Cemented vs. uncemented glenoid fixation in total shoulder arthroplasty for osteoarthritis: a New Zealand Joint Registry study.

BACKGROUND: Total shoulder arthroplasty (TSA) is commonly performed for shoulder osteoarthritis (OA). Uncemented metal-backed (MB) glenoid components were introduced in an attempt to avoid glenoid loosening. New Zealand and Australian Joint Registry studies have shown significantly higher revision rates when uncemented MB glenoids are used. We used the New Zealand Joint Registry (NZJR) to compare all-cause revision rates and functional scores for TSA and investigated the trends of glenoid fixation used in New Zealand. METHODS: The NZJR was accessed for all primary TSA undertaken for OA from January 2000 to December 2017. Patient demographics were collated. All-cause revision rates were reported as rate per 100 component-years. Analysis was repeated excluding the uncemented SMR L2 glenoid (LimaCorporate), as it was a potential confounder. Kaplan-Meier survival analysis was performed. Oxford Shoulder Scores at both 6 months and 5 years were analyzed. RESULTS: A total of 2613 TSAs were performed for OA during the study period, representing 85.0% of all TSAs in New Zealand. Overall, 62.1% of the patients were female. In addition, 69.6% of glenoids were cemented and 30.4% uncemented. The most common uncemented MB glenoid was SMR 86.6% (LimaCorporate), and cemented was Global (DePuy) 49.8%. The revision rate for TSA with uncemented glenoids was significantly higher at 2.03 compared with cemented at 0.41 per 100 component-years (P < .001). Hazard ratio 5.0 for revision of uncemented glenoids. No significant difference was found in Oxford Scores at 6 months (39.7 vs. 40.3, P = .13) or 5 years (42.1 vs. 42.8, P = .22). The most common mode of failure was glenoid loosening in cemented glenoids (44.4%), and component failure in uncemented (34.8%). Revision for rotator cuff, deep infection, and instability were comparable between groups. When excluding SMR L2, uncemented MB glenoid all-cause revision rates remained significantly higher than cemented (1.42 vs. 0.41 per 100 component-years, P < .001). SMR L1 uncemented MB glenoids had a higher revision rate than the non-SMR uncemented glenoids (1.61 vs. 0.18 per 100 component-years, P = .009). Uncemented glenoid use peaked in New Zealand in 2011 at 46.7% of TSAs but declined to 20.1% in 2017. CONCLUSIONS: In the NZJR, primary TSAs undertaken for OA have a significantly higher all-cause revision rate when the glenoid component is uncemented. Uncemented glenoids have a 5.0 times higher revision rate. Excluding SMR L2 glenoids from the analysis, the significantly higher revision rate remained for uncemented glenoids. These data reaffirm that uncemented MB glenoids are associated with higher revision rates.