Comparison of clinical and radiological outcomes between opening-wedge and closing-wedge high tibial osteotomy: A comprehensive meta-analysis

High tibial osteotomy (HTO) has been widely used for clinical treatment of osteoarthritis of the medial compartment of the knee, and both opening-wedge and closing-wedge HTO are the most commonly used methods. However, it remains unclear which technique has better clinical and radiological outcomes in practice. To systematically evaluate this issue, we conducted a comprehensive meta-analysis by pooling all available data for the opening-wedge HTO and closing-wedge HTO techniques from the electronic databases including PubMed, Embase, Wed of Science and Cochrane Library. A total of 22 studies encompassing 2582 cases were finally enrolled in the meta-analysis. There was no significant difference regarding surgery time, duration of hospitalization, knee pain VAS, Lysholm score and HSS knee score (clinical outcomes) between the opening-wedge and closing-wedge HTO groups (P > 0.05). However, the opening-wedge HTO group showed wider range of motion than the closing-wedge HTO group (P = 0.003). Moreover, as for Hip-Knee-Ankle angle and mean angle of correction, no significant difference was observed between the opening-wedge and closing-wedge HTO groups (P > 0.05), while the opening-wedge HTO group showed greater posterior tibial slope angle (P < 0.001) and lesser patellar height than the closing-wedge HTO group (P < 0.001). On light of the above analysis, we believe that individualized surgical approach should be introduced based on the clinical characteristics of each patient.


Introduction
Knee osteoarthritis (OA) is a common and multi-factorial arthritic disorder, which may lead to joint dysfunction, i.e. reduction of joint motion and physical disability, as a result of degeneration, destruction, and loss of articular cartilage of knee joint [1]. Most patients with knee OA would develop varus deformity, and varus malalignment further overloads the medial tibiofemoral compartment, causing degenerative changes in the articular cartilage at last. Both a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 conservative methods and surgical treatment were used for treatment of pain and dysfunction in knee OA. Surgical treatment included various techniques, i.e. high tibial osteotomy (HTO), arthroscopic surgery and total knee arthroplasty. Among them, HTO is the most frequently used method in young and more active patients [2].
In previous studies, various techniques of HTO were documented and evaluated. Among them, closing-wedge HTO and opening-wedge HTO, which were stabilized by a locking plate [3,4], were two of the most frequently used techniques. For the closing-wedge HTO, a wedgeshaped cut is made in the lateral; while for the opening-wedge HTO, the medial tibia is cut. Many previous studies have compared the advantages and disadvantages between the closingwedge and the opening-wedge HTO [5][6][7]. For example, the closing-wedge HTO was more advantageous in patients with a medial compartment OA with varus deformity for decrease of the posterior tibial slope [8]. However, several disadvantages have been also reported for the closing-wedge HTO, such as the need for fibular osteotomy, bone stock loss and nervous system complications [4,5]. Different from the closing-wedge HTO, the opening-wedge HTO which has been performed more recently and doesn't need a fibula osteotomy, caused fewer co-morbidities related to the fibular ostetomy [4]. Compared to the closing-wedge HTO, the opening-wedge HTO has several advantages, i.e. easier adjustment of alignment correction, more rapid rehabilitation, as reported previously [9,10].
In detail, the advantages, disadvantages and sometimes complications were compared between the opening-wedge and closing-wedge HTO [7,8,11]; however, the results remained inconsistent. In the present study, we performed a comprehensive meta-analysis of the clinical and radiological outcomes between the opening-wedge and closing-wedge HTO, to explore whether the opening-wedge HTO was superior to the closing-wedge HTO. Our study would provide reasons for clinical guidance of surgical choice.

Eligibility criteria
Eligible studies must meet the following criteria: 1) articles published in English and peerreviewed journals; 2) randomized controlled trials (RCTs) and non-RCTs design comparing clinical and/or radiographic outcomes between closing-wedge HTO and opening-wedge HTO; 3) patient samples were independent from each other (if samples from different studies overlapped, only the studies with the largest sample size were included).
Animal study, review, commentary and meeting abstracts without detailed and novel data were excluded from the meta-analysis.

