Dataset Viewer
doc_id
stringlengths 13
18
| text
stringlengths 1.54k
193k
|
|---|---|
s13018-025-05901-1 | Cui et al. Journal of Orthopaedic Surgery and Research (2025) 20:503 https://doi.org/10.1186/s13018-025-05901-1
Journal of Orthopaedic Surgery and Research
Noninvasive prediction of failure of the conservative treatment in lateral epicondylitis by clinicoradiological features and elbow MRI radiomics based on interpretable machine learning: a multicenter cohort study
Jianing Cui1, Ping Wang1, Xiaodong Zhang2, Ping Zhang3, Yuming Yin4 and Rongjie Bai1*
Abstract Objectives To develop and validate an interpretable machine learning model based on clinicoradiological features and radiomic features based on magnetic resonance imaging (MRI) to predict the failure of conservative treatment in lateral epicondylitis (LE).
Methods This retrospective study included 420 patients with LE from three hospitals, divided into a training cohort (n = 245), an internal validation cohort (n = 115), and an external validation cohort (n = 60). Patients were categorized into conservative treatment failure (n = 133) and conservative treatment success (n = 287) groups based on the outcome of conservative treatment. We developed two predictive models: one utilizing clinicoradiological features, and another integrating clinicoradiological and radiomic features. Seven machine learning algorithms were evaluated to determine the optimal model for predicting the failure of conservative treatment. Model performance was assessed using ROC, and model interpretability was examined using SHapley Additive exPlanations (SHAP).
Results The LightGBM algorithm was selected as the optimal model because of its superior performance. The combined model demonstrated enhanced predictive accuracy with an area under the ROC curve (AUC) of 0.96 (95% CI: 0.91, 0.99) in the external validation cohort. SHAP analysis identified the radiological feature “CET coronal tear size” and the radiomic feature “AX_log-sigma-1-0-mm-3D_glszm_SmallAreaEmphasis” as key predictors of conservative treatment failure.
Conclusions We developed and validated an interpretable LightGBM machine learning model that integrates clinicoradiological and radiomic features to predict the failure of conservative treatment in LE. The model demonstrates high predictive accuracy and offers valuable insights into key prognostic factors.
Keywords Lateral epicondylitis, Machine learning, Magnetic resonance imaging, Radiomics, SHapley additive explanation
Correspondence: Rongjie Bai [email protected]
Full list of author information is available at the end of the article
© The Author(s) 2025. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit h t t p : / / c r e a t i v e c o m m o n s . o r g / l i c e n s e s / b y - n c - n d / 4 . 0 /.
Page 2 of 12
Introduction Lateral epicondylitis (LE) is caused by a chronic degen- erative lesion at the origin of the common extensor ten- don (CET) of the elbow and is one of the most common conditions leading to lateral elbow pain and limitation of upper limb function [1]. Conservative treatment has a high success rate in the early stages of LE, but some patients still require surgical management after long- term conservative treatment fails [2]. There is a lack of standardization of treatment decisions for LE, which relies heavily on the experience of the clinician [2]. Without timely and appropriate treatment, patients are at higher risk of progressive tendon degeneration and potential rupture. The longer the symptoms persist, the lower the likelihood of successful treatment and recovery [3, 4]. Therefore, early and accurate prediction of failure of conservative treatment and early surgical intervention is essential to improve the prognosis of patients with LE.
Current research has recognized the importance of accurate early prognostic prediction and has found that persistent pain, CET abnormalities, and muscle edema are associated with failure of conservative treatment [4–7]. However, the predictive performance of traditional models still needs to be improved. In addition, although magnetic resonance imaging (MRI) is recognized as a high-precision detection method [8]. Currently, deter- mining the extent of elbow injury primarily relies on the individual radiologist’s work experience and interpreta- tion skills. It is more difficult to capture subtle changes in the tissue features shown on MRI that may be correlated with treatment outcomes.
At present, medical image data has grown explosively, in which there is a large amount of interference and use- less information, which invariably brings challenges to the professional ability of doctors and restricts the accu- racy of judgment. Therefore, it is particularly important to conduct scientific, accurate, and effective information mining. In recent years, radiomics is an advanced quan- titative image analysis method that utilizes computer algorithms to extract a large number of quantitative fea- tures from medical images to detect information not vis- ible to human observers. Radiomics has been effective in improving the accuracy of clinical diagnosis and progno- sis [9–11], and has been progressively applied to muscu- loskeletal radiology in recent years, demonstrating high performance and efficiency in the diagnosis of struc- tural tears in the anterior cruciate ligament, meniscus, and supraspinatus tendon [12–14]. However, radiomics studies on lateral elbow tendon and ligament injuries are relatively few and short-lived. Furthermore, the pro- cess developed in machine learning algorithms is opaque and has been likened to a “black box”. That is, research- ers cannot know how the algorithm works. This is gener- ally regarded as unfortunate because it is not conducive
to clinical application. The Shapley Additive exPlanations (SHAP) is an emerging interpretability method that uses Shapley values to measure how much each feature con- tributes to the model’s prediction [15–17]. The SHAP analysis provides case-specific (local) and population- level (global) explanations for “black-box” problems while maintaining a high level of consistency and diagnostic accuracy compared to other explainable AI methods [18]. Therefore, we aimed to construct a machine learning model combining MRI radiomics and clinicoradiological features to predict the failure of conservative treatments and validate the generality of the model through external validation, followed by further interpretation and visual- ization of the model using the SHAP method.
Materials and methods The Institutional Review Board (IRB) approval was obtained for this study, and informed consent was waived owing to the study’s retrospective nature.
Patient selection
This is a multicenter study in includes 3 hospitals. A total of 420 consecutive LE patients between January 2017 and January 2023 were retrospectively recruited in this study. The inclusion criteria were as follows: (1) The patient presented with typical symptoms of lateral elbow pain and was diagnosed with lateral epicondylitis by physical examination; (2) the patient underwent an MRI scan of the elbow. Exclusion criteria were as follows: (1) systemic comorbidities; (2) history of elbow surgery or arthroscopic procedure; (3) insufficient clinical informa- tion; (4) loss of follow-up; (5) poor MRI quality. Figure 1 depicts the study workflow. Initially, all patients with LE were treated conservatively, including rest, nonsteroidal anti-inflammatory drugs (NSAIDs), force-resistant stent protection, and physical therapy. Failure of conservative treatment is defined as no improvement in pain or dys- function after 6 to 12 months of nonsurgical treatment. Patients were categorized by experienced orthopedic surgeons specializing in upper limb corrections into the conservative treatment success group and the conser- vative treatment failure group based on whether or not they received surgical treatment. The same classification criteria were used in all three hospitals. Patients from our hospital were randomly assigned to the training set, and the internal validation was set at a ratio of 7:3, while patients from the other two hospitals were used as the external validation set. The clinical variables included age, affected side, and duration of pain.
MR imaging examinations and analysis
MRI of the elbows of patients at the three hospitals was performed using a 3.0T MRI scanner (Philips Medi- cal Systems) equipped with a dedicated extremity coil.
Page 3 of 12
Fig. 1 Flowchart depicts the patient’s enrollment process in the study. MRI: magnetic resonance imaging
The MRI protocol of the elbow included T1-weighted image (T1WI) sequences in coronal and proton den- sity-weighted imaging with fat suppression (PD-FS) sequences in axial, coronal, and sagittal planes. The MRI protocols of the three hospitals are detailed in Table S1.
An experienced musculoskeletal radiologist reviewed the elbow MRI blinded to clinical data. CET abnor- malities are categorized as tendinosis and tendon tears. Tendinosis was characterized by increased signal inten- sity within the tendon, lower than the fluid signal, and/ or an increased tendon thickness. Determination of the presence of a tear based on fluid signal intensity in the tendon region [5]. The maximum extent of the CET tear was measured on coronal and axial PD-FS images. Lat- eral collateral ligament (LCL) complex insufficiency was classified into three grades: Grade 0 was normal, Grade 1 was a partial tear, peri-ligamentous edema, thickening or thinning, and Grade 2 was a complete tear [5]. Joint effusions, muscle edema, subchondral cysts, subchondral bone marrow edema, and cartilage defects, were classi- fied as present or absent.
Radiomics workflow
Selection, Model Construction, and Model evaluation (Fig. 2).
Scar segmentation
ROI were manually outlined on coronal and axial PD-FS images by a musculoskeletal radiologist using ITK SNAP software (version 3.6.0, www.itksnap.org). The ROI region is the portion of the CET and LCL complex originating at the lateral epicondyle of the humerus and contains 4–5 consecutive MRI slices (Fig. 2). The seg- mentation protocol is described in the Supplementary material.
Feature extraction and feature selection
A total of 2260 features were initially extracted from the coronal and axial PD-FS images, covering shape, first- order statistics, and texture matrices. After conducting robustness testing, which included Pearson correlation and intraclass correlation coefficient (ICC) for absolute agreement, and additional selection procedures, the final radiomics features were retained for combined model construction [19]. Detailed procedures are available in the Supplementary Material.
The comprehensive radiomics workflow encompassed four stages: MRI preparation and Regions of interest (ROI) Segmentation, Feature Extraction and subsequent
Model construction, evaluation, and interpretability
We constructed two models to evaluate the incremen- tal value of MRI radiomics in predicting the failure of
Page 4 of 12
Fig. 2 Workflow of the radiomic analysis in this study. ICC: intraclass correlation coefficient for absolute agreement; ANOVA: analysis of variance; LASSO: least absolute shrinkage and selection operator; MI: mutual information; RFE: recursive feature elimination; ROC: receiver operating characteristic curve; AUC: area under the ROC curve; SHAP: Shapley additive explanations
conservative treatment of LE. Clinicoradiological fea- tures comprise conventional MRI findings (e.g., CET tear size and muscle edema) and traditionally clinically collected parameters (e.g., age, sex, and pain duration). In contrast, radiomic features are quantitative features extracted from MRI through computational algorithms that capture imaging characteristics beyond human visual assessment (e.g., first-order statistics and texture matrices). The clinicoradiological model was developed based on univariable and multivariable logistic regres- sion analyses of clinicoradiological characteristics. The combined model was based on univariable and multi- variable logistic regression analyses of clinicoradiologi- cal model features and selected radiomics features. These features are used in seven machine learning algorithms to develop machine learning models, including support vec- tor machine (SVM), logistic regression (LR), K-Nearest Neighbors (KNN), Decision Tree (DT), Random Forest (RF), XGBoost, and Light Gradient Boosting Machine (LightGBM). Each machine learning model’s predictive performance was evaluated by comparing the accuracy, sensitivity, and specificity in predicting the failure of con- servative treatments.
SHAP analysis
After identifying the best predictive model, we turned our focus to understanding the contribution of each vari- able to the prediction. SHAP quantifies the contribution
of each feature to a model’s prediction, thereby improv- ing global and local interpretability [20, 21]. Specifi- cally, global interpretation aims to assign meaning to each model feature. This process involves calculating the SHAP value for each feature (higher magnitude indicates greater influence on model results), plotting a SHAP beeswarm plot (showing the distribution of SHAP val- ues across individual patients), and plotting a SHAP bar chart (showing the features’ mean absolute SHAP values across patients, ranked). The purpose of local interpreta- tion is to emphasize the impact of features on determin- ing the outcomes of the model in individual patients by constructing SHAP force plots that show the features that contribute most to the predicted probability of an individual patient in the model.
Statistical analysis
Continuous variables were expressed as median and interquartile range (IQR), while categorical variables were expressed as counts and percentages. Differences between the groups were analyzed using the Mann– Whitney U test, chi-square test, or Kruskal-Wallis H-test as appropriate. Univariable analysis of candidate vari- ables was performed using univariable logistic regression. Parameters that demonstrated statistical significance (P < 0.05) in the univariable analysis were then examined using a backward stepwise multivariable logistic regres- sion analysis. DeLong’s test was used to compare the
Page 5 of 12
AUC of the combined model and the clinicoradiological model. A p-value less than 0.05 indicated statistical sig- nificance. All figures presented in this work were created using R software (Version 4.2.1, available on the R Proj- ect website). For statistical analysis and machine learning analysis, the R packages “ggstatsplot” (version 0.9.1.9000), “ggcor”, the open-source Python packages SciPy, scikit- learn, glmnet, caret, and SHAP were used.
Result Patient population
A total of 483 patients with LE underwent elbow MRI in 3 hospitals, of whom 63 were excluded. Therefore, 420 patients were included in the final study, including 245 patients in the training set, 115 in the internal validation set, and 60 in the external validation set (Fig. 1). There were no significant differences in clinical and MRI char- acteristics between the training, internal validation, and external validation sets (all P > 0.05) (Table 1). The con- servative treatment failure group had fewer men than the conservative treatment success group (P = 0.031). Pain duration was longer in the conservative treatment fail- ure group compared to the conservative treatment suc- cess group (both P < 0.001). CET tears were significantly larger in the conservative treatment failure group (both P < 0.001). The conservative treatment failure group had significantly higher CET and LCL complex abnormal- ity grades (both P < 0.001). In addition, muscle oedema, cartilaginous defect, and subchondral cyst were more common in the conservative treatment failure group (P< 0.001, P = 0.009, and P< 0.001) (Table S2).
Radiomics feature extraction, selection, and development of models
A total of 473 radiomics features demonstrated robust- ness against resolution resampling (ICC > 0.9). The remaining radiomics features were reduced by four feature selection methods, and finally, we selected 12 radiomics features using the RFE method.
Clinicoradiological model- Univariable logistic regres- sion analysis revealed significant variables. After mul- tivariable analysis, only duration of pain[OR: 1.146 (1.082–1.213), P <0.001], CET coronal tear size [OR: 1.760 (1.424–2.175), P <0.001], Grade 1 of LCL complex abnormality [OR: 2.658 (1.272–5.553), P = 0.009], and presence of muscle oedema [OR: 9.840 (2.766–35.013), P = 0.002] were independent predictors of conservative treatment failure (Table 2).
Combined model- Univariate analysis was based on the 4 features of the clinicoradiological model and selected 12 radiomics features using the RFE method. In univari- able logistic analysis, 13 characteristics (4 clinicoradio- logical and 9 radiomics) had a P value of less than 0.05 and were therefore included in the multivariable logistic analysis. The final analysis identified 7 factors associated with the prognosis of conservative treatment, including 3 clinicoradiological features and 4 radiomics features (all P<0.05) (Table S3).
Model evaluation and model interpretability
Among the 7 different machine learning models, we selected LightGBM as our algorithm for constructing the combined model owing to its superior performance in
Table 1 Clinical and MRI characteristics of patients in the training and validation sets Characteristics Sex, n (male, %) Age (years) Laterality, n (left, %) Duration of pain (months) CET abnormality, n (%) Tendinosis Tear Size of CET tear (mm) Coronal plane Axial plane LCL complex abnormality, n (%) Grade 0 Grade 1 Grade 2 Muscle oedema, n (%) Joint effusion, n (%) Cartilaginous defect, n (%) Subchondral cyst, n (%) Subchondral marrow oedema, n (%) Data are presented as number (percentage) for dichotomous variables and median (interquartile range) for continuous variables. MRI: Magnetic resonance imaging; LCL: lateral collateral ligament; CET: common extensor tendon. *Statistically significant difference
Training (n = 245) 100(41) 47(41,53) 73(30) 10(8,12)
Internal (n = 115) 53(46) 47(41,52) 37(32) 10 (8,13)
External (n = 60) 25(42) 48.50(41,54.75) 18(30) 10 (7.25,12.75)
171(70) 74(30)
85(74) 30(26)
39(65) 21(35)
0(0,2.15) 0 (0,2.87)
0 (0,1) 0 (0,1.93)
0 (0,3.37) 0 (0,4.19)
136(56) 64(56) 34(57) 24(10) 226(92) 76(31) 96(39) 19(8)
104(42) 50(44) 23(38) 12(10) 107(93) 30(26) 38(33) 9(8)
5(2) 1(1) 3(5) 6(10) 57(95) 16(27) 20(33) 1(2)
P value 0.636 0.697 0.897 0.708 0.570
0.360 0.212 0.487
0.982 0.756 0.572 0.448 0.224
Page 6 of 12
Table 2 Clinical and MRI features associated with failed Conservative treatment for lateral epicondylitis Characteristics
Univariable analysis 95%CI 0.310–0.969 0.975–1.038 0.416–1.393 1.109–1.232 3.082–10.136
OR 0.548 1.006 0.761 1.169 5.589
P value 0.039* 0.720 0.377 <0.001* <0.001*
OR -
Sex, n (male, %) Age (years) Laterality, n (left, %) Duration of pain (months) CET abnormality, n (%) Size of CET tear (mm) Coronal plane Axial plane LCL complex abnormality, n (%) Grade 0 Grade 1 Grade 2 Muscle oedema, n (%) Joint effusion, n (%) Cartilaginous defect, n (%) Subchondral cyst, n (%) Subchondral marrow oedema, n (%) MRI: Magnetic resonance imaging; LCL; lateral collateral ligament; CET: common extensor tendon; OR; odds ratio; CI: confidence interval
1.156 -
1.757 1.519
1.476–2.091 1.326–1.739
<0.001* <0.001*
1.728 -
3.489 16.923 10.679 2.596 1.693 1.462 2.091
1.961–6.206 1.815-157.692 3.814–29.904 0.734–9.191 0.957–2.993 0.845–2.528 0.813–5.377
<0.001* 0.013* <0.001* 0.139 0.070 0.174 0.126
2.522 4.203 9.530
Multivariable analysis
95%CI -
1.093–1.223 -
1.397–2.138 -
1.202–5.288 0.297–59.490 2.621–34.653
P value -
<0.001* -
<0.001* -
0.014* 0.288 <0.001*
Table 3 Performance of different machine learning combined models in the external validation set 95% CI Machine learning Model 0.64–0.92 SVM 0.66–0.92 LR 0.45–0.76 KNN 0.31–0.54 Decision Tree 0.86–0.99 RF 0.82–0.99 XGBoost 0.91–0.99 LightGBM AUC: area under the curve; CI: confidence interval; PPV: positive predictive value; NPV: negative predictive value; SVM: Support Vector Machine; LR: logistic regression; KNN: K-Nearest Neighbors; RF: Random Forest; LightGBM: Light Gradient Boosting Machine
AUC 0.79 0.80 0.61 0.43 0.93 0.91 0.96
Accuracy 0.75 0.73 0.43 0.33 0.85 0.92 0.87
Sensitivity 0.68 0.68 0.74 0.68 0.84 0.74 0.74
Specificity 0.78 0.76 0.29 0.17 0.85 1 0.93
PPV 0.59 0.57 0.33 0.28 0.73 1 0.82
NPV 0.84 0.84 0.71 0.54 0.92 0.89 0.88
Table 4 Accuracy and predictive value of the clinicoradiological model and combined model
Accuracy
Sensitivity
Specificity
PPV
NPV
Training set Clinicoradiological model Combined model Internal validation set Clinicoradiological model Combined model External validation set Clinicoradiological model Combined model PPV: positive predictive value; NPV: negative predictive value
0.784 0.841
0.835 0.757
0.867 0.867
0.363 0.726
0.595 0.487
0.790 0.73
0.976 0.893
0.949 0.885
0.902 0.927
0.892 0.768
0.846 0.668
0.790 0.824
0.772 0.881
0.832 0.784
0.902 0.884
the external validation set (Table 3). The implementation of the best machine learning model source code is avail- able in a public repository with the URL: h t t p s : / / g i t h u b . c o m / y x z 9 5 / E l b o w M R I m o d e l. Table 4 shows the predictive performances of the clinicoradiological and combined models. In the external validation set, the combined model significantly outperformed the clinicoradiological
model (AUC: 0.96 [0.91–0.99] vs. 0.83 [0.67–0.96] by DeLong’s test) (Fig. 3).
