Hemiarthroplasty versus total shoulder arthroplasty in B2 glenoids with an intact rotator cuff: a long-term matched cohort analysis

BACKGROUND: Walch B2 glenoids present unique challenges to the shoulder arthroplasty surgeon, particularly in young, active patients who may wish to avoid the restrictions typically associated with an anatomic total shoulder arthroplasty (TSA). Long-term data are limited when comparing hemiarthroplasty (HA) and TSA for patients with an intact rotator cuff. The purpose of our study was to compare the long-term outcomes of HA vs. TSA in a matched analysis of patients with B2 glenoids, primary osteoarthritis (OA), and an intact rotator cuff. METHODS: A retrospective review was performed of all patients who underwent HA or TSA between January 2000 and December 2011 at a single institution. Inclusion criteria were primary OA, Walch B2 glenoid morphology, an intact rotator cuff intraoperatively, at least 2 years of clinical follow-up, or revision within 2 years of surgery. Fifteen HAs met inclusion criteria and were matched 1:2 with 30 TSAs using age, sex, body mass index, and implant selection. Clinical outcomes including range of motion (ROM), visual analog scale (VAS) for pain, subjective shoulder value score, American Shoulder and Elbow Surgeons (ASES) score, complications, and revisions were recorded. Postoperative radiographs were reviewed to assess for stem loosening, humeral head subluxation, glenoid loosening, and glenoid erosion. RESULTS: A total of 15 HAs and 30 TSAs met inclusion criteria at a mean follow-up of 9.3 years. The mean age at the time of surgery was 60.2 years for HA and 65.4 years for TSA (P = .08). Both cohorts had significant improvements in ROM, subjective shoulder value, and VAS pain scores (P < .001). TSA had higher postoperative ASES scores compared to HA (P = .03) and lower postoperative VAS pain scores (P = .03), although the decrease in pain from preoperatively to final follow-up was not significantly different between HA and TSA (P = .11). HAs were more likely to have posterior humeral subluxation (P < .001) and stem lucencies (P = .02). Revisions occurred in 11.1% of the cohort with no difference for HA and TSA (P = .73). CONCLUSIONS: At nearly 10 years of follow-up, HA and TSA both showed significant improvements in ROM and pain when performed for primary glenohumeral OA in B2 glenoids with intact rotator cuffs. Compared to HA, TSAs had less posterior humeral subluxation, less stem lucencies, higher ASES scores, and lower postoperative VAS pain scores. However, our study failed to demonstrate a difference in ROM, complication, or revision rates between HA and TSA.