Fixation of femoral allograft/prosthesis composites after 25%, 50% and 75% resection.

The relative linear and angular displacements of proximal femoral reconstructions were compared within six different replacement techniques during ex vivo axial compression, mediolateral bending, and axial torsion in dogs. Each femur was osteotomized at 25%, 50%, or 75% of its length and the proximal portion subsequently replaced using one of six techniques. The reconstruction techniques included various combinations of proximal and distal fixation methods (graft fixation/distal fixation): (1) an allograft/prosthesis composite (APC) press-fit proximally and cemented distally (press-fit/cement); (2) APC cemented proximally and distally (cement/cement); (3) APC cemented proximally and the host bone/graft interface double plated (cement/plates); (4) APC cemented proximally and secured distally with bicortical screws (cement/screws); (5) APC secured proximally and distally with bicortical screws (screws/screws); (6) Segmental proximal femoral replacement cemented into the distal femur without an allograft (no graft/cement). For axial compression and mediolateral bending, the combined resection lengths revealed no differences in linear and angular displacements, respectively, between reconstruction methods. During axial torsion, the cement/cement technique allowed larger angular displacements than all but the press-fit/cement technique which had larger displacements than the cement/screws, screws/screws, and no graft/cement groups (p < 0.0001). Overall, the measured implant stability was solid and consistent as evidenced by small amounts of relative displacement and small error values.