SHAP provided a quantitative explanation for the combined model. Figure 4 shows global interpret- ability, whereas Fig. 5 shows interpretability (explanation of an individual case). The feature impor- tance rankings of the 7 predictors were shown in Fig. 4A. The “CET tear size in the coronal plane” and
local
Page 7 of 12
Fig. 3 The ROC curves in the training, internal, and external validation sets. (A) ROC curve of the clinical model. (B) ROC curve of the combined model. ROC: receiver operating characteristic
Page 8 of 12
Fig. 4 SHAP summary plots of the combined model. (A) The SHAP bar chart. Ranking the importance of each feature in the final model output. Higher SHAP values indicate a more significant contribution of the model to a particular feature. (B) The SHAP beeswarm plot. A comprehensive visualization of the cumulative influence of each feature. Each dot represents an individual sample, and the color represents the value of each feature. The x-axis rep- resents the SHAP values, and a positive SHAP value indicates that it has positive effects on the predictions of the model and vice versa. SHAP: Shapley additive explanation
“AX_log-sigma-1-0-mm-3D_glszm_SmallAreaEmphasis” ranked in the top two in terms of their average impact on the combined model’s prediction of prognosis for con- servative treatment. Figure 4B shows the distribution of the contribution of each feature to the model output. We can see that increases in the “CET tear size in the coro- nal plane” and “AX_log-sigma-1-0-mm-3D_glszm_Smal- lAreaEmphasis” have a positive impact and push the prediction toward conservative treatment failure. Fig- ure 5 shows two typical examples of correctly predicting whether a patient will fail conservative treatment. The SHAP force plot shows the positive and negative impacts
of all the features in the combined model on the pre- dicted outcome in a single case.
Discussion While most LE patients recover within 6–12 months with conservative treatment, some patients suffer for 24 months or more, significantly affecting daily life [2, 22]. Improved prognostic tools are needed to predict treat- ment outcomes. In this study, we developed a machine learning model that integrates clinicoradiological and elbow MRI radiomic features to predict the failure of conservative treatment in patients with LE. The com- bined model demonstrated robust performance in the
Page 9 of 12
Fig. 5 Individual visualization of the combined model by SHAP force plots explains how the model predicts the outcome of two patients with lateral epi- condylitis who were treated conservatively. Patient 1 shows an example of a patient who underwent surgery after conservative treatment failed. Patient 2 shows an example of a patient with effective conservative treatment. Images Ba−d, Da−e, Fa−d, and Ha−e were consecutive layers of the original DICOM image in coronal and axial views, respectively. Images Aa−d, Ca−e, E a−d, and Ga−e were manual annotation diagrams. The red area is the ROI area. The base value represents the average prediction of the combined model, and f (x) represents its final prediction probability. The width of the bar indicates how much the feature affects the conservative treatment outcome; the closer to the f (x) value, the wider the bar of the feature, and the greater the effect on the conservative treatment outcome. The red color indicates high feature values, which are usually associated with an increased likelihood of conserva- tive treatment failure. Conversely, the blue color indicates low feature values, which are usually associated with a decreased likelihood of conservative treatment failure. The SHAP force plot for Patient 1 shows that pain duration was the feature that contributed most positively to predicting patient failure of conservative treatment in the combined model. The SHAP force diagram for Patient 2 shows that “AX_log-sigma-1-0-mm-3D_glszm_SmallAreaEmpha- sis” was the feature that contributed most negatively to predicting patient failure of conservative treatment in the combined model
Page 10 of 12
external validation set and provided high interpretability through SHAP analysis, which offers valuable insights into the factors influencing the model’s predictions.
Several studies have recognized that accurate early prognostic prediction and the development of a rational treatment plan are essential for treating LE and have ana- lyzed the factors associated with the failure of conserva- tive treatment. Elisa et al. [4] compared demographic and disease-specific characteristics of the non-surgical and surgical treatment groups and found that characteristics such as radial tunnel syndrome and symptom duration longer than 12 months were independent predictors of surgical treatment. Studies also analyze the relationship between MRI and prognosis in patients with LE [5, 7, 23, 24]. A study on 60 patients with LE developed a model that included persistent pain, CET abnormalities in the longitudinal plane, and muscle edema to predict opera- tive treatment [5]. Ikeda et al. [7]concluded that treat- ment strategies including surgery should be considered for patients with severe MRI findings of CET and LCL complex. In contrast, Kessler et al. [24]concluded that the size of the CET defect size on MRI did not predict the success or failure of conservative treatment. Therefore, the clinical generalizability of its existing studies has been questioned due to limited patient samples, lack of inter- nal and external validation, and controversial results of studies. In addition, the relationship between radiomics and the prognosis of LE has not been previously evalu- ated. Given that CET in patients with LE undergoes his- tologically progressive degenerative changes, the severity of which can only be definitively assessed by postopera- tive pathology. Radiomics is a field that extracts a large number of quantitative features from medical images to capture subtle changes in CET and LCL complex. Because radiomics extracts high-dimensional features on a spatial scale, it is easier to detect hidden information and reflect the underlying pathophysiology associated with lateral epicondylitis than traditional imaging fea- tures, thus providing a more objective and reproducible assessment method.
Radiomics has emerged as a powerful tool for evalu- ating tendon and ligament injuries, offering high diag- nostic efficacy [12]. Cheng et al. [13] found that a machine-learning-based multi-sequence MRI radiomics model demonstrated superior performance in diagnosing anterior cruciate ligament tears. Zhan et al. [14] found that MRI-based radiomics could accurately determine the presence and extent of tears in the supraspinatus tendon. Droppelmann et al. [25] found that ultrasound image- based machine learning algorithms have high diagnos- tic accuracy for CET tears in patients with LE, with an AUC of 0.991. However, the application of radiomics to the elbow joint has been limited. To our knowledge, this is the first study to assess the ability of radiomic features
to predict failure of conservative treatment for LE with further external validation, providing clinicians with a tool to make timely decisions regarding surgical interven- tion, thereby minimizing patient discomfort and reduc- ing healthcare costs.
There are two main reasons why the ROI of our study included not only CET but also the LCL complex. (1) LE is a degenerative process that occurs at the origin of the CET [1]. However, previous studies have shown that CET is not always an isolated lesion of LE, and LCL complex injury often occurs with progressive injury of CET [23, 26]. This is further supported by our study, where the rate of LCL complex abnormalities was significantly higher in the conservative treatment failure group. (2) Given the close anatomical relationship between the LCL complex and CET, both originating from the lateral epicondyle, it is challenging to distinguish between the tendon and lig- ament on MRI.
In this research, the performance of five machine learn- ing algorithms was compared. Eventually, LightGBM was chosen as the optimal algorithm, and it demonstrated excellent performance and generalizability in external validation. Our findings demonstrate that integrating clinicoradiology and radiomics features is critical for optimizing predictive performance. While traditional clinicoradiological features are effective in capturing macroscopic patterns of disease progression, they exhibit limited sensitivity in detecting subtle pathophysiologic changes. Radiomics analysis, while capable of quantify- ing imperceptible tissue features, lacks strong clinical relevance. Our combined model compensates for these respective incorporating quantitative imaging features while retaining clinically routine fea- tures, ultimately achieving higher predictive accuracy.
limitations by
Lack of interpretability is a significant challenge asso- ciated with machine learning predictive models based on radiomics, which limits their application in clinical practice. We calculated the SHAP values to interpret the combined model and address this issue. SHAP recog- nizes specific patterns learned through complex machine learning algorithms and provides both global and local interpretability for radiomics models. SHAP enables interpretation and visualization of LightGBM models in a clinician-friendly manner through SHAP plots. From a global perspective, the SHAP summary plots provide an intuitive and concise graph that represents the range and importance distribution of the feature’s impact on the output of the combined model (the more important the feature, the higher the SHAP value, and the wider the range of dots). In this study, clinicians could see how the features’ value impacted the assessment through the dots’ range and color. At the same time, clinicians were able to simultaneously identify the feature that contributed the most to the model as the radiologic feature “CET tear size
Page 11 of 12
on the coronal plane.” In LE, the CET is the primary site of stress injury caused by repetitive contractions of the forearm extensor muscles. Several studies have shown that patients with more extensive CET tears treated con- servatively have a poorer prognosis and require further operative treatment. Furthermore, the radiomics feature “AX_log-sigma-1-0-mm-3D_glszm_SmallAreaEmpha- sis” had the second highest average impact on the com- bined model predictions. It describes the distribution of small-sized zones, which may correspond to areas of inflammation or tissue damage [27]. In our study, the ‘AX_log-sigma-1-0-mm-3D_glszm_SmallAreaEmpha- sis’ of the conservative treatment success group was sig- nificantly lower than that of the conservative treatment failure group, which implies that the conservative treat- ment success group was observed in axial images with CET and LCL complex areas that were smaller and finer textured.
By understanding the influence of features on the com- bined model and their underlying pathophysiological mechanisms, clinicians can utilize the SHAP force plot for a local interpretation of individual patient predic- tions. The clinician simply compares the SHAP value for a given case to a baseline threshold, with values above the baseline threshold indicating that the patient belongs to the conservative treatment failed group. This visualiza- tion method intuitively displays feature contributions through color-coded arrows (red indicates an increased probability that the patient will require surgery) and scale lengths representing the relative degree of impact. As shown in Fig. 5A (red), ‘CET tear size’ has a positive effect on the assessment of conservative treatment fail- ure, and the length of the arrow for ‘CET tear size’ is lon- ger than that for ‘duration of pain’ representing a greater effect on the model than that for ‘duration of pain’.
models. Future prospective multicenter studies using standardized imaging protocols would be valuable in validating our findings and improving the robustness of our conclusions. Additionally, we recommend investigat- ing ways to combine imaging modalities such as X-ray, CT, and ultrasound with clinical features to develop a more comprehensive predictive model that can signifi- cantly improve the clinical applicability of the model in different healthcare settings. Second, while our model is based on radiomic features, incorporating deep learn- ing approaches may further enhance predictive accuracy. Finally, some patients had relatively severe symptoms as they had received poor results from conservative treat- ment in other hospitals, and then came to our hospital.
In conclusion, we developed and validated a noninva- sive and robust LightGBM machine learning model that combines MRI radiomics and clinicoradiological features to predict the failure of conservative treatment in LE. The SHAP provides a bridge for personalized prediction, which may aid clinical decision-making for the individu- alized treatment of LE.
Abbreviations MRI SHAP CET LCL T1WI PD-FS ROI AUC ROC IQR ICC
Magnetic resonance imaging SHapley Additive exPlanations Common extensor tendon Lateral collateral ligament T1-weighted image Proton density-weighted imaging with fat suppression Regions of interest Area under the curve Receiver operating characteristic curve Interquartile range Intraclass correlation coefficients
Supplementary Information The online version contains supplementary material available at h t t p s : / / d o i . o r g / 1 0 . 1 1 8 6 / s 1 3 0 1 8 - 0 2 5 - 0 5 9 0 1 - 1.
Our predictive model enables clinicians to make more accurate and timely diagnostic assessments, facilitating prompt intervention and optimized management strate- gies. For high-risk patients identified by the model, cli- nicians may consider early surgical intervention with intensified monitoring to prevent disease progression. Conversely, low-risk patients can be managed with stan- dard conservative protocols while avoiding unnecessary surgical procedures. This reduces pain and further dam- age and enhances patients’ quality of life by expediting the rehabilitation process and enabling faster return to daily activities and physical function.
Supplementary Material 1
Acknowledgements Not applicable.
Author contributions RB was responsible for searching for research related to the subject content, designing the research proposal, and writing the introductory section of the manuscript. JC and PW performed the MRI image analysis, and ROI outlining, and wrote the discussion and introduction sections of the manuscript. XZ and PZ collected MR images from other hospitals, reviewed the image features, and revised the manuscript. YY was responsible for the statistical processing and writing the methods section.
Our study has several limitations. First, as a retro- spective study, our study inevitably presents potential selection bias. Although we used standardized image preprocessing and harmonization techniques, subtle differences in MRI protocols, scanner parameters, and clinical evaluation criteria among the three hospitals may affect the generalizability of feature extraction and
Funding This study was supported by the National Natural Science Foundation of China (Grant numbers 82171921, 81771809) and the Beijing Natural Science Foundation of China (Grant number 7202063).
Data availability The data that support the findings of this study are available from Beijing Jishuitan Hospital but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly
Page 12 of 12
available. Data are however available from the authors upon reasonable request and with permission of Beijing Jishuitan Hospital.
11. Pak E, Choi KS, Choi SH, et al. Prediction of prognosis in glioblastoma using radiomics features of dynamic Contrast-Enhanced MRI. Korean J Radiol. 2021;22(9):1514–24.
Declarations
12. Fritz B, Yi PH, Kijowski R, Fritz J. Radiomics and deep learning for disease detection in musculoskeletal radiology: an overview of novel MRI- and CT- Based approaches. Invest Radiol. 2023;58(1):3–13.
Ethics approval and consent to participate This study was approved by the ethics committee of Beijing Jishuitan Hospital. For this retrospective analysis informed consent is not required.
13. Cheng Q, Lin H, Zhao J, Lu X, Wang Q. Application of machine learning-based multi-sequence MRI radiomics in diagnosing anterior cruciate ligament tears. J Orthop Surg Res. 2024;19(1):99.
Consent for publication Not applicable.
14. Zhan J, Liu S, Dong C, et al. Shoulder MRI-based radiomics for diagnosis and severity staging assessment of surgically treated supraspinatus tendon tears. Eur Radiol. 2023;33(8):5587–93.
Competing interests The authors declare no competing interests.
Author details 1Department of Radiology, Beijing Jishuitan Hospital, Capital Medical University, Beijing 100035, China 2Department of Radiology, The Third Affiliated Hospital, Southern Medical University, Zhongshan Avenue West, Tianhe District, Guangzhou 510515, China 3Department of Radiology, Beijing Geriatric Hospital, Beijing 100095, China 4Department of Radiology, Pomona Valley Hospital Medical Center, Pomona, CA 91767, USA
Received: 19 February 2025 / Accepted: 9 May 2025
15. Ponce-Bobadilla AV, Schmitt V, Maier CS, Mensing S, Stodtmann S. Practical guide to SHAP analysis: explaining supervised machine learning model predictions in drug development. Clin Transl Sci. 2024;17(11):e70056. 16. Rodríguez-Pérez R, Bajorath J. Interpretation of compound activity predic- tions from complex machine learning models using local approximations and Shapley values. J Med Chem. 2020;63(16):8761–77.
17. Quinn KN, Wilber H, Townsend A, Sethna JP. Chebyshev approxima- tion and the global geometry of model predictions. Phys Rev Lett. 2019;122(15):158302.
18. Taiyeb Khosroshahi M, Morsali S, Gharakhanlou S, et al. Explainable artificial intelligence in neuroimaging of Alzheimer’s disease. Diagnostics (Basel). 2025;15(5):612.
19. Du R, Lee VH, Yuan H, et al. Radiomics model to predict early progression of nonmetastatic nasopharyngeal carcinoma after intensity modulation radia- tion therapy: A multicenter study. Radiol Artif Intell. 2019;1(4):e180075. 20. Chen Y, Chen S, Tang W, et al. Multiparametric MRI radiomics with machine learning for differentiating HER2-Zero, -Low, and -Positive breast cancer: model development, testing, and interpretability analysis. AJR Am J Roent- genol. 2025;224:e2431717.
References 1. Wolf JM. Lateral epicondylitis. N Engl J Med. 2023;388(25):2371–7. 2.
Karjalainen T, Buchbinder R. Is it time to reconsider the indications for surgery in patients with tennis elbow. Bone Joint J. 2023;105–B:109–11. Smidt N, Lewis M, VAN DER Windt DA, Hay EM, Bouter LM, Croft P. Lateral epi- condylitis in general practice: course and prognostic indicators of outcome. J Rheumatol. 2006;33(10):2053–59. Knutsen EJ, Calfee RP, Chen RE, Goldfarb CA, Park KW, Osei DA. Factors associ- ated with failure of nonoperative treatment in lateral epicondylitis. Am J Sports Med. 2015;43(9):2133–7. Jeon JY, Lee MH, Jeon IH, Chung HW, Lee SH, Shin MJ. Lateral epicondylitis: associations of MR imaging and clinical assessments with treatment options in patients receiving Conservative and arthroscopic managements. Eur Radiol. 2018;28(3):972–81.
3.
4.
5.
6. Miyamura S, Temporin K, Miyata S, Miyake T, Shimada K. Arthroscopic debridement for refractory lateral epicondylitis results for substantial improvement in tendinosis scores and good clinical outcomes: qualita- tive and quantitative magnetic resonance imaging analysis. Arthroscopy. 2022;38(12):3120–9. Ikeda K, Ogawa T, Ikumi A, et al. Magnetic resonance imaging predicts outcomes of Conservative treatment in patients with lateral epicondylitis. J Orthop Sci. 2024;29(3):795–801. 7.
21. Ning C, Ouyang H, Xiao J, et al. Development and validation of an explain- able machine learning model for mortality prediction among patients with infected pancreatic necrosis. EClinicalMedicine. 2025;80:103074. 22. Soeur L, Desmoineaux P, Devillier A, Pujol N, Beaufils P. Outcomes of
arthroscopic lateral epicondylitis release: should we treat earlier. Orthop Traumatol Surg Res. 2016;102(6):775–80.
23. Cha YK, Kim SJ, Park NH, Kim JY, Kim JH, Park JY. Magnetic resonance imag- ing of patients with lateral epicondylitis: relationship between pain and severity of imaging features in elbow joints. Acta Orthop Traumatol Turc. 2019;53(5):366–71.
24. Kessler RE, Day MS, Tyler TF, et al. Predictive value of magnetic resonance imaging in outcomes of nonsurgical treatment of lateral epicondylitis. JSES Int. 2022;6(2):305–8.
25. Droppelmann G, Tello M, García N, Greene C, Jorquera C, Feijoo F. Lateral elbow tendinopathy and artificial intelligence: binary and multilabel find- ings detection using machine learning algorithms. Front Med (Lausanne). 2022;9:945698.
26. Kholinne E, Liu H, Kim H, Kwak JM, Koh KH, Jeon IH. Systematic review of elbow instability in association with refractory lateral epicondylitis: myth or fact. Am J Sports Med. 2021;49(9):2542–50.
27. Zwanenburg A, Vallières M, Abdalah MA, et al. The image biomarker stan- dardization initiative: standardized quantitative radiomics for High-Through- put image-based phenotyping. Radiology. 2020;295(2):328–38.
8. Dewan AK, Chhabra AB, Khanna AJ, Anderson MW, Brunton LM. MRI of the elbow: techniques and spectrum of disease: AAOS exhibit selection. J Bone Joint Surg Am. 2013;95(14):e991–13. Zhou Y, Wu D, Yan S, et al. Feasibility of a Clinical-Radiomics model to predict the outcomes of acute ischemic stroke. Korean J Radiol. 2022;23(8):811–20. 10. Kim KM, Hwang H, Sohn B, et al. Development and validation of MRI-Based radiomics models for diagnosing juvenile myoclonic epilepsy. Korean J Radiol. 2022;23(12):1281–9.
9.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
s13018-024-05015-0 | Sun et al. Journal of Orthopaedic Surgery and Research (2024) 19:626 https://doi.org/10.1186/s13018-024-05015-0
Journal of Orthopaedic Surgery and Research
The efficacy of patellar denervation on prognosis and kneeling capacity after unicompartmental knee arthroplasty: a randomized clinical trial
Ying-Jin Sun1†, Ning Liu1†, Long Huang1, Xiang-Yang Chen1, Ju-Tai Wu1* and Shuo Feng1*
Abstract Objective The aim of this study was to investigate the effect of patellar denervation (PD) on pain, function and ability to kneel after unicompartmental knee arthroplasty (UKA).
Methods Patients with medial knee osteoarthritis who underwent UKA were prospectively selected. Patients were randomly divided into PD and non-PD groups based on whether patellar denervation was performed. Clinical assessment was performed using the Hospital for Special Surgery (HSS) knee score, Kujiala score, visual analogue scale (VAS) and forgotten joint score (FJS-12), as well as postoperative complications were recorded. The patients’ postoperative self-perception and actual ability to perform different kneeling positions were assessed in the two groups.
Results UKA patients treated with PD achieved better Kujiala scores and FJS-12 scores, reduced anterior knee pain and improved kneeling ability postoperatively, validating the effectiveness of PD in UKA. Perception and actual performance of kneeling remained mismatched in PD patients, but performance during different kneeling activities was generally better than in non-PD patients.
Trial registration Clinical Trial Registration: ChiCTR1900025669.
Conclusion Patellar denervation can safely and effectively improve patellofemoral joint function, pain and kneeling ability in the early postoperative period after UKA.
