as an anesthetizing agent and 5 mg/kg xylazine as a relaxing agent, then sacrificed and the small intestine was removed by a midline abdominal incision. The small intestine cleaned carefully using a syringe filled with warm (37 °C) saline (pH = 7.4), then the intestinal segments (∼7.8 cm 2 ) were separated. 2.9.2. Study design Accurately, 4 mL of EPL-loaded optimized NLCs system (equivalent to 5 mg of drug) were filled in an intestinal segment (of area 7.8 cm 2 ) via micropipette, then the two sides of the intestine were tangled with a thread. Each intestinal sac was placed in a bottle containing 500 mL of normal saline (pH = 7.4) maintained at 37 ± 0.5 °C. The bottle was placed in a thermostatically controlled shaking water bath (GFL, Gesellschatt Laboratories, Berlin, Germany) maintained at 37 ± 0.5 °C and 100 rpm. Samples (3 mL) were withdrawn at specific time intervals (0.5 h, 1 h, 2 h, 4 h, 6 h, 8 h, and 24 h) and replaced by equal volume of fresh medium. In order to determine the amount of EPL permeated through the intestine, samples were diluted with equal amount of acetonitrile (HPLC grade) followed by sonication for 10 min in order to detect the total drug permeated (as free and vesicles-encapsulated drug). Samples were analyzed according to a previously validated HPLC method (Rane et al., ). For comparison, EPL permeation from EPL aqueous suspension containing EPL 1.25 mg/mL was examined using the same procedure mentioned above. Chromatographic separation was achieved on RP C 18 column (250 mm × 4.6 mm, 5 μm). The system consisted of an Agilent 1260 infinity series connected to a DAD detector with a quaternary pump and an injector with a 20 μl loop. Manual injections were carried out using a 100 μl Hamilton syringe. The mobile phase consisted of a mixture acetonitrile and ammonium acetate buffer 50 mM adjusted at pH 7.0 (at ratio 55:45 v/v). The determination was performed using ultraviolet detector (Shimadzu, Tokyo, Japan) at 240 nm and injection volume was 5 μL.
Animals Male albino rabbits (2.5–3 kg) were housed in a controlled temperature and humidity room (25 °C, 55% air humidity with free access to water ad libitum). Experiments were approved by the Research Ethics Committee (REC; PT 2682) at Faculty of Pharmacy, Cairo University (Cairo, Egypt). One day before the experiment, rabbits were fasted overnight with free access to water. Rabbits received an intramuscular injection of ketamine (35 mg/kg) as an anesthetizing agent and 5 mg/kg xylazine as a relaxing agent, then sacrificed and the small intestine was removed by a midline abdominal incision. The small intestine cleaned carefully using a syringe filled with warm (37 °C) saline (pH = 7.4), then the intestinal segments (∼7.8 cm 2 ) were separated.
Study design Accurately, 4 mL of EPL-loaded optimized NLCs system (equivalent to 5 mg of drug) were filled in an intestinal segment (of area 7.8 cm 2 ) via micropipette, then the two sides of the intestine were tangled with a thread. Each intestinal sac was placed in a bottle containing 500 mL of normal saline (pH = 7.4) maintained at 37 ± 0.5 °C. The bottle was placed in a thermostatically controlled shaking water bath (GFL, Gesellschatt Laboratories, Berlin, Germany) maintained at 37 ± 0.5 °C and 100 rpm. Samples (3 mL) were withdrawn at specific time intervals (0.5 h, 1 h, 2 h, 4 h, 6 h, 8 h, and 24 h) and replaced by equal volume of fresh medium. In order to determine the amount of EPL permeated through the intestine, samples were diluted with equal amount of acetonitrile (HPLC grade) followed by sonication for 10 min in order to detect the total drug permeated (as free and vesicles-encapsulated drug). Samples were analyzed according to a previously validated HPLC method (Rane et al., ). For comparison, EPL permeation from EPL aqueous suspension containing EPL 1.25 mg/mL was examined using the same procedure mentioned above. Chromatographic separation was achieved on RP C 18 column (250 mm × 4.6 mm, 5 μm). The system consisted of an Agilent 1260 infinity series connected to a DAD detector with a quaternary pump and an injector with a 20 μl loop. Manual injections were carried out using a 100 μl Hamilton syringe. The mobile phase consisted of a mixture acetonitrile and ammonium acetate buffer 50 mM adjusted at pH 7.0 (at ratio 55:45 v/v). The determination was performed using ultraviolet detector (Shimadzu, Tokyo, Japan) at 240 nm and injection volume was 5 μL.
