Takedown of Ankle Fusions and Conversion to Total Ankle Replacements.

Abstract

With ankle replacements gaining credibility, there is a small subset of patients who might benefit from a conversion of an ankle fusion to a replacement. The objective of this study is to present clinical and radiographic results of patients who had their ankle fusion converted to total ankle arthroplasty (TAA). Patients presented to the senior author with ongoing ankle pain following fusion or increasing pain after a period of relative comfort after an ankle fusion. Outcomes were evaluated preoperatively and postoperatively with the Veterans Rand Health Survey (VR-12), Ankle Osteoarthritis Scale (AOS), and visual analog scale (VAS) pain scale. A patient satisfaction survey was also distributed. All ankle fusion conversions between April 2010 and December 2019 were included. Fifty-one patients (30 females) with the mean age of 62.6 (range, 37-83) years were followed. Mean follow-up was 4.2±2.3 (range, 2-11.5) years. Two patients were lost to follow-up (1 is deceased). Pre- and postoperative mean patient-reported outcome scores were all significantly different between groups except VR-12 mental subscale scores. VR-12 physical scores improved from 28.7±8.9 preoperatively to 38.9±9.9 (P < .001) postoperatively; VR-12 mental score was stable. AOS pain and disability subscales similarly improved: 55.9±24.4 to 27.9±25.4 (P < .001) and 61.7±21.1 to 31.1±25.7 (P < .001), respectively. VAS pain improved from a mean of 64.5±27.3 to 29.4±27.7 (P < .001). There was no tibiotalar dorsiflexion or plantarflexion with the ankle fusion. Initial postoperative visit revealed that average dorsiflexion was 10.9±5.93 degrees and average plantarflexion was 14.1±5.22 degrees. At the latest follow-up, dorsiflexion improved significantly to 15.5±6.33 degrees (P < .001), with no significant improvement in plantarflexion (P = .980). In this single-surgeon longitudinal study of 51 patients with a painful, malaligned, or nonhealed ankle fusion treated with an ankle replacement, we found highly satisfactory functional outcomes at an average of 4.2 years. Continued long-term follow-up will reveal whether the longevity of these replacements is comparable to primary replacements.