Revision Rates of New Total Ankle Replacements vs Old Total Ankle Replacement: A Comprehensive Cohort Study

Research Type: Level 2 - Prospective comparative study, Meta-analysis of Level 2 studies or Level 1 studies with inconsistent results Introduction/Purpose: Advancements in total ankle replacement (TAR) implants have been evident in recent years, as supported by various studies. Concurrently, there has been an exponential increase in the number of patients undergoing TAR procedures. While national registries have compared outcomes across different types of TAR implants, limited literature exists on comparative studies conducted by the same group of surgeons. Such studies enhance internal validity by minimizing inter-surgeon variability. Methods: This prospective cohort study examined total ankle replacements (TAR) conducted by five specialized foot and ankle surgeons at a single tertiary care center from 2003 to 2023. We excluded revision prostheses and patients who declined registry participation. Each implant’s follow-up began at surgery, with revision defined as component replacement, amputation, or arthrodesis. TAR implants were divided into “old” (Agility, Hintegra, STAR, Mobility) and “new” (Infinity, InBone II, Invision, Zimmer Biomet). We recorded patient demographics such as age, diabetes status, and osteoarthritis etiology, then compared these variables between implant categories. A Kaplan-Meier analysis evaluated revision rates, defining treatment failure as any revision. Patient-reported outcome measures (PROMs), the Ankle Osteoarthritis Scale (AOS) and both physical (SF36PC) and mental (SF36MC) components of the Short Form-36 were assessed preoperatively and one year postoperatively. All clinical data were systematically collected and carefully validated. A negative binomial regression model accounted for confounders and tested for interactions. Results: This study included 631 TAR. Diabetes (p=0.31), sex (p=0.55), and age (p=0.07) showed no statistical differences, while inflammatory etiologies were more frequent in old prostheses (24.73% vs. 11.21%, p< 0.01). Kaplan-Meier analysis showed a 10-year revision rate of 14.13% for old implants versus 10.51% for new implants, with a significant log-rank test (p < 0.01). Patient-reported outcomes (AOS, SF36PC, SF36MC) were statistically similar preoperatively and at one-year follow-up. Mean follow-up time differed significantly between old (15.41 years) and new (5.49 years) prostheses (p < 0.01). Negative binomial regression indicated a higher incidence rate ratio (7.83) for old implants and lower ratios for older age (0.98), postraumatic (0.45), and inflammatory (0.29) etiologies, confirming an elevated failure risk for older prostheses. All differences were deemed significant if p< 0.05. Conclusion: In our cohort, newer TAR designs demonstrated greater survivorship compared to older designs, particularly when accounting for differences in follow-up time. Up to 10 years post-implantation, newer designs exhibited superior performance, suggesting ongoing advancements in TAR technology.