Results and discussion 3.1. Selection of EPL-loaded nanostructured lipid matrix EPL-loaded NLCs were successfully prepared using the emulsification solvent evaporation method. All the prepared systems were homogenous and showed no signs of precipitation or phase separation. GMS was chosen as solid lipid to prepare EPL-loaded NLCs, owing to its inherent self-emulsifying property and low cytotoxicity. Solid lipids of long hydrocarbon chain (more than C 12 ) are characterized by low HLB values and high solubilizing power so they are preferred as lipid excipients (Patil-Gadhe & Pokharkar, ). Miglyol ® 812N was selected as the liquid lipid, being a medium chain triglyceride with a unique class of saturated lipids, in addition to its ability to act as emulsifier and suspender. It is well known that medium chain triglycerides excipients composed of C 6–12 fatty acids, such as Miglyol ® 812N are easily absorbed from the intestine (Furuse et al., ). Regarding the surfactants, Pluronic ® F127, Solutol ® HS15, and Cremophor ® RH40 were evaluated as nonionic surfactants in the preparation of EPL-loaded NLCs, due to their low toxicity and high hydrophilicity and compatibility. Further, the used solvent mixture (ethanol:acetone mixture) was selected based on its ability to solubilize drug, solid lipid, and liquid lipid results in their thorough and uniform drug–lipid association (Hu et al., , ). 3.2. Analysis of the factorial design For analysis of model parameters, the model having high values of adjusted R 2 and predicted R 2 , within 0.2 numerical value difference between each of other was considered to ensure the validity of the analyzed model. The model should have nonsignificant lack of fit with adequate precision more than 4 (Huang et al., ). shows the regression analysis for the studied factors and the best model selected through the design expert ® software (design expert software version 7). 3.3. Effect of variables on the properties of the prepared EPL-loaded NLCs 3.3.1. Particle size of EPL-loaded NLCs shows the results PS of EPL-loaded NLCs characterization. The average PS of the prepared EPL-loaded NLCs ranged from 100.85 ± 0.49 to 346.60 ± 16.82 nm. The 2FI model was the most suitable one fitting ( p value˂.0001) with non-significant lack of fit ( p value = .5240, ) and adequate precision of 18.635 indicating the model ability to navigate the design space (Elbary et al., ). All the studied factors had significant effect on the PS . ANOVA results showed that increasing the liquid lipid to solid lipid ratio led to a significant decrease in the PS of the prepared EPL-loaded NLCs ( p = 0.001), . This is could be attributed to the nature of the used liquid oil (Miglyol ® 812N), which is considered as an emulsifier and suspender. The low viscosity of Miglyol ® 812N allows the rapid diffusion of the organic phase through the aqueous phase during the solvent evaporation, where the surfactant molecules move faster and, therefore, produce smaller droplets (Wang et al., ). These results are in accordance with previously published results by Pezeshki et al. (Hasan et al., ), in their study on the preparation of betacarotene-loaded nanoemulsion. The authors reported that the smallest PS was observed upon using Miglyol ® 812N compared to other oils (corn oil and octyl octanoate). The mechanism of PS reduction by Miglyol ® 812N could be attributed to the reduction of the interfacial tension owing to its surfactant property (Sanad et al., ). ANOVA results also showed that increasing the concentration of the surfactant is associated with a significant increase in the PS ( p = .0479), . The surfactant concentration affects the PS in a paradoxical manner, where increasing the surfactant concentration leads to faster adsorption on particle surface making a mechanical barrier against crystallization and, thus, a smaller final PS. However, excess surfactant would increase the PS as well (Hasani et al., ), where above the critical micelle concentration, the surfactant molecules are oriented in a micelle form rather than to be adsorbed on other surfaces. Hence, the surfactant adsorption on particle surface is reduced. As a result, particle agglomeration is increased leading to larger PS. Also above the critical micelle concentration, the micelle solubilization may solubilize some of the drug particles and thereby enhances the particle growth by the Ostwald ripening process (Deng et al., ). A similar finding was reported by Shamma et al. (Aburahma et al., ), in their study on the fabrication of NLCs for follicular delivery of spironolactone. The authors reported a concentration-dependent increase in PS with the increase in surfactant concentration (Tween 80) and related this finding to the hydrophobic interaction between the nonpolar alkyl chains of the surfactant and the solid lipid molecules resulting in larger PS. Similar results were also obtained by Kumbhar & Pokharkar , in their study on the preparation of bicalutamide-loaded NLCs, where an increase in PS was reported upon increasing the Pluronic ® F127 concentration. Concerning the surfactant type, ANOVA results showed that the type of surfactant had a significant effect on the PS ( p = .0001). The PS of EPL-loaded NLCs prepared using Pluronic ® F127 were significantly smaller compared to NLCs prepared using Solutol ® HS15 and Cremophor ® RH40. as illustrated in . Cremophor ® RH40 is a bulky surfactant possessing polyethylene glycols and glycerol ethoxylates of long-chain ricinoleic acid. This could have imparted poor stabilizing efficiency and affect its ability to produce small PS (Shete & Patravale, ). On the other hand, Pluronic ® F127 imparted better surface coverage and enabled smaller PS. Pluronic ® F127 is a stabilizer with a big hydrophilic head group, which provides a better steric hindrance and thus reduces the PS. The results are in agreement with Weerapol et al. who reported that the molecular structure of the surfactant has a significant effect on the final droplet size . 3.3.2. Polydispersity index of EPL-loaded NLCs shows the PDI values of the prepared EPL-loaded NLCs. PDI values ranged from 0.17 ± 0.18 to 0.62 ± 0.03 indicating that most of the measured NLCs systems have an acceptable homogenous particle distribution. The 2FI model was the most suitable model fitting ( p value ˂ .0001) with non-significant lack of fit ( p value= .4571) and adequate precision of 22.008 indicating an adequate signal . ANOVA results showed that only the liquid lipid to solid lipid ratio had a significant negative effect on the PDI ( p value˂.0001), Increasing the ratio was associated with significant reduction in the PDI value, . Generally, the addition of liquid lipid (Miglyol ® 812N) favored the formation of homogeneous particles (Aburahma et al., ). The higher content Miglyol ® 812N reduces the viscosity inside NLCs, and reduces the surface tension resulting in the formation of smaller and smoother surface particles with low PDI (Jenning et al., ; Uprit et al., ). Similar results were obtained by several researchers in the literature (Hu et al., ; Agrawal et al., ), where the PDI of the prepared NLCs was decreased by increasing the liquid lipid (oleic acid) content. 