Keywords Unicompartmental knee arthroplasty, Patellar denervation, Anterior knee pain, Kneel
†Ying-Jin Sun and Ning Liu contributed equally to this work.
Correspondence: Ju-Tai Wu [email protected] Shuo Feng [email protected] 1Department of Orthopedic Surgery, Affiliated Hospital of Xuzhou Medical University, 99 Huaihai Road, Xuzhou, Jiangsu 221002, China
© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creativecommons.org/licenses/by-nc-nd/4.0/.
Page 2 of 8
Introduction Unicompartmental knee arthroplasty (UKA) is a mini- mally invasive procedure for the treatment of end-stage osteoarthritis of the medial compartment of the knee, allowing for the preservation of more anatomical struc- tures and proprioception [1–4]. Kneeling is a movement that Asians often perform in their daily life and religious activities and plays a crucial role in postoperative sat- isfaction and quality of life [5, 6]. Although it has been shown that knee flexion range after UKA did not corre- late well with the ability to kneel, kneeling was still a dif- ficult challenge for many early UKA patients [7, 8]. More than 50% of patients with knee arthroplasty consider kneeling to be the most important and common activ- ity in their lives, and more than 70% of patients consider kneeling to be the most difficult activity [8]. Artz et al. [7] found that 32% of UKA patients were still unable to kneel 2 years postoperatively due to pain, and that only 11% of patients were able to kneel easily.
② a complete lateral intercompartment gap and ante- rior cruciate ligament; and ③ the Iwano grade [17] for patellofemoral joint imaging was stage I-II. ④ Flexion contracture < 15°, knee mobility ≥ 90°, and internal and external valgus deformity < 15°. The exclusion criteria were ① acute osteonecrosis, and inflammatory arthri- tis such as rheumatoid arthritis, ankylosing spondyli- tis, gouty arthritis or infectious arthritis; ② a history of knee surgery in the past or during the follow-up period; ③ waist, hip or foot pain on the operative side; and ④ incomplete clinical or follow-up data.
Sample size calculation
The sample size was calculated using G*Power (version 3.1.9.7). The power (1-β) was 80%, the effect size was 0.5, and the two-tailed ɑ was 0.05. The number of samples from at least 53 patients in each group was calculated. We assumed that the loss to follow-up rate was 20%. Therefore, 66 patients were included in each group.
Several neurohistological studies have suggested that two nociceptive afferent fibres extending around the patella located in the vastus medialis and vastus latera- lis were closely associated with anterior knee pain (AKP) [9, 10]. Patellar denervation (PD) is used to reduce AKP by damaging the peripatellar innervation through peri- patellar electrocautery [11, 12]. The use of PD in total knee arthroplasty has been shown to reduce the risk of postoperative AKP by 32–70% [13–15]. Therefore, more and more orthopaedic surgeons are applying denervation techniques to knee arthroplasty, but reports on its use in UKA are rare [11, 12, 16].
Method of randomization
Subjects were recruited and randomly assigned by two independent physicians. A randomized grouping scheme was kept in 132 opaque envelopes, which were opened in order of enrollment, and the grouping of patients was determined according to the allocation scheme in the envelopes. The envelopes were not opened until the patient entered the operating room. Patients were cat- egorized into PD group and non-PD group based on whether they received PD.
There is a lack of evidence that UKA combined with PD is effective in improving joint function and reducing pain during flexion and kneeling. Therefore, the primary objective of this study was to investigate the effects of PD on pain, joint function, and kneeling ability after UKA. The secondary objective was to validate the efficacy and safety of PD for use in UKA. We hypothesised that peri- patellar electrocautery denervation would be an effective and safe method to improve joint function, reduce pain, and improve patients’ ability to kneel postoperatively.
Materials and methods Patient selection
This study was a randomized clinical trial. Based on the selection criteria, we recruited 132 patients who under- went UKA at our hospital between September 2019 and December 2022. This study followed the principles of the Declaration of Helsinki, and all patients were provided preoperative informed consent. This study was approved by the Ethics Committee and has been registered in the Chinese Clinical Trial Registry (ChiCTR1900025669).
Surgical treatment
All surgeries were performed by the same experienced orthopedic team. All patients underwent Oxford mobile- bearing medial UKA (Zimmer Biomet, America) accord- ing to the manufacturer’s instructions. The patients were placed in the supine position under general anaesthesia, a tourniquet was applied to the thigh, the leg was sus- pended in a lower limb brace, the hip was flexed at 30° with mild abduction, the calf was naturally lowered, and the knee could be flexed to a minimum of 110°. The joint cavity was opened via a medial parapatellar approach, part of the infrapatellar fat pad was resected, and the osteophytes were removed. According to the principle of flexion-extension balance, tibial plateau and distal femur osteotomies were carried out sequentially, and after installing the trial mould to test the joint move- ment and stability, the appropriate type of prosthesis was selected, fixed and installed with bone cement. In the PD group, the end of the electrotome was bent 3 mm. In the extended knee position, the medial aspect of the patella was slightly externally turned. Circumferential electro- cautery denervation of the patella was performed using an electrocautery knife, especially on the medial-superior
The inclusion criteria of the study were as follows: ① diagnosis of osteoarthritis of the medial compartment;
Page 3 of 8
and lateral-superior edges of the patella. The depth of electrocautery was 2 to 3 mm. The soft tissues around the knee joint of the two groups were routinely injected with “cocktail” analgesic solution. After the instruments and dressings were counted, one drainage tube was placed, and the wound was closed layer by layer.
Postoperative management
Standardized protocols were used for all postoperative management. From 12 h to 2 weeks after the operation, an oral anticoagulant (apexaban)were taken to prevent thrombosis. The drainage tube was removed within 24 h after surgery, and the patients were encouraged to walk with the aid of walkers. The patients were instructed to exercise the ankle pump, actively contract the quadri- ceps femoris and perform straight leg elevation exercises. Patients were asked to follow up in an outpatient clinic at 1, 6 and 12 months after surgery.
Clinical evaluation
all parameters, and P < 0.05 was considered to indicate a statistically significant difference.
Results There are 58 patients (58 knees) in the PD group and 62 patients (62 knees) in the non-PD group included in the analysis (Fig. 1). All baseline characteristics were comparable between the two groups, and there were no significant differences in preoperative demograph- ics, American Society of Anesthesiologists (ASA) grade, Charlson comorbidity index (CCI), patellar morphology, degree of patellofemoral joint degeneration and clinical scores were not significantly different (Table 1).
Clinical outcomes
The clinical outcomes showed statistically significant dif- ferences in postoperative Kujala scores, VAS and FJS-12 scores between the two groups; there was no statisti- cally significant difference in postoperative HSS scores between the two groups (Table 2).
Patellar function was evaluated using the Kujiala score [18], and knee function was evaluated using the Hospital for Special Surgery (HSS) knee score [19]. The HSS score and Kujala score ranged from 0 to 100, with higher scores indicating better function. Visual analogue scale (VAS) was used to evaluate the degree of postoperative knee pain [20]. The degree of postoperative prosthesis self- adaptation was assessed using the Forgotten Joint Score (FJS-12) [21, 22] at the last follow-up after surgery, with higher scores indicating a better subjective feeling. Com- plications that occurred during the perioperative period and follow-up were recorded.
Kneeling outcomes
There was statistically significant difference in the actual ability to kneel in the postoperative period between the two groups (Table 3). The highest percentage of “good” actual kneeling ability was “90° kneeling on the cushion in the PD group (55.2%) ” and the lowest percentage of “good” actual ability was “120° kneeling on the floor in the non-PD group (16.1%)”. The highest percentage of actual inability to kneel was “on the floor in the non-PD group (24.2%)”, and the lowest percentage was “on the cushion in the PD group (6.9%)”.
Kneeling evaluation
At 12 months after surgery, an independent investiga- tor assessed the patients’ perceived and actual ability to kneel. Patients were asked in advance if they thought they could kneel at 90° or 120° on the cushion and floor, and their perception of kneeling was assessed accord- ing to whether they answered “yes” or “no”. We assessed patients’ actual ability to kneel with reference to question 7 of the Oxford Knee Score (score 0–4) [23]. To simplify the statistical analysis, a score of 0 indicates “impossible”, 1–2 indicates “poor”, and 3–4 indicates “good”.
There was significant difference between the perceived and actual performance of the two groups (Figs. 2 and 3). The highest concordance between patients’ actual kneel- ing performance being “good” and perceiving that they could kneel was “120° kneeling on the cushion in the PD group (89.7%) ”, and the worst concordance was “90° kneeling on the floor in the non-PD group (69.2%) ”. The highest concordance between patients’ actual inability to kneel and their perceived inability to kneel was “kneeling on the cushion in the PD group (100%)”, and the worst concordance was “kneeling cushion in the non-PD group (80%)”.
Data analyses
Postoperative complications
The statistical software used was SPSS 26.0. Continuous variables are expressed as the mean ± standard deviation (SD) and were analysed using Student’s t test. Noncontin- uous variables are expressed as numbers and percentages (%) and were analysed using the Chi-square test. Mann- Whitney U test was used to compare the grade data of the two groups. α = 0.05 was used as the test criterion for
During the perioperative period, the intervention group and conventional group had 3 cases of lower extremity vein thrombosis (1 and 2, respectively), 1 case of hema- toma (1 and 0, respectively) and 3 cases of poorly healed skin incisions (1 and 2, respectively). The patients were cured after symptomatic treatment, and none of them underwent secondary surgical revision. The incidence of postoperative complications was 5.2% (3/58) in the
Page 4 of 8
Fig. 1 CONSORT flow diagram showing the enrollment of the patients, the allocation of treatment, and the completion of the study
PD group and 6.5% (4/62) in the non-PD group. The difference between the two groups was not statistically significant (P = 0.537). There were no postoperative com- plications, such as infection, fracture, patellar necrosis, or implant loosening, in either group during the follow-up period.
Discussion The findings in this study were consistent with the hypothesis. UKA patients treated with PD achieved bet- ter Kujiala scores and FJS-12 scores, reduced anterior knee pain and improved kneeling ability postoperatively, validating the effectiveness of PD in UKA. Perception and actual performance of kneeling remained mismatched in PD patients, but performance during different kneeling activities was generally better than in non-PD patients. No postoperative complications due to peripatellar elec- trocautery were found in UKA patients who underwent PD treatment, indicating the safety of PD in UKA.
PD can effectively reduce postoperative AKP, but the pathogenesis of postoperative AKP has not been fully elucidated, and it is generally believed that substance-p nerve fibres in the peripatellar soft tissue are the main cause. Substance P, an nociceptive neurotransmitter found in afferent nerve fibres, is predominantly found in the the patellar retinaculum, fat pad, periosteum, and cartilage affected by degenerative diseases [9, 24]. There- fore, pain receptor desensitization through peripatellar electrocautery can reduce pain transmission. In addition, peripatellar soft tissue traction or bony structure com- pression may cause nerve fibre edema and degeneration, as well as intraoperative traction and suturing may easily produce peripheral neuralgia, increasing the risk of post- operative pain. The denervation technique can partially terminate the potential pain pathway and reduce the occurrence of pain.
The effectiveness of PD is currently controversial. Our study revealed that patients who underwent PD had
Page 5 of 8
Table 1 Demographic data
PD (n = 58) 62.3 ± 7.2
Non-PD (n = 62) 63.4 ± 7.3
P-Value 0.413 0.220
Age (y) Sex (%) Man Woman BMI (kg/m²) CCI (%) Grade I Grade II ASA grade (%) Grade I Grade II Grade III Wiberg classification (%) Type I Type II Type III Iwano stage (%) Stage I Stage II Preoperative HSS score Preoperative Kujala score BMI, body mass index; CCI, Charlson comorbidity index; ASA, American Society of Anesthesiologists
23 (39.7) 35 (60.3) 25.8 ± 3.4
18 (29.0) 44 (71.0) 26.2 ± 2.7
0.401 0.219
36 (62.1) 22 (37.9)
45 (72.6) 17 (27.4)
0.156
28 (48.3) 22 (37.9) 8 (13.8)
40 (64.5) 18 (29.0) 4 (6.5)
0.276
10 (17.2) 35 (60.3) 13 (22.4)
18 (29.0) 30 (48.4) 14 (22.6)
0.378
38 (61.3) 24 (38.7) 53.8 ± 6.1 48.5 ± 6.0
40 (69.0) 18 (31.0) 52.6 ± 6.8 46.7 ± 5.8
0.321 0.110
Table 2 Comparison of postoperative clinical outcomes between the two groups
PD (n = 58)
Non-PD (n = 62)
P-Value
HSS score At 6 months At 12 months Kujala score At 6 months At 12 months VAS At 6 months At 12 months FJS-12 Cushion Kneeling perception (Yes / No) Floor Kneeling perception (Yes / No) HSS, the hospital for special surgery; FJS-12, the Forgotten Joint Score. Bold values indicate statistically significant values (P < 0.05)
88.2 ± 4.2 90.3 ± 4.2
86.8 ± 4.8 89.3 ± 4.3
0.102 0.171
80.5 ± 4.2 83.8 ± 3.9
76.7 ± 5.2 81.5 ± 4.4
< 0.001 0.004
1.3 ± 0.6 1.2 ± 0.4 73.8 ± 6.5 34/24
1.7 ± 0.8 1.4 ± 0.6 70.4 ± 5.8 32/30
0.007 0.027 0.003 0.441
26/32
25/37
0.618
better patellar scores, greater joint forgetting and less pain. However, some studies have suggested that these improvements may not be durable. However, some stud- ies have suggested that these improvements may not be long-lasting. A meta-analysis by Yuan et al. [15] showed that the effective effect of PD was limited to 12 months postoperatively, but a meta-analysis by Duan et al. [25] demonstrated that the effect of PD can extend beyond 12 months. As severe patellofemoral joint degenera- tion (PFJD) is widely considered a contraindication to
Table 3 Comparison of postoperative kneeling scores for question 7 of the Oxford knee score
Impossible Poor
90° Kneeling (Cushion) PD Non-PD 120° Kneeling (Cushion) PD Non-PD 90° Kneeling (Floor) PD Non-PD 120° Kneeling (Floor) PD Non-PD Bold values indicate statistically significant values (P < 0.05)
4 (6.9) 10 (16.1)
22 (37.9) 29 (46.8)
4 (6.9) 10 (16.1)
25 (43.1) 33 (53.2)
7 (12.1) 15 (24.2)
31 (53.4) 34 (54.8)
7 (12.1) 15 (24.2)
34 (58.6) 37 (59.6)
Good
32 (55.2) 23 (37.1)
29 (50.0) 19 (30.6)
20 (34.5) 13 (21.0)
17 (29.3) 10 (16.1)
P-Value 0.029
0.018
0.038
0.031
UKA, all the patients included in this study did not have severe PFJD. However, a prospective study by Suwanko- monkul et al. [12] showed that UKA patients with severe PFJD who received PD also had favourable outcomes in the short term. It is worth noting that Pongcharoen et al. [26] conducted a prospective comparison of patients with severe and non-severe PFJD treated with UKA, and the results showed that patients with severe PFJD had poorer knee joint scores. In addition, Hamilton et al. [27] reported up to 15 years of follow-up in UKA patients that severe PFJD may have a negative impact on descending stairs. Considering the unsustainable effectiveness of PD and the long-term adverse outcomes of severe PFJD, UKA in combination with PD is not recommended as the preferred option for patients with bicompartmental lesions.
PD patients showed significant improvement in kneel- ing ability after surgery. The results also further confirm the findings of the Artz et al. [7] study that there is a correlation between kneeling ability, pain and function. The causes of kneeling difficulties are multifaceted; in addition to pain and function, numbness, fear of injury to the prosthesis, comorbidities and third-party recom- mendations can all contribute to limitations in kneeling [28–31]. Many healthcare professionals advise patients not to kneel due to concerns about the safety of kneel- ing, but there is no clinical evidence of an association between kneeling and prosthesis loosening. This and other studies have found discrepancies between patients’ self-perceived ability to kneel and their actual ability to observe [32, 33]. Although some patients may perceive an inability to kneel, many patients actually have the poten- tial to kneel after appropriate rehabilitation training and instruction [28, 30, 34]. Therefore, while PD improves the ability to kneel by improving the patient’s function and reducing AKP, there is still a need for acquired help to build confidence and overcome kneeling difficulties
Page 6 of 8
Fig. 2 Matching self perception and actual ability of PD patients in four kneeling positions
through a comprehensive rehabilitation programme, individualized education and advice, and effective physiotherapy.
In this study, the application of PD in UKA was not found to increase the occurrence of related complica- tions. However, during electrocautery denervation in the peripatellar region, it may cause damage to the nutrient vessels around the patella, increasing the risk of patellar necrosis and fracture [10, 35]. Therefore, the operator should be familiar with the distribution of the peripatel- lar nerves and master the range and depth of electro- cautery. Before peripatellar electrocautery, we bent the metal tip of the electrotome tip by 3 mm. This not only prevents the patella from completely turning over, but also controls the depth of electrocautery and avoids damage caused by nerve regeneration and excessive electrocautery.
observation of long-term outcomes or complications. Second, this study transformed the five-categorical vari- ables into three-categorical variables in assessing the ability to kneel. Although the multicategorical transfor- mation can simplify model building and interpretation, it may cause limitations in the reliability and interpret- ability of the results. In addition, this study was evaluated only through four simple kneeling activities and lacked multidimensional tests of kneeling ability. Therefore, there is still a need for long-term prospective studies to validate PD in UKA and more comprehensive tests to assess kneeling ability in the future.
Conclusion Patellar denervation can safely and effectively improve patellofemoral joint function, pain and kneeling ability in the early postoperative period after UKA.
This study inevitably has several limitations. First, the short follow-up period of this study did not allow for
Page 7 of 8
Fig. 3 Matching self perception and actual ability of non-PD patients in four kneeling positions
Abbreviations UKA PD HSS VAS FJS-12 AKP PFJD ASA CCI
Unicompartmental Knee Arthroplasty Patellar Denervation Hospital for Special Surgery Visual Analogue Scale Forgotten Joint Score Anterior Knee Pain Patellofemoral Joint Degeneration American Society of Anesthesiologists Charlson Comorbidity Index
Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate This study has been approved by the Ethics Committee of Xuzhou Medical University Affiliated Hospital.
Author contributions SYJ and LN analyzed the data and wrote the manuscript. HL was responsible for data collection and collation. CXY provided supervision and guidance on the content of research topics. FS and WJT made significant contributions to the conception and design of the study, critically revised the manuscript and provided crucial theoretical support to help the team interpret the result. All authors read and approved the final manuscript.
Consent for publication All eligible participants signed an informed consent form prior to surgery.
Competing interests The authors declare no competing interests.
Received: 14 July 2024 / Accepted: 19 August 2024
Funding This study was funded by the Youth medical science and technology innovation project of Xuzhou Health Commission (XWKYHT20210577).
Page 8 of 8
References 1.
Jenny J-Y. Minimally invasive unicompartmental knee arthroplasty. Eur J Orthop Surg Traumatol. 2018;28:793–7. https://doi.org/10.1007/ s00590-017-2107-5.
2. Ge J, Hernigou P, Guo W, Zhang N, Liu C, Zhang Q. Minimally invasive small incision surgical technique for unicompartmental knee arthroplasty. Int Orthop. 2023;47:2717–25. https://doi.org/10.1007/s00264-023-05908-5. Al-Dadah O, Hing C. Unicompartmental knee arthroplasty: General trends of clinical practice. Knee. 2023;41:A1–2. https://doi.org/10.1016/j. knee.2023.03.013. Knifsund J, Niinimaki T, Nurmi H, Toom A, Keemu H, Laaksonen I, et al. Functional results of total-knee arthroplasty versus medial unicompart- mental arthroplasty: two-year results of a randomised, assessor-blinded multicentre trial. BMJ Open. 2021;11:e046731. https://doi.org/10.1136/ bmjopen-2020-046731. Barker KL, Hannink E, Pemberton S, Jenkins C. Knee arthroplasty patients predicted Versus actual recovery: what are their expectations about time of recovery after surgery and how long before they can do the tasks they want to do? Arch Phys Med Rehabil. 2018;99:2230–7. https://doi.org/10.1016/j. apmr.2018.03.022. Scott CEH, Bugler KE, Clement ND, MacDonald D, Howie CR, Biant LC. Patient expectations of arthroplasty of the hip and knee. J Bone Joint Surg Br. 2012;94:974–81. https://doi.org/10.1302/0301-620X.94B7.28219. Artz NJ, Hassaballa MA, Robinson JR, Newman JH, Porteous AJ, Murray JRD. Patient reported Kneeling ability in fixed and mobile bearing knee arthroplasty. J Arthroplasty. 2015;30:2159–63. https://doi.org/10.1016/j. arth.2015.06.063. 5.