Data extraction
For each eligible study, two independent investigators extracted the following data: 1) first author and publication year; 2) type of study design; 3) sample size, age and gender distribution; 4) follow-up periods; 5) clinical and radiographic outcomes; 6) other data if essential. Any disagreement of data was subject to the third author. Moreover, if essential data was not available directly from the manuscript, the corresponding author was contacted, or we calculated them, i.e. standard deviations and confidence intervals, from already provided data.
In detail, clinical outcomes included, but not limited to, time course of surgical operation, duration of hospital stays, various sorts of knee scoring (Hospital for Special Surgery (HSS) knee score, knee pain visual analog score, Lysholm knee score and Wallgren-Tegner score), knee range of motion (ROM), as well as patient satisfaction (score). Radiographic outcomes included, but not limited to, mean angle of correction, hip-knee-ankle (HTA) angle, posterior tibial slope angle, and patellar height (PH).

Assessment of methodological quality
The revised Jadad scale by the Cochrane Non-randomized Studies Methods Working Group were used to assess methodological quality [12]. A total of 7 points were assigned to this scoring method, including the randomization process (2 points), allocation concealment (2 points), appropriateness of blinding (2 points), and a description of dropouts and withdrawals (1 point). A score of 4-7 points indicated high quality, and 0-3 indicating poor quality. Two investigators independently extracted the data using a standardized form. The final score was based on the consensus assessment from two investigators. If disagreements persisted, a third author was consulted.

Statistical analysis
The Stata 12.0 statistical software package (http://www.stata.com/) and the Review Manager 5.3 software (The Cochrane Collaboration) were applied to conduct publication bias analysis, meta-analysis and sensitivity analysis. Potential publication bias was determined by both the Egger regression test for a funnel plot [13] and the Begg-Mazumdar test, which is based on Kendall's-τ [14]. For dichotomous variables, the odds ratio (OR) with 95% confidence interval (95% CI) was derived for each outcome; for continuous variables, we calculated the standardized mean difference and 95% CI. We performed the meta-analysis using a fixed-effect model if no significant heterogeneity was present; otherwise, a random-effect model was applied. To assess heterogeneity between studies, we performed a chi-square test and estimated the I 2 statistic. For all analyses, P < 0.05 was considered statistically significant. The study was approved by the ethics committee of Lishui People's Hospital of Wenzhou Medical University. All the protocols and experimental procedures were in accordance with the Declaration of Helsinki and other international rules (S1 Table).

Literature search and eligible studies
According to our literature search approaches, a total of 513 references were indexed. After excluding 487 studies for reasons specified in Fig 1 and S1 Fig, a total of 22 studies (7 RCTs and 15 non-RCTs) encompassing 2582 cases, including 1274 knees which were underwent opening-wedge HTO and 1308 ones with closing-wedge HTO operation, were finally enrolled in the final meta-analysis [5-8, 10, 15-31]. Table 1 listed the detailed information of recruited studies for this meta-analysis, i.e. sample size, age and gender proportion, fixation approaches, etc. It is worth note that four studies [29,[32][33][34] were removed from the current meta-analysis due to sample duplication within other studies [17,28,31].

Clinical outcomes
As for clinical outcomes, the following several aspects were compared between opening-wedge and closing-wedge HTO, i.e. surgery time, duration of hospital stays, HSS knee score (> 1 year follow-up), knee pain VAS, range of motion (flexion angle) and Lysholm knee score. Other parameters, including patient satisfaction (score), Wallgren-Tegner score, Insall-Salvati index and complete weight bearing (day), were not pooled in our meta-analyses due to limited studies (N < 2) or sample size. Notably, since no significant difference was observed for pre-operative characteristics between the opening-wedge and closing-wedge HTO groups, only the post-operative values of clinical outcomes were compared.

