[Complications from the hinged GUEPAR prosthesis after resection of knee tumor. 102 cases]

The purpose of this study was to present the complications which occurred in knee replacement with the GUEPAR prosthesis, after bone tumor resection. We tried to point out complications in relation to prosthetic design and surgical technique.

Between 1972 and 1993, 90 patients had a knee resection, for 80 malignant and 10 benign tumors. There were 51 males and 49 females, aged 12 to 75 years (mean age 35). Fifty-six distal femur resections and 34 proximal tibial resections were performed. Length of resection averaged 16 cm (9 to 30). The reconstruction was always achieved with a cemented, custom-made GUEPAR prosthesis. Including revisions, there was a total of 102 prostheses in 90 patients. A patellar resurfacing was performed in 64 cases. An allograft reconstruction was associated in 39 knees. In all tibial resections and in two extra-articular femoral resections, the extensor mechanism had to be reconstructed. Several reconstruction techniques were associated, in which 19 medial gastrocnemius transfers. The prosthesis design was slightly modified with time. Ten patients received radiotherapy, and 55 had chemotherapy.

Results and complications were retrospectively assessed, with an average follow-up of 4.3 years (1 to 22). Six patients were lost for follow-up, 62 patients were alive, with no evolutive disease, 13 had an evolutive disease and 10 were deceased from disease. Nineteen patients had distant metastasis and 17 had local recurrences. Apart from intraoperative complications, late mechanical complications included: 13 aseptic loosening, 2 femoral shaft fractures, 18 knee contractures, 5 femoral stem fractures and, 18 intra-articular instabilities related to wear of the hinge-axis. In the 39 allograft-composite prostheses, only 15 had a favorable evolution. There were 15 extensor mechanism failures and 13 knees had persistent infection. There was a total of 94 reoperations in the 90 patients. In 28 cases, the initially implanted prosthesis was removed. There were also 18 revisions, 7 amputations and 3 arthrodeses. Survivorship analysis showed a 60 per cent probability for the initial prosthesis not to be revised at 10 year-follow-up, apart from oncologic complications.

Results with allograft-composite reconstruction were not better than with massive prosthesis. When needed, soft tissue coverage and patellar tendon augmentation would better be performed with gastrocnemius plasty. Polyethylene and steel bushes were not solution for hinge axis wear.