Fixation stability of femoral components in a canine hip replacement model.

A canine hip replacement model was used to compare fixation stability in cemented and cementless femoral components. Parameters of comparison were the load-induced positional changes of each prosthesis relative to its proximal femoral cortex, hereafter called relative displacements. Identical femoral components, with the proximal third of their stem porous-coated, were implanted in the right femurs of 10 large, mixed-breed dogs. Five were tightly fit to allow porous ingrowth, and five were cemented into the medullary canal. Four months after implantation, all femurs were harvested. A prosthesis was implanted in the left (normal) femur of each dog ex vivo with fixation identical to the contralateral limb to simulate acute postoperative fixation. Eddy current transducers measured relative displacements under application of static loads, serially applied in the axial, mediolateral, and craniocaudal directions. Thereafter, the femurs were transversely sectioned and morphologically analyzed to correlate bony apposition at the implant surface with relative displacements. We observed no difference in relative displacements between acute and 4-month-cemented groups (e.g., 0.0059 +/- 0.0021 vs. 0.0060 +/- 0.0048 mm, respectively, for 100-N axial loading measured at midstem). With cementless implantation, relative displacements of the acute group were significantly larger (p = 0.007) than those of the 4-month group (e.g., 0.236 +/- 0.257 vs. 0.097 +/- 0.129 mm, respectively, for 100-N axial loading measured at midstem). Cementless components implanted for 4 months were not significantly different than cemented components, but a trend suggested that they were still not as stable as cemented components, particularly for craniocaudal loads. Relative displacements of the 4-month, porous ingrowth group were approximately proportional to the percentage of bony apposition raised to the -1.44 power (r = 0.94).