3.3.3. Zeta potential of EPL-loaded NLCs The ZP provides information about the magnitude of the charge on the surface of the particles in an aqueous dispersion and allows predicting the long-term physical stability of the formulations (Kovačević et al., ). Results of ZP of the prepared EPL-loaded NLCs are presented in . Results show that the prepared EPL-loaded NLCs were all negatively charged having ZP in the range of −20.200 ± 1.273 to −36.750 ± 0.495 mV. Generally, the lipid nanocarriers are negatively charged because of their high content of fatty acids and triglycerides (Schwarz & Mehnert, ; Shah et al., ). Patil et al. . reported that GMS used in the formulation of carvedilol-loaded NLCs resulted in particles with negative ZP. In addition, the used liquid lipid, Miglyol ® 812N, carry negative charge at the carboxylic groups imparting a negative surface charge (Harisa & Badran, ). These results are in agreement with those obtained in a previous study by Hou et al. . The ZP has a negative predicted R 2 value indicates that the overall mean is a better predictor of this response and it was not significantly affected by the studied factors. 3.3.4. Entrapment efficiency of EPL-loaded NLCs The EE% of EPL in prepared EPL-loaded NLCs are presented in . The EE% ranged from 34.31 ± 1.26 to 70.64 ± 0.00% w/w. The Quadratic model showed the highest R 2 , thus, it was selected for analyzing the effect of variables on the EE% ( p value˂.0001) with non-significant lack of fit ( p value = .8118) and adequate precision of 18.648 . ANOVA results showed that the liquid lipid concentration had a significant negative impact on EPL EE%. Increasing the liquid lipid:solid lipid ratio is associated with a significant reduction in the EE% of EPL in the prepared NLCs ( p = .0001), . The ratio of components (solid lipid and liquid lipid) has been reported to significantly impact the final structure of the lipid particles and hence the loading of the molecule (Uprit et al., ). At high ratios of oil, the drug molecules might be expulsed from the inner core of the solid matrix toward the surface of the particle, because the limits of solubility could be exceeded. The same can be said about the use of high ratios of stabilizing agents, which may generate solubilized systems rather than particulate systems where the molecules of the lipid components could not form solid structures (Bunjes et al., ). Another explanation is that, during NLCs fabrication, all or part of the lipid matrix re-transformed to a less or unstable α-polymorphic form or to the metastable β’-polymorphic form then to the most stable β-form (Westesen et al., ). During this transformation, there is a subsequent reduction in the number of defects in the lipid matrix, as in the case of the configuration of β’/β forms leading to diminishing the drug encapsulation. The rate of this conversion is higher in short- and medium chain ones (like Miglyol ® 812N) compared to for long-chain triglycerides (Souto & Müller, ). ANOVA results also showed that the type of surfactant had a significant impact on the EE% of EPL in the prepared EPL-loaded NLCs ( p < .0001), . The highest EE% was observed when using PL ® F127 as surfactant followed by Solutol ® HS15 and the lowest one observed with Cremophor ® RH40. An obvious variation in emulsification power was evident upon using the previously mentioned surfactants despite the fact that all of them have HLB values greater than 10. This is because various factors rather than HLB value had also a great impact on the magnitude of EE% like the length and structure of hydrocarbon chain of surfactants. Cremophor ® RH40 is composed mainly of castor oil, a triglyceride in which each of its three hydroxyl groups esterifies with a long-chain unsaturated fatty acid (hydroxylated 12-hydroxy, 9-octadecenoic acid) known as ricinoleic acid. Glycerol ethoxylates and polyethylene glycols derivatives of long-chain ricinoleic acid were found to impart poor stabilizing efficiency confirmed by the lower EE% at higher concentration especially in the presence of low lipid concentration (Abd El-Halim et al., ). On contrary, Pluronic ® F127 being block copolymer is composed of polyoxypropylene oxide as a hydrophobic segment and polyoxyethylene oxide as a hydrophilic segment (Dai et al., ; Ghosh et al., ). This huge polymeric grid of polyoxyethylene oxide (hydrophilic segment) orients in external phase, while polyoxypropylene oxide (hydrophobic segment) settles at interface and leading to more steric stabilization of NLCs permitting larger surface area for drug housing (Chen et al., ). Another explanation, Pluronic ® F127 has a higher nominal molecular weight (12,500) g/mol (Bohorquez et al., ) than that of Cremophor ® RH40 (2500 g/mol) (Zeng et al., ), therefore, Pluronic ® F127 can form rigid dispersion with low tendency of polymorphic transition. It is known that polymorphic transition reduces drug loading capacity in NLCs due to increased ordering and packing density of the lipid crystals, thus, leading to expulsion of drug from the crystal matrix and consequently low entrapment (Bohorquez et al., ). Unlike Pluronic ® F127, Cremophor ® RH40 is semisolid at room temperature, so the formed surfactant layer is relatively flexible, and, thus, likely induced rapid polymorphic transition.
Selection of EPL-loaded nanostructured lipid matrix EPL-loaded NLCs were successfully prepared using the emulsification solvent evaporation method. All the prepared systems were homogenous and showed no signs of precipitation or phase separation. GMS was chosen as solid lipid to prepare EPL-loaded NLCs, owing to its inherent self-emulsifying property and low cytotoxicity. Solid lipids of long hydrocarbon chain (more than C 12 ) are characterized by low HLB values and high solubilizing power so they are preferred as lipid excipients (Patil-Gadhe & Pokharkar, ). Miglyol ® 812N was selected as the liquid lipid, being a medium chain triglyceride with a unique class of saturated lipids, in addition to its ability to act as emulsifier and suspender. It is well known that medium chain triglycerides excipients composed of C 6–12 fatty acids, such as Miglyol ® 812N are easily absorbed from the intestine (Furuse et al., ). Regarding the surfactants, Pluronic ® F127, Solutol ® HS15, and Cremophor ® RH40 were evaluated as nonionic surfactants in the preparation of EPL-loaded NLCs, due to their low toxicity and high hydrophilicity and compatibility. Further, the used solvent mixture (ethanol:acetone mixture) was selected based on its ability to solubilize drug, solid lipid, and liquid lipid results in their thorough and uniform drug–lipid association (Hu et al., , ).
Analysis of the factorial design For analysis of model parameters, the model having high values of adjusted R 2 and predicted R 2 , within 0.2 numerical value difference between each of other was considered to ensure the validity of the analyzed model. The model should have nonsignificant lack of fit with adequate precision more than 4 (Huang et al., ). shows the regression analysis for the studied factors and the best model selected through the design expert ® software (design expert software version 7).
