Glenoid and scapulohumeral characteristics and their influence on preoperative range of motion and patient-determined outcomes

Abstract

BackgroundIt is not clear if glenoid and scapulohumeral characteristics influence preoperative range of motion (ROM) and patient-determined outcomes. It is important to understand these interactions when planning and performing total shoulder arthroplasty in efforts of improving patient satisfaction and implant longevity. MethodsA retrospective review of patients that had three-dimensional computed tomography imaging for total shoulder arthroplasty was performed. Patients were separated into 2 groups determined by the presence (rotator cuff tear arthropathy [RCTA]) or absence (osteoarthritis [OA]) of an irreparable rotator cuff tear. Using the computed tomography measurements, shoulders were stratified by glenoid version (anteverted, normal, and retroverted), glenoid inclination (inferior, normal, and superior), and scapulohumeral subluxation (anterior, centered, and posterior) based on criteria determined from a review of the orthopedic literature. The Western Ontario Osteoarthritis Scale and the American Shoulder and Elbow Surgeons scores and ROM were determined preoperatively. ResultsIn OA patients (n = 154), version was associated with scapulohumeral subluxation (P < .0001). Retroverted glenoids had less flexion (96° vs. 108°; P = .049) and external rotation (15° vs. 21°; P = .04) compared with normal version. Inferiorly inclined glenoids had greater posterior subluxation (77%) than those with normal (67%; P = .001) and superior inclination (68%; P = .01). There were no relationships between excessive inclination or subluxation on ROM. In RCTA patients (n = 115), retroverted glenoids had greater superior inclination compared with normal glenoids (12.1° vs. 8.4°; P = .049). Version was associated with scapulohumeral subluxation (anteverted = mean 34% subluxation; normal version = 56.4% subluxation; retroverted = 71.2% subluxation; P < .0001). Retroverted glenoids had less flexion compared with normal version (70° vs. 90°; P = .048), less abduction (62°) than normal glenoids (88°; P = .03) and anteverted glenoids (115°; P = .03), and less abduction/internal rotation (7°) than normal (22°; P = .03) and anteverted glenoids (36°; P = .04). Superiorly inclined glenoids have more posterior subluxation than normally inclined glenoids (64% vs. 56.6%; P = .02). There was no relationship between inclination and ROM. Patients with posterior subluxation had less external rotation compared with those with a centered humeral head (10° vs. 22°; P = .009) and less abduction/internal rotation compared with anterior subluxation (12° vs. 35°; P = .02). There was no relationship between version, inclination, or subluxation with preoperative Western Ontario Osteoarthritis Scale or American Shoulder and Elbow Surgeons in patients with OA (P > .17) or RCTA (P > .31). ConclusionsAn interaction between version, inclination, and scapulohumeral subluxation in patients with OA and RCTA was found. Retroverted glenoids had decreased ROM measurements. RCTA shoulders with posterior scapulohumeral subluxation had decreased ROM. There was no relationship between glenoid and scapulohumeral morphology and patient-determined outcome scores. Level of evidenceLevel III; Retrospective Case-Control Prognosis Study