Total Shoulder Arthroplasty Versus Reverse Shoulder Arthroplasty in Primary Glenohumeral Osteoarthritis With Intact Rotator Cuffs: A Meta-Analyses

Abstract

Traditional practice favors total shoulder arthroplasty (TSA) for the treatment of primary glenohumeral osteoarthritis (PGHO) with an intact rotator cuff; however, the indications for reverse shoulder arthroplasty (RSA) have expanded to include PGHO. The purpose of this systematic review is to compare the mean differences in the range of motion and patient-reported outcomes between the TSA and RSA with an intact rotator cuff and to analyze the subgroup of the Walch type B2 glenoid. This IRB-exempt, PROSPERO-registered systematic review strictly followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P) guidelines. A literature search of five databases revealed 493 articles, of which 10 were included for quantitative synthesis. Level III evidence studies with the diagnosis of PGHO and ≥2 years of follow-up were included. Studies without preoperative and postoperative data were excluded. The Newcastle-Ottawa scale was used to evaluate the methodologic quality of the included studies. Preoperative and postoperative range of motion and patient-reported outcomes were collected. The random-effects model was employed, and p < 0.05 was considered statistically significant. There were a total of 544 and 329 studies in the TSA group and RSA group, respectively. The mean age in the TSA group and RSA groups were 65.36 ± 7.06 and 73.12 ± 2.40, respectively (p = 0.008). The percentages of males in the TSA and RSA groups were 73.2% and 51.1%, respectively (p = 0.02). The mean differences in forward elevation, external rotation in adduction, internal rotation scale, visual analog scale (VAS), American Shoulder and Elbow Surgeons (ASES) score, and Single Assessment Numeric Evaluation (SANE) scores were improved for both groups with no significant differences between the two. There were 9.6 times the revisions in the TSA group (8.8% vs. 0.91%; p = 0.014) and 1.5 times the complications in the TSA group (3.68% vs. 2.4%; p = 0.0096). Two hundred and forty-two glenoids were identified as Walch type B2 (126 in the TSA group and 116 in the RSA group). The mean ages in the B2 subgroup were 68.20 ± 3.25 and 73.03 ± 1.49 for the TSA and RSA, respectively (p = 0.25). The percentages of males in the B2 subgroup were 74.6% and 46.5% for the TSA and RSA groups, respectively (p = 0.0003). The ASES, SANE, forward elevation, and external rotation in the adduction results were descriptively summarized for this subgroup, with average mean differences of 49.0 and 51.2, 45.7 and 66.1, 77.6° and 58.6°, and 38.6° and 34.1° for the TSA and RSA groups, respectively. In the setting of primary glenohumeral osteoarthritis with an intact rotator cuff, the RSA has a similar range of motion and clinical outcomes but lower complication and revision rates as compared to the TSA. This may hold true in the setting of the B2 glenoid, although a high-powered study on this subgroup is required. Anatomic shoulder arthroplasty maintains an important role in select patients. Further studies are required to better elucidate the role of glenoid bone loss and posterior humeral head subluxation with regard to implant choice.

Keywords: degenerative joint disease; evidence-based practice; glenohumeral arthritis; shoulder replacement; shoulder surgery; surgical decision-making; surgical outcomes.