MoM bearing remains controversial in RHA and has been largely abandoned in THA due to early catastrophic failures such as ARMD associated with potential major bone loss and severe damages to the soft tissue at the time of revision [
24,
25]. Conversely, RHA is still performed in some centers and reported with excellent functional outcome and survivorship up to 99.7% at 10 years [
16,
26]. However, series that compared survivorship of MoM RHA and large head THA with consistent long-term follow-up are lacking. Therefore, this study aimed to identify and compare the mode of failure and survivorship of RHA and THA at a minimum ten year follow-up with a particular attention to specific complications related to MoM bearing. The most important finding of the present study was that survivorship of RHA was significantly higher than survivorship of large head MoM THA at five, ten and 15 years after surgery. Importantly, this difference in survivorship was explained by a 2.93-fold increase in failure rate due to ARMD in THA. Indeed, ion levels and dissociation of Co/Cr ratio were significantly higher in THA. Therefore, these results emphasize the potential role of trunnionosis as the main mechanism of failure in THA and not only the MoM bearing by itself.
The overall 15-year survivorship of 83% in RHA and 73% in THA observed in the current study was in agreement with the previous series [
15,
27‐
35]. Indeed, Ng et al. and Althuizen et al. reported a high failure rate for Durom® MoM THA with a ten year revision rate ranging from 14 to 31% [
36,
37]. In our series, Co level and Co/Cr ratio were significantly higher in THA compared to RHA. These findings suggest an additional source of ion production in THA other than the head-cup bearing interface [
4,
38,
39]. A similar result was described by Ridon et al. and Johnson et al. in series comparing RHA and THA constructs performed with the same acetabular component [
15,
40]. In addition, Garbuz et al. showed that patients with a large head MoM THA presented with a 46-fold and tenfold increase in Co and Cr levels respectively compared to RHA [
18]. Goldberg et al. demonstrated that Cr release remains localized around the taper of the femoral stem, while Co is released into the blood, leading to a higher blood level of Co and dissociation of the Co/Cr ratio in THA [
38]. Therefore, this Co/Cr ratio dissociation could be supposed to be a direct consequence of fretting corrosion (i.e., trunnionosis) at the head-neck junction. Moreover, trunnionosis was also described as the result of a local interplay between the head/taper engagement levels, and the horizontal lever arm and load offset applied on the trunnion [
41‐
43]. Consequently, the contribution of trunnionosis to metal ion production and ARMD could be considered as a natural phenomenon in large head MoM THA that is not influenced by implant positioning or edge loading at the MoM bearing level [
43]. Therefore, we believe that ion production at the Morse taper interface could explain the significantly higher incidence of ARMD and lower implant survivorship in the THA group of the current study. In addition, the metal debris production due to trunnionosis might increase third body wear at the MoM bearing interface and therefore increase the risk of ARMD in THA [
43]. By definition, RHA is not affected by potential trunnionosis. This could explain the significantly lower ion levels and ARMD rate, and higher survivorship compared to THA we observed in the current long-term follow-up study. Taking altogether, these results suggest that RHA may be a valid bone preserving option in carefully selected patients [
16]. Our study presented with some limitations. First, this total joint registry study was observational and focused on revision rate and survivorship. This study did not aim to evaluate functional or radiological outcomes. Second, the indication for a RHA or THA procedure was at the senior surgeon’s discretion, with RHA mainly performed in younger and more active patients presenting with primary hip osteoarthritis. Therefore, the two groups were not matched for age, sex, patient’s functional demand, or indication. Third, no analysis of the implants was performed after RHA or THA revision. Especially, no trunnion analysis such as tribo-corrosion was performed on the explanted femoral stem neck and head. Only gross macroscopic assessment of the trunnion was mentioned in operative reports.