3.
4.
20. Hawker GA, Mian S, Kendzerska T, French M. Measures of adult pain: visual Analog Scale for Pain (VAS Pain), Numeric Rating Scale for Pain (NRS Pain), McGill Pain Questionnaire (MPQ), short-form McGill Pain Questionnaire (SF-MPQ), Chronic Pain Grade Scale (CPGS), short Form-36 Bodily Pain Scale (SF-36 BPS), and measure of intermittent and constant Osteoarthritis Pain (ICOAP). Arthritis Care Res (Hoboken). 2011;63(Suppl 11):S240–252. https:// doi.org/10.1002/acr.20543.
21. Thienpont E, Opsomer G, Koninckx A, Houssiau F. Joint awareness in different types of knee arthroplasty evaluated with the Forgotten Joint score. J Arthro- plasty. 2014;29:48–51. https://doi.org/10.1016/j.arth.2013.04.024.
22. Behrend H, Giesinger K, Giesinger JM, Kuster MS. The forgotten joint as the ultimate goal in joint arthroplasty: validation of a new patient-reported out- come measure. J Arthroplasty. 2012;27:430–e4361. https://doi.org/10.1016/j. arth.2011.06.035.
23. Dawson J, Fitzpatrick R, Murray D, Carr A. Questionnaire on the perceptions of patients about total knee replacement. J Bone Joint Surg Br. 1998;80:63–9. https://doi.org/10.1302/0301-620x.80b1.7859.
24. Witoński D, Wagrowska-Danielewicz M. Distribution of substance-P nerve fibers in the knee joint in patients with anterior knee pain syndrome. A pre- liminary report. Knee Surg Sports Traumatol Arthrosc. 1999;7:177–83. https:// doi.org/10.1007/s001670050144.
25. Duan G, Liu C, Lin W, Shao J, Fu K, Niu Y, et al. Different factors Conduct Anterior knee Pain following primary total knee arthroplasty: a systematic review and Meta-analysis. J Arthroplasty. 2018;33:1962–e19713. https://doi. org/10.1016/j.arth.2017.12.024.
8. Weiss JM, Noble PC, Conditt MA, Kohl HW, Roberts S, Cook KF et al. What functional activities are important to patients with knee replacements? Clin Orthop Relat Res 2002:172–88. https://doi. org/10.1097/00003086-200211000-00030
9. Wojtys EM, Beaman DN, Glover RA, Janda D. Innervation of the human knee joint by substance-P fibers. Arthroscopy. 1990;6:254–63. https://doi. org/10.1016/0749-8063(90)90054-h.
26. Pongcharoen B, Reutiwarangkoon C. The comparison of anterior knee pain in severe and non severe arthritis of the lateral facet of the patella follow- ing a mobile bearing unicompartmental knee arthroplasty. Springerplus. 2016;5:202. https://doi.org/10.1186/s40064-016-1914-1.
27. Hamilton TW, Pandit HG, Maurer DG, Ostlere SJ, Jenkins C, Mellon SJ, et al. Anterior knee pain and evidence of osteoarthritis of the patellofemoral joint should not be considered contraindications to mobile-bearing unicompart- mental knee arthroplasty: a 15-year follow-up. Bone Joint J. 2017;99–B:632–9. https://doi.org/10.1302/0301-620X.99B5.BJJ-2016-0695.R2.
10. Maralcan G, Kuru I, Issi S, Esmer AF, Tekdemir I, Evcik D. The innervation of
28. Hassaballa MA, Porteous AJ, Newman JH. Observed kneeling ability after
patella: anatomical and clinical study. Surg Radiol Anat. 2005;27:331–5. https://doi.org/10.1007/s00276-005-0334-7.
11. Budhiparama NC, Hidayat H, Novito K, Utomo DN, Lumban-Gaol I, Nelissen RGHH. Does circumferential patellar denervation result in decreased knee Pain and Improved patient-reported outcomes in patients undergoing Non- resurfaced, simultaneous bilateral TKA? Clin Orthop Relat Res. 2020;478:2020– 33. https://doi.org/10.1097/CORR.0000000000001035.
12. Suwankomonkul P, Arirachakaran A, Kongtharvonskul J. Short-term improve- ment of patellofemoral pain in medial unicompartmental knee arthroplasty with patellar denervation: a prospective comparative study. Musculoskelet Surg. 2022;106:75–82. https://doi.org/10.1007/s12306-020-00675-7. 13. Alomran A. Effect of patellar denervation on mid-term results after non- resurfaced total knee arthroplasty. A randomised, controlled trial. Acta Orthop Belg. 2015;81:609–13.
total, unicompartmental and patellofemoral knee arthroplasty: perception versus reality. Knee Surg Sports Traumatol Arthrosc. 2004;12:136–9. https:// doi.org/10.1007/s00167-003-0376-5.
29. Fletcher D, Moore AJ, Blom AW, Wylde V. An exploratory study of the long- term impact of difficulty kneeling after total knee replacement. Disabil Rehabil. 2019;41:820–5. https://doi.org/10.1080/09638288.2017.1410860.
30. White L, Stockwell T, Hartnell N, Hennessy M, Mullan J. Factors prevent- ing kneeling in a group of pre-educated patients post total knee arthro- plasty. J Orthop Traumatol. 2016;17:333–8. https://doi.org/10.1007/ s10195-016-0411-1.
31. MacDonald B, Kurdin A, Somerville L, Ross D, MacDonald S, Lanting B. The effect of sensory deficit after total knee arthroplasty on patient satisfac- tion and kneeling ability. Arthroplast Today. 2021;7:264–e2672. https://doi. org/10.1016/j.artd.2020.11.020.
14. van Jonbergen H-PW, Reuver JM, Mutsaerts EL, Poolman RW. Determi- nants of anterior knee pain following total knee replacement: a systematic review. Knee Surg Sports Traumatol Arthrosc. 2014;22:478–99. https://doi. org/10.1007/s00167-012-2294-x.
15. Yuan M-C, Ding Z-C, Ling T-X, Zhou Z. Patellar denervation with Electrocau- tery reduces anterior knee Pain within 1 year after total knee arthroplasty: a Meta-analysis of Randomized controlled trials. Orthop Surg. 2021;13:14–27. https://doi.org/10.1111/os.12735.
16. Xie X, Pei F, Huang Z, Tan Z, Yang Z, Kang P. Does patellar denervation reduce post-operative anterior knee pain after total knee arthroplasty? Knee Surg Sports Traumatol Arthrosc. 2015;23:1808–15. https://doi.org/10.1007/ s00167-015-3566-z. Iwano T, Kurosawa H, Tokuyama H, Hoshikawa Y. Roentgenographic and clinical findings of patellofemoral osteoarthrosis. With special reference to its relationship to femorotibial osteoarthrosis and etiologic factors. Clin Orthop Relat Res 1990:190–7. 17.
32. Narkbunnam R, Rojjananukulpong K, Ruangsomboon P, Chareancholvanich K, Pornrattanamaneewong C. The association between perception of patients and their actual ability to do floor activities after mobile-bearing unicompartmental knee arthroplasty: a prospective, cross-sectional study. Jt Dis Relat Surg. 2023;34:245–52. https://doi.org/10.52312/jdrs.2023.877. 33. Schai PA, Gibbon AJ, Scott RD. Kneeling ability after total knee arthroplasty. Perception and reality. Clin Orthop Relat Res 1999:195–200. Jenkins C, Barker KL, Pandit H, Dodd CAF, Murray DW. After partial knee replacement, patients can kneel, but they need to be taught to do so: a single-blind randomized controlled trial. Phys Ther. 2008;88:1012–21. https:// doi.org/10.2522/ptj.20070374.
35. Russell RD, Huo MH, Jones RE. Avoiding patellar complications in total knee replacement. Bone Joint J. 2014;96–B:84–6. https://doi. org/10.1302/0301-620X.96B11.34305.
18. Kujala UM, Jaakkola LH, Koskinen SK, Taimela S, Hurme M, Nelimarkka O. Scoring of patellofemoral disorders. Arthroscopy. 1993;9:159–63. https://doi. org/10.1016/s0749-8063(05)80366-4.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
19. Ranawat CS, Insall J, Shine J. Duo-condylar knee arthroplasty: hospital for special surgery design. Clin Orthop Relat Res 1976:76–82. |
s13018-019-1155-4 | Yan et al. Journal of Orthopaedic Surgery and Research (2019) 14:111 https://doi.org/10.1186/s13018-019-1155-4
RESEARCH ARTICLE
Open Access
Influence of distal radius fractures involving the intermediate column on forearm rotation Bingshan Yan, Yanchao Chen and Wangping Yin*
Abstract
Background: The objective of the study was to compare the radiologic and clinical outcome of patients with distal radius fractures involving the intermediate column and distal radial metaphyseal fractures treated conservatively. Methods: Two cohorts of 52 matched patients with distal radius fractures treated conservatively, one with a fracture involving the intermediate column and the other with no intermediate column fracture, were retrospectively analyzed by examining the data. Patients were matched for age, sex, fracture side, and AO fracture type. The two groups were analyzed for differences in wrist motion; grip strength; Gartland and Werley score; Disabilities of the Arm, Shoulder and Hand (DASH) score; and visual analogue scale (VAS) score at 12 months. The differences in continuous variables were compared using the paired t test. Linear regression analyses or Pearson correlation analyses were used to evaluate the associations of radiological parameters with clinical outcomes. Results: The analysis showed significant differences in the range of motion (ROM) for pronation (p = 0.000) and supination (p = 0.008) in the paired groups. There was a significant difference in DASH scores (p = 0.024) in the paired groups. Using Pearson correlation analysis, negative correlations (r = − 0.360, p = 0.000) were observed between articular step-off and ROM for pronation. Linear regression analyses also indicated that ROM for pronation had negative relationships (β = − 6.327, p = 0.001) with articular step-off. Conclusions: Distal radius fractures involving the intermediate column had an adverse effect on forearm rotation after distal radius fractures treated conservatively.
Keywords: Distal radius fracture, Triangular fibrocartilage complex, A three-column theory, The intermediate column
Background Distal radius fractures are one of the most common injur- ies of all extremity fractures, with 41–50% involving ar- ticular surfaces of the distal radius end [1]. Currently, there are many classifications of distal radius fractures. Each fracture classification has its own advantages. Rikli and Regazzoni proposed a three-column biomechanical construction forming the distal radius and the distal ulna to guide surgical fixation [2]. In their theory, the inter- mediate column consists of the lunate fossa and the sig- moid notch, which is vital to transmit the load in the wrist. Anatomically, the sigmoid notch of the distal radius,
Correspondence: [email protected] Department of Orthopedic Surgery, Jinshan Hospital, Fudan University, Shanghai 201508, People’s Republic of China
where the triangular fibrocartilage complex (TFCC) ori- gins from, articulates with the convex ulnar head [3]. This means that any incongruency of the sigmoid notch and the ulnar head may lead to pain or dysfunction of the dis- tal radioulnar joint (DRUJ) [4].
Many studies have reported that distal radius fractures need anatomical reconstruction, stable fixation, and early function as any other intra-articular fractures [5], espe- cially in fractures of the intermediate column of the wrist. It is widely believed that there is a close correlation be- tween anatomical results and functional outcomes [2]. But, as far as we know, there is little direct evidence of the correlation between the fractures of intermediate column and the wrist outcome. Thus, we designed a 1:1 matched retrospective case-control study, which analyzed clinical
© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Yan et al. Journal of Orthopaedic Surgery and Research (2019) 14:111
data of 52 distal radius fractures involving the intermediate column (C1, C2) using the AO classification matched with 52 cases with distal radial metaphyseal fractures (A2, A3). All cases were treated conservatively. Our null hypothesis was that fractures of the intermediate column of the wrist would affect wrist function, especially forearm rotation.
Materials and methods This investigation got our hospital review board approval. Inclusion criteria were as follows: the patient with distal radius fractures classified as distal radius fractures involv- ing the intermediate column fracture (C1, C2) or not (A2, A3) and no history of wrist medical morbidities. Distal ra- dius fractures involving the intermediate column have three types: only the radiocarpal joint involved (rarely), in- volving both the radiocarpal joint and the radioulnar joint, or only the radioulnar joint involved. Exclusion criteria in- cluded patients above the age of 70years, distal radius fractures to return for follow-up evaluation, or skeletally immature patients. Open fractures or multiple fractures were also excluded. The study used a 1:1 matched case-control design. Distal radius fractures were classified according to the AO/ASIF system. Two groups of 104 patients were matched on the following factor: sex, age within 5-year groups, fracture side, and fracture type. The case group was distal radius fractures involving the intermediate column fracture (C1, C2), and the control group was extra-articular fracture of the distal radius (A2, A3). Type A2 matched type C1, as well as type A3 matched type C2. All patients were under- gone closed reduction and casting.
treated operatively,
refusal
Patients revisited for clinical assessments at 3 and 12 months. The primary parameter of clinical assessments for the injured and uninjured sides was active wrist and forearm ROM (extension, flexion, radial deviation, ulnar deviation, supination, and pronation). Active wrist and forearm ROMs were evaluated following a standard guideline [6]. An independent author performed every ROM measure three times. The mean of three measured values was adopted for the analysis. The data we re- corded was the percentage of the affected side compared to the contralateral side. Grip strength was measured and recorded in the same way.
X-rays were used to evaluate the alignment of the distal part of the radius on posteroanterior and lateral radio- graphs using the picture archiving and communication system. X-rays taken at immediate post-reduction and at the last follow-up were adopted to measure radiographic parameters [7]. Radiographic parameters included radial inclination, ulnar inclination, palmar tilt, and articular step-off. In this study, we defined the articular step-off as an independent factor and used it to distinguish whether there is a fracture involving the intermediate column. All X-rays were also measured three times by an attending
Page 2 of 5
orthopedic surgeon. A mean of measurements was used for the analysis. At the last follow-up, grip strength, Dis- abilities of the Arm, Shoulder and Hand (DASH) scores, Gartland and Werley scores, and VAS scores were used to evaluate wrist function [8].
Statistical analysis Continuous variables included age, ROM, articular step-off, grip strength, VAS, and DASH. The mean, standard devi- ation, and 95% confidence intervals of continuous variables were calculated. The differences in continuous variables were compared using the paired t test or Wilcoxon signed-rank test. Categorical variables were evaluated using the chi-squared test or Fisher exact test. Pearson correlation analysis was performed to evaluate the associations of ar- ticular step-off and ROM. Linear regression analyses were used to evaluate the associations of radiological parameters with clinical outcomes. p values less than 0.05 were consid- ered statistically significant. SPSS Statistical Software Pack- age for Windows (version 23.0) was used for the statistical analysis in a personal computer.
Results Between January 2014 and October 2017, 104 patients with a mean age of 52.5± 9.3 years (range, 25–66) satisfied the inclusion criteria. There were 42 males and 62 females. Eighteen patients had a type A2 fracture, 34 type A3, 18 type C1, and 34 type C2. Fifty-four patients were accom- panied by ulnar styloid fractures. The mean follow-up time was 15 months (range, 12 to 24months). Using a paired Student’s t test for continuous variables and the paired chi-squared test for categorical variables, there were no statistically significant differences between two matched cohorts on the basis of age, sex, fracture side, and ulnar styloid fracture (Table 1).
Grip strength, DASH, VAS, Gartland and Werley scores, radiographic parameters, and wrist and forearm active ROMs at the final follow-up for the two cohorts are summarized in Table 2. There were significant
Table 1 Descriptive data of the matched cohorts
Case group
Control group
p value
Sex
0.781
Male
21
21
Female
31
31
Age (years)
52.7 ± 9.3
52.4 ± 9.4
0.454
Fracture side
0.163
Left
24
24
Right
28
28
Ulnar styloid fracture
1.00
Yes
27
27
No
25
25
Yan et al. Journal of Orthopaedic Surgery and Research (2019) 14:111
Table 2 Radiographic and clinical outcomes in the matched cohorts
Parameters
Volar tilt (degrees)
Radial inclination (degrees)
Intra-articular step-off (mm)
Range of motion (% of the contralateral side)
Flexion
Extension
Pronation
Supination
Radial deviation
Ulnar deviation
Ulnar styloid fractures
Grip strength (% of the contralateral side)
DASH
VAS
Gartland and Werley scores
differences in ROM for pronation (p = 0.000) and supin- ation (p = 0.008) between the matched cohorts. We also observed a significant difference in DASH scores (p = 0.024) between the matched cohorts. No significant dif- ferences were found in any other variable. We found sta- (r = − 0.360, p = 0.000) tistically negative correlations between articular step-off and ROM for pronation. In a stepwise forward regression, the results also revealed only ROM for pronation had significant relationships (β = − 6.327, p = 0.001) with articular step-off. At 1 year, the mean value of intra-articular step-off was 2.0 mm in the case cohort. With respect to the volar angulation, radial inclination of the distal radial fracture, there were no significant differences between the matched cohorts at the final follow-up (Table 2).
Discussion The purpose of this study was to analyze the clinical and radiologic outcomes of distal radius fractures involving the intermediate column (C1, C2) and distal radius frac- tures (A1, A2) treated conservatively. The null hypothesis was that distal radius fractures involving the intermediate column have a significantly poorer radiologic and clinical outcome. The analysis of the data supported the hypoth- esis. In this study, the ROM for pronation and supination revealed significant differences between the matched co- horts. It also suggested that distal radius fractures involv- ing the intermediate column had a worse result in DASH scores. So far, there were few reports specialized on frac- tures of the intermediate column. This type of fracture was included in some other researches. From the view of the three-column theory, the die-punch fracture should be classified as the intermediate column fracture. In a study
Page 3 of 5
Case group
Control group
p
7.5 ± 4.9
7.4 ± 4.2
0.672
20.5 ± 3.6
20.1 ± 3.1
0.153
2.0 ± 0.9
73.1 ± 9.6
73.9 ± 8.9
0.450
73.8 ± 7.5
73.1 ± 8.5
0.586
77.0 ± 7.8
82.8 ± 7.3
0.000
72.0 ± 7.4
74.9 ± 8.7
0.008
81.9 ± 6.3
83.2 ± 7.1
0.328
78.6 ± 6.4
77.2 ± 4.9
0.244
27
27
1.000
86.2 ± 5.4
87.7 ± 4.8
0.175
13.0 ± 5.8
11.0 ± 4.2
0.024
1.0 ± 1.0
1.0 ± 1.00
1.000
3.71 ± 1.52
3.63±1.59
0.794
of investigating results of AO type C fractures of distal ra- dius treated by volar locking plate, LV et al. reported that die-punch fractures had poorer clinical outcomes at an early postoperative period [1].
As mentioned before, articular step-off represented a fracture involving the intermediate column in the study. We found statistically negative correlations be- tween articular step-off and ROM for pronation in the matched cohorts. This meant that the more the displacement of articular step-off, the worse the rota- tion function of the forearm. Our result is not contra- dictory to the previously published study: Swart et al. found after operatively treated distal radius fractures, supination improved more quickly, usually within the first 3 to 6 months [9].
There are many plausible explanations for these re- sults. In view of anatomy, the sigmoid notch of the ra- dius constitutes the intermediate column. The sigmoid notch which serves as an anchor for the TFCC probably plays a role in DRUJ stability [3]. It makes sense: when the displaced fracture fragment involving the sigmoid notch apparently changed the tension of the TFCC, there would produce a rotation dysfunction of the fore- arm. Moreover, the displaced fragment could cause ar- the DRUJ and jeopardize the ticular incongruity of rotation of the forearm [4, 10]. Ishikawa et al. reported that the malposition of the ulnar head might result in re- stricted pronation because of the change of the tension of the surrounding soft tissues [11]. Adams confirmed that displaced distal radial fractures could place im- mense stress on the TFCC, which may result in a de- creased pronation and/or supination [12]. Delclaux et al. found the patient with distal radius fracture malunion
Yan et al. Journal of Orthopaedic Surgery and Research (2019) 14:111
basically improved their wrist functions of pronation/su- pination after the corrective osteotomy [13].
