Survival of opening versus closing wedge high tibial osteotomy: A meta-analysis

This meta-analysis was designed to compare the longevity of the survivorship of opening versus closing wedge high tibial osteotomy (HTO). All studies reporting survival rates in patients who underwent open or closed wedge HTO with more than 5-year follow-up duration were included in the meta-analysis. Survival time was considered as time to conversion to TKA. Twenty three studies were included in meta-analysis, 20 of which were of level IV evidence. The pooled 5-year survival rates were 95.1% (95% CI: 93.1 to 97.1%) in open wedge HTO and 93.9% (95% CI: 93.1 to 94.6%) in closed wedge HTO. Although there was 1.2% greater survival rate in open wedge HTO than in closed wedge HTO, this difference did not reach statistical significance (P = 0.419). Pooled 10-year survival rates were 91.6% (95% CI: 88.5 to 94.8%) in open wedge HTO and 85.4% (95% CI: 84.0 to 86.7%) in closed wedge HTO, indicating that open wedge HTO had 6.2% greater survival rate 10 years after surgery than did closed wedge HTO (P = 0.002). No difference in 5-year survivorship was found between open- and closed-wedge HTO. However, the survival rate was higher in open-wedge HTOs than in closed wedge HTO at 10 years.

All 23 studies included in this meta-analysis had a low risk of selection bias. None assessed possible confounding factors. Of these 23 studies, 16 were considered high quality, with > 5 points on the NOS. Inter-rater reliabilities (к values) for all items of the NOS ranged from 0.68 to 0.88, indicating at least a more than substantial agreement between the two investigators. In general, publication bias did not need to be evaluated if fewer than 10 studies were included. Therefore, we only assessed the publication bias of 5-and 10-year survival rates of closed wedge HTO. Funnel plots showed that the mean survival rate of closed wedge HTOs were relatively symmetric at 5 years ( Fig. 2A), but skewed left asymmetrically at 10 years (Fig. 2B), indicating a lack of publication bias at the 5 year survival rate but some publication bias at the 10 year survival rate among the included studies. Egger's test also confirmed these trends of publication biases, with no significant publication bias in survival rates at 5 years (P = 0.109), but some publication bias at 10 years (P = 0.012).  Long-term survival rate. Of the 23 studies, five included 467 knees that underwent medial opening wedge HTOs and described the survival rate at 10 years after surgery, and 16 included 2496 knees that underwent lateral closing wedge osteotomy and described the survival rate at 10 years after surgery. The pooled 10-year survival rates were 91.6% (95% CI: 88.5 to 94.8%) in open wedge HTOs and 85.4% (95% CI: 84.0 to 86.7%) in closed wedge HTOs, respectively, indicating that open wedge HTOs had 6.2% greater survival rate at 10 years after surgery than did closed wedge HTOs (P = 0.002, Fig. 4A and B).

