Abstract
Allograft-prosthesis composite (APC) can restore capsular and ligamentous tissues of the knee sacrificed in a tumor extirpation. We asked if performing APC would restore knee stability and allow the use of nonconstrained arthroplasty while preventing aseptic loosening. We retrospectively compared 50 knee APCs performed with non-constrained revision knee prosthesis (Group 1) with 36 matched APCs performed with a constrained prosthesis (Group 2). In Group 1, the survival rate was 69% at five and 62% at ten years. Sixteen reconstructions were removed due to complications: eight deep infections, three fractures, two instabilities, one aseptic loosening, one local recurrence, and one nonunion. In Group 2, the survival rate was 80% at five and 53% at ten years. Nine reconstructions were removed: 3 due to deep infections, 3 to fractures, and 3 to aseptic loosening. In both groups, we observed more allograft fractures when the prosthetic stem does not bypass the host-donor osteotomy (P > 0.05). Both groups had mainly good or excellent MSTS functional results. Survival rate and functional scores and aseptic loosening were similar in both groups. A rotating-hinge APC is recommended when host-donor soft tissue reconstruction fails to restore knee instability. The use of a short prosthetic stem has a statistical relationship with APC fractures.