Purpose of the study: To compare femoral reconstruction using megaprosthesis versus allograft prosthesis composite.
Material and methods: Forty-one consecutive proximal femoral reconstructions with an allograft-prosthesis composite (21 cases) or a megaprosthesis (20 cases) after tumor resection were reviewed in a retrospective study. The following criteria were considered: functional outcome; long term survival; complications. Chi-square test and Wilcox tests were used to compare groups. The medium and long-term survival curves for these reconstructions were made using the Kaplan-Meier standard methods. The failure of prosthesis was defined as revision for mechanical failure (either aseptic loosening or dislocation), for infection or local recurrence. The comparison of the curves was performed using the Log-Rank test.
Results: Infection (10 p. 100) and instability, in both groups, and loosening, in the megaprosthesis group, were the common causes of failure. There was difference between functional results in the two groups (limping and crutches using was more lower in allograft-prosthesis composite group). Survival analysis showed a 5 and 10-year survival of 77 +/- 12 p. 100 for the patients with composites. Five and ten - year survival were 73 +/- 11 p. 100 and 0 p 100 respectevely for those with megaprostheses. No significant difference was noted between survival of these two groups but a tendancy (p =0.09). Radiological allograft resorption was noted for more than 50 p. 100 of allograft composite prosthesis without modification of functional result or symptomatic loosening.
Discussion: The functional results seem better in the composite group when compared to the megaprosthesis group. Reconstruction of the abductor mechanism is essential to stabilize the prosthesis and to decrease the limp. When the great trochanter cannot be preserved, we used suture of gluteus medius tendon to tensor of fascia lata, which is re-enforced using a piece of biceps femoris. The dislocation rate was approximately the same in our two groups. Several authors reported a lower dislocation rate with composite reconstructions than massive prosthesis. The rate of infection is similar to other reported series. In our study it has been possible to show a tendancy for superior survival of the composite reconstruction. When the review was later than 5 years the radiological appearance of the graft in our series was often concerning with resorption or fragmentation present in six of the eight cases. This radiological appearance is not as yet responsible for any revision or any change in the functional result however it does remain a worry.
Conclusion: Composite reconstructions probably allow a better functional result when considering proximal reconstruction of the femur. The radiological appearance of these allografts in the long term is however worry some without any evidence so far of worsening functional level or any evidence of prosthetic loosening. It would seem to us that the current level of knowledge would advocate the use of massive allografts together with prosthesis. This does seem still to remain the best choice for proximal femoral reconstruction.