Radiographic failure patterns of polished cemented stems.

Changes in cemented femoral component geometry and surface finish have been made to improve predicted results. We studied the failure rates and radiographic failure patterns of four polished femoral stems with different geometries to determine if design changes over time improved our results. We retrospectively reviewed 910 patients with 1031 total hip arthroplasties performed from 1993 to 2003 with a minimum 2-year followup. All arthroplasties used the same cementing technique with polished femoral stems with a surface roughness of 4 microinches but with differing geometries. By 2005, seven stems (0.6%) were revised for aseptic loosening. All loosening started at the cement-bone interface. The time from the onset of loosening to revision averaged 3.3 years. In all but one patient the bone damage was minimal and amenable to simple revision femoral techniques. We found no obvious difference in loosening rates or radiographic failure patterns among the four stem geometries. The patients showed less bone damage and progressed slower than previous reports of failed rough femoral stems.