Effect of variables on the properties of the prepared EPL-loaded NLCs 3.3.1. Particle size of EPL-loaded NLCs shows the results PS of EPL-loaded NLCs characterization. The average PS of the prepared EPL-loaded NLCs ranged from 100.85 ± 0.49 to 346.60 ± 16.82 nm. The 2FI model was the most suitable one fitting ( p value˂.0001) with non-significant lack of fit ( p value = .5240, ) and adequate precision of 18.635 indicating the model ability to navigate the design space (Elbary et al., ). All the studied factors had significant effect on the PS . ANOVA results showed that increasing the liquid lipid to solid lipid ratio led to a significant decrease in the PS of the prepared EPL-loaded NLCs ( p = 0.001), . This is could be attributed to the nature of the used liquid oil (Miglyol ® 812N), which is considered as an emulsifier and suspender. The low viscosity of Miglyol ® 812N allows the rapid diffusion of the organic phase through the aqueous phase during the solvent evaporation, where the surfactant molecules move faster and, therefore, produce smaller droplets (Wang et al., ). These results are in accordance with previously published results by Pezeshki et al. (Hasan et al., ), in their study on the preparation of betacarotene-loaded nanoemulsion. The authors reported that the smallest PS was observed upon using Miglyol ® 812N compared to other oils (corn oil and octyl octanoate). The mechanism of PS reduction by Miglyol ® 812N could be attributed to the reduction of the interfacial tension owing to its surfactant property (Sanad et al., ). ANOVA results also showed that increasing the concentration of the surfactant is associated with a significant increase in the PS ( p = .0479), . The surfactant concentration affects the PS in a paradoxical manner, where increasing the surfactant concentration leads to faster adsorption on particle surface making a mechanical barrier against crystallization and, thus, a smaller final PS. However, excess surfactant would increase the PS as well (Hasani et al., ), where above the critical micelle concentration, the surfactant molecules are oriented in a micelle form rather than to be adsorbed on other surfaces. Hence, the surfactant adsorption on particle surface is reduced. As a result, particle agglomeration is increased leading to larger PS. Also above the critical micelle concentration, the micelle solubilization may solubilize some of the drug particles and thereby enhances the particle growth by the Ostwald ripening process (Deng et al., ). A similar finding was reported by Shamma et al. (Aburahma et al., ), in their study on the fabrication of NLCs for follicular delivery of spironolactone. The authors reported a concentration-dependent increase in PS with the increase in surfactant concentration (Tween 80) and related this finding to the hydrophobic interaction between the nonpolar alkyl chains of the surfactant and the solid lipid molecules resulting in larger PS. Similar results were also obtained by Kumbhar & Pokharkar , in their study on the preparation of bicalutamide-loaded NLCs, where an increase in PS was reported upon increasing the Pluronic ® F127 concentration. Concerning the surfactant type, ANOVA results showed that the type of surfactant had a significant effect on the PS ( p = .0001). The PS of EPL-loaded NLCs prepared using Pluronic ® F127 were significantly smaller compared to NLCs prepared using Solutol ® HS15 and Cremophor ® RH40. as illustrated in . Cremophor ® RH40 is a bulky surfactant possessing polyethylene glycols and glycerol ethoxylates of long-chain ricinoleic acid. This could have imparted poor stabilizing efficiency and affect its ability to produce small PS (Shete & Patravale, ). On the other hand, Pluronic ® F127 imparted better surface coverage and enabled smaller PS. Pluronic ® F127 is a stabilizer with a big hydrophilic head group, which provides a better steric hindrance and thus reduces the PS. The results are in agreement with Weerapol et al. who reported that the molecular structure of the surfactant has a significant effect on the final droplet size . 3.3.2. Polydispersity index of EPL-loaded NLCs shows the PDI values of the prepared EPL-loaded NLCs. PDI values ranged from 0.17 ± 0.18 to 0.62 ± 0.03 indicating that most of the measured NLCs systems have an acceptable homogenous particle distribution. The 2FI model was the most suitable model fitting ( p value ˂ .0001) with non-significant lack of fit ( p value= .4571) and adequate precision of 22.008 indicating an adequate signal . ANOVA results showed that only the liquid lipid to solid lipid ratio had a significant negative effect on the PDI ( p value˂.0001), Increasing the ratio was associated with significant reduction in the PDI value, . Generally, the addition of liquid lipid (Miglyol ® 812N) favored the formation of homogeneous particles (Aburahma et al., ). The higher content Miglyol ® 812N reduces the viscosity inside NLCs, and reduces the surface tension resulting in the formation of smaller and smoother surface particles with low PDI (Jenning et al., ; Uprit et al., ). Similar results were obtained by several researchers in the literature (Hu et al., ; Agrawal et al., ), where the PDI of the prepared NLCs was decreased by increasing the liquid lipid (oleic acid) content. 3.3.3. Zeta potential of EPL-loaded NLCs The ZP provides information about the magnitude of the charge on the surface of the particles in an aqueous dispersion and allows predicting the long-term physical stability of the formulations (Kovačević et al., ). Results of ZP of the prepared EPL-loaded NLCs are presented in . Results show that the prepared EPL-loaded NLCs were all negatively charged having ZP in the range of −20.200 ± 1.273 to −36.750 ± 0.495 mV. Generally, the lipid nanocarriers are negatively charged because of their high content of fatty acids and triglycerides (Schwarz & Mehnert, ; Shah et al., ). Patil et al. . reported that GMS used in the formulation of carvedilol-loaded NLCs resulted in particles with negative ZP. In addition, the used liquid lipid, Miglyol ® 812N, carry negative charge at the carboxylic groups imparting a negative surface charge (Harisa & Badran, ). These results are in agreement with those obtained in a previous study by Hou et al. . The ZP has a negative predicted R 2 value indicates that the overall mean is a better predictor of this response and it was not significantly affected by the studied factors. 