Short-term results in total hip arthroplasty (THA) with large-diameter metal-on-metal (MoM) bearings were encouraging, but high failure rates have been reported in the long term, notably implicating corrosion due to modularity. Several studies compared resurfacing (to which modularity does not apply) versus large-diameter MoM THA; but, to our knowledge, none compared the same bearing in the two situations with more than 10 years’ follow-up. We therefore conducted a retrospective case-control study, using a single cup model (Durom™, Zimmer, Warsaw, USA) for both resurfacing (R) and large-diameter THA, to determine the role of modularity in failure of large-diameter MoM bearings. The study compared (1) metallic ion levels, and (2) survival, functional scores and complications rates between R and THA.

Large-diameter MoM bearing failure implicates not bearing wear but head-neck junction modularity in larger-diameter MoM THA.

Eighty-three THAs and 90 Rs were included between February 2004 and March 2006. All patients had clinical and radiologic follow-up with chromium (Cr) and cobalt (Co) ion blood assay.

In the THA group, 24 of the 83 patients (28.9%) underwent revision for adverse reaction to metal debris (ARMD), versus none in the R group. Ten-year all-cause survival was significantly better in R (97.7%; 95% CI, 96.2–99.2) than THA (67.1%; 95% CI, 60.9–73.3). Median blood ion level was higher in THA (with a difference between Co and Cr: 5.75μg/L (range, 3.82–19.2) versus 1.75μg/L (range, 1.34–2.94) respectively) than in R (no difference: 0.89μg/L (range, 0.67–2.89) and 1.07μg/L (range, 0.67–1.65) respectively). In the THA group, there were positive correlations between Co and Cr elevation and implant revision (both p<0.0001). Co/Cr ratio was significantly higher in THA (2.57) than R (0.88) (p<0.0001), and higher again in the 24 cases of THA revision (4.67). There was no significant difference in mean PMA score (THA: 17.08±1.82 (range, 7–18); R: 17.50±0.74 (range, 15–18)), whereas mean Oxford score was better in R (14.32±2.5 (range, 12–24)) than THA (18.17±8.05 (range, 12–42)) (p=0.02).

The present study confirmed the incontrovertible implication of modularity in failure of large-diameter MoM THA, by analyzing the same bearing in THA and in resurfacing. Trunnionosis was observed in the 24 cases of revision, with the THA adaptation ring inducing serious metallic ion release (with dissociated Co/Cr ratio), accounting for the high rate of revision.

III, case-control study.