Effects of prosthesis surface roughness on the failure process of cemented hip implants after stem-cement debonding

Retrieval studies suggest that the loosening process of the cemented femoral components of total hip arthroplasties is initiated by failure of the bond between the prosthesis and the cement mantle. Finite element (FE) analyses have demonstrated that stem-cement debonding has stress-producing effects in the cement mantle. High interface friction, which corresponds to a degree of surface roughness, reduces these stresses. In experiments, however, debonded rough stems produced more cement damage than polished ones; in the Swedish Hip Register polished stems were clinically superior with respect to stems with a mat surface finish. The purpose of the present study was to investigate this contradiction. For this purpose, global and local FE models with debonded stem-cement interfaces were used to study the effects of prosthesis surface roughness on the cement stresses on a global scale and microscale, respectively. Similar to earlier numerical studies, the global FE model predicted that an increased surface roughness of the stem reduced the stresses in the cement mantle. The local model provided insight in the load-transfer mechanism on a microscale and could explain the experimental and clinical findings. The local cement peak stresses around the asperities of the surface roughness profile increased with increasing surface roughness and decreased again beyond a particular roughness value. Cement abrasion is caused by localized stresses in combination with micromotion. From this study it can be concluded that to minimize cement abrasion, debonded stems should either have a polished microstructure to minimize the local cement stresses or have a profiled macrostructure to minimize micromotions at the stem-cement interface.