In vitro comparison of the effects of rough and polished stem surface finish on pressure generation in cemented hip arthroplasty

Abstract

Background and purpose: High pressures around implants can cause bone lysis and loosening. We investigated how pressures are generated around cemented femoral stems.

Method: We compared the pressures generated by rough and polished tapered stems at their cement interfaces, in an in vitro model, before and after 1 million load cycles.

Results: At the start of the study, the loading of both polished and rough stems generated interface pressures that were not statistically significantly different. After 1 million load cycles, the rough stems generated greater interface pressures than at the start (p = 0.03), with maximum pressure wave amplitudes of 450,000 Pa or 3,375 mm Hg. The pressures generated by polished stems were similar before and after 1 million load cycles, and were lower than the pressures generated by the rough stems (p = 0.01). Stem loading caused micromotion between the stem and cement. Polished stems migrated distally in the cement but retained rotational and axial stability. The rough stems also migrated distally and wore the cement mantle, leading to increased rotational instability.

Interpretation: The change in the rotational micromotion of the rough stem is likely to be the principal cause of the increased stem pump output and to be a key factor in the longevity of cemented femoral implants.

Figure 1.

Cross-sectional view of HAPS apparatus with model stem in situ and showing sampling sites at the cement mantle-stem interface.

Figure 2.

Typical pressure waves generated by a polished stem in a conforming cement mantle at 2 different sites, together with the stem loading schedule. The ΔPmax is the maximum pressure range developed over 1 load cycle (measured in pascals). The typical pressure wave is either dominated by a sine wave (A) or a bi-modal spiked waveform (B). The peaks of the sine wave (arrows 1 and 2) tend not to be synchronized with the peak of the applied axial load; however, the peaks in the bi-modal spiked waveform are synchronised with the torque loading.

Figure 3.

Comparison of the pressure waves of one load cycle generated by polished and rough stems taken at 5 sampling sites at the start of loading (dashed lines) and after 1 million load cycles (solid lines).

Figure 4.

Average of median ΔPmax recorded at the start of the study and after completing 1 million cycles with both polished (A) and rough (B) stems (3 mantles were studied in each group).