Discussion
In the present study, we performed a comprehensive meta-analysis of clinical and radiographic outcomes between the opening-wedge and closing-wedge HTO groups. For most of the included indexes, no significant difference was observed between these two groups. However, for some outcomes, i.e., postoperative range of motion (flexion angle), posterior tibial slope angle and patellar height, there was indeed significant difference between the opening-wedge and closing-wedge HTO groups.
As for clinical outcomes, only the range of motion (flexion angle) showed significant difference between the opening-wedge HTO and closing-wedge HTO groups (P = 0.003), and the former exhibited wider range of motion than the latter (SMD = 0.33, 95% CI = 0.11-0.56). As we know, a portion of patients undergoing the opening-wedge HTO were required to receive range-of-motion exercise during the early postoperative period, and in Song et al. study [6,21], all the opening-wedge HTO patients received the range-of-motion exercise, which was not the case for the closing-wedge group. In our meta-analysis, two studies from Song et al. constitute the largest portion of the overall sample size, which might lead to biased result regarding the range of motion (flexion angle).
As for the patellar height, we pooled data from three distinct measuring methods, including Caton index, Insall-Salvati index and Blackburne-Peel index. Our analyses showed that the patellar height, as measured by Caton index and Insall-Salvati index, was significantly decreased after the opening-wedge HTO than that after the closing-wedge HTO. Careful examination of the raw data from previous studies, we observed that the patellar height was decreased after the opening-wedge HTO, while increased after the closing-wedge HTO. The possible explanation of this controversial result was that the opening-wedge HTO may lower the tibial tuberosity leading to decreased patella height, while the closing-wedge HTO could elevate the tibial tuberosity as a result of a proximalisation of the proximal tibia [35]. Different from our meta-analysis, Takeuchi et al. found that patellar height was maintained after opening-wedge HTO and speculated that this result was due to range-of-motion exercises during the early postoperative period [36]. However, in Song et al. study, in which range-of-motion exercise was recommended after opening-wedge HTO, most patients showed a significant decrease in patellar height after surgery as well, suggesting that range-of-motion exercise was not the main cause of decrease in patellar height [21]. Intriguingly, slight but not significant reduction of patella height as measured by Blackburne-Peel index was observed in the opening-wedge HTO group when compared with that of the closing-wedge group (P = 0.377). This inconsistency may be resulted from the limited sample size, which could cause publication bias and insufficient statistical power.
Moreover, the result of the meta-analysis showed that posterior tibial slope angle after opening-wedge HTO was significantly increased than that after closing-wedge HTO (SMD = 1.31, 95% CI = 0.71-1.91). Similar with the patellar height, the posterior slope was increased following the opening-wedge HTO, while decreased after the closing-wedge HTO. It is one frequently reported problem associated with the opening-wedge HTO that posterior tibial slope might be increased. Several theories have been put forward to explain the etiology of changes of posterior tibial slope following HTO. The most widely accepted theory is that the unique anatomic geometry of the proximal tibia is responsible for the change of posterior tibial slope angle following HTO [37]. This geometric shape requires a smaller opening gap at the anterior part of the osteotomy site than the posteromedial part, as the former is closer to the hinge point of osteotomy if the posterior tibial slope angle remains unchanged [38].
Several limitations should be presented to interpret the present results. Firstly, different surgical approaches were included and mixed in the current meta-analysis, which may confound the meta-analysis result. Secondly, though most of the included studies were of high-quality, a small portion of the included studies were of low-quality as indicated by the revised Jaded scale ( Table 1). Insufficient method description, low reliability of radiographic results, and study design in these studies might lead to between-group heterogeneity. Thirdly, although all the eligible studies were included in this meta-analysis, the sample size was relatively small, which may induce type-II statistic error. Updated meta-analysis of this issue is essential for a more solid conclusion in future.

Conclusions
Except for the range of motion (flexion angle), posterior tibial slope angle and patellar height, both the opening-wedge and closing-wedge HTO had satisfactory and comparable clinical and radiographic outcomes. On light of the above analysis, we believe that individualized surgical approach should be introduced based on the clinical characteristics of each patient.