Humeral loosening in reverse shoulder arthroplasty: an analysis of 2,342 cases

Humeral loosening is a rare complication in reverse shoulder arthroplasty (RSA) representing approximately 1% of total complications. The purpose of this study is to identify patients who underwent RSA and were revised because of loosening of the humeral component, identify patients who are at increased risk, and report on their surgical outcomes.A retrospective review of all patients who received a primary RSA or revision RSA (rRSA) by a single surgeon from 2002-2021 identified a total of 1591 primary RSA and 751 rRSA procedures. These procedures were then organized based on indication for surgery. Further analysis was performed to identify RSAs that were subsequently revised because of aseptic loosening of the humeral component. A total of 41 surgeries met the inclusion criterion for the study, which was any RSA or rRSA that was revised because of loosening of the humeral component. Exclusion criterion was revision for a reason other than humeral loosening, neurogenic arthritis, or revision for loosening that was not originally implanted by the senior author (9 surgeries). Ultimately, 32 surgeries met criteria for further analysis. These 32 surgeries were organized by indication for preceding RSA or rRSA and were assessed for an association between indication for RSA or rRSA and eventual revision for humeral loosening. Additionally, these procedures were compared to a "control cohort" of procedures that were not revised and that had minimum 2 years' follow-up. To assess outcomes for these patients, pre- and postoperative Simple Shoulder Test, American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form (ASES), and ranges of motion were compared. Twenty procedures had sufficient data for outcomes analysis and were followed for an average of 54 months (range: 24-132 months).In primary RSAs, procedures performed for fracture or fracture sequelae were associated with revision for humeral-sided loosening (P = .009). In rRSAs, procedures performed for failed hemiarthroplasty for CTA were associated with rerevision for humeral-sided loosening (P = .009). Nine percent of patients revised for humeral loosening required rerevision for recurrent humeral loosening. Analysis of patients with 2-year clinical follow-up for humeral-sided loosening showed improvement in ASES pain (P = .014), ASES function (P = .042), and total ASES scores (P = .007).Humeral loosening is rare in RSA. In our cohort, 0.7% of the primary RSAs performed and 2.8% of the rRSAs performed were eventually revised for humeral loosening. Revisions of RSA for humeral loosening yield modest clinical improvements. Rerevision for humeral loosening was 6.3% of patients in our cohort.