gms | German Medical Science

Objectives: Component positioning, particularly the glenoid, is important for the longevity of total shoulder arthroplasty (TSA) and remains the primary source of loosening. The aim of the study was to identify if initial glenoid component malposition affects wear and loosening at two years postoperative. Furthermore, does glenoid component malposition influence clinical and patient-reported outcomes.

Methods: This prospective study included patients who participated in a randomized clinical trial using uncemented trabecular-metal (TM) and cemented polyethylene (PE) glenoid implants. Patients 18-79 years of age with a primary diagnosis of glenohumeral osteoarthritis who underwent a TSA between June 2012 and December 2016 were eligible. Radiographic images were taken preoperatively and two years post-operatively. Glenoid wear, inclination and version were assessed pre-operatively. Post-operative assessment included glenoid component loosening, placement and version, and humeral head position. Clinical assessments included: range of motion (ROM) and the EQ-5D, SF-12, Western Ontario Osteoarthritis Score (WOOS) and American Shoulder and Elbow Surgeon (ASES) Score.Descriptive statistics were calculated for radiological and clinical variables and fit proportional odds logistic regression models were performed to model for loosening.

Results: Ninety-two patients with an average age of 69.9 ± 6.2 years were included in this study. Glenoid component position improved on average little but significantly in version (-19.4 ± 8.6° to -17.7 ± 8.5°; p<0.05) and inclination (11.5 ± 7.1° to 5.9 ± 6.3°; p<0.01) from preoperative to postoperative. Glenoid component offset and humeral head centering in both superior-inferior and anterior-posterior dimensions remained unchanged throughout the follow-up. Loosening according to Lazarus classification was noticed in 21 cases (Lazarus score of 1 in 15 and 2 in 6 cases) at 2 years. None of these cases required revision for loosening. Trends seen with the combined data were also observed when analysis was conducted on TM and PE glenoid subgroups. Clinical assessments (EQ-5d, SF-12, WOOS and ASES) and ROM showed overall continuous improvements from preoperative scores (p < 0.05). Outcome scores did not correlate with loosening at 2 years.

Conclusion: Glenoid component positioning during total shoulder arthroplasty was improved with respect to the inclination compared to the pre-operative state. There was no significant change after initial position for glenoid inclination, version or humeral head centering through a 2-year follow-up. Radiologic loosening (21 out of 92) of the glenoid is not negatively reflected in clinical scores, ROM or humeral head position at the 2-year follow-up.