Abstract
BACKGROUND: Proximal humeral bone resorption (PHBR) after reverse total shoulder arthroplasty (RTSA) is not fully understood, partly because prior two-dimensional (2D) radiographic studies may under-/overestimate three-dimensional (3D) bone loss. We aimed to quantify 1-year postoperative bone loss using volumetric 3D computed tomography (CT) and identify associated risk factors. METHODS: We performed a retrospective cohort study of 63 shoulders that received primary RTSA for cuff tear arthropathy (CTA) between April 2019 and April 2023. Mimics 26.0 software (Materialise, Leuven, Belgium) and 3-Matic 18.0 software (Materialise, Leuven, Belgium) were used to segment four proximal-humeral regions and the humeral shaft using 3D reconstruction and segmentation. Bone formation and resorption were volumetrically estimated. Paired t-tests were used to compare preoperative and 1-year postoperative bone volumes, and univariable and multivariable linear regression analyses were performed to identify factors associated with PHBR. RESULTS: Bone resorption was detected in all four proximal‑humeral regions and the shaft, with the greatest loss at the greater tuberosity region (mean ± SD, 1,468.9 ± 1,161.1 mm(3)) and the least at the teres minor region (423.9 ± 540.2 mm(3)). Shoulders reconstructed with onlay-cup implants demonstrated significantly more proximal bone loss than inlay-cup implants (P < .05). A higher canal‑fill ratio and a smaller neck-shaft angle (NSA) were associated with bone resorption (both P < .05). Female patients were also associated with greater bone loss (P < .05), whereas humeral‑stem length and surface‑coating type showed no significant effect (all P > .05). CONCLUSIONS: 3D CT volumetric analysis demonstrated substantial proximal humeral bone loss within the first postoperative year after RTSA, most pronounced at the greater tuberosity. An onlay-cup design, higher canal filling ratio, smaller NSA, and female sex were associated with greater regional bone resorption, whereas stem length and surface‑coating type were not associated. LEVEL OF EVIDENCE: Retrospective comparative study, LEVEL III.