Catastrophic failure of femoral stem modular junction when combined with metal-on-metal bearing in comparison to ceramic-on-ceramic: A retrospective cohort study

Background: Metal ion release from total hip arthroplasty's (THA) metal-on-metal (MoM) bearing surfaces or head-neck modular junction (trunnionosis) has been identified as a major cause of adverse reaction to metal debris (ARMD). No study has compared the effects of these bearing couples when combined with modular neck femoral stems (MNFS) (i.e. did a modular CoCr have a higher effect than large MoM bearings on whole blood ion concentrations and ARMD). Therefore we did a retrospective comparative analysis of prospectively collected data aiming to: (1) assess the difference in metal ion release between a group of MoM and CoC bearings implanted with a non-cemented MNFS; (2) compare the ARMD and the related revision rates between the two bearing types; (3) compare the patients' HOOS and Harris Hip scores. Hypothesis: Metal ion levels and complications rate will be higher in MoM group. Methods: Thirty hips received the same short MNFS with modular CoCr. Seventeen had CoC bearing (16 = 36 mm) and 13 large diameter head (LDH) MoM bearing (mean = 50 mm, 44-54 mm) Both groups had whole blood Cobalt (Co) and Chromium (Cr) concentrations measurements at last follow up or before revision surgery (if any) and were compared to 41 control individuals without implants. ARMD, revision rates and other complications were tensed. Functional outcome differences between operative groups were assessed with HOOS and Harris Hip scores. Results: Mean whole blood Co and Cr were 1.9 mu g/L (0.1-10.6 min-max) and 1.2 mu g/L (0.2-4.6 min-max) in the CoC group and 12.7 mu g/L (2.5-40.7 min-max) and 8.2 mu g/L (2.8-40.6 min-max) for MoM group. Controls presented whole blood Co and Cr levels of 0.2 mu g/L (0.04-0.2 min-max) and 0.6 mu g/L (0.1-0.7 min-max) respectively. CoC had significantly higher whole blood metal ion concentrations than controls, while MoM had significantly higher concentrations than both (p < 0.001). MoM had a 9/13 (69%) revision rate due to ARMD (with stem-neck junction corrosion and wear were present in all cases) while the CoC had none. Functional scores at last follow up did not statistically differ between groups (Harris: CoC = 94.1 +/- 8.4 vs MoM 91.8 +/- 14.8 (p = 0.22) and CoC HOOS 90.2 +/- 10.7 vs 75.6 +/- 21.7 (p = 0.08)). Conclusion: In our study, MNFS with CoCr modular neck released metal ion systemically, as shown in CoC THAs, but when combined with MoM LDH bearings, that elevation was significantly higher. In MNFS MoM, these high metal ion levels translated into a dramatic ARMD related revision rate where stem-neck junction corrosion and wear were present in all cases. More research is required to understand the effects of MoM bearing on neck-stem junctions, and its mechanisms of corrosion. (C) 2020 Elsevier Masson SAS. All rights reserved.