A multicenter, randomized controlled trial comparing a second-generation uncemented trabecular metal-backed vs. cemented polyethylene glenoid component in total shoulder arthroplasty: 5-year results

Background: Previously, we reported early (2-year) findings from a randomized controlled trial comparing a second-generation uncemented trabecular metal-backed (TM) glenoid vs. cemented polyethylene glenoid (POLY) in patients undergoing a total shoulder arthroplasty. The purpose of the current study is to report disease-specific quality of life, clinical, patient-reported, and radiographic outcomes at midterm (5-year) from this trial. Methods: Five surgeons from 3 centers participated. Patients 18-79 years with a primary diagnosis of glenohumeral osteoarthritis were screened for eligibility. Randomization to an uncemented TM or cemented POLY glenoid was performed intra-operatively after adequate bone stock was confirmed. Study intervals were baseline, 2- and 5-year postoperative. The primary outcome was the Western Ontario Osteoarthritis Shoulder (WOOS) quality of life score. Secondary outcomes included the American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form, EuroQol-5 Dimensions, and 12-Item Short Form Health Survey scores and clinical and radiographic examinations. Radiographic images were reviewed for metal debris according to Endrizzi. Mixed effects repeated measures analysis of variance for within- and between-group comparisons were performed. Results: Of the 104 patients who consented, 93 were randomized (46 TM; 47 POLY). There were no differences between groups at baseline (TM: mean age 66.5 years [standard deviation (SD) 6.4], 24 male and 22 female; and POLY: mean age 68.4 years [SD 5.5], 23 male / 24 female). Mean (SD) WOOS scores at baseline and 2 and 5 years were as follows: TM, 32 (21), 92 (13), and 93 (11); POLY, 27 (15), 93 (11), and 93 (10), respectively. No statistical or clinically relevant differences were noted with patientreported outcomes between groups. Metal debris was observed in 11 (23.9%) patients, but outcomes were not negatively impacted, and debris severity was minor (grades 1 and 2). Complication rates were similar between groups (TM: 7 of 46 [15.2%], and POLY: 8 of 47 [17.0%]; P 1 / 4 .813). No aseptic glenoid failures were reported, but 1 patient in the TM group required revision because of infection. Conclusions: Our short -term (2-year) findings were maintained with longer follow-up. At 5 years postoperation, there were no statistically or clinically significant differences between an uncemented second-generation TM glenoid and a cemented polyethylene glenoid with respect to disease-specific quality of life or patient-reported outcomes. No glenoid implant failures were reported, and complication