Besides fracture morphology and fracture displace- ment, the presence of ulnar styloid fractures is supposed to be a susceptibility factor for DRUJ instability associ- ated with distal radius fractures [14]. Some authors have suggested that ulnar styloid fractures accompanied by distal radius fracture are associated with wrist outcomes and increase the risk of DRUJ instability [15, 16]. How- ever, more and more researches have supported that ulnar styloid fracture does not appear to influence on wrist function or outcome in patients with distal radial fractures [17–19]. Due to the ongoing debate in ulnar styloid fracture, we controlled this possible confounder between two groups. There was no significant difference (p = 1.00) in the ulnar styloid fracture.
Chung et al. demonstrated that after operatively treated distal radius fractures, age was significantly related to long-term outcomes. They showed the older patients were vulnerable to have poorer clinical outcomes [20]. Lar- ouche et al. reported that patients with distal radius frac- tures older than 55years did not gain a better outcome in the surgical group as compared with the conservative group [21]. Hohmann et al. supported that elderly patients with minor malunion of the distal radius had no adverse effect on patient-perceived outcomes regardless of whether these fractures were treated operatively or con- servatively [22]. The above findings indicated that age was vital to influence patient-perceived outcomes after the treatment of distal radius fractures. However, our results showed there was no statistical difference between two matched cohorts in age.
This study has several potential shortcomings. First, this is a retrospective study, which has its drawbacks such as selection bias and limited clinical data. Second, due to lack of a very good or excellent inter- and intra-observer repro- ducibility in the existing classifications of distal radius fractures [23], AO subclassifications based on X-ray and/ or CT scans may also have inconsistencies, leading to changes in our conclusion. Ma et al. described that X-ray results might be false negative in diagnosis of die-punch fracture. The X-ray-missed diagnosis rate was 11.1%, and the misdiagnosis rate was 15.6% [24]. Third, the issue of how to avoid variability in ROM measurement still cannot be completely solved. Hohmann et al. reported there was accurate, valid, and reliable use of a goniometer according to a standardized protocol advised by the American Soci- ety for Hand Therapists [22]. Therefore, we used the goni- ometer to measure wrist ROM following the standard guideline to reduce variability. And we supported that clinical examination comparing between the injured and uninjured contralateral sides was a reasonable way to re- duce individual differences in the assessment of objective indicators of wrist function after distal radius fractures.
Page 4 of 5
Conclusions The results prove that fractures of the intermediate col- umn play an important role in determining restricted forearm rotation after distal radius fractures treated con- servatively. The data also suggest that patients with dis- tal radius fractures involving the intermediate column have lower DASH scores. Based on these findings, we think fractures involving the intermediate column of the distal end deserve more attentions from the surgeon. However, there still have several risks of bias in this study. A further investigation is needed to better under- stand the complex relationship between anatomy, func- tion, and radiologic parameters in patients with fractures involving the intermediate column of the distal end.
Abbreviations DASH: Disabilities of the Arm, Shoulder and Hand; DRUJ: Distal radioulnar joint; ROM: Range of motion; TFCC: Triangular fibrocartilage complex; VAS: Visual analogue scale
Acknowledgements Not applicable.
Funding This study was supported by Jinshan Health Bureau, Shanghai, China (JSKJ- KTMS-2016-10).
Availability of data and materials The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.
Authors’ contributions BY wrote the manuscript. YC and WY did the measurements. BY carried out the statistical analysis. YC and WY supervised the study. All authors read and approved the final manuscript.
Ethics approval and consent to participate Ethics approval was given (Committee of Jinshan Hospital, Fudan University, Shanghai, P.R. China, reference no. 2017-09-01), and participants in this study gave written consent to participate.
Consent for publication Not applicable.
Competing interests The authors declare that they have no competing interests.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Received: 4 February 2019 Accepted: 12 April 2019
References 1.
LV Z, Zhao Y, Zhang YWQ, Ding W, Shen Y. Results of type C fractures of distal radius treated by volar locking plate (VLP), an observational comparative study. Int J Clin Exp Med. 2017;10(10):14721–9. Rikli DA, Regazzoni P. Fractures of the distal end of the radius treated by internal fixation and early function. A preliminary report of 20 cases. J Bone Joint Surg Br. 1996;78-B:588–92. Palmer AK, Werner F. The triangular fibrocartilage complex of the wrist- anatomy and function. J Hand Surg Am. 1981;6(2):153–62. Xing SG, Chen YR, Xie RG, Tang JB. In vivo contact characteristics of distal radioulnar joint with malunited distal radius during wrist motion. J Hand Surg Am. 2015;40(11):2243–8.
2.
3.
4.
Yan et al. Journal of Orthopaedic Surgery and Research (2019) 14:111
5.
6.
7.
8.
9.
Jakob M, Rikli DA, Regazzoni P. Fractures of the distal radius treated by internal fixation and early function. J Bone Joint Surg Br. 2000;82(3):340–4. Driessens S, Diserens-Chew T, Burton C, Lassig E, Hartley C, McPhail S. A retrospective cohort investigation of active range of motion within one week of open reduction and internal fixation of distal radius fractures. J Hand Ther. 2013;26(3):225–30 quiz 31. Zhang X, Hu C, Yu K, Bai J, Tian D, Xu Y, et al. Volar locking plate (VLP) versus non-locking plate (NLP) in the treatment of die-punch fractures of the distal radius, an observational study. Int J Surg. 2016;34:142–7. Kim JK, Yi JW, Jeon SH. The effect of acute distal radioulnar joint laxity on outcome after volar plate fixation of distal radius fractures. J Orthop Trauma. 2013;27(12):735–9. Swart E, Nellans K, Rosenwasser M. The effects of pain, supination, and grip strength on patient-rated disability after operatively treated distal radius fractures. J Hand Surg Am. 2012;37(5):957–62.
10. Rhee PC, Medoff RJ, Shin AY. Complex distal radius fractures: an anatomic algorithm for surgical management. J Am Acad Orthop Surg. 2017;25(2):77–88. Ishikawa J, Iwasaki N, Minami A. Influence of distal radioulnar joint subluxation on restricted forearm rotation after distal radius fracture. J Hand Surg Am. 2005;30(6):1178–84.
10. Rhee PC, Medoff RJ, Shin AY. Complex distal radius fractures: an anatomic algorithm for surgical management. J Am Acad Orthop Surg. 2017;25(2):77–88. Ishikawa J, Iwasaki N, Minami A. Influence of distal radioulnar joint subluxation on restricted forearm rotation after distal radius fracture. J Hand Surg Am. 2005;30(6):1178–84.
12. Adams BD. Effects of radial deformity on distal radioulnar joint mechanics. J Hand Surg Am. 1993;18(3):492–8.
13. Delclaux S, Trang Pham TT, Bonnevialle N, Apredoaei C, Rongieres M, Bonnevialle P, et al. Distal radius fracture malunion: importance of managing injuries of the distal radio-ulnar joint. Orthop Traumatol Surg Res. 2016;102(3):327–32. Fujitani R, Omokawa S, Akahane M, Iida A, Ono H, Tanaka Y. Predictors of distal radioulnar joint instability in distal radius fractures. J Hand Surg Am. 2011;36(12):1919–25. 14.
15. May MM, Lawton JN, Blazar PE. Ulnar styloid fractures associated with distal radius fractures: incidence and implications for distal radioulnar joint instability. J Hand Surg Am. 2002;27(6):965–71.
16. Ruch DS, Lumsden BC, Papadonikolakis A. Distal radius fractures: a comparison of tension band wiring versus ulnar outrigger external fixation for the management of distal radioulnar instability. J Hand Surg Am. 2005; 30(5):969–77. Kim JK, Koh YD, Do NH. Should an ulnar styloid fracture be fixed following volar plate fixation of a distal radial fracture? J Bone Joint Surg Am. 2010;92(1):1–6. Souer JS, Ring D, Matschke S, Audige L, Marent-Huber M, Jupiter JB, et al. Effect of an unrepaired fracture of the ulnar styloid base on outcome after plate-and-screw fixation of a distal radial fracture. J Bone Joint Surg Am. 2009;91(4):830–8. Kazemian GH, Bakhshi H, Lilley M, Emami Tehrani Moghaddam M, Omidian MM, Safdari F, et al. DRUJ instability after distal radius fracture: a comparison between cases with and without ulnar styloid fracture.
16. Ruch DS, Lumsden BC, Papadonikolakis A. Distal radius fractures: a comparison of tension band wiring versus ulnar outrigger external fixation for the management of distal radioulnar instability. J Hand Surg Am. 2005; 30(5):969–77. Kim JK, Koh YD, Do NH. Should an ulnar styloid fracture be fixed following volar plate fixation of a distal radial fracture? J Bone Joint Surg Am. 2010;92(1):1–6. Souer JS, Ring D, Matschke S, Audige L, Marent-Huber M, Jupiter JB, et al. Effect of an unrepaired fracture of the ulnar styloid base on outcome after plate-and-screw fixation of a distal radial fracture. J Bone Joint Surg Am. 2009;91(4):830–8. Kazemian GH, Bakhshi H, Lilley M, Emami Tehrani Moghaddam M, Omidian MM, Safdari F, et al. DRUJ instability after distal radius fracture: a comparison between cases with and without ulnar styloid fracture.
17.
18.
Int J Surg. 2011;9(8):648–51.
20. Chung KC, Kotsis SV, Kim HM. Predictors of functional outcomes after surgical treatment of distal radius fractures. J Hand Surg Am. 2007;32(1):76–83. Larouche J, Pike J, Slobogean GP, Guy P, Broekhuyse H, O'Brien P, et al. Determinants of functional outcome in distal radius fractures in high- functioning patients older than 55 years. J Orthop Trauma. 2016;30(8):445–9.
20. Chung KC, Kotsis SV, Kim HM. Predictors of functional outcomes after surgical treatment of distal radius fractures. J Hand Surg Am. 2007;32(1):76–83. Larouche J, Pike J, Slobogean GP, Guy P, Broekhuyse H, O'Brien P, et al. Determinants of functional outcome in distal radius fractures in high- functioning patients older than 55 years. J Orthop Trauma. 2016;30(8):445–9.
22. Hohmann E, Meta M, Navalgund V, Tetsworth K. The relationship between radiological alignment of united distal radius fractures and functional and patient-perceived outcomes in elderly patients. J Orthop Surg (Hong Kong). 2017;25(1):2309499016684976. Evans S, David M, Quraishi MK, Hanif UK, Sadique H, Machani B. The use of plain radiographs in the classification of distal radius fractures. J Orthop. 2014;11(3):142–4. 23.
24. Ma Y, Yin Q, Rui Y, Gu S, Yang Y. Image classification for die-punch fracture of intermediate column of the distal radius. Radiol Med. 2017; 122(12):928–33.
Page 5 of 5 |
s13018-025-05457-0 | Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49 https://doi.org/10.1186/s13018-025-05457-0
Journal of Orthopaedic Surgery and Research
CORRESPONDENCE
Open Access
Meta-analysis of the efficacy of lateral unicompartmental knee arthroplasty and total knee arthroplasty in the treatment of isolated lateral compartment knee osteoarthritis Yong-le Wang1, Yue Li1, Jingle Zhuge1 and Xi-yong Li1*
Abstract Objectives To systematically evaluate the efficacy of lateral unicompartmental knee arthroplasty (LUKA) and total knee arthroplasty (TKA) in the treatment of isolated lateral compartment knee osteoarthritis (LCKO), and to provide guidance and a basis for selecting surgery in clinical practice.
Methods Inclusion and exclusion criteria for literature were established, appropriate effect indicators were selected, and PubMed, Web of Science, Embase, Medline, Cochrane Library, and CNKI databases were searched using a com- puter. The Newcastle Ottawa scale (NOS) was used to evaluate the quality of the literature. After data extraction, Meta- analysis was performed using Revman5.4 software.
Results A total of 8 studies were included in this meta-analysis, involving 490 patients, with 204 in the TKA group and 286 in the LUKA group. Meta-analysis found that, there was no significant statistical difference in operation time between the LUKA group and the TKA group. Nevertheless, LUKA offers advantages such as reduced intraoperative blood loss and shorter hospitalization time. At the same time, the LUKA group also had advantages over the TKA group in postoperative visual analogue scale (VAS) score, knee joint range of motion, Oxford Knee Score (OKS)score and Hospital for Special Surgery (HSS) score.
Conclusion When LUKA and TKA are used to treat isolated LCKO, LUKA has the advantages of less intraoperative blood loss, shorter hospital stay, lower postoperative VAS score, and better knee function score and range of motion.
Keywords Osteoarthritis, Lateral compartment, Lateral unicompartmental knee arthroplasty, Total knee arthroplasty, Meta
Introduction Osteoarthritis (OA) is a joint disease that affects a large population globally [1]. Among them, knee OA is the most common type of OA, and lateral compartment
Correspondence: Xi-yong Li [email protected] 1 Department of Orthopaedics, Wenzhou TCM Hospital of Zhejiang Chinese Medical University, No. 9, Jiaowei Road, Liuhongqiao, Wenzhou 325000, People’s Republic of China
knee osteoarthritis (LCKO) is a specific type of knee OA characterized by degeneration of the lateral com- partment of the knee. A major cause of LCKO is val- gus alignment, which leads to abnormal loading on the lateral compartment, accelerating cartilage wear and degeneration. In cases of knee valgus, the pressure on the lateral compartment intensifies, whereas the medial compartment experiences a corresponding decrease in pressure. This imbalance in pressure distribution is one of the main reasons for lateral compartment degenera- tion [2]. The terminal treatment for knee OA typically
© The Author(s) 2025. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by- nc- nd/4. 0/.
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
involves joint replacement [3]. When the lesion in the knee is confined to a single compartment, performing total knee arthroplasty (TKA) may unnecessarily dam- age other healthy compartments [4]. Although TKA is currently the most common and effective treatment option for advanced OA, the choice of knee implant types is not entirely uniform, including posterior sta- bilized (PS) implants, cruciate retaining (CR) implants, bi-cruciate substituting (BCS) implants, and bicruciate- retaining (BCR) implants. However, studies have found no significant differences in the clinical function of these four types of implants during long-term follow- up [5]. Among them, BCR implants retain better post- operative proprioception because they preserve both the anterior and posterior cruciate ligaments, yet the majority of clinical practices still use implants that sac- rifice the anterior cruciate ligament. Unicompartmental knee arthroplasty (UKA) addresses unicompartmental issues while preserving the inherent ligament structure of the knee and retaining more bone mass, which is beneficial for postoperative revision [6]. Although UKA also has two types of implants, mobile bearing and fixed bearing, the study found no significant difference in postoperative outcome indicators between the two [7]. However, isolated lateral unicompartmental degen- eration accounts for only one-eighth of unilateral knee OA cases, and lateral unicompartmental knee arthro- plasty (LUKA) accounts for less than one-tenth of all UKA procedures [8]. Early studies pointed out that the unique anatomical and kinematic characteristics of the lateral compartment of the knee, coupled with inade- quate prosthesis design, increase the complexity of sur- gical procedures. Additionally, LUKA requires precise bone cutting to adjust the lower limb alignment, cor- rect valgus alignment, and optimize the position and angle of prosthesis implantation, thereby improving surgical success rates and patient satisfaction [9]. Com- pared to medial UKA, the clinical outcomes of LUKA are less satisfactory, with a higher complication rate. Therefore, this technique has relatively low acceptance among clinicians and patients, with many researchers opting for TKA to treat LCKO [10]. In recent years, advances in surgical techniques, prosthesis design, sur- gical indications, and rehabilitation methods have led to significant progress in the treatment of LCKO with LUKA [11]. For example, Giordano et al. found that even obese patients with a BMI > 30 kg/m2 still had good clinical outcomes after LUKA [12]. The purpose of this Meta-analysis is to compare the efficacy and safety of LUKA and TKA in the surgical treatment of LCKO, providing evidence-based medical evidence for the selection of surgical treatment options for LCKO.
Page 2 of 12
Materials and methods Search strategy
The current meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist [13]. PubMed, Web of Science, Embase, Medline, Cochrane Library, and CNKI databases were searched by computer. The search keywords were Total Knee Arthroplasty, Unicompart- mental Knee Arthroplasty, and Isolated Lateral Osteoar- thritis. The search strategy was (((Knee Arthroplasties) OR (Knee Replacement) or (Total Knee Replacement) OR (Total Knee Arthroplasty) OR (TKA)) AND ((Uni- compartmental Knee Arthroplasty) OR (UKA) OR (Uni- compartmental Knee Replacement) OR (Partial Knee Arthroplasty) OR (Partial Knee Replacement)) AND ((Lateral Osteoarthritis) OR (Isolated Lateral Osteoar- thritis))). The search time was from the establishment of the electronic database to November 2024.
Inclusion and exclusion criteria
Inclusion Criteria: (1) The study types included published clinical controlled studies with complete data and no lan- guage restrictions. (2) Included studies included patients with confirmed diagnosis of LCKO requiring surgical treatment. (3) The intervention measures were LUKA and TKA.
Exclusion criteria: (1) Systematic reviews, overviews, case reports or letters. (2) Articles with duplicate reports. (3) Patients who underwent medial unicompartmental knee arthroplasty. (4) Articles that diagnosed isolated LCKO and underwent high osteotomy or other surgeries. The outcome measures included (1) operation time, (2) intraoperative blood loss, (3) hospital stay, (4) postopera- tive visual analogue scale (VAS) score, (5) postoperative range of motion, (6) Oxford Knee Score (OKS), and (7) Hospital for Special Surgery (HSS) score.
Data extraction and quality evaluation
Two independent researchers extracted data separately, adhering to a strict standard protocol. When disagree- ments arose, they were resolved through discussion or jointly evaluated with a more senior researcher until a consensus was reached. The extracted information includes the first author, publication year, study coun- try, participant sex, number of cases and controls, and investigators’ conclusions. At the same time, the quality of the literature was evaluated according to the Newcas- tle–Ottawa-Scale (NOS). NOS adopted a semi-quantita- tive principle to quantitatively evaluate the selection of research subjects, comparability and results, with a full score of 9 points [14].
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Fig. 1 Flow diagram of literature searching
Statistical methods
Meta-analysis of data extracted from the included stud- ies were conducted using Review Manager 5.4 software. When the effect indicator was a binary variable, the odds ratio (OR) and 95% confidence interval (CI) were calcu- lated; when the effect indicator was a continuous variable, the mean difference (MD) and 95% CI were calculated. Heterogeneity was evaluated using chi-square-based Q and I2 values. P > 0.10 or I2 < 50% indicated no noteworthy heterogeneity among the included studies, which neces- sitated the use of a fixed-effect model. When significant heterogeneity was present, a random-effects model was employed. When the heterogeneity of the studies was high, sensitivity analysis was performed. P < 0.05 indi- cated that the difference was statistically significant.
Page 3 of 12
Results Basic characteristics of the included literature
According to the search strategy, a total of 2418 relevant articles were retrieved. According to the exclusion crite- ria, 2399 articles were excluded, and 19 relevant papers were retrieved from major databases via strict imple- mentation of the inclusion criteria. After full-text read- ing, 8 articles were finally included in this meta-analysis, including 490 patients, including 204 in the TKA group and 286 in the LUKA group. The baseline conditions of patients such as age and gender were compared in each included article, which were comparable (P > 0.05). The literature screening process and results are shown in Fig. 1, and the basic characteristics of the included litera- ture studies are shown in Table 1.
