Background: The use of a megaprosthesis has become the method of choice for reconstruction after bone tumor resection at the knee. However, the long-term survival of megaprostheses is poor. In this study, we sought to identify factors that were associated with implant failure and amenable to interventions designed to improve implant survival.
Methods: A retrospective review of the charts of ninety-one patients who had undergone resection of a tumor of the knee followed by reconstruction with a custom-made megaprosthesis was performed. The distal part of the femur was resected in fifty-six patients and the proximal part of the tibia, in thirty-five patients. The reconstruction was performed with an allograft-prosthesis composite in thirty-three patients and with metal or plastic sleeves in fifty-eight patients. Reconstruction of the extensor mechanism was necessary in all thirty-five patients with a tibial tumor.
Results: The median duration of follow-up was sixty-two months. The extensor mechanism was significantly less likely to rupture when partial continuity had been preserved at the time of the resection. Intra-axial laxity (an arc of motion of >5 degrees in the frontal plane) was significantly more common when the prosthesis had an antirotation pin than when it did not have an antirotation pin (p = 0.0023). There was mechanical failure of ten allograft-prosthesis composites and ten sleeve reconstructions. Thirty-six patients had removal of at least one component of the prosthesis. When revision due to local tumor recurrence was excluded, the median duration of prosthetic survival was 130 months following the distal femoral resections and 117 months following the proximal tibial resections. The median duration of survival was 117 months for the allograft-prosthesis composites and 138 months for the sleeve reconstructions. Body weight and activity level were independent predictors of early revision.
Conclusions: The long-term survival of the knee megaprostheses in this study was poor. Mechanical failure was multifactorial and the leading cause of revision. Use of allograft-prosthesis composites and use of bushings or an antirotation pin appeared to have no mechanical benefits. We recommend that weight control programs and advice about adapting their activity level be offered to patients preoperatively.