End-stage ankle arthritis may be managed surgically with either ankle fusion or total ankle replacement (TAR). The results of total ankle replacement have improved over the recent decades, but challenges remain. Peri-implant osteolysis has been major problem, as it compromises the stability of the implant components and can lead to aseptic loosening and implant failure. 

For this retrospective study, 164 ankles (34 Scandinavian Total Ankle Replacement (STAR) and 130 Ankle Evolutive System (AES)) operated on in single institution during 1997–2008 were followed clinically and radiologically. Histological samples were collected from ankles revised due to periprosthetic osteolysis. Analysis from the data covering the years 1997–2006 of the Finnish Arthroplasty Registry was conducted. 

The peri-implant osteolysis was quite common in the AES total ankle implants: 70% of the ankles exhibited osteolysis at the latest follow-up. Dual-coating of the implant was associated with a 3.1-fold risk of osteolysis and of significantly earlier development of osteolysis compared to single-coating. Histology revealed a foreign-body reaction characterized by extensive soft and bone tissue necrosis. RANK/RANKL-mediated osteoclast and multinuclear foreign body giant cells contributed to peri-implant osteolysis, and there was increased expression of danger signals in the peri-implant tissues, suggesting an auto-inflammation mechanism behind osteolysis. 

The annual incidence of TAR according to the Finnish Arthroplasty Registry was 1.5 per 105 inhabitants and overall implant survival was 83% at 5 years when any revision was the end point. The most common reasons for revision were aseptic loosening (39%) and instability (39%). In the registry study, there was no difference in the survival rates between the STAR and AES designs, nor was there any association between age, gender, diagnosis, or hospital volume and TAR survival. 

The survival of the STAR implant was satisfactory, 93.8% (95% CI 77.5% to 98.4%) at 5 years, and 87.2% (95% CI 69.4% to 95.0%) at 10 and 15 years. There was no statistically significant association between implant survival and patient age, gender, BMI, or diagnosis. The overall rate of revisions was 44%, which includes all postoperative revisions for osteolysis, component and insert exchanges, and conversions to arthrodesis. 

The survival of the AES implant was strongly affected by osteolysis and malalignment, and inferior compared to previously published results. The 5-year survival was 87.3% (95% CI 80.0% to 92.0%), and the 10-year survival 74.9% (95% CI 65.4% to 82.2%). Postoperative alignment of ≥10º of varus predicted a poorer outcome and was statistically significant for implant survival (p=0.0005). The revision rate for all revisions was 57% including all postoperative revisions for osteolysis, component exchanges, and conversions to arthrodesis. Osteolysis was the main reason for revisions and failure. 

The survival of the STAR total ankle replacement was satisfactory in the long-term, but the results of the AES total ankle implants were strongly influenced by aggressive and early-emerging osteolysis. Future studies should focus on examining the mechanism behind the osteolytic process in TAR to avoid similar problems for implant development in the future.