Risk Factors for Early Mechanical Failure and Revision in Modern Total Ankle Replacements

Abstract

Category: Ankle Arthritis; Ankle Introduction/Purpose: Total ankle replacements (TAR) have rapidly advanced in terms of volume, technique, design, and indications. However, they are still at risk for failure and need for revision, even early in the postoperative period. Although recently there has been extensive research on TAR, many studies are limited in their power and sample size to assess risk factors for failure/revision. Using a large multi-surgeon retrospective database, we sought to describe the incidence and etiology of mechanical failure across modern implants. We also aimed to identify risk factors for early mechanical failure. We hypothesized that greater body mass index (BMI) and ipsilateral hindfoot fusion would be predictors of mechanical failure, and that implants with greater extent of tibial fixation (stems and keels) would be protective against mechanical failure. Methods: This is a single-center retrospective review of TAR patients in a prospective database. Adults undergoing primary TAR using any implant (Cadence, Inbone, Infinity, Invision, Salto, STAR, Vantage, Zimmer) who were minimum 2 years post-surgery were included. Exclusion criteria included revision TAR or patients undergoing ankle fusion takedown and conversion to TAR. Five surgeons contributed patients. Demographics, etiology of arthritis, operative data, and postoperative complications, reoperations, and revisions were collected. Revisions and reoperations were classified by standard criteria. Cause of revision was recorded. There were 560 TARs eligible for 2-year follow-up (mean 3.9 +- 1.1 years postoperative), with mean follow-up of 2.2 +- 1.4 years (range 1-71 months). The cohort had mean age 63.8 +- 9.1 years, BMI 29.4 +- 5.2 kg/m2, and 57% were male. Kaplan- Meier survival plots were created. Risk factors for revision due to mechanical failure were analyzed using multivariable logistic regression with partial Wald chi square statistics. Results: Most implants were low-profile (n=444, 79.3%), with 28 implants having tibial keels (5%) and 88 with tibial/talar stems (15.7%). There were 25 (4.5%) revisions of tibial and/or talar components at mean 1.6 years after index surgery (Figure). Preoperative coronal deformity did not impact revision rates (p=0.433). Patients who underwent revision for mechanical failure had significantly higher rates of concomitant hindfoot fusion (20% vs. 8.4%, p=0.047) and greater BMI (31.4 kg/m2 vs. 29.3 kg/m2, p=0.04). In multivariable logistic regression, hindfoot fusion was just short of statistical significance (odds ratio [OR] 2.96, p=0.056), and BMI was not significant (OR 1.05, p=0.417). Stemmed or keeled implants did not significantly decrease revision rates compared to low-profile implants, although the revision rate of stemmed implants was substantially low (1/88, 1.1%). Conclusion: Among 560 modern TAR implants, short-term survivorship was 95% at 2-year follow-up, which is consistent with existing studies. Patients who underwent revision for early mechanical failure had significantly higher BMI and rates of hindfoot arthrodesis. However, these variables fell short of statistical significance in multivariable regression, suggesting that there are either additional factors contributing to revision, or a larger cohort is needed. In addition, these results are likely biased by surgeons' tendencies to use stemmed implants in patients with hindfoot fusions. Further research is warranted to determine if stemmed implants provide a protective effect in these high-risk patients.