Total ankle replacement (TAR) is increasingly used in the treatment of end-stage ankle arthropathy, but much debate exists about the clinical result. The goals of present study are: 1) to provide an overview of the clinical outcome of 58 TAR's in a single centre and 2) to assess the association between radiological characteristics and clinical outcome.

We reviewed a prospective included cohort of 58 TAR's in 54 patients with a mean age of 66.9 (range 54–82) and a mean follow-up of 21.6 months (range 1.45–66.0). The TAR's where performed by a single surgeon in a single centre (MUMC) between 2010 and 2015, using the CCI ankle replacement. A standard surgical protocol and standardized post-op rehabilitation was used. Patients were followed-up pre-op and at 1 day, 6 weeks, 3–6–12 months and yearly thereafter post-op. The AOFAS and range of motion (ROM) were assessed and all complications, re-operations and the presence of pain were recorded. Radiographic assessment consisted of the estimation of prosthesis alignment, migration, translation and radiolucent lines using the Rippstein protocol (1). The clinical outcome was compared with a systematic review of TAR outcome.

Ten intra-operative complications occurred and 9 were malleolar fractures. Post-operative complications occurred in 20 out of the 54 patients (37.0%). Impingement (5/54 patients), deep infection (4/54 patients), delayed wound healing (3/54 patients) and minor nerve injuries (3/54 patients) were the most frequently recorded. 18 patients (31.0%) underwent one or more re-operations and 12 of these 18 patients underwent a component revision (mostly the PE insert) or a conversion to arthrodesis. Despite the complications and revisions, the functional outcome improved. Radiologically 15.8% of the TAR's were positioned in varus and 1.8% in valgus. Migration in the frontal and sagittal plane is seen in 3 and 2 TAR's respectively. Radiolucency is significantly increasing with the follow-up time (p=0.009). Migration in the frontal plane is significantly associated with conversion to arthrodesis (p=0.005) and migration in the sagittal plane to revision of a component or conversion to arthrodesis (p=0.04). Finally, pain is significantly associated with re-operations (p=0.023) and complications (p=0.026). Remarkable is that the clinical outcome is independent of the direct post-op alignment of the TAR.

The complication-, re-operation and revision or conversion to arthrodesis rates makes the clinical outcome of TAR still questionable favourable. Especially the complication and re-operation incidences are greater than found in the systematic review. However, it is remarkable that the minor complications and re-operations not related to the TAR are not often mentioned in the literature. Radiographic characteristics could be of value in predicting this clinical outcome and thereby influence the post-operative handling. In conclusion, our results show relatively high incidences of complications (37.0%) and re-operations (31.0%) when minor complications and re-operations are included. TAR clinical outcome can be predicted by radiographic migration characteristics and pain.