BACKGROUND: A single-center study of 144 THAs revised specifically for periprosthetic joint infection (PJI) observed that trabecular metal (TM) acetabular components had a reduced risk of rerevision for subsequent infection compared with non-TM implants. It was suggested that TM was protective against infection after revision and that TM may be useful when revising THAs for PJI. Three registry studies have subsequently assessed the effect of TM on future infection. In the National Joint Registry (NJR) for England and Wales, we earlier reported lower revision rates for infection when TM (versus non-TM) was used in primary THA, but no difference in rerevision rates for infection when TM was used for all-cause revision THAs. The latter findings in all-cause revisions were also confirmed in a study from the Swedish and Australian registries. It is possible that TM only reduces the risk of infection when it is specifically used for PJI revisions (rather than all-causes). However, to date, the registry analyses have not had large enough cohorts of such cases to assess this meaningfully. QUESTIONS/PURPOSES: (1) In revision THAs performed for PJI, are rerevision rates for all-cause acetabular indications lower with TM acetabular components compared with non-TM designs? (2) In revision THAs performed for PJI, are rerevision rates of any component for infection lower with TM acetabular components compared with non-TM designs? METHODS: A retrospective observational study was performed using NJR data from England and Wales, which is the world's largest arthroplasty registry and contains details of over two million joint replacement procedures. The registry achieves high levels of patient consent (92%) and linked procedures (ability to link serial procedures performed on the same patient and hip; 94%). Furthermore, recent validation studies have demonstrated that when revision procedures have been captured within the NJR, the data completion and accuracy were excellent. Of 11,988 revisions performed for all causes, 794 were performed for PJI in which the same cementless acetabular component produced by one manufacturer was used. Acetabular components were either TM (n = 541) or non-TM (n = 253). At baseline the two groups were comparable for sex, age, body mass index, and American Society of Anesthesiologists (ASA) grade. Outcomes after revision THA (rerevision for all-cause acetabular indications and rerevision of any component for infection) were compared between the groups using Fine and Gray regression analysis, which considers the competing mortality risk. Regression models were adjusted for the propensity score, with this score summarizing many of the potential patient and surgical confounding factors (age, sex, ASA grade, surgeon grade, approach, and type of revision procedure performed). RESULTS: There was no difference in 5-year cumulative acetabular component survival rates between TM (96.3%; 95% confidence interval [CI], 94%-98%) and non-TM components (94.4%, 95% CI, 90%-97%; subhazard ratio, 0.78, 95% CI, 0.37-1.65; p = 0.509). There was no difference in 5-year cumulative implant survival rates free from infection between TM (94.8%; 95% CI, 92%-97%) and non-TM components (94.4%, 95% CI, 90%-97%; subhazard ratio, 0.97, 95% CI, 0.48-1.96; p = 0.942). CONCLUSIONS: We found no evidence to support the notion that TM acetabular components used for PJI revisions reduced the subsequent risk of all-cause rerevision or the risk of rerevision for infection compared with non-TM implants from the same manufacturer. We therefore advise caution against recent claims that TM components may protect against infection. LEVEL OF EVIDENCE: Level III, therapeutic study.
Clin orthop relat res
1382 - 1389