A comparative study of modern total ankle replacement and ankle arthrodesis for ankle osteoarthritis at different follow-up times: a systematic review and meta-analysis.

ABSTRACT

PURPOSE: Total ankle replacement (TAR) or ankle arthrodesis (AA) is the main surgical treatment for end-stage ankle osteoarthritis. However, the therapeutic effect of the two surgical procedures at different follow-up times remains controversial. The purpose of this meta-analysis is to compare the short-term, medium-term, and long-term safety and efficiency of the two modern surgical treatments. METHODS: We conducted a comprehensive search in PubMed, EMBASE, Cochrane library databases, Web of Science, and Scopus. The main results were the patient's reported outcome measure (PROM) score, satisfaction, complications, reoperation, and surgery success rate. Different follow-up times and implant designs were used to evaluate the source of heterogeneity. We used a fixed effects model for meta-analysis and I2 statistic for evaluating heterogeneity. RESULTS: Thirty-seven comparative studies were included. In the short term, TAR significantly improved clinical scores (AOFAS score WMD = 7.07, 95% Cl 0.41-13.74, I2 = 0.0%; SF-36 PCS score WMD = 2.40, 95% Cl 2.22-2.58, I2 = 0.0%; SF-36 MCS score WMD = 0.40, 95% Cl 0.22-0.57, I2 = 0.0%; VAS for pain WMD = - 0.50, 95% Cl - 0.56-0.44, I2 = 44.3%) and had the lower incidence of revision (RR = 0.43, 95% CI 0.23-0.81, I2 = 0.0%) and complications (RR = 0.67, 95% Cl 0.50-0.90, I2 = 0.0%). In the medium term, there were still higher improvements in both the clinical scores (SF-36 PCS score WMD = 1.57, 95% Cl 1.36-1.78, I2 = 20.9%; SF-36 MCS score WMD = 0.81, 95% Cl 0.63-0.99, I2 = 48.8%) and the patient satisfaction (RR = 1.24, 95% Cl 1.08-1.41, I2 = 12.1%) in the TAR group, but its total complications rate (RR = 1.84, 95% Cl 1.26-2.68, I2 = 14.9%) and revision rate (RR = 1.58, 95% CI 1.17-2.14, I2 = 84.6%) were significantly higher than that of the AA group. In the long term, there was no significant difference in clinical score and satisfaction, and a higher incidence of revision (RR = 2.32, 95% Cl 1.70-3.16, I2 = 0.0%) and complications (RR = 3.18, 95% Cl 1.69-5.99, I2 = 0.0%) was observed in TAR than in AA. The result of the third-generation design subgroup was consistent with that of the above pooled results. CONCLUSION: TAR had advantages over AA in the short term due to better performance in terms of PROMs, complications, and reoperation rates, but its complications become a disadvantage in the medium term. In the long term, AA seems to be favored because of lower complications and revision rates, although there is no difference in clinical scores.