Clinical outcomes related to glenosphere overhang in reverse shoulder arthroplasty using a lateralized humeral design

Previous studies have demonstrated that decreased impingement-free range of motion (ROM) can adversely influence clinical outcomes following reverse shoulder arthroplasty (RSA). Inferior placement of the glenosphere is thought to minimize impingement and its associated sequelae. This study evaluated the relationship between inferior overhang of the glenosphere and clinical outcomes in patients undergoing primary RSA using a lateralized humeral implant design.

By use of a prospectively collected shoulder arthroplasty database, all primary RSAs performed at our institution between 2007 and 2015 with a single implant design (lateralized humerus and medialized glenoid) and minimum 2-year follow-up were evaluated. Glenosphere overhang in relation to the inferior rim of the glenoid was measured in millimeters on postoperative Grashey radiographs of the shoulder and categorized into tertiles (low, <7.1 mm; medium, 7.1 to 9.9 mm; and high, >9.9 mm). Clinical outcomes of interest comprised the changes between preoperative and postoperative values in the following ROM and outcome score measures: active forward elevation (aFE), active external rotation, American Shoulder and Elbow Surgeons score, Constant-Murley score, Shoulder Pain and Disability Index score, and Simple Shoulder Test score. Random-effects linear models were used to assess univariate and multivariable associations between overhang tertile and change in patient outcomes. Differences in outcomes were further compared using the minimal clinically important difference (MCID).

The study identified 284 shoulders in 265 patients. The median follow-up period was 36 months (range, 24-108 months). The median glenosphere inferior overhang was 8.4 mm, with an interquartile range of 6.3-10.6 mm. Plots demonstrated nonlinear relationships between overhang and outcome scores and between overhang and ROM. Patients with high overhang experienced a significantly greater improvement in aFE compared with patients with low overhang (P = .019), which exceeded the MCID. No other differences in ROM and outcome scores between overhang groups exceeded the MCID. For other outcome scores and ROM measurements, there was no significant relationship with glenosphere overhang. Increased overhang was associated with a significantly lower incidence of scapular notching (P = .005).

Patients undergoing RSA using a lateralized humerus design with greater inferior overhang of the glenosphere demonstrated a significantly greater improvement in aFE and lower rate of notching compared with those with low overhang. No ideal glenosphere overhang range was identified to maximize function in this study.