Discussion
This study estimated the approximate survival rates of open-and closed-wedge HTO by pooling the results of previous studies, most of which were cases series that did not compare two techniques directly. Although the methodological quality of the pooled studies was insufficient to adjust for possible confounders, the present study showed that the survival rate of open-wedge HTO was higher than that of closed-wedge HTO at 10 years. This meta-analysis was undertaken to compare the survivorship of open-and closed-wedge HTO on the treatment of symptomatic medial knee osteoarthritis with varus leg alignment. Survival rates of HTO were not significantly different at 5 years follow-up. Over time, both techniques exhibited decreased survival rates and closed-wedge HTOs decreased more than open-wedge HTOs. There are several possible reasons for the superior survival rate of open-wedge HTOs at 10 years. First, the open-wedge HTO is thought to allow a more accurate correction than closed-wedge HTO because it allows fine-tuning of the desired correction in both coronal and sagittal planes 10, 34-36 . A higher degree of precision can theoretically result in better mechanical alignment and possibly superior survivorship 10, 36 . Smith et al. 12 reported in a previous meta-analysis that there was a statistically significant difference in the mechanical axis with a more precise correction following open-wedge HTOs. Sun et al. 6 recently performed a meta-analysis that showed that open-wedge HTOs have a higher accuracy than closed-wedge osteotomy in cases of overcorrection and undercorrection, even though there was no statistically significant difference in the postoperative mechanical axis. Second, the dynamics of knee alignment is a possible reason for the inferior result of closed-wedge HTO. The presence of a lateral tibial thrust and a high knee adductor moment are well known risk factors of HTO for survivorship. One of the main disadvantages of closed-wedge HTO is the extensive lateral approach that inevitably affects the proximal tibiofibular joint and lateral collateral ligament. For this reason, closed-wedge HTOs are thought to result in a higher adductor knee moment and a higher possibility of persistent lateral thrust than open-wedge HTO postoperatively. Naudie et al. 25 reported that preoperative lateral tibial thrust is significantly correlated with the failure of HTOs in uni-and multi-variate analysis. Prodromos et al. 37 studied gait analysis after HTO, and showed that patients with a low knee adductor moment had better clinical results. Also, a medial opening-wedge HTO is a well-established procedure for the correction of proximal tibial vara with medial compartment osteoarthritis (OA). Proximal tibial vara has been reported in over 85% of cases with medial compartment OA and varus malalignment of the limb resulting from  OA may be attributable to the loss of cartilage and bony height of medial proximal tibia 38,39 . From a biomechanical aspect, opening the depressed medial proximal tibia is thought be a more reasonable procedure in terms of correcting the deformed lesion than closing the intact lesion of the proximal tibia. Finally, the development of fixation devices for use in an open-wedge HTO is another possible reason for the higher survival rate. Traditionally, an open-wedge HTO is associated with complications including implant failure, lateral cortical fracture, and delayed union or nonunion 6,34,[40][41][42] . Since angle-stable locking plates were introduced, implant related complications have been reduced markedly because of the HTO's increased stability 32,34,[40][41][42][43][44] .
Our meta-analysis shows that the survival rate with open-and closed-wedge HTO was 95.1% (95% CI: 93.1 to 97.1%) and 93.9% (95% CI: 93.1 to 94.6%) at 5 years, respectively, and 91.6% (95% CI: 88.5 to 94.8%) and 85.4% (95% CI: 84.0 to 86.7%) at 10 years, respectively. Our study revealed the only closed-wedge HTO survival rate at 15 years [74.8% (95% CI 72.5 to 77.2)] because studies of the 15 year survival rate with open-wedge HTO were limited. To our knowledge, no specific meta-analysis to date has quantified the survival rate of open-and closed-wedge HTO. The quantified survival rates found in the present study provide useful information not only for orthopedic surgeons but also for patients suffering from medial compartment OA of knee. If conservative treatment of medial compartment OA fails, surgical options include HTO, unicompartment arthroplasty (UKA), or TKA 21 . Therefore, these quantified results provide more information to orthopedic surgeons for choosing the appropriate treatment methods, although the final decision should be made after considering all factors. In addition, our study's comparison assists surgeons in choosing between two different HTO methods, unless a patient has a clear indication for one method over another. Similarly, patients are most concerned about clinical improvement and procedure survival when choosing between the two different HTO methods. Although clinical outcomes of open-and closed-wedge HTO have been compared in a few studies, previous meta-analyses showed no differences in most studies 6,12 . However, the present meta-analysis shows that open-wedge HTO has a longer survival rate than closed-wedge HTO at 10 years follow-up and provides a quantitative survival rate for both open-and closed-wedge HTOs. According to these results, patients who are candidates for HTO can get a clearer understanding of the consequences of open-and closed-wedge HTO with regard to survivorship. These results are important for surgeons and patients alike. Patients often demand a lucid explanation of the specific surgical procedure and have access to enormous amounts of information through the internet regarding any recommended surgery.
The current study has some limitations. First, differences in study designs are a limitation of this study. Most of the studies included in this meta-analysis were observational studies that were of variable methodological quality resulting in some inherent heterogeneity. Second, the number of studies regarding the survivorship of open-wedge HTOs is smaller than that of closed-wedge HTOs because closed-wedge HTOs were introduced earlier. If more studies report the survivorship of open-wedge HTOs in the future, it would help overcome this limitation. A third limitation is the heterogeneity of the fixation devices and wedge components used. This is a major limitation since biomechanical studies have demonstrated that any differences between open-and closed-wedge HTOs may be because of the nature of the osteotomy procedure and the fixation device used 12,32,34,[40][41][42][43][44][45] .
In conclusion, this is the first meta-analysis that shows no difference in the 5-year survivorship of open-and closed-wedge HTOs, and the survival rate was higher in open-wedge HTOs at 10 years. In addition, we estimated the long-term survival rate of both open-and closed wedge HTOs, providing useful information to surgeons and patients. However, the clinical evidence about long term survival rate at 10 years should be interpreted with caution, given our finding that there may be some publication bias in the 10 year survival rate among the included studies. Randomized control trials with a robust design need to be conducted to draw definitive conclusions regarding which of the two techniques yields superior long-term survival rates. Data extraction. Two investigators independently recorded data from each study using a predefined data extraction form. Any disagreement unresolved by discussion was resolved by consensus or by discussion with a third investigator.

Methods
Variables recorded included: (1) type of HTO (i.e., opening and/or closing wedge HTO and sample size; (2) numbers and percentages of surviving procedures without conversion to TKA at last follow-up; and (3) follow-up duration. Studies were excluded if (1) they dealt with a different type of high tibial osteotomy (i.e. dome osteotomy); and (2) their follow-up duration was < 5 years.
SCientifiC RepORTS | 7: 7296 | DOI:10.1038/s41598-017-07856-8 Assessment of methodological quality. Two investigators independently assessed the methodological quality of each study using the Newcastle-Ottawa Scale (NOS), as recommended by the Cochrane Non-Randomized Studies Methods Working Group. In this analysis, the NOS star system, which awards stars depending on the level of bias, was adjusted to a scale that included only low (one star), high, and unclear bias. Each study was judged on three criteria: the selection of the study groups, the comparability of the groups, and the ascertainment of either the exposure or outcome of interest for case-control and cohort studies. Studies of high quality were defined as a score > 5 points. Disagreements in scores were resolved by discussion and consensus between the two reviewers.
Statistical analysis. The main outcome of the meta-analysis was to compare the 5-year and 10-year procedure survival rates between opening-and closing-wedge HTOs, with continuous variables reported as the mean survival rate and the 95% confidence interval (CI). These values were analyzed with a random effects model. Interrater reliability in assessing methodological quality was evaluated by kappa (к), with values of ≤0.40, 0.41-0.60, 0.61-0.80, and 0.81-1.00 indicating no, moderate, substantial, and almost perfect agreement, respectively. Heterogeneity among the studies was determined by estimating the proportion of between-study inconsistencies due to actual differences between studies, rather than due to random error or chance, using theI 2 statistic, with values of 25%, 50%, and 75% considered low, moderate, and high, respectively. All statistical analyses were performed using RevMan version 5.2 and Stata/MP 13.0. Publication bias was also assessed using funnel plots and Egger's test.