3.3.4. Entrapment efficiency of EPL-loaded NLCs The EE% of EPL in prepared EPL-loaded NLCs are presented in . The EE% ranged from 34.31 ± 1.26 to 70.64 ± 0.00% w/w. The Quadratic model showed the highest R 2 , thus, it was selected for analyzing the effect of variables on the EE% ( p value˂.0001) with non-significant lack of fit ( p value = .8118) and adequate precision of 18.648 . ANOVA results showed that the liquid lipid concentration had a significant negative impact on EPL EE%. Increasing the liquid lipid:solid lipid ratio is associated with a significant reduction in the EE% of EPL in the prepared NLCs ( p = .0001), . The ratio of components (solid lipid and liquid lipid) has been reported to significantly impact the final structure of the lipid particles and hence the loading of the molecule (Uprit et al., ). At high ratios of oil, the drug molecules might be expulsed from the inner core of the solid matrix toward the surface of the particle, because the limits of solubility could be exceeded. The same can be said about the use of high ratios of stabilizing agents, which may generate solubilized systems rather than particulate systems where the molecules of the lipid components could not form solid structures (Bunjes et al., ). Another explanation is that, during NLCs fabrication, all or part of the lipid matrix re-transformed to a less or unstable α-polymorphic form or to the metastable β’-polymorphic form then to the most stable β-form (Westesen et al., ). During this transformation, there is a subsequent reduction in the number of defects in the lipid matrix, as in the case of the configuration of β’/β forms leading to diminishing the drug encapsulation. The rate of this conversion is higher in short- and medium chain ones (like Miglyol ® 812N) compared to for long-chain triglycerides (Souto & Müller, ). ANOVA results also showed that the type of surfactant had a significant impact on the EE% of EPL in the prepared EPL-loaded NLCs ( p < .0001), . The highest EE% was observed when using PL ® F127 as surfactant followed by Solutol ® HS15 and the lowest one observed with Cremophor ® RH40. An obvious variation in emulsification power was evident upon using the previously mentioned surfactants despite the fact that all of them have HLB values greater than 10. This is because various factors rather than HLB value had also a great impact on the magnitude of EE% like the length and structure of hydrocarbon chain of surfactants. Cremophor ® RH40 is composed mainly of castor oil, a triglyceride in which each of its three hydroxyl groups esterifies with a long-chain unsaturated fatty acid (hydroxylated 12-hydroxy, 9-octadecenoic acid) known as ricinoleic acid. Glycerol ethoxylates and polyethylene glycols derivatives of long-chain ricinoleic acid were found to impart poor stabilizing efficiency confirmed by the lower EE% at higher concentration especially in the presence of low lipid concentration (Abd El-Halim et al., ). On contrary, Pluronic ® F127 being block copolymer is composed of polyoxypropylene oxide as a hydrophobic segment and polyoxyethylene oxide as a hydrophilic segment (Dai et al., ; Ghosh et al., ). This huge polymeric grid of polyoxyethylene oxide (hydrophilic segment) orients in external phase, while polyoxypropylene oxide (hydrophobic segment) settles at interface and leading to more steric stabilization of NLCs permitting larger surface area for drug housing (Chen et al., ). Another explanation, Pluronic ® F127 has a higher nominal molecular weight (12,500) g/mol (Bohorquez et al., ) than that of Cremophor ® RH40 (2500 g/mol) (Zeng et al., ), therefore, Pluronic ® F127 can form rigid dispersion with low tendency of polymorphic transition. It is known that polymorphic transition reduces drug loading capacity in NLCs due to increased ordering and packing density of the lipid crystals, thus, leading to expulsion of drug from the crystal matrix and consequently low entrapment (Bohorquez et al., ). Unlike Pluronic ® F127, Cremophor ® RH40 is semisolid at room temperature, so the formed surfactant layer is relatively flexible, and, thus, likely induced rapid polymorphic transition.
Particle size of EPL-loaded NLCs shows the results PS of EPL-loaded NLCs characterization. The average PS of the prepared EPL-loaded NLCs ranged from 100.85 ± 0.49 to 346.60 ± 16.82 nm. The 2FI model was the most suitable one fitting ( p value˂.0001) with non-significant lack of fit ( p value = .5240, ) and adequate precision of 18.635 indicating the model ability to navigate the design space (Elbary et al., ). All the studied factors had significant effect on the PS . ANOVA results showed that increasing the liquid lipid to solid lipid ratio led to a significant decrease in the PS of the prepared EPL-loaded NLCs ( p = 0.001), . This is could be attributed to the nature of the used liquid oil (Miglyol ® 812N), which is considered as an emulsifier and suspender. The low viscosity of Miglyol ® 812N allows the rapid diffusion of the organic phase through the aqueous phase during the solvent evaporation, where the surfactant molecules move faster and, therefore, produce smaller droplets (Wang et al., ). These results are in accordance with previously published results by Pezeshki et al. (Hasan et al., ), in their study on the preparation of betacarotene-loaded nanoemulsion. The authors reported that the smallest PS was observed upon using Miglyol ® 812N compared to other oils (corn oil and octyl octanoate). The mechanism of PS reduction by Miglyol ® 812N could be attributed to the reduction of the interfacial tension owing to its surfactant property (Sanad et al., ). ANOVA results also showed that increasing the concentration of the surfactant is associated with a significant increase in the PS ( p = .0479), . The surfactant concentration affects the PS in a paradoxical manner, where increasing the surfactant concentration leads to faster adsorption on particle surface making a mechanical barrier against crystallization and, thus, a smaller final PS. However, excess surfactant would increase the PS as well (Hasani et al., ), where above the critical micelle concentration, the surfactant molecules are oriented in a micelle form rather than to be adsorbed on other surfaces. Hence, the surfactant adsorption on particle surface is reduced. As a result, particle agglomeration is increased leading to larger PS. Also above the critical micelle concentration, the micelle solubilization may solubilize some of the drug particles and thereby enhances the particle growth by the Ostwald ripening process (Deng et al., ). A similar finding was reported by Shamma et al. (Aburahma et al., ), in their study on the fabrication of NLCs for follicular delivery of spironolactone. The authors reported a concentration-dependent increase in PS with the increase in surfactant concentration (Tween 80) and related this finding to the hydrophobic interaction between the nonpolar alkyl chains of the surfactant and the solid lipid molecules resulting in larger PS. Similar results were also obtained by Kumbhar & Pokharkar , in their study on the preparation of bicalutamide-loaded NLCs, where an increase in PS was reported upon increasing the Pluronic ® F127 concentration. Concerning the surfactant type, ANOVA results showed that the type of surfactant had a significant effect on the PS ( p = .0001). The PS of EPL-loaded NLCs prepared using Pluronic ® F127 were significantly smaller compared to NLCs prepared using Solutol ® HS15 and Cremophor ® RH40. as illustrated in . Cremophor ® RH40 is a bulky surfactant possessing polyethylene glycols and glycerol ethoxylates of long-chain ricinoleic acid. This could have imparted poor stabilizing efficiency and affect its ability to produce small PS (Shete & Patravale, ). On the other hand, Pluronic ® F127 imparted better surface coverage and enabled smaller PS. Pluronic ® F127 is a stabilizer with a big hydrophilic head group, which provides a better steric hindrance and thus reduces the PS. The results are in agreement with Weerapol et al. who reported that the molecular structure of the surfactant has a significant effect on the final droplet size .
