Early radiographic and clinical outcomes of primary short stem anatomic total shoulder arthroplasty with a peripherally enhanced fixation glenoid: a multicenter study

Background

Glenoid component loosening remains the most common reason for revision of anatomic total shoulder arthroplasty (aTSA). We assessed early clinical and radiographic outcomes following aTSA using a press-fit short stem and a peripherally enhanced fixation glenoid.

Methods

275 consecutive patients with end-stage glenohumeral arthritis and Walch A- or B-type glenoid morphology who underwent primary aTSA in 2017-2018 at two high-volume shoulder arthroplasty institutions were evaluated, and patient-reported outcomes (PROMs) and radiographic findings were studied in those with completed baseline and minimum 2-year follow-up, respectively. Patient demographics, glenoid morphology, body mass index (BMI), Charlson Comorbidity Index (CCI), range of motion, American Shoulder and Elbow Surgeons (ASES) score, and Simple Assessment Numeric Evaluation score were collected. Radiographic analysis of glenoid and humeral components was performed. Multivariable logistic, equal adjacent odds ordinal, and beta regression were respectively used to identify predictors of glenoid radiolucent lines, humeral calcar resorption, and total ASES score.

Results

Patients were 43% female, with a mean age of 66, a median BMI of 30, and median follow-up of 28.4 months. ASES and Simple Assessment Numeric Evaluation scores improved by respective medians of 54.4 and 55.0 points, forward elevation by median 35°, and external rotation by median 30° (all P < .001 for preoperative to postoperative change). Postoperative radiographs of 177 cases showed 10 (5.7%) glenoid osteolysis, 51 (28.8%) glenoid radiolucent lines, and 81 (45.8%) calcar resorptions. The follow-up duration (median 40.1 vs. 27.2 months; P < .001), BMI (median 27.5 vs. 30.7; P < .001), and Charlson Comorbidity Index (Q3 0 vs. 1; P = .02) were associated with glenoid osteolysis in bivariate analyses. In multiple logistic regression, surgeon (C vs. A/B) was the only statistically significant predictor of glenoid radiolucent lines [OR 0.27, 95% CI (0.1, 0.8)]. By descending importance, Surgeon C [OR 6.5 (2.0, 20.5)], humeral canal filling ratio [upper vs. lower quartile OR 2.3 (1.3, 4.0)], mediolateral humeral head deviation [upper vs. lower quartile OR 1.9 (1.0, 3.5)], and glenoid osteolysis [OR 13.5 (2.6, 71.6)] significantly predicted greater calcar resorption. Longer follow-up duration marginally statistically significantly predicted lower ASES score [upper vs. lower quartile OR 0.8 (0.6, 1.0)].

Conclusions

Following aTSA with a peripherally enhanced fixation glenoid, pain, range of motion, and patient-reported outcomes significantly improved at a minimum of 2 years with only 5.7% glenoid osteolysis despite heterogeneous preoperative glenoid pathologies.