In the current study, models of pelvises reconstructed with femur and tibia after hindquarter amputation were successfully established, and FE analysis was used to examine their biomechanical properties. We analyzed the displacement of intact and reconstructed pelvises under a physiological load of 0 to 500 N, and the results demonstrated that the reconstructed pelvises had excellent stability. This implies that patients would be able to sit and walk with a prosthesis shortly after surgery if this method of reconstruction is used after hindquarter amputation. The lesser displacement of the intact pelvis under 0 to 500 N indicated that its stability was better compared with that of the reconstructed pelvises, suggesting that with this reconstruction method, full weight bearing should be avoided until bone fusion has occurred. The use of pedicled autografts for the reconstruction of pelvic defects reportedly leads to a more rapid bone union[23, 24].
There are many methods that have been used to reconstruct the pelvis after hemipelvectomy that include allografting and prostheses[25–28]. Regardless of the method, restoration of pelvic ring continuity increases the stability of skeletal support for weight bearing in both sitting and standing positions and advances in reconstructive surgery have led to the ability to reconstruct pelvic ring defects. Few studies, however, have provided a detailed analysis of the stresses resulting from reconstruction of the pelvic ring. Ji et al. reported a FE analysis of the reconstruction of type II + III pelvic resection with a modular hemipelvic endoprosthesis. The authors found no difference in stresses along the bilateral arcuate lines between a reconstructed pelvis and normal pelvis, but found that the stress distribution on the prosthesis along the sciatic notch was significantly greater than on the unaffected side, and that the posterior side column between the point of iliac fixation and the acetabulum was subjected to the greatest stress.
Different stress distributions were found between the intact and reconstructed pelvises in this study. An unbalanced stress distribution was found in the resected pelvis, and the higher stress distribution noted in the reconstructed pelvis was related to the stress block of the two lag screws. FE modeling showed that maximum von Mises stresses in the reconstructed pelvis were in the area of the connection of the femur and sacrum, left end of the screws, and right SIJ. These stresses could lead to degeneration of the SIJ on the normal side, development of scoliosis, and a contralateral pelvic tilt. The use of larger diameter screws (7.3 or 7.5 mm) would prevent too much stress at the SIJ on the normal side. The stress concentration on the left end of the screws was related to their small contact surface with the femur, and larger screws would increase the interface with the SIJ, sacrum, and femur, thus decreasing the maximum stress on the SIJ on the normal side.
The stress distribution in the reconstructed bone was different for the two reconstructed pelvic models. In the femoral model, the von Mises stress was 13.9 MPa, whereas in the tibial model, the maximum stress at the tibia was only 6.41 MPa, significantly less than that in the femoral model and less than 80 MPa, the yield stress of cortical bone. It should be noted that in the tibial reconstructed model, the von Mises stress was concentrated on the tibial shaft, which could lead to partial or complete fracture of the tibia and reconstruction failure. From a biomechanical point of view, distal femur reconstruction of the pelvis is suitable after hindquarter amputation.
This study had several limitations. Only an axial compressive load was applied in this study, but in vivo forces applied to the pelvis are more complex. Further mechanical evaluation of the reconstructed pelvis regarding flexion, extension, bending, and rotation are required. Moreover, the FE models of the pelvis in the current study also need further modifications. In the current models, the SIJ was regarded as fused, which is an over simplification of the real SIJ and could lead to some deviations in the results. In addition, the model would be more accurate with the addition of elastic tissues such as ligaments and muscles.