Abstract
Stress shielding and load transfer to the femur following total hip arthroplasty have been studied extensively. A number of models have addressed the effects of surface finish of double-tapered, non-collared cemented stems on load transfer to the femur. However, a great number of cemented femoral stem designs in wide use today are not double tapered, and many, such as the Charnley, have collars. The effects of surface finish of such stems on load transfer to the femur are not completely understood. In this study, we measured the effects of surface finish of a straight, non-tapered cemented femoral stem, with and without a collar, in two stem sizes, on load transfer to the femur, using an in vitro laboratory model.
Eight types of straight stems were fabricated, with polished or rough surfaces, with and without a collar, and in two sizes. All stems were based on the same template, and varied only in the desired combination of parameters studied. Three each of the eight unique stem types (total of 24 specimens) were cyclically loaded for 77,000 cycles at 1 Hz, alternating between walking and stair-climbing load profiles. Surface strains were measured at ten locations in each femur during designated initial and final periods.
Of the three design variables, stem surface finish had the greatest effect on femoral surface strains. Specifically, compared to rough stems, with polished stems, mean proximal medial compressive strains were smaller, whereas mean distal medial compressive strains were greater. In contrast, on the anterior surface, mean proximal anterior tensile strains were greater, whereas mean distal anterior strains were smaller. All femoral surface strains increased with cyclic loading, however, strains increased at a greater rate with polished stems than with rough stems. Proximal medial strains were somewhat increased with the presence of a collar, however, these differences were small (< 100 μ∊) and/or not statistically significant. Similarly, distal medial strains were increased with the presence of a collar but, again, the differences were not consistent (p > 0.16). Compared to large stems, with small stems, proximal medial compressive strains were greater. The results emphasize the importance cemented femoral stem surface roughness and the manner in which this changes stem-cement bond strength, affecting the distribution of stresses in the femur. This is an important consideration in the design of femoral stems.