Polydispersity index of EPL-loaded NLCs shows the PDI values of the prepared EPL-loaded NLCs. PDI values ranged from 0.17 ± 0.18 to 0.62 ± 0.03 indicating that most of the measured NLCs systems have an acceptable homogenous particle distribution. The 2FI model was the most suitable model fitting ( p value ˂ .0001) with non-significant lack of fit ( p value= .4571) and adequate precision of 22.008 indicating an adequate signal . ANOVA results showed that only the liquid lipid to solid lipid ratio had a significant negative effect on the PDI ( p value˂.0001), Increasing the ratio was associated with significant reduction in the PDI value, . Generally, the addition of liquid lipid (Miglyol ® 812N) favored the formation of homogeneous particles (Aburahma et al., ). The higher content Miglyol ® 812N reduces the viscosity inside NLCs, and reduces the surface tension resulting in the formation of smaller and smoother surface particles with low PDI (Jenning et al., ; Uprit et al., ). Similar results were obtained by several researchers in the literature (Hu et al., ; Agrawal et al., ), where the PDI of the prepared NLCs was decreased by increasing the liquid lipid (oleic acid) content.
Zeta potential of EPL-loaded NLCs The ZP provides information about the magnitude of the charge on the surface of the particles in an aqueous dispersion and allows predicting the long-term physical stability of the formulations (Kovačević et al., ). Results of ZP of the prepared EPL-loaded NLCs are presented in . Results show that the prepared EPL-loaded NLCs were all negatively charged having ZP in the range of −20.200 ± 1.273 to −36.750 ± 0.495 mV. Generally, the lipid nanocarriers are negatively charged because of their high content of fatty acids and triglycerides (Schwarz & Mehnert, ; Shah et al., ). Patil et al. . reported that GMS used in the formulation of carvedilol-loaded NLCs resulted in particles with negative ZP. In addition, the used liquid lipid, Miglyol ® 812N, carry negative charge at the carboxylic groups imparting a negative surface charge (Harisa & Badran, ). These results are in agreement with those obtained in a previous study by Hou et al. . The ZP has a negative predicted R 2 value indicates that the overall mean is a better predictor of this response and it was not significantly affected by the studied factors.
Entrapment efficiency of EPL-loaded NLCs The EE% of EPL in prepared EPL-loaded NLCs are presented in . The EE% ranged from 34.31 ± 1.26 to 70.64 ± 0.00% w/w. The Quadratic model showed the highest R 2 , thus, it was selected for analyzing the effect of variables on the EE% ( p value˂.0001) with non-significant lack of fit ( p value = .8118) and adequate precision of 18.648 . ANOVA results showed that the liquid lipid concentration had a significant negative impact on EPL EE%. Increasing the liquid lipid:solid lipid ratio is associated with a significant reduction in the EE% of EPL in the prepared NLCs ( p = .0001), . The ratio of components (solid lipid and liquid lipid) has been reported to significantly impact the final structure of the lipid particles and hence the loading of the molecule (Uprit et al., ). At high ratios of oil, the drug molecules might be expulsed from the inner core of the solid matrix toward the surface of the particle, because the limits of solubility could be exceeded. The same can be said about the use of high ratios of stabilizing agents, which may generate solubilized systems rather than particulate systems where the molecules of the lipid components could not form solid structures (Bunjes et al., ). Another explanation is that, during NLCs fabrication, all or part of the lipid matrix re-transformed to a less or unstable α-polymorphic form or to the metastable β’-polymorphic form then to the most stable β-form (Westesen et al., ). During this transformation, there is a subsequent reduction in the number of defects in the lipid matrix, as in the case of the configuration of β’/β forms leading to diminishing the drug encapsulation. The rate of this conversion is higher in short- and medium chain ones (like Miglyol ® 812N) compared to for long-chain triglycerides (Souto & Müller, ). ANOVA results also showed that the type of surfactant had a significant impact on the EE% of EPL in the prepared EPL-loaded NLCs ( p < .0001), . The highest EE% was observed when using PL ® F127 as surfactant followed by Solutol ® HS15 and the lowest one observed with Cremophor ® RH40. An obvious variation in emulsification power was evident upon using the previously mentioned surfactants despite the fact that all of them have HLB values greater than 10. This is because various factors rather than HLB value had also a great impact on the magnitude of EE% like the length and structure of hydrocarbon chain of surfactants. Cremophor ® RH40 is composed mainly of castor oil, a triglyceride in which each of its three hydroxyl groups esterifies with a long-chain unsaturated fatty acid (hydroxylated 12-hydroxy, 9-octadecenoic acid) known as ricinoleic acid. Glycerol ethoxylates and polyethylene glycols derivatives of long-chain ricinoleic acid were found to impart poor stabilizing efficiency confirmed by the lower EE% at higher concentration especially in the presence of low lipid concentration (Abd El-Halim et al., ). On contrary, Pluronic ® F127 being block copolymer is composed of polyoxypropylene oxide as a hydrophobic segment and polyoxyethylene oxide as a hydrophilic segment (Dai et al., ; Ghosh et al., ). This huge polymeric grid of polyoxyethylene oxide (hydrophilic segment) orients in external phase, while polyoxypropylene oxide (hydrophobic segment) settles at interface and leading to more steric stabilization of NLCs permitting larger surface area for drug housing (Chen et al., ). Another explanation, Pluronic ® F127 has a higher nominal molecular weight (12,500) g/mol (Bohorquez et al., ) than that of Cremophor ® RH40 (2500 g/mol) (Zeng et al., ), therefore, Pluronic ® F127 can form rigid dispersion with low tendency of polymorphic transition. It is known that polymorphic transition reduces drug loading capacity in NLCs due to increased ordering and packing density of the lipid crystals, thus, leading to expulsion of drug from the crystal matrix and consequently low entrapment (Bohorquez et al., ). Unlike Pluronic ® F127, Cremophor ® RH40 is semisolid at room temperature, so the formed surfactant layer is relatively flexible, and, thus, likely induced rapid polymorphic transition.