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Page 4 of 12
Table 1 Main characteristics and Quality score of literatures included
Study
Study design
Country
Group
Patients
Agea (years)
Gender (M/F)
Outcomes
NOS Scale
Gou 2024 [15]
Prospective
China
LUKA
30
65.41 ± 4.27
11/19
(1)(3)(4)
8
TKA
24
65.28 ± 4.16
8/16
Sheng 2024 [16]
Retrospective
China
LUKA
25
71.4 ± 9.8
7/18
(1)(2)(3)(4)(5) (7)
8
TKA
25
69.8 ± 10.5
7/18
Tang 2023 [17]
Prospective
China
LUKA
30
64.4 ± 6.7
12/18
(1)(2)(4)(5)(6)(7)
8
TKA
30
66.3 ± 6.2
10/20
Tu 2020 [18]
Retrospective
China
LUKA
121
70.2 ± 8.8
38/83
(5)(6)(7)
7
TKA
35
69.7 ± 6.9
9/26
Walker 2014 [19]
Retrospective
Germany
LUKA
20
NA
5/17
(5)(6)
7
TKA
20
NA
5/17
Wang 2020 [20]
Retrospective
China
LUKA
20
59.50 ± 6.13
9/11
(1)(2)(4)(5)(6)
8
TKA
19
62.74 ± 7.89
7/12
Yu 2024 [21]
Retrospective
China
LUKA
23
66.91 ± 5.82
9/14
(5)(6)(6)
8
TKA
23
69.30 ± 5.46
9/14
Zhou 2021 [22]
Retrospective
China
LUKA
15
69.13 ± 9.09
NA
(1)(3)(5)(6)
6
TKA
26
68.06 ± 8.07
NA
F: female, M:male, NA: not available, a: Mean ± SD, (1) operation time, (2) intraoperative blood loss, (3) hospital stay, (4) postoperative VAS score, (5) postoperative range of motion, (6) OKS score, and (7) HSS score
Fig. 2 Forest plot of surgical time for LUKA and TKA in the treatment of LCKO
Fig. 3 Forest plot of blood loss during LUKA and TKA in the treatment of LCKO
Quality evaluation of included literature
This study included 2 prospective studies and 6 retro- spective studies. The quality of the included literature was evaluated according to the NOS scoring scale. The NOS scores were divided into three levels: low, medium, and high quality, i.e., < 5 points, 5–7 points, and ≥ 8 points. The specific evaluation results are shown in Table 1.
Results of meta‑analyses Operation time
This study included 5 comparisons of the operation time between LUKA and TKA. A total of 244 patients were included, including 124 in the TKA group and 120 in the LUKA group. The heterogeneity test results were: I2 = 97%, P < 0.18, and the random effects model was
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Fig. 4 Forest plot of hospitalization time for LUKA and TKA in the treatment of LCKO
Fig. 5 Forest plot of VAS scores after LUKA and TKA in the treatment of LCKO
used for analysis. The results showed that the operation time of the two groups was similar, and there was no sta- tistically significant difference [95% CI (− 3.48, 18.56), P = 0.18]. The specific results are shown in Fig. 2.
Intraoperative blood loss
This study included 3 comparisons of intraoperative blood loss between LUKA and TKA. A total of 149 patients were included, including 74 in the TKA group and 75 in the LUKA group. The heterogeneity test results: I2 = 100%, P = 0.03, and the random effects model was used for analysis. The results showed that the dif- ference in intraoperative blood loss between LUKA and TKA was statistically significant [95% CI (8.83, 171.50), P = 0.03]. The results showed that compared with TKA, LUKA had less intraoperative blood loss. The specific results are shown in Fig. 3.
Hospital stay
This study included 3 comparisons of hospital stay between LUKA and TKA. A total of 145 patients were included, including 75 in the TKA group and 70 in
Page 5 of 12
the LUKA group. The heterogeneity test results were: I2 = 20%, P < 0.001, and the fixed effect model was used for analysis. The results showed that the difference in hospital stay between the two groups was statistically significant [95% CI (2.03, 2.89), P < 0.001]. The specific results are shown in Fig. 4.
Postoperative VAS score
The VAS scoring standard is the pain level scoring stand- ard, which uses the visual analog method to judge the severity of pain [23]. This study included a total of 4 com- parisons of VAS scores after TKA and LUKA. A total of 203 patients were included, including 98 in the TKA group and 105 in the LUKA group. The heterogeneity test results: I2 = 95%, P < 0.001, and the random effects model was used for analysis. The results showed that the VAS after LUKA was statistically different from that after TKA [95% CI (0.31, 1.74), P < 0.001]. Among them, the VAS score after LUKA was significantly lower one week after surgery, and the difference was statistically signifi- cant [95% CI (0.09, 1.97), P = 0.03], while at 1 month after surgery, there was no significant statistical difference
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Fig. 6 Forest plot of range of motion after LUKA and TKA in the treatment of LCKO
between the two sides [95% CI (− 0.41, 2.46), P = 0.16]. The specific results are shown in Fig. 5.
Postoperative range of motion
A total of 6 articles recorded the index of postopera- tive knee range of motion, with a total of 397 patients, including 161 in the TKA group and 236 in the LUKA group. The results showed that the overall heterogene- ity was 85%, and the random effects model was used. The results showed that the postoperative range of motion of LUKA was statistically different from that of TKA [95% CI (− 9.28, 4.65), P = 0.007]. Among them, the range of motion of the knee joint in the LUKA group was better than that in the TKA group, both at 1 month after sur- gery and 3 months after surgery, and the difference was statistically significant at 1 month after surgery [95% CI (− 9.96, − 8.02), P < 0.001] and 3 months after surgery [95% CI (− 8.52, − 1.38), P = 0.007]. The specific results are shown in Fig. 6.
OKS score
A total of 5 articles compared the OKS scores after TKA and LUKA, with a total of 347 patients, includ- ing 136 in the TKA group and 211 in the LUKA group. The final heterogeneity was 96%, so we also used a random effects model. The results showed that in the
Page 6 of 12
comparison of OKS scores after surgery, the LUKA group was significantly lower than the TKA group, and the difference was statistically significant [95% CI (1.67, 3.63), P < 0.001]. Subgroup analysis showed that the OKS scores at 1 month after surgery, 3 months after surgery, and 6 months after surgery were all lower in the LUKA group than in the TKA group, and the dif- ferences were statistically significant, [95% CI (4.61, 5.17), P < 0.001], [95% CI (4.27, 5.24), P < 0.001], [95% CI (3.05, 3.55), P < 0.001]. The specific results are shown in Fig. 7.
HSS score
A total of 4 articles involved the HSS score index after LUKA and TKA, with a total of 312 patients, includ- ing 113 in the TKA group and 199 in the LUKA group. Statistical analysis showed high heterogeneity (I2 = 89%, P = 0.003), and the random effects model was used for analysis. The results showed that the difference in HSS scores between the two groups was statistically sig- nificant [95% CI (− 6.53, − 1.30), P = 0.003]. The specific results are shown in Fig. 8.
Publication bias and sensitivity analyses
The RevMan 5.4 software was used to analyze the publi- cation bias and sensitivity of various effect indicators. It can be seen that the scatter points on both sides of the
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Fig. 7 Forest plot of OKS scores after LUKA and TKA in the treatment of LCKO
Fig. 8 Forest plot of HSS scores after LUKA and TKA in the treatment of LCKO
funnel plot are generally symmetrical, indicating no sig- nificant publication bias and relatively stable and reli- able data. However, due to the small number of included literature, the bias results of individual effect indicators may have certain errors. The specific results are shown in Figs. 9, 10.
Discussion This is the first meta-analysis comparing LUKA and TKA for the treatment of LCKO. The results of this study indi- cate that the LUKA group has advantages over the TKA
Page 7 of 12
group in terms of less intraoperative blood loss, higher postoperative functional scores, and greater range of motion when treating LCKO. There is no statistically sig- nificant difference between the LUKA and TKA groups in terms of surgical duration (Fig. 11).
Knee osteoarthritis (KOA) is a common degenerative disease of the knee joint that affects joint function and quality of life. The majority of knee joint degeneration primarily involves a single compartment [24]. Isolated LCKO accounts for approximately 10% of all KOA cases [25]. Isolated LKCO is often associated with lateral
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Fig. 9 Bias risk diagram
condylar hypoplasia, lateral tibial bone loss, and valgus remodeling of the femoral and tibial metaphyses [26]. Additionally, there may be contraction of the lateral collateral ligament, lateral capsule, and iliotibial band, which serve to protect and stabilize the joint. Contrac- tion of these lateral structures can limit the range of motion of the knee joint, leading to increased intra- articular pressure. Simultaneously, the contracted ilioti- bial band can also exert abnormal pressure on the knee joint, resulting in postoperative knee instability [27]. Valgus alignment of the knee joint is a crucial factor in the development of LCKO as it increases the load on the lateral compartment, leading to accelerated carti- lage wear in the lateral compartment and subsequently progressing to LCKO. The etiology of valgus alignment can be attributed to various factors, including genetics, disease, and lifestyle habits. Therefore, when treating LCKO, it is essential to fully consider the impact of val- gus alignment and select appropriate surgical strategies to restore normal alignment and function of the knee joint [2].
In the terminal stage of KOA, knee replacement sur- gery is typically employed. TKA, for instance, is widely used to treat KOA due to its proven efficacy and mature technology. However, studies have found that approxi- mately 10% of patients undergoing TKA develop valgus deformity [28]. LUKA, a surgical procedure for treating arthritis confined to the lateral compartment of the knee joint, not only preserves bone and ligaments but also reli- ably restores normal knee kinematics and function [29]. Appropriate osteotomy plays a crucial role in both LUKA and TKA surgeries. TKA involves the replacement of the entire knee joint, necessitating a more comprehensive
Page 8 of 12
osteotomy. Besides the lateral compartment, osteotomies are also required for the medial compartment and patel- lofemoral joint to ensure proper prosthesis installation and knee stability [30]. Unlike TKA, LUKA only involves the replacement of the lateral compartment, requiring extremely high precision in osteotomy to ensure correct prosthesis placement and knee stability. More impor- tantly, osteotomy can precisely adjust the valgus align- ment of the knee joint, ensuring a tight fit between the newly implanted prosthesis and surrounding bone tissue, thereby restoring the normal biomechanical characteris- tics of the knee joint [9]. Compared with TKA, LUKA has the advantages of being less invasive, faster in recovery, and earlier in resumption of daily activities and sports31. Although Migliorini et al. found that the revision rate and complications after UKA surgery were generally low [32], there are also risks of increased medial condylar wear and prosthesis dislocation after LUKA [33]. Therefore, the aim of this study is to compare the efficacy and safety of LUKA and TKA in the treatment of isolated LCKO.
This Meta-analysis included 8 studies and found no statistically significant difference in surgical duration between the LUKA and TKA groups. This finding is not entirely consistent with some previous studies. Ma et al. suggested that UKA has significant advantages over TKA in reducing surgical incision size and surgical duration [34]. The lack of a significant difference in surgical dura- tion between the two groups in this study may be related to the surgeons’ lack of familiarity with the LUKA pro- cedure. Due to its anatomical and biomechanical char- acteristics, LUKA demands higher technical proficiency [9]. However, in terms of intraoperative blood loss and hospital stay, the LUKA group was found to be superior
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Fig. 10 Summary of risk of bias
to the TKA group, aligning with the findings of Ma et al. [34]. The surgical approach of LUKA, which involves only lateral osteotomy, results in less intraoperative trauma. This less invasive surgery not only reduces intraoperative blood loss and surgical duration but also lowers the risk of postoperative infection and other complications [35].
Page 9 of 12
Many scholars posit that early LUKA procedures, con- strained by surgical techniques and prosthesis design, led to poor postoperative outcomes and higher complication rates. Consequently, TKA was often chosen to address issues in the lateral compartment of the knee joint [36]. However, with advancements in prosthesis and surgical techniques in recent years, studies have found that LUKA can achieve better clinical outcomes, such as range of motion, in appropriate patients compared to TKA [35]. Regarding pain scores, Hannah A. Wilson et al. found no significant difference between the UKA and TKA groups [37], whereas this study found that the LUKA group had lower postoperative VAS scores compared to the TKA group. Pongcharoen et al. found no significant differ- ence in the OKS between the UKA and TKA groups [38], which contrasts with our study results. Our study found that the LUKA group had higher postoperative OKS and HSS scores compared to the TKA group. This may be attributed to the smaller amount of bone resection and intraoperative trauma in LUKA, which accelerates post- operative healing and preserves more of the patient’s pro- prioception due to less disruption of the knee’s original structure. Studies have shown that UKA, by preserving the anterior cruciate ligament and thus maintaining nor- mal knee structure, results in better postoperative range of motion compared to TKA [39]. David et al. also found that the UKA group had better range of motion com- pared to the TKA group [31]. This aligns with our study findings, where we observed that the LUKA group had better knee range of motion compared to the TKA group.
Limitations There are some potential limitations in this Meta-anal- ysis: (1) Most of the included studies are retrospective case–control studies, with only 2 prospective studies and a lack of randomized controlled trials, which may impact the results; (2) Some effect indicators are based on a limited number of studies, resulting in insufficient evidence strength; (3) The analysis results for some effect indicators exhibit considerable heterogeneity. In future research, it is necessary to include more comprehensive effect indicators (such as hospitalization costs, postop- erative inflammatory markers, and postoperative imaging indicators), larger sample sizes, studies from more coun- tries, and more randomized controlled trials to provide stronger evidence to support these findings.
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Page 10 of 12
Fig. 11 Funnel plot of publication bias. a Operation time, b intraoperative blood loss, c hospital stay, d postoperative VAS score, e postoperative range of motion, f OKS score, and g HSS score
Conclusion In conclusion, LUKA demonstrates advantages over TKA in terms of intraoperative blood loss, hospital stay, post- operative VAS score, postoperative functional score, and postoperative range of motion in the treatment of LCKO.
the raw data. XYL and JLZG were responsible for the acquisition and sorting of data. XYL performed the interpretation of the data. All authors have read and approved the final manuscript.
Funding Not applicable.
Acknowledgements Not applicable.
Availability of data and materials No datasets were generated or analysed during the current study.
Author contributions YL and YLW were responsible for the design of the current study, and both performed the statistical analysis. YL and YLW confirmed the authenticity of all
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
Declarations
Ethics approval and consent to participate Not applicable.
Consent to participate Informed consent was obtained from all individual participants included in the study.
Consent for publication Not applicable.
Competing interests The authors declare no competing interests.
Received: 20 November 2024 Accepted: 3 January 2025
References 1. Macías-Hernández SI, Morones-Alba JD, Miranda-Duarte A, Coronado- Zarco R, de LA Soria-Bastida M, Nava-Bringas T, et al. Glenohumeral osteoarthritis: overview, therapy, and rehabilitation. Disabil Rehabil. 2017;39:1674–82.
2. Andriacchi TP. Valgus alignment and lateral compartment knee osteoar- thritis: a biomechanical paradox or new insight into knee osteoarthritis? Arth Rheum. 2013;65:310–3.
3. Abramoff B, Caldera FE. Osteoarthritis: pathology, diagnosis, and treat- ment options. Med Clin North Am. 2020;104:293–311.
4. Bisogno MR, Scuderi GR. Management of extensor mechanism disruption after total knee arthroplasty. Orthop Clin North Am. 2022;53:277–86. Sun K, Wu Y, Wu L, Shen B. Comparison of clinical outcomes among total knee arthroplasties using posterior-stabilized, cruciate-retaining, bi-cruciate substituting, bi-cruciate retaining designs: a systematic review and network meta-analysis. Chin Med J (Engl). 2023;136:1817–31. Smith E, Lee D, Masonis J, Melvin JS. Lateral unicompartmental knee arthroplasty. JBJS Rev. 2020;8:e0044. 6.
5.
7. Migliorini F, Maffulli N, Cuozzo F, Elsner K, Hildebrand F, Eschweiler J, et al. Mobile bearing versus fixed bearing for unicompartmental arthroplasty in monocompartmental osteoarthritis of the knee: a meta-analysis. J Clin Med. 2022;11:2837. Scott CEH, Nutton RW, Biant LC. Lateral compartment osteoarthritis of the knee: biomechanics and surgical management of end-stage disease. Bone Jt J. 2013;95:436–44.
8.
9. Ollivier M, Abdel MP, Parratte S, Argenson J-N. Lateral unicondylar knee arthroplasty (UKA): contemporary indications, surgical technique, and results. Int Orthop. 2014;38:449–55.
10. Walker T, Aldinger PR, Streit MR, Gotterbarm T. Lateral unicompartmental knee arthroplasty: a challenge. Oper Orthop Traumatol. 2017;29:17–30.
11. Wen T, Xue H, Ma T, Yang T, Xue L, Tu Y. A standardized technique for lateral unicompartmental knee arthroplasty. J Orthop Surg (Hong Kong). 2023;31:10225536231164028.
12. Giordano L, Maffulli N, Morenghi E, Quaglia A, Prospero E, Rosa F, et al. A BMI above 30 results in satisfying outcomes in patients undergoing fixed- bearing lateral unicompartmental knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. 2023;31:1106–12.
13. Page MJ, Moher D, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. PRISMA 2020 explanation and elaboration: updated guidance and exemplars for reporting systematic reviews. BMJ. 2021;372:n160.
14. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assess- ment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol. 2010;25:603–5.
15. Gou X, Mao Y, Wang G. Clinical efficiency of fixed-bearing unicom- partmental knee arthroplasty versus total knee arthroplasty for lateral compartment knee osteoarthritis and the effect on recovery of motor function. Altern Ther Health Med. 2024;30:114–21.
Page 11 of 12
16. Dong S, Qiong S, Xinwen B, et al. Comparison of the efficacy of two replacement surgeries in the treatment of lateral compartment osteoar- thritis of the knee. J Clin Orthop. 2024;27(05):667–71.
17. Yueping T, Jianfeng G, Jin Hu. Comparison of early efficacy of uni- compartmental knee replacement and total knee replacement in the treatment of lateral compartment knee osteoarthritis. Chin J Tradit Chin Med Traumatol. 2023;31(05):40–4. https:// doi. org/ 10. 20085/j. cnki. issn1 005- 0205. 230508.
18. Tu Y, Ma T, Wen T, Yang T, Xue L, Xue H. Does unicompartmental knee replacement offer improved clinical advantages over total knee replace- ment in the treatment of isolated lateral osteoarthritis? A matched cohort analysis from an independent center. J Arthroplasty. 2020;35:2016–21.
19. Walker T, Gotterbarm T, Bruckner T, Merle C, Streit MR. Total versus unicompartmental knee replacement for isolated lateral osteoarthritis: a matched-pairs study. Int Orthop. 2014;38:2259–64.
20. Wang Feng Ju, Xiaocong WB, et al. Lateral unicompartmental knee replacement and total knee arthroplasty for the treatment of lat- eral unicompartmental knee osteoarthritis. Chin J Tissue Eng Res. 2021;25(12):1836–41.
21. Yu H, Meng J, Xu Y, Xu K, Lin C, Bao J. Comparison of early curative effects of lateral unicompartmental knee arthroplasty with total knee arthro- plasty and the legacy constrained condylar knee arthroplasty in the treatment of isolated lateral osteoarthritis: a single-center retrospective study. J Orthop Surg Res. 2024;19:597.
22. Zhou Xiaoqiang Yu, Xiao SC, et al. Comparison of early efficacy of fixed- platform unicompartmental knee replacement and total knee replace- ment in the treatment of lateral compartment knee osteoarthritis. J Pract Orthop. 2021;27(10):918–23. https:// doi. org/ 10. 13795/j. cnki. sgkz. 2021. 10. 012.
23. Karcioglu O, Topacoglu H, Dikme O, Dikme O. A systematic review of the pain scales in adults: which to use? Am J Emerg Med. 2018;36:707–14. 24. Vincent TL, Alliston T, Kapoor M, Loeser RF, Troeberg L, Little CB. Osteo- arthritis pathophysiology: therapeutic target discovery may require a multifaceted approach. Clin Geriatr Med. 2022;38:193–219.
25. Plancher KD, Briggs KK, Chinnakkannu K, Dotterweich KA, Commaroto SA, Wang KH, et al. Isolated lateral tibiofemoral compartment osteoarthritis: survivorship and patient acceptable symptom state after lateral fixed- bearing unicompartmental knee arthroplasty at mean 10-year follow-up. J Bone Jt Surg Am. 2022;104:1621–8.
26. Lombardi AV, Dodds KL, Berend KR, Mallory TH, Adams JB. An algorithmic approach to total knee arthroplasty in the valgus knee. J Bone Jt Surg Am. 2004;86(2):62–71.
27. Xie K, Lyons ST. Soft tissue releases in total knee arthroplasty for valgus deformities. J Arthroplasty. 2017;32:1814–8.
