Midterm radiographic outcomes of anatomic total shoulder arthroplasty in biplanar glenoid deformities

Background Optimal management of retroversion in anatomic total shoulder arthroplasty (aTSA) remains controversial, and limited attention has been directed to the impact of glenoid inclination. Prior biomechanical study suggests that residual glenoid inclination generates shear stresses that may lead to early glenoid loosening. Combined biplanar glenoid deformities may complicate anatomic glenoid reconstruction and affect outcomes. The goal of this matched-cohort analysis was to assess the relationship between biplanar deformities and midterm radiographic loosening in aTSA. Methods The study cohort was identified via an institutional repository of 337 preoperative CT scans from 2010 to 2017. Glenoid retroversion, inclination, and humeral head subluxation were assessed via 3D-planning software. Patients with retroversion >= 20 degrees and inclination >= 10 degrees who underwent aTSA with eccentric reaming and nonaugmented components were matched by age, sex, retroversion, and Walch classification to patients with retroversion >= 20 degrees only. Primary outcome was glenoid component Lazarus radiolucency score. Results Twenty-eight study subjects were matched to 28 controls with retroversion only. No difference in age (61.3 vs. 63.6 years, P = .26), sex (19 [68%] vs. 19 [68%] male, P > .99), or follow-up (6.1 vs. 6.4 years, P = .59). Biplanar deformities had greater inclination (14.5 degrees vs. 5.3 degrees, P < .001), retroversion (30.0 degrees vs. 25.6 degrees, P = .01), and humeral subluxation (86.3% vs. 82.1%, P = .03). Biplanar patients had greater postoperative implant superior inclination (5.9 degrees [4.6 degrees] vs. 3.0 degrees [3.6 degrees], P = .01) but similar rate of complete seating 24 [86%] vs. 24 [86%] P > .99). At final follow-up, biplanar subjects had higher Lazarus radiolucent scores (2.4 [1.7] vs. 1.6 [1.1], P = .03) and a higher proportion of patients with glenoid radiolucency (19 [68%] vs. 11 [39%], P = .03). No difference in complete component seating (86% vs. 86%, P = .47) or initial radiolucency grade (0.21 vs. 0.29, P = .55) on immediate postop radiographs. Biplanar patients demonstrated a greater amount of posterior subluxation at immediate postoperative (3.5% [1.3%] vs. 1.8% [0.6%]; P = .03) and final follow-up (7.6% [2.8%] vs. 4.0% [1.8%]; P = .04). At final radiographic follow-up, biplanar subjects had higher Lazarus radiolucent scores (2.4 [1.7] vs. 1.6 [1.1], P = .03; ICC = 0.82). Bivariate regression analysis demonstrated that biplanar deformity was the only significant predictor (OR 3.3, P = .04) of glenoid radiolucency. Discussion Biplanar glenoid deformity resulted in time-zero glenoid implant superior inclination and increased midterm radiographic loosening and posterior subluxation. Attention to glenoid inclination is important for successful anatomic glenoid reconstruction. Future research is warranted to understand the long-term implications of these findings and impact of using augmented implants or reverse shoulder arthroplasty to manage biplanar deformities. Level of evidence Level III Retrospective Matched Cohort Comparison Prognosis Study