Optimization and validation of EPL-loaded NLCs The optimization process depends on the collection all the responses into one variable in order to elucidate the desired levels for each of the studied factors (Pandya et al., ). The optimum NLCs system with smallest PS, highest EE%, and PDI less than 0.5 with desirability value equal 0.905 was suggested through Design expert ® software. This optimized EPL-loaded NLCs system is composed of liquid to solid lipid ratio equal 2:1 in presence of 0.43%w/v PL-F127 as a surfactant. Accordingly, this NLCs system was prepared and evaluated for the PS, PDI, ZP, and EE%. shows the predicted values for all responses of the optimum formula, compared to actual values of the prepared optimized one and the calculated residual value. Results show that the actual values close to the predicted values indicating the success of the design. 4.1. In-vitro comparative drug release study shows the release profile of EPL from the optimized EPL-loaded NLCs system compared to that of EPL drug suspension (containing equivalent drug dose as the optimized system). As shown in , the optimized NLCs system succeeded to control the release of EPL with about 52.31% EPL released after 2 h compared to 16.34% EPL released from the aqueous suspension at the same time. By the end of the release time (24 h), 85.064% and 37.65% EPL were released from the optimized NLCs system and the aqueous suspension, respectively. This difference in the release pattern is due to the fact that most NLCs exhibit a biphasic release pattern (Abd El-Halim et al., ; Gordillo-Galeano & Mora-Huertas, ; Yu et al., Joshi and Patravale ); starts with a spurt stage due to the rapid release of the surface drug in the outside shell, while the second prolonged phase (up to 24 h) is due to the presence of the encapsulated drug within the lipid matrices. Moreover, the nature of the stabilizing agent (Pluronic ® F127) used and its arrangement (Abdelbary & Haider, ; Dan, ; Alvarez-Trabado et al., ), greatly affect the drug release pattern by providing large surface area for drug accommodation, consequently hinders the drug expulsion from the NLCs and imparts great stability (Ghosh et al., ). 4.2. Transmission electron microscopy (TEM) TEM images of the optimized EPL-loaded NLCs, presented in , revealed that the particles were non-aggregated with smooth spherical shape and narrow size distribution. These results are in accordance to previously obtained studies showing that NLCs are spherical in shape (Saupe et al., ; Araújo et al., ; Alam et al., ). Also, it can be noticed that some of the obtained NLCs particles had an empty core and a drug-enriched shell demonstrated by the dark shade around the particles confirming the ability of the used surfactant to solubilize the drug and concentrate it on the surface of the nanoparticles. Results of the PS obtained from the images are in close agreement with those obtained by the dynamic light scattering. 4.3. Fourier infrared spectroscopy and X-ray diffraction presents the IR spectra of the tested samples. The IR spectrum of pure EPL shows the main characteristic functional groups at 2988.65 cm −1 (C–H stretching), 1778.08 cm −1 (anhydride O–C=O stretching), 1726.22 cm −1 (C=O ester stretching), and (1657.64 cm −1 ) C=O stretching. IR spectra of the physical mixture and the freeze dried EPL-loaded NLCs system showed no chemical changes, where all drug functional groups were retained, which confirmed the compatibility of EPL with other formulation additives. presents the diffraction patterns of the examined samples. The diffraction pattern of pure EPL exhibited high-intensity crystallinity peaks at 10.19, 12.24, 14.55, 18.12, and 22.61° 2 θ . In the diffractogram of the physical mixture, the characteristic peaks of GMS were present at 11.9, 14.58, and 18.00° 2 θ , confirming its crystallinity. The crystalline peaks of EPL in physical mixture at 10.19, 14.55, and 22.61° 2 θ were evident but with lower intensities, confirming the presence of EPL in crystalline form. On the contrary, the diffractogram of optimized EPL-loaded NLCs system showed absence of EPL constructive peaks, signifying that the drug lost its crystallinity where it was transformed to an amorphous state or dispersed within the nanocarriers. On the other hand, the two characteristic peaks of GMS (14.58, 18.00° 2 θ ) were still present in the diffractogram of optimized NLCs system confirming that the carrier crystallized into its stable polymorphic form. 4.4. Ex-vivo permeation study The HPLC method showed a linear response in the range of 0.2–2 μg/ml ( R 2 = 0.9915). shows the ex-vivo permeation profiles of EPL through the rabbit intestine in 0.9% saline (pH = 7.4) from the optimized EPL-loaded NLCs (containing 5 mg EPL) compared to that of EPL aqueous suspension (containing an equivalent dose). As shown in , great difference in the permeation profiles of optimized EPL-loaded NLCs and dug aqueous suspension through the intestine was observed, where the cumulative amount of drug permeated from the optimized EPL-loaded NLCs was two-fold higher that that from EPL aqueous suspension after 24 h. The optimized EPL-loaded NLCs showed significantly higher cumulative amount of EPL permeated after 24 h (Q 24h optimized NLCs system∼292.56 ± 11.61 μg/cm 2 ) compared to (Q 24h suspension∼139.35 ± 0.00 μg/cm 2 ) ( p < .05). The low transfer through the intestine was observed with drug suspension due to the poor EPL aqueous solubility (BCS class II) (Khames, ). The transport of drug from optimized-loaded NLCs system through intestinal membrane could be achieved as free molecules or as the intact vesicles encapsulating the drug that was acetonitrile diluted samples were used for the analysis (acetonitrile is capable to disrupt the vesicle wall releasing the encapsulated drug) (Ibrahim et al., ). Several studies reported the uptake of intact vesicles by the intestinal epithelial cells (M-cells) in the Payer's patch (Poonia et al., ; Aburahma, ) and the intestinal enterocytes (Niu et al., ), confirming that the main transport mechanism is supposed to be through vesicular uptake as most of the drug is encapsulated into the vesicles, where the free nanosized molecules are transported and absorbed via paracellular or transcellular pathways through the intestinal membrane. These results indicate that the optimized NLCs system was capable to incorporate high amounts of drug and, hence, achieve high drug permeation in the ex-vivo study.
