Long-term clinical outcomes of allograft-prosthetic reconstruction for tumours of the extremities

PURPOSE: Allograft-prosthetic composites (APC) are used to reconstruct large periarticular defects following tumour resection, with potential advantages especially restoration of bone stock and ligamentous reattachment. While short- and mid-term outcomes have been reported on extensively, long-term clinical results remain limited. This study evaluated the incidence of mechanical and non-mechanical complications, risk factors for complications, and the cumulative incidence of reconstruction failure following APC reconstruction for extremity tumours with a minimum follow-up of ten years. METHODS: We retrospectively reviewed 64 APC with at least ten years follow-up in our centre. Predominant diagnoses were osteosarcoma (40%) and chondrosarcoma (28%). Reconstructions involved the proximal femur (39%), distal femur (22%), proximal tibia (23%) and proximal humerus (16%). Median follow-up was 24.5 years (95%CI 23.6-25.4). RESULTS: Instability occurred in nine reconstructions (14%). Non-union was observed in nine reconstructions (14%). Implant loosening occurred in seven reconstructions (11%) after a median of 14 years (range 2-18 years). Allograft collapse occurred in 13 reconstructions (20%) after a median of three years (range 1-15). Infection developed in five reconstructions (8%). Cumulative incidence of mechanical failure at five, ten and 25 years was 15.6% (95%CI 6.6-24.6), 21.9% (95%CI 11.6-32.1) and 28.6% (95%CI 17.2-39.9), respectively. CONCLUSIONS: APC are associated with a considerable risk of both early and late complications. Non-union and infection predominate in the early postoperative period, whereas aseptic loosening and fractures are the main causes of late failure, occurring up to 18 years after surgery. These findings suggest that the routine use of APC for periarticular reconstruction after tumour resection should be reconsidered.