28. Alesi D, Meena A, Fratini S, Rinaldi VG, Cammisa E, Lullini G, et al. Total knee arthroplasty in valgus knee deformity: is it still a challenge in 2021? Musculoskelet Surg. 2021;106:1.
29. Johal S, Nakano N, Baxter M, Hujazi I, Pandit H, Khanduja V. Unicompart- mental knee arthroplasty: the past, current controversies, and future perspectives. J Knee Surg. 2018;31:992–8.
30. Rossi R, Cottino U, Bruzzone M, Dettoni F, Bonasia DE, Rosso F. Total knee arthroplasty in the varus knee: tips and tricks. Int Orthop. 2019;43:151–8. 31. Dalury DF, Fisher DA, Adams MJ, Gonzales RA. Unicompartmental knee arthroplasty compares favorably to total knee arthroplasty in the same patient. Orthopedics. 2009;32:253.
32. Migliorini F, Driessen A, Oliva F, Maffulli GD, Tingart M, Maffulli N. Better outcomes and reduced failures for arthroplasty over osteotomy for advanced compartmental knee osteoarthritis in patients older than 50 years. J Orthop Surg Res. 2020;15:545.
33. Fiocchi A, Condello V, Madonna V, Bonomo M, Zorzi C. Medial vs lateral unicompartmental knee arthroplasty: clinical results. Acta Bio Medica : Atenei Parmensis. 2017;88:38.
34. Ma J, Zhang L, Wang C, Xu K, Ren Z, Wang T, et al. The mid-term out- comes of mobile bearing unicompartmental knee arthroplasty versus total knee arthroplasty in the same patient. Front Surg. 2023;10:1033830.
35. van der List JP, Chawla H, Zuiderbaan HA, Pearle AD. Patients with isolated lateral osteoarthritis: unicompartmental or total knee arthroplasty? Knee. 2016;23:968–74.
36. Baumann F, Bahadin Ö, Krutsch W, Zellner J, Nerlich M, Angele P, et al. Proprioception after bicruciate-retaining total knee arthroplasty is
Wang et al. Journal of Orthopaedic Surgery and Research (2025) 20:49
comparable to unicompartmental knee arthroplasty. Knee Surg Sports Traumatol Arthrosc. 2017;25:1697–704.
37. Patient relevant outcomes of unicompartmental versus total knee replacement: systematic review and meta-analysis - PMC [Internet]. [cited 2024 Dec 4]. Available from: https:// pmc. ncbi. nlm. nih. gov/ artic les/ PMC63 83371/
38. Pongcharoen B, Timjang J. The outcomes of mobile bearing unicompart- mental knee arthroplasty and total knee arthroplasty on anteromedial osteoarthritis of the knee in the same patient. Arch Orthop Trauma Surg. 2020;140:1783–90.
39. Lombardi AV, Berend KR, Walter CA, Aziz-Jacobo J, Cheney NA. Is recovery faster for mobile-bearing unicompartmental than total knee arthro- plasty? Clin Orthop Relat Res. 2009;467:1450–7.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations.
Page 12 of 12 |
s13018-021-02602-3 | Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468 https://doi.org/10.1186/s13018-021-02602-3
RESEARCH ARTICLE
Open Access
Proprioceptive elbow training reduces pain and improves function in painful lateral epicondylitis—a prospective trial
B. Schiffke-Juhász1*, K. Knobloch2, P. M. Vogt3 and L. Hoy4
Abstract
Background: In painful epicondylitis, previous studies reported deficiencies in elbow proprioception. In line, proprioceptive training of the lower limb has been reported substantial beneficial in a number of indications. Therefore, we have asked if a specified proprioceptive training using training devices that are capable of activating the deep musculature in the upper limb is able to reduce the symptoms of epicondylitis.
Materials and methods: We included 71 patients with painful lateral epicondylitis > 3 months. Interventions: Group A: Proprioceptive training intervention with a Flexibar® (9 min daily for 12 weeks). Group B: at least 40 min running or walking/week with the XCO® in addition to the proprioceptive training with the Flexibar® (9 min daily for 12 weeks), follow-up for 12 weeks. Primary end point: Pain on visual analogue scale (VAS, 0-10); secondary end points: DASH-Score (0 = very good, 100 = very poor), grip strength according to Jamar dynamometer (kg), vibration sensation measured with a 128 Hz tuning fork. Results: The pain on VAS in group A was reduced significantly. 3.6 ± 2.0 to 2.4 ± 2.1 (−33%, p = 0.013), and from 3.7 ± 2.4 to 2.2 ± 1.9 (−41%, p = 0.004) in group B after 12 weeks. There was no significant difference between A and B (p = 0.899). In both groups, there was a significant improvement of the DASH-Score (A: 32 ± 15 to 14 ± 12, −56%, p < 0.001; B: 27 ± 12 to 12 ± 11, −55%, p = 0.001) without any difference between groups A and B (p = 0.339). Grip strength improvement in group A from 24 ± 12 to 33 ± 11 kg (+38%, p < 0.001), and from 29 ± 14 to 34 ± 11 kg (+15%, p < 0.001) in group B. In line, vibration sensation improved in both groups (A: 6.3 ± 0.6 to 6.5 ± 0.5, p = 0.0001; B: 6.3 ± 0.7 to 6.6 ± 0.5, p = 0.003). Conclusion: A 12-week proprioceptive training with the Flexibar® improves pain, quality of life, grip strength and vibration sensation in patients with painful lateral epicondylitis.
Level of evidence: Ib, randomised clinical trial Trial registration: German Clinical Trials Register, DRKS00024857, registered on 25 March 2021—retrospectively registered, http://apps.who.int/trialsearch/
Keywords: Epicondylitis humeri radialis, Tennis elbow, Proprioception, Proprioceptive exercise
Correspondence: [email protected] 1Herzogin Elisabeth Hospital, Braunschweig, Germany Full list of author information is available at the end of the article
© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 2 of 13
Background The epicondylitis humeri radialis, often simply re- ferred to as tennis elbow, appears with an incidence of 1-3% in the population [1, 2], and occurs most fre- quently between the third and fifth decade of life [1]. Amongst women between the 40th and 50th year of life the disease may occur with a prevalence of 10% [2]. Thereby, the epicondylitis humeri radialis is not only a very widespread disease but also affects more often the working population. Even though the aeti- ology of the disease has still not been fully identified and it seems to be determined multifactorially, a number of studies have shown that there are various risk factors, for example, repetitive movements, force- ful work or activities which require an unnatural pos- ture of hands and arms [3, 4]. Such stresses cannot only be found amongst active tennis players but also within a lot of professional activities, so that only a small part of the patients with lateral epicondylitis ex- ercises the name-giving sport at all [1]. Another risk factor relating to a low social support at work could be identified amongst female patients [3]. Further- more, there seems to be a relation with a current or previous nicotine abuse [3, 4]. further studies were able to demonstrate reduced deficiencies in proprioception in patients with symptomatic lateral epicondylitis [5].
In addition,
in 1936, a the symptoms after spontaneous consolidation of about 8-12 months on the condition of functional physical rest [1], most patients expect a quicker ther- apeutical intervention, so that the question concern- ing an appropriate therapy is further relevant. At this point, a very wide spectrum of therapeutical options is available for therapists, which were repetitively compared against each other within various studies. it could be shown that there is no In this context, universal solution for the therapy of the lateral epi- from various condylitis, but that patients benefit therapeutic approaches. Whereas surgical treatment only play a role in cases of failure of the conservative therapeutic options [1, 6], there are, besides rarely ap- plied conservative therapeutic options, often treat- ments using orthoses, injections and physiotherapy in the foreground. In case of a treatment using orthoses, it can be chosen from a lot of different types of orth- oses, whereby a number of studies could not prove relevant differences in the outcome of the different types of orthoses [7], though Garg et al. could indeed find a slight advantage in the pain reduction of using wrist extension splints rather than using usual epicon- dylitis braces [8]. However, the pain relief is here tied to wearing the orthosis, whereby the affected muscles functional only
Even though Cyriax already described,
get
a
relief but no sustainable
optimisation. Another frequently used therapeutic op- tion is a local injection, most common with a medical preparation of cortisone (rarer used are PRP, Botox, or others). In doing so, it is essential to take into ac- count not only the therapeutic effect but also the risk potential due to systemic complications (for example multifocal osteonecrosis or other well-known systemic side effects) and the possibility of local complications (for example tendon or fascial ruptures) [9]. As dem- onstrated in various studies, the therapeutic effect of injection treatment seems to be only superior in the short-term follow-up to the physiotherapeutic options [10]. With regard to the use of PRP, Ang Li et al. found inferiority to corticosteroids after 4-8 weeks, but superiority of PRP in the long term (after 24 weeks) [11]. Ruiz et al. were also able to show im- provements in symptoms for the use of botulinum toxin, but accompanied by a certain weakness of the 3rd finger, without other side effects [12]. Another study by Newcomer et al. ultimately identified re- habilitation as the first-line therapy for short-lasting symptoms [13]. However, it should be noted that the repertoire of physiotherapeutic options with over 40 different methods [6] is extremely wide spread, so that even at this point a calculated decision regarding a suitable therapeutic option becomes necessary. As deficiencies in proprioception can play a role in dif- ferent tendon diseases which was explicitly proven by Juul-Kristensen et al. for patients with lateral epicon- dylitis compared to healthy control subjects [5], we aspired to design a training concept to improve the proprioceptive capabilities regarding the affected area. In this context, we have chosen to examine not only the training with one or two appropriate training de- vices to run a prospective, randomised clinical trial with regard to the improvement of the proprioceptive capabilities but also the clinical development of the symptoms.
Hypothesis The combination of a proprioceptive and ballistic train- ing is effective for patients with lateral epicondylitis re- garding pain reduction and functional improvement.
Materials and methods Agreement of Hochschule Hannover, Germany.
the
ethics
committee: Medizinische
Inclusion criteria Patients suffering epicondylitis humeri radialis for at least 3 months, verified by clinical examination with ex- ercise related pain, at the age of 18 to 65 years. There- fore used diagnosis criteria were pain on palpation of
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 3 of 13
the lateral epicondyle as well as the declaration of pain during provocation tests.
Exclusion criteria Epicondylitis humeri ulnaris, missing consent, beginning of different therapeutic treatments during the study phase.
Intervention The intervention in ‘Group A’ consisted in an at least 12-week training with the Flexibar®. The Flexibar® is a swing bar with rubber weights at the ends. The device is actively vibrated. The alternating vibration causes an additional reflexive tensioning of the deep muscles. We requested a daily training of at least 9 min, whereby the exercises shown in Fig. 2 ought to be performed in each case for 30 s per limb.
Patient characteristics Figure 1 illustrates the course of the study in form of a CONSORT flow chart. This includes an evaluation of 108 patients to verify clinically the existence of a lateral epicondylitis by medical history as well as performing significant clinical tests (pain on palpation of the lateral epicondyle, painful extension of fingers against resist- ance). Seventy-one patients with lateral epicondylitis were subsequently randomised by lottery into two groups. The lottery has taken place by drawing an opaque lot with the designation ‘Group A’ or ‘Group B’. The patients’ characteristics of both groups are shown in an overview in Tables 1 and 2.
Test subjects in ‘Group B’ were requested to do the same training programme, but received an additional other training device—the XCO® Walking & Running. The XCO® Walking & Running is a metal tube filled with a granulate (Fig. 3). The granulate is set in mo- tion by the swing of the arm whilst walking and each time the granulate hits the ends of the tube, a so- called ‘reactive impact’ occurs, which in turn should lead to a reflexive, stabilising tensioning of the deep muscles. With that training device, the patients had to complete 2 further training sessions per week. Each training session should include at least 20 min of walking or running with the training device. Thereby, the XCO®-tubes should be held one in each
Fig. 1 Flow chart of the study protocol
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Table 1 General characteristics
Group A (Flexibar®)
Basic characteristics
Gender
m = 15; f = 24
Age [years]
47 ± 8
Weight [kg]
72 ± 15
Height [cm]
173 ± 8
BMI
24.2 ± 4.3
Underlying diseases
Nicotine abuse
15%
Diabetes
3%
Hypercholesterolemia
8%
Hypertension
8%
Heart disease
0%
Intake of acetylsalicylic acid
5%
Intake of cortisone (spray)
3%
Diseases of the elbow
Family disposition
neg. = 29; pos. = 10
Pain of right elbow
74%
Pain of left elbow
5%
Bilateral pain of elbow
21%
Morning stiffness
33%
Swelling
15%
Pressure pain
90%
Duration of pain
> 27 weeks
Pain in the morning [VAS]
2.7 ± 2.2
Daily maximum of pain [VAS]
5.2 ± 1.8
Intake of antibiotics (Ciprobay, Tavanic)
3%
Previous therapies
Massage
51%
Cross friction
38%
Heat application
18%
Cold application
44%
Shock wave
15%
Eccentric training
8%
Sclerotherapy
0%
Injection of corticosteroids
56%
Injection of Traumeel
10%
Surgical treatment
3%
Bandage
79%
Taping
10%
Kinesiology taping
10%
Collateral tendon diseases
Pain of the patella tendon
5%
Pain of the Achilles tendon
0%
Sulcus ulnaris syndrome
3%
Group B (Flexibar®+XCO®)
m = 17; f = 15
47 ± 8
76 ± 12
175 ± 10
24.7 ± 3.5
13%
0%
9%
13%
3%
6%
13%
neg. = 27; pos. = 5
69%
13%
19%
28%
13%
88%
> 27 weeks
2.5 ± 1.8
4.7 ± 2.2
0%
41%
16%
22%
38%
13%
0%
3%
63%
0%
6%
72%
16%
9%
9%
6%
3%
Page 4 of 13
Chi-square test
0.217
0.173
0.300
0.006
0.352
0.728
0.362
0.800
0.499
0.266
0.838
0.103
0.341
0.539
0.539
0.539
0.585
0.695
0.522
0.485
0.465
0.076
0.362
0.424
0.060
0.102
0.371
0.111
0.066
0.126
0.475
0.025
0.311
0.169
0.176
0.050
0.369
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 5 of 13
Table 2 Sports and profession
Group A (Flexibar®)
Group B (Flexibar®+XCO®)
Chi-square test
Sports
Constant sporting activity
80%
91%
0.369
Weekly training sessions
1.6 ± 0.7 à 1.7 ± 1.3 h
1.5 ± 0.6 à 1.6 ± 0.7 h
0.226
Professional groups
0.271
Employee
54%
56%
Engineer
5%
13%
Independent
3%
3%
Housewife
5%
0%
Sports teacher/therapist
0%
6%
Pensioner
5%
0%
Civil servant
18%
9 %
Craftsman
5%
6%
Lawyer
0%
6%
Physician
3%
0%
Use of PC
Professional use of PC
87%
91%
0.648
Hours of professional use of PC per day
4.9 ± 2.6
5.3 ± 2.7
0.545
Private use of PC
92%
88%
0.768
Hours of private use of PC per day
0.9 ± 0.5
1.0 ± 0.9
0.221
hand and be moved with a powerful movement of the arms to achieve the effect described.
– Measuring of the strength development according to
Jamar [kg]
Patients of both groups were asked to keep a diary concerning their pain levels and also training sessions during the whole period of the study.
Before the start of the training phase and after termin- ating the 12 weeks period, we performed each time an examination, whereby different measurement parameters were collected and the patients had to fill in the DASH- Score (score for the disabilities of the arm, shoulder and hand). We measured the strength development accord- ing to Jamar, the 2-point discrimination at the distal phalanx of each finger, both radial and ulnar, the vibra- tion sensation (by 128 Hz tuning fork) on top of the acromion, at the epicondylus humeri radialis and ulnaris, as well as at the distal ending of the radius and the ulna.
Primary goal parameter
– Measuring the vibration sensation in 1/8 steps by
using a 128 Hz tuning fork
– Measuring the 2-point discrimination at the distal phalanx of each finger both radial and ulnar
Statistics The input and processing of the statistical data was per- formed with SPSS 17.0. The patients’ characteristics were processed using cross tables and chi-square tests. For the analysis of the measured values, we initially cal- culated averages: average of the strength development, average of the 2-point discrmina- tion for each hand, and average of the vibration sensa- tion at the different localisations, etc. We subsequently calculated the changes of the paired t test, comparing before and after. For the comparison between both groups, we used the unpaired t test.
the 3 attempts of
– Course of pain on the visual analogue scale (VAS 0-
10)
Secondary goal parameter
– DASH-Score (from 0 = no limitations to 100 =
strong limitations)
– Compliance whilst performing the training period
Results Randomisation of the patients to ‘Group A’ (Flexibar®) and ‘Group B’ (Flexibar® + XCO®) resulted in a distribu- tion without relevant differences between the groups. The detailed list of basic characteristics, underlying dis- eases, diseases of the elbow and collateral tendon dis- eases can be found in Table 1. The table also shows that the patients in both groups reported symptoms for more
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Fig. 2 Instructions for the training. Arm in front of the body: 1a Hand in a neutral position, 1b hand in pronation, 1c hand in supination. Arm besides the body: 2a Hand in a neutral position, 2b hand in pronation, 2c hand in supination. Arm above the head: 3a Hand in a neutral position, 3b hand in pronation, 3c hand in supination
Page 6 of 13
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 7 of 13
Fig. 3 XCO® Walking & Running
than 27 weeks on average when they entered the study. In addition, the treatments already received in advance can also be taken from here. In addition, Table 2 shows the sporting activity, professional activity and PC use.
Primary goal parameter Course of pain on the visual analogue scale (VAS 0-10) In ‘Group A’ (Flexibar®), there was a reduction of pain from 3.6 ± 2.0 to 2.4 ± 2.1, which is corresponding with a decrease of 1.2 ± 2.5 points on the pain scale and a p value of p = 0.013. In ‘Group B’ (Flexibar® + XCO®), we found a reduction of pain from 3.7 ± 2.4 to 2.2 ± 1.9 and therefore a decrease of 1.5 ± 2.4 points on the pain scale and a p value of p = 0.004. See also Fig. 4. There was no significant difference between ‘Group A’ (Flexibar®) and ‘Group B’ (Flexibar® + XCO®) (p = 0.899).
points. This corresponds with an improvement of 44% in ‘Group A’ (Flexibar®) and 44 % in ‘Group B’ (Flexibar® + XCO®). We were thus able to observe a significant im- provement relating to the symptomatic limb with a p value of p = 0.001 in both groups. This is corresponding with a change for the better of 56% in both groups (Fig. 5) and without significant difference between the groups (p = 0.677).
Compliance whilst performing the training period The patients in ‘Group A’ (Flexibar®) trained with the Flexibar® on average over a period of 91 ± 12 days for 798 ± 223 min (corresponding on average to 106% of the required training), which complies with a weekly training of 61 ± 17 min and a daily training of 9 ± 3 min. On average, the patients in this group missed 12 ± 11 training sessions (corresponding on average to 15 % of the required training sessions).
Secondary goal parameter DASH-Score (from 0 = no limitations to 100 = strong limitations) Concerning the DASH-Score, an improvement was achieved from 32 ± 15 points before the period of train- ing to 14 ± 12 points after the training phase in ‘Group A’ (Flexibar®). In ‘Group B’ (Flexibar® + XCO®), there was an improvement from 27 ± 12 points to 12 ± 11
The patients in ‘Group B’ (Flexibar® + XCO®) trained with the Flexibar® on average over a period of 95 ± 17 days for 825 ± 192 min (corresponding on average to 109% of the required training), which complies with a weekly training of 61 ± 14 min and a daily training of 9 ± 2 min. On average, the patients in this group missed 10 ± 10 training sessions (corresponding on average to
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 8 of 13
Fig. 4 Course of pain on VAS
12% of the required training sessions). In the same period, they trained with the XCO® on average 648 ± 446 min (corresponding on average to 135% of the re- quired practice time), which complies with a weekly training of 47 ± 27 min and 27 ± 12 min per training session. This practice time was spread on average over 2 ± 1 training sessions per week. In doing so, the patients in ‘Group B’ missed 6 ± 9 training sessions with the
XCO® (corresponding on average to 25 % of the required training sessions) (Fig. 6).
Measuring of the strength development according to Jamar [kg] The measurements were taken concerning the symp- tomatic limb by using the 3-attempt-method, each time with extended as well as with a flexed elbow.