In-vitro comparative drug release study shows the release profile of EPL from the optimized EPL-loaded NLCs system compared to that of EPL drug suspension (containing equivalent drug dose as the optimized system). As shown in , the optimized NLCs system succeeded to control the release of EPL with about 52.31% EPL released after 2 h compared to 16.34% EPL released from the aqueous suspension at the same time. By the end of the release time (24 h), 85.064% and 37.65% EPL were released from the optimized NLCs system and the aqueous suspension, respectively. This difference in the release pattern is due to the fact that most NLCs exhibit a biphasic release pattern (Abd El-Halim et al., ; Gordillo-Galeano & Mora-Huertas, ; Yu et al., Joshi and Patravale ); starts with a spurt stage due to the rapid release of the surface drug in the outside shell, while the second prolonged phase (up to 24 h) is due to the presence of the encapsulated drug within the lipid matrices. Moreover, the nature of the stabilizing agent (Pluronic ® F127) used and its arrangement (Abdelbary & Haider, ; Dan, ; Alvarez-Trabado et al., ), greatly affect the drug release pattern by providing large surface area for drug accommodation, consequently hinders the drug expulsion from the NLCs and imparts great stability (Ghosh et al., ).
Transmission electron microscopy (TEM) TEM images of the optimized EPL-loaded NLCs, presented in , revealed that the particles were non-aggregated with smooth spherical shape and narrow size distribution. These results are in accordance to previously obtained studies showing that NLCs are spherical in shape (Saupe et al., ; Araújo et al., ; Alam et al., ). Also, it can be noticed that some of the obtained NLCs particles had an empty core and a drug-enriched shell demonstrated by the dark shade around the particles confirming the ability of the used surfactant to solubilize the drug and concentrate it on the surface of the nanoparticles. Results of the PS obtained from the images are in close agreement with those obtained by the dynamic light scattering.
Fourier infrared spectroscopy and X-ray diffraction presents the IR spectra of the tested samples. The IR spectrum of pure EPL shows the main characteristic functional groups at 2988.65 cm −1 (C–H stretching), 1778.08 cm −1 (anhydride O–C=O stretching), 1726.22 cm −1 (C=O ester stretching), and (1657.64 cm −1 ) C=O stretching. IR spectra of the physical mixture and the freeze dried EPL-loaded NLCs system showed no chemical changes, where all drug functional groups were retained, which confirmed the compatibility of EPL with other formulation additives. presents the diffraction patterns of the examined samples. The diffraction pattern of pure EPL exhibited high-intensity crystallinity peaks at 10.19, 12.24, 14.55, 18.12, and 22.61° 2 θ . In the diffractogram of the physical mixture, the characteristic peaks of GMS were present at 11.9, 14.58, and 18.00° 2 θ , confirming its crystallinity. The crystalline peaks of EPL in physical mixture at 10.19, 14.55, and 22.61° 2 θ were evident but with lower intensities, confirming the presence of EPL in crystalline form. On the contrary, the diffractogram of optimized EPL-loaded NLCs system showed absence of EPL constructive peaks, signifying that the drug lost its crystallinity where it was transformed to an amorphous state or dispersed within the nanocarriers. On the other hand, the two characteristic peaks of GMS (14.58, 18.00° 2 θ ) were still present in the diffractogram of optimized NLCs system confirming that the carrier crystallized into its stable polymorphic form.
Ex-vivo permeation study The HPLC method showed a linear response in the range of 0.2–2 μg/ml ( R 2 = 0.9915). shows the ex-vivo permeation profiles of EPL through the rabbit intestine in 0.9% saline (pH = 7.4) from the optimized EPL-loaded NLCs (containing 5 mg EPL) compared to that of EPL aqueous suspension (containing an equivalent dose). As shown in , great difference in the permeation profiles of optimized EPL-loaded NLCs and dug aqueous suspension through the intestine was observed, where the cumulative amount of drug permeated from the optimized EPL-loaded NLCs was two-fold higher that that from EPL aqueous suspension after 24 h. The optimized EPL-loaded NLCs showed significantly higher cumulative amount of EPL permeated after 24 h (Q 24h optimized NLCs system∼292.56 ± 11.61 μg/cm 2 ) compared to (Q 24h suspension∼139.35 ± 0.00 μg/cm 2 ) ( p < .05). The low transfer through the intestine was observed with drug suspension due to the poor EPL aqueous solubility (BCS class II) (Khames, ). The transport of drug from optimized-loaded NLCs system through intestinal membrane could be achieved as free molecules or as the intact vesicles encapsulating the drug that was acetonitrile diluted samples were used for the analysis (acetonitrile is capable to disrupt the vesicle wall releasing the encapsulated drug) (Ibrahim et al., ). Several studies reported the uptake of intact vesicles by the intestinal epithelial cells (M-cells) in the Payer's patch (Poonia et al., ; Aburahma, ) and the intestinal enterocytes (Niu et al., ), confirming that the main transport mechanism is supposed to be through vesicular uptake as most of the drug is encapsulated into the vesicles, where the free nanosized molecules are transported and absorbed via paracellular or transcellular pathways through the intestinal membrane. These results indicate that the optimized NLCs system was capable to incorporate high amounts of drug and, hence, achieve high drug permeation in the ex-vivo study.
Conclusion In this study, EPL-loaded NLCs were successfully prepared by emulsification-solvent evaporation method using d -optimal design. The optimized system exhibited a small particle size (134 ± 15.60 nm) with narrow size homogenous distribution (PDI = 0.313 ± 0.09), and high entrapment efficiency (76 ± 6.56%). Morphological examination by TEM confirmed a small smooth spherical nano-carrier structure. The ex-vivo permeation study confirmed the ability of the optimized system to cross the intestinal barrier, with two-fold higher drug permeated after 24 h compared to the drug aqueous suspension. These promising results pave the way for the potential use of the prepared EPL-loaded NLCs as s successful drug delivery system for oral treatment of CSCR.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.