Five Hundred Fifty-five Retrieved Metal-on-metal Hip Replacements of a Single Design Show a Wide Range of Wear, Surface Features, and Histopathologic Reactions

Background In 2010, a widely used metal-on-metal hip implant design was voluntarily recalled by the manufacturer because of higher than anticipated failure rates at 5 years. Although there was a large published range of revision rates, numerous studies had reported a higher risk of revision for excessive wear and associated adverse tissue reactions when compared with other metal-on-metal total hips. The reasons for this were suggested by some to be related to cup design features.

Questions/purposes From retrievals of ASR metal-on-metal implants and tissue samples obtained at revision surgery, we asked the following questions: (1) What were the common and uncommon surface features? (2) What were the common and uncommon linear and volumetric wear characteristics? (3) Were there common taper corrosion characteristics? (4) What aseptic lymphocytic vasculitis-associated lesion (ALVAL) features were present in the tissues?

Methods Five hundred fifty-five ASRs, including 23 resurfacings, were studied at one academic research center. Features of wear (eg, light and moderate scratching), damage (eg, deposits, gouges), and bone attachment on the porous coating were semiquantitatively ranked from 0 (none) to 3 (> 75%) based on the amount of a feature in each region of interest by the same experienced observer throughout the study. Visible features of head taper corrosion were ranked (Goldberg score) from 1 (none) to 4 (severe) by the same observer using a previously published scoring method. An experienced tribologist measured component wear depth using a coordinate measuring machine and quantified wear volume using previously validated methods. All available tissues were sampled and examined for features of ALVAL and scored from 0 to 10 by a single observer using a method they previously developed and published. A score from 0 to 4 is considered low, 5 to 8 is considered moderate, and 9 or 10 is considered high with regard to the risk of metal hypersensitivity features in the tissues.

Results The most common bearing surface features were light and moderate scratches and removal or postremoval damage. Discoloration and deposits were commonly observed on femoral heads (55% [305 of 553]) and less commonly on cups (30% [165 of 546]). There was no evidence of impingement or dislocation damage. There was typically a small amount of bone attachment in at least one of eight designated regions of interest (84% [460 of 546]); extensive or no bone attachment was uncommon. Edge wear was highly prevalent. The maximum wear of 469 cups (88%) occurred near the edge, whereas the maximum wear of 508 femoral heads (94%) occurred between the pole and 45° from the pole. The median combined head-cup wear volume was 14 mm³(range, 1-636 mm³). One hundred sixty-nine pairs (32%) had a combined wear volume of < 10 mm³, 42 pairs (8%) had volumetric wear of > 100 mm³, and 319 pairs (60%) had wear volume between 10 and 100 mm³. Seventy-four percent of tapers (390 of 530) received a Goldberg score of 4, 22% (116 of 530) a score of 3, < 5% (24 of 530) a score of 2, and none received a score of 1. The most frequent ALVAL score was 5 out of 10 (35 of 144 hips [24%]) and ranged from 2 (one hip) to 10 (nine hips); 92 of 144 (64%) had a moderate score, 17 of 144 (12%) had a high score, and 35 (24%) had a low score.

Conclusions Although edge wear was prevalent, in most cases, this was not associated with high wear. The increased diameter and decreased coverage angle of the ASR design may have resulted in the observed high incidence of edge wear while perhaps decreasing the risk for impingement and dislocation.

Clinical Relevance The role of bearing wear in the revisions of metal-on-metal implants is controversial, because it is known that there is a large range of in vivo wear rates even within the same implant type and that patient variability affects local tissue responses to wear debris. The observations from our study of 555 retrieved ASR implant sets indicate that there was a wide range of wear including a subset with very high wear. The results suggested that the failure of the ASR and ASR XL was multifactorial, and the failure of different subgroups such as those with low wear may be the result of mechanisms other than reaction to wear debris.