The Effect of Deformity and Hindfoot Arthritis on Midterm Outcomes of Ankle Replacement and Fusion

Abstract

Category: Ankle Arthritis Introduction/Purpose: Background: In 2014, Daniels et al reported the largest prospective comparison of clinical outcomes for Ankle Fusion (AF) and Replacement (TAR). Outcomes were comparable at 5.5 years mean F/U but were not stratified for potentially important factors including ipsilateral: Intra-articular deformity Significant foot deformity Hindfoot arthritis (or prior fusion) These factors define the Type and complexity of end stage ankle arthritis (ESAA) as described by the COFAS Classification. Since the effect of COFAS Type on outcomes is unknown, it remains unknown if TAR or AF might outperform the other in specific Types of ESAA and therefore be preferentially indicated. Purpose: Determine if outcomes for TAR and AF: vary with COFAS Type are different for ankles of the same COFAS Type. Methods: Prospective data from the COFAS multicenter database (3 sites, 5 surgeons) was used to compare outcomes in 890 consecutive non-randomized ankles with minimum 2 year F/U. TAR prostheses utilized were STAR, Hintegra, Infinity and InBone II. AF’s utilized open or arthroscopic techniques. Ankles were stratified according to the COFAS Classification of ESAA (1-isolated ankle arthritis; 2 - intra-articular ankle deformity (varus/valgus); 3 - foot or tibial deformity; 4 - hindfoot arthritis/fusion), creating 8 groups for comparison (AF1 – AF4; TAR1 – TAR4) A linear mixed-effects regression model (adjusted for significant variables identified by univariate analysis: Pre-operative Ankle Osteoarthritis Scale [AOS], age, diabetes status, BMI, surgeon) was used to compare the primary (AOS at latest F/U [AOS-LFU]) and secondary (SF-36 Physical Component Score at latest F/U [PCS-LFU]) outcome measures between groups. Reoperation & revision rates according to the COFAS Coding System were also compared. Results: There were 349 AF and 541 TAR cases with mean F/U of 5.4 years. Age averaged 55.9 (AF) and 64.3 (TAR) years. Analysis of raw AOS and PCS scores showed no difference pre-operatively between all 8 groups; all improved from pre-operative to LFU (p<0.05). Improvement was greater for TAR than AF in all COFAS Types, leading to significantly lower (better) AOS-LFU for TAR than AF in all COFAS Types (p <0.05). Multivariate analysis (see Figure) showed increasing (worsening) AOS-LFU for AF when progressing from Type 1 to 4, leading to better outcomes for TAR compared to AF in all Types, reaching statistical significance in Types 3 and 4. Revision rates were (6.3% [AF], 7.4% [TAR], NS); Non-revision reoperation rates were (14.6% [AF], 21.3% [TAR], p=0.01). Conclusion: At mid-term, patient reported outcomes for AF worsen with increasing ESAA complexity (as defined by COFAS Type), while outcomes for TAR are not affected by ESAA complexity. Both TAR and AF have similar revision rates, though non-revision reoperation rates are significantly higher for TAR. Both TAR and AF yield similar outcomes in non-complex ESAA (Types 1 and 2). However, for Complex ESAA (Types 3 and 4), TAR leads to significantly better outcomes, suggesting TAR may be indicated over AF in the presence of significant hindfoot deformity or hindfoot arthritis in patients who are otherwise candidates for either procedure.