Fig. 5 DASH-Score before and after the 12 weeks of training
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 9 of 13
Fig. 6 Synopsis of the compliance of the patients
In ‘Group A’ (Flexibar®), we found an improvement of the strength development concerning the symptomatic limb when testing with an extended elbow from 24 ± 12 kg to 33 ± 11 kg and when testing with a 90° flexed elbow from 26 ± 12 kg to 32 ± 9 kg. This is correspond- ing with an increase of the strength development of 38% in ‘Group A’ (extended) or 23% (flexed). Therefore, (Flexibar®), a significant improvement was achieved in both test positions with a p value of p = 0.001 (extended) and p = 0.004 (flexed).
development of 17% (extended) or 3% (flexed). There- fore, in ‘Group B’ (Flexibar® + XCO®), a significant im- provement was achieved with a p value of p = 0.005 when testing with an extended elbow, whilst the testing with a flexed elbow showed only a slight improvement with a p value of p = 0.372 not (Fig. 7).
Neither the measurement with the elbow extended (p = 0.735) nor the measurement with the elbow flexed (p = 0.677) showed significant differences between ‘Group A’ (Flexibar®) and ‘Group B’ (Flexibar® + XCO®).
In ‘Group B’ (Flexibar® + XCO®), we found an im- provement of the strength development concerning the symptomatic limb when testing with an extended elbow from 29 ± 14 kg to 34 ± 11 kg and when testing with a 90° flexed elbow from 31 ± 13 kg to 32 ± 11 kg. This is strength corresponding with an increase of
the
Measuring the vibration sensation in 1/8 steps by using a 128 Hz tuning fork In ‘Group A’ (Flexibar®), we could find an improvement of the vibration sensation for the symptomatic limbs on average from 6.3/8 ± 0.6/8 to 6.5/8 ± 0.5/8 and therefore
Fig. 7 Strength development according to Jamar
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 10 of 13
about 3%. In ‘Group B’ (Flexibar® + XCO®), we could find an improvement for the symptomatic limbs on aver- age from 6.3/8 ± 0.7/8 to 6.6/8 ± 0.5/8 and therefore about 5%. In both groups, a significant improvement of the vibration sensation was achieved with a p value of p = 0.001 in ‘Group A’ (Flexibar®) and p = 0.003 in ‘Group B’ (Flexibar® + XCO®) (Fig. 8), but without significant difference between the groups (p = 0.091).
Measuring the 2-point discrimination at the distal phalanx of each finger both radial and ulnar In ‘Group A’ (Flexibar®), we could find an improvement of the 2-point discrimination for the symptomatic limbs on average from 5.2 mm ± 0.7 mm to 4.3 mm ± 0.6 mm and therefore about 18%. In ‘Group B’ (Flexibar® + XCO®), we could find an improvement of the 2-point discrimination for the symptomatic limbs on average from 5.0 mm ± 0.7 mm to 4.4 mm ± 0.6 mm and there- fore about 12%. In conclusion, in both groups, we mea- sured a 2-point discrimination with a p value of p = 0.001 (Fig. 9), but there was no significant difference between the groups (p = 0.959).
significant
improvement of
the
Discussion and conclusions The primary purpose of any therapeutic effort for treat- ing patients with lateral epicondylitis is the reduction of subjectively experienced pain by the patient which we put in the focus of our study as our primary goal param- eter. In doing so, we were able to notice a continuous decline of the experienced pain intensity over the whole period of the study (see Fig. 4), so that both groups lastly achieved a significant reduction of the pain (34 % in ‘Group A’ (Flexibar®) and 40% in ‘Group B’ (Flexibar® +
XCO®)). However, we could not determine a significant difference between both groups, so that we have to con- clude that a more intensive training with different de- for the vices does not bring any additional benefit development of pain. In addition, we could verify a sig- nificant improvement of functionality of the upper limb in both groups, which we established by the DASH- Score (see Fig. 5). In both groups, a reduction of the point value was recorded by 56%, which corresponds to an even more remarkable improvement than the pain re- duction. We could not find a significant difference be- tween the both practice groups here either. As both parameters (VAS and DASH) are certainly mostly rele- vant for the contentment of the patients but despite everything have to be regarded as subjective parameters, we have added objectively measurable parameters to our analysis—the strength development, the vibration sensa- tion and the 2-point discrimination. With regard to the measurement of strength development, it is hardly pos- sible to objectively test the muscles that are primarily stressed by the training, so we used the grip strength ac- cording to Jamar as a measurable parameter. Since the muscles tested here are also stressed to a certain extent (firm grip of the equipment during the exercise), but are not the actual target muscle group of the treatment, an- other explanation for the improvement in the values could be assumed as follows: The most likely explan- ation is pain reduction and the thus facilitated exercise performance. In particular, we were able to observe sig- nificant improvements when performing the exercise with the elbow extended—which corresponds to the pos- ition that was initially already more painful. These values can therefore be interpreted as an improvement in func- tionality or discomfort rather than a pure increase in
Fig. 8 Development of the vibration sensation
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 11 of 13
Fig. 9 Development of the 2-point discrimination
strength. In ‘Group A’ (Flexibar®), we found a significant improvement in both testings with the elbow extended and with the elbow flexed, whilst in ‘Group B’ (Flexibar® + XCO®), only the testing with the elbow extended showed a significant improvement and the testing with the elbow flexed only showed a slight improvement. The fact that ‘Group A’ (Flexibar®)—with the less intensive training—showed a better result here indicates that grip strength is not essential for the development of the first two goal parameters and that an increase in strength corresponding to the training intensity at least does not affect the tested muscle group. Furthermore, we also wanted to directly objectify the intended and expected effect on proprioceptive capabilities. For this purpose, we decided to test vibration sensation and 2-point dis- crimination in order to obtain measurable and compar- able values. Again, we could prove a statistically significant improvement in both groups, both for the vi- bration sensation and also for the 2-point discrimination as an indication that the intervention was able to achieve the intended effect on the proprioception capability. Also here, a more intensive proprioceptive training with two different devices did not result in a bigger improve- ment of the proprioceptive capabilities. It should not be overlooked that this therapeutic option is bound to a sufficient compliance of the patients. Within our study, we asked the patients to keep a training diary that gave us the chance to evaluate the compliance. We could find that the training with the Flexibar® was rarely missed and sometimes even done more than required, whilst about 25% of the training sessions with the XCO® were missed. This is probably due to the fact that the flexibar exercises are easier to integrate into the daily routine of a working patient than jogging or walking exercises. The
compliance as a very important factor of the success of this therapeutic options is also for certain one of the most relevant limitations. As a lot of patients hope for an immediate relief of their symptoms without any ac- tive support by themselves, a reduced compliance—be- yond the setting of a clinical trial with a voluntary participation—is possible, whereby the therapeutic suc- cess could ultimately be reduced. Another limitation could be the correct execution of the exercise perform- ance which is difficult to be controlled during autono- mous training. Within the study, we gave exact exercise instructions to our patients at the beginning of the train- ing period and also checked the exercise performance probatory at the beginning as well as at the end of the training period. Only 1-2 patients of each group showed an incorrect exercise performance, but we could also de- tect significant differences in the intensity of the execu- tion; the influence of which on the outcome could not be measured. These differences may be attributed to the great heterogeneity of the collective of probands, but should also be seen in the public patient collective. Whether the collected results we have obtained can also be transferred to similar training devices of other manu- facturers cannot be said with certainty, since we have not carried out a comparative study.
Overall, we were able to prove that the performance of a proprioceptive and ballistic training with the Flexibar® resulted in significant improvement of the subjectively experienced symptoms of the patients. However, a fur- ther benefit could not be shown by an additional train- ing with the XCO®. Furthermore, we were able to measure other improvements with objective measure- ment parameters—in this case, the strength develop- 2-point ment,
the
vibration
sensation
and
the
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 12 of 13
discrimination—which demonstrated the effectiveness of this training.
Within comparative research in online databases, a large number of hits are found in search of clinical train- ing studies for patients with lateral epicondylitis. Peter- son et al. [14] demonstrated, for example, that an active training in patients with lateral epicondylitis can cause a more rapid relief of pain than a purely wait-and-see be- haviour. In another clinical trial, Viswas et al. [15] found a superiority of a controlled exercise programme com- pared with a treatment with Cyriax physiotherapy. In addition, studies comparing a stretching with a consist- ent use of an orthosis [16] demonstrated an advantage of the active to the passive intervention. Various studies addressed the effect of an eccentric training of patients with lateral epicondylitis. At this, Söderberg et al. [17] could prove a significantly higher increase of a pain-free hand grip by combining an eccentric training with the use of an orthosis than the control group which received only the orthosis. Peterson et al. [18] finally compared two active training methods—eccentric versus concen- tric training—and found an advantage of the eccentric training over the concentric training for patients with lateral epicondylitis. However, all of the abovementioned studies have in common that no attention was paid to the reduced proprioception capability in the area of the elbow, which has been proven by Juul-Kristensen et al. [5]. However, examinations of other joints have already shown that proprioceptive training can have a positive effect on the functionality and reduction of symptoms in diseased joints. For comparison, other studies can be used which also detected a reduced proprioceptive cap- ability in patients with symptomatic gonarthrosis [19]. Further studies demonstrated the effectiveness of pro- prioceptive training compared with a control group [20] and other results even show an advantage of a proprio- ceptive training compared with an isometric training [21]. Concerning the beneficial effects of a propriocep- tive training for patients with lateral epicondylitis, only a few studies have been carried out yet. However, Tripp et al. [22] demonstrated in their study an improvement of the proprioceptive capabilities in the area of the elbow by performing a vibration training using vibrating dumb- bells and varying frequencies. This trial used a collective of probands without a symptomativ lateral epicondylitis though. In the course of our study, we were able to prove that patients with lateral epicondylitis, which are more likely to have proprioceptive deficits as shown in other trials [5], active vibration training could cause an improvement of the proprioception itself as well as an improvement of grip strength, functionality and in par- ticular pain reduction. In this setting, evaluated training devices can be used by the patients in the home environ- training is possible ment,
without a commitment to a location or a fixed date and therefore the integration into the daily work routine of the primarily concerned working patient population is possible without difficulties.
One limitation of the study is that it was not possible to recruit a sufficient number of probands suffering per- manent symptoms if there had been the chance that they could have been randomised to a group without treat- ment. Thus, a comparative assessment with the spontan- eous outcome is not possible. Furthermore, we cannot make any statement about de- vices from other manufacturers, because we did not test them and it is not certain whether the results could be transferred in this way.
in our
collective
Abbreviations DASH-Score: Score for the disabilities of the arm, shoulder and hand; VAS: Visual analogue scale
Acknowledgements Not applicable.
Authors’ contributions The study design was worked out by Berit Schiffke-Juhasz and Prof. Dr. med. K. Knobloch together. The clinical investigations and data collection were conducted by Berit Schiffke-Juhasz. Subsequently, the data set was analysed with the support of the statistician Dr. rer. nat. habil. L. Hoy and the results were interpreted by Berit Schiffke-Juhasz and Prof. Dr. med. K. Knobloch. The preparation of the article, as well as the graphics and tables, was done by Berit Schiffke-Juhasz and the contents were again coordinated with Prof. Dr. med. K. Knobloch. Prof. Dr. med. P. M. Vogt, as head of the department, was responsible for supervising the study. The authors read and approved the final manuscript.
Funding No other funding agency was involved anymore.
Availability of data and materials The datasets generated and analysed during the current study are not publicly available due still contain data that have not yet been analysed but are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate We have the agreement of the ethics committee from Medizinische Hochschule Hannover, Germany.
Consent for publication No personal data is disclosed in our article. The photos used in Fig. 2 show the main author herself and may be published.
Competing interests For our study, we received the training devices for our patients from Flexi- Sports® and the training devices remained in the property of the patients after the study. Apart from that, we received no further grants or support from Flexi-Sports® or another company or institution. Conversely, Flexi- Sports® has not demanded the results of our trial to publish them for adver- tising. So, we can declare that there is no conflict of interest for us.
so that an independent
Author details 1Herzogin Elisabeth Hospital, Braunschweig, Germany. 2Sportpraxis Knobloch, Hannover, Germany. 4Medizinische Hochschule Hannover, Hannover, Germany. 3Formerly Medizinische Hochschule Hannover, Hannover, Germany.
Schiffke-Juhász et al. Journal of Orthopaedic Surgery and Research (2021) 16:468
Page 13 of 13
Received: 30 March 2021 Accepted: 4 July 2021
22.
Tripp BL, Faust D, Jacobs P. Elbow joint position sense after neuromuscular training with handheld vibration. J Athl Train. 2009;44(6):617–23*9.
References 1.
Stöckle U. Ellenbogenchirurgie. Standardverfahren, Tipps und Tricks. 1. Auflage; 2010. Verhaar JA. Tennis elbow. Anatomical, epidemiological and therapeutic aspects. Int Orthop. 1994;18(5):263–7. https://doi.org/10.1007/BF00180221. Haahr JP, Andersen JH. Physical and psychosocial risk factors for lateral epicondylitis: a population based care-referent study. Occup Environ Med. 2003;60(5):322–9. https://doi.org/10.1136/oem.60.5.322. Shiri R, Viikari-Juntura E, Varonen H, Heliövaara M. Prevalence and determinants of lateral and medial epicondylitis: a population study. Am J Epidemiol. 2006;164(11):1065–75. https://doi.org/10.1093/aje/kwj325. Juul-Kristensen B, Lund H, Hansen K, Christensen H, Denneskiold-Samsøe B, Bliddal H. Poorer elbow proprioception in patients with lateral epicondylitis than healthy controls: a cross-sectional study. J Shoulder Elbow Surg. 2008;17(1 Suppl):72S–81S. Sevie TL, Wilson JK. Treating lateral epicondylitis. Sports Med. 1999;28(5): 375–80. https://doi.org/10.2165/00007256-199928050-00006. Van De Streek MD, Van Der Schans CP, De Greef MH, Postema K. The effect of a forearm/hand splint compared with an elbow band as a treatment for lateral epicondylitis. Prosthet Orthot Int. 2004;28(2):183–9. https://doi.org/1 0.1080/03093640408726703. Garg R, Adamson GJ, Dawson PA, Shankwiler JA, Pink MM. A prospective randomized study comparing a forearm strap brace versus a wrist splint fort he treatment of lateral epicondylitis. J Shoulder Elbow Surg. 2010;19(4):508– 12. https://doi.org/10.1016/j.jse.2009.12.015. Nichols AW. Complications associated with the use of corticosteroids in the treatment of athletic injuries. Clin J Sport Med. 2005;15(5):370–5.
2.
3.
4.
5.
6.
7.
8.
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
9.
10. Bisset L, Beller E, Jull G, Brooks P, Darnell R, Vincenzino B. Mobilisation with movement and exercise, corticosteroid injection, or wait and see for tennis elbow: randomised trial. BMJ. 2006;333(7575):939. https://doi.org/10.1136/ bmj.38961.584653.AE. Li A, Wang H, Yu Z, Zhang G, Feng S, Liu L, et al. Platelet-rich plasma vs. corticosteroids for elbow epicondylitis: a systematic review and meta- analysis. Medicine (Baltimore). 2019;98(51):e18358. 11.
12. Ruiz AG, Diaz GV, Fernández BR, Ruiz De Vargas CE. Effects of ultrasound- guided administration of botulinum toxin (incobotulinumtoxin A) in patients with lateral epicondylitis. Toxins (Basel). 2019;11(1):46. https://doi. org/10.3390/toxins11010046.
13. Newcomer KL, Laskowski ER, Idank DM, McLean TJ, Egan KS. Corticosteroid injection in early treatment of lateral epicondylitis. Clin J Sport Med. 2001; 11(4):214–22. https://doi.org/10.1097/00042752-200110000-00002.
14. Peterson M, Butler S, Eriksson M, Svärdsudd K. A randomized controlled trial of exercise versus wait-list in chronic tennis elbow (lateral epicondylosis). Ups J Med Sci. 2011;116(4):269–79. https://doi.org/10.3109/03009734.2011.600476.
15. Viswas R, Ramachandran R, Korde AP. Comparison of effectiveness of supervised exercise program and Cyriax physiotherapy in patients with tennis elbow (lateral epicondylitis): a randomized clinical trial. ScientificWorldJournal. 2012;2012:939645. Sölveborn SA. Radial epicondylalgia (‘tennis elbow’): treatment with stretching or forearm band. A prospective study with long-term follow-up including range-of-motion measurements. Scand J Med Sci Sports. 1997;7(4):229–37. Söderberg J, Grooten WJ, Ang BO. Effects of eccentric training on hand strength in subjects with lateral epicondylalgia: a randomized-controlled trial. Scand J Med Sci Sports. 2012;22(6):797–803. https://doi.org/10.1111/j.1 600-0838.2011.01317.x.
15. Viswas R, Ramachandran R, Korde AP. Comparison of effectiveness of supervised exercise program and Cyriax physiotherapy in patients with tennis elbow (lateral epicondylitis): a randomized clinical trial. ScientificWorldJournal. 2012;2012:939645. Sölveborn SA. Radial epicondylalgia (‘tennis elbow’): treatment with stretching or forearm band. A prospective study with long-term follow-up including range-of-motion measurements. Scand J Med Sci Sports. 1997;7(4):229–37. Söderberg J, Grooten WJ, Ang BO. Effects of eccentric training on hand strength in subjects with lateral epicondylalgia: a randomized-controlled trial. Scand J Med Sci Sports. 2012;22(6):797–803. https://doi.org/10.1111/j.1 600-0838.2011.01317.x.
16.
18. Peterson M, Butler S, Eriksson M, Svärdsudd K. A randomized controlled trial of eccentric vs. concentric graded exercise in chronic tennis elbow (lateral elbow tendinopathy). Clin Rehabil. 2014;28(9):862–72. https://doi.org/10.11 77/0269215514527595.
19. Chen Y, Yu Y, He CQ. Correlations between joint proprioception, muscle strength, and functional ability in patients with knee osteoarthritis. Sichuan Da Xue Xue Bao Yi Xue Ban. 2015;46(6):880–4. Tabacs J, Krowchuk NM, Garland SJ, Carpenter MG, Hunt MA. Dynamic balance training improwes physical function in individuals with knee osteoarthritis: a pilot randomized controlled trial. Arch Phys Med Rehabil. 2017;98:1586. 21. Ojoawo AO, Olaogun MO, Hassan MA. Comparitive effects of proprioceptive and isometric exercises on pain intensity and difficulty in patients with knee osteoarthritis: a randomized control study. Technol Health Care. 2016;24(6): 853–63. https://doi.org/10.3233/THC-161234.
19. Chen Y, Yu Y, He CQ. Correlations between joint proprioception, muscle strength, and functional ability in patients with knee osteoarthritis. Sichuan Da Xue Xue Bao Yi Xue Ban. 2015;46(6):880–4. Tabacs J, Krowchuk NM, Garland SJ, Carpenter MG, Hunt MA. Dynamic balance training improwes physical function in individuals with knee osteoarthritis: a pilot randomized controlled trial. Arch Phys Med Rehabil. 2017;98:1586. 21. Ojoawo AO, Olaogun MO, Hassan MA. Comparitive effects of proprioceptive and isometric exercises on pain intensity and difficulty in patients with knee osteoarthritis: a randomized control study. Technol Health Care. 2016;24(6): 853–63. https://doi.org/10.3233/THC-161234. |
s13018-022-03255-6 | "Nan et al. Journal of Orthopaedic Surgery and Research (2022) 17:407 https://doi.org/10.1186/s13018(...TRUNCATED) |
s13018-019-1110-4 | "Durrant et al. Journal of Orthopaedic Surgery and Research (2019) 14:107 https://doi.org/10.1186/s1(...TRUNCATED) |
s13018-018-0976-x | "Bai et al. Journal of Orthopaedic Surgery and Research (2018) 13:283 https://doi.org/10.1186/s13018(...TRUNCATED) |
s13018-014-0116-1 | "Ma et al. Journal of Orthopaedic Surgery and Research 2014, 9:116 http://www.josr-online.com/conten(...TRUNCATED) |
s13018-021-02429-y | "Chung and Kim Journal of Orthopaedic Surgery and Research (2021) 16:303 https://doi.org/10.1186/s13(...TRUNCATED) |
End of preview. Expand
in Data Studio
- Downloads last month
- 84