Superior internal rotation after anatomic vs. reverse total shoulder arthroplasty performed with lesser tuberosity...

Recent studies report superior internal rotation (IR) and lower anterior shoulder pain after anatomic total shoulder arthroplasty (aTSA) compared to reverse TSA (rTSA) for cuff-intact osteoarthritis. However, these differences may reflect variations in subscapularis management rather than implant design, as lesser tuberosity osteotomy (LTO) is routinely performed during aTSA but less commonly during rTSA. This is an important distinction given the higher healing rates after LTO and impact of subscapularis healing on IR after rTSA. We compared outcomes between aTSA and rTSA in cuff-intact osteoarthritis patients when subscapularis management was standardized using LTO.

We conducted a retrospective review of consecutive primary rTSA or aTSA cases performed with an LTO (2019-2023). The inclusion criteria included glenohumeral arthritis, intact rotator cuff, and minimum 2-year follow-up. Demographics, range of motion (ROM), and patient-reported outcomes including Single Assessment Numeric Evaluation, American Shoulder and Elbow Surgeons score, Anterior Shoulder Pain and Dysfunction Score, and visual analog scale for pain were compared. IR was assessed with the vertebral level method and converted to ordinal scores based on the Constant score. Pre-operative glenoid morphology and retroversion were determined from advanced imaging. LTO healing was evaluated radiographically at minimum 3 months. Multivariable analysis was performed with age, sex, Walch classification, retroversion, and LTO healing included as covariates (α = 0.05).

Forty-six LTO-rTSA and 160 LTO-aTSA patients were included at a mean follow-up 3.9 ± 1.6 years (range, 2.0-6.9 years). aTSA patients were younger (66.4 vs. 71.5 years, P < .001) but groups were otherwise similar with respect to sex, glenoid morphology, retroversion, and pre-operative ROM and function. LTO healing/non-displacement occurred in 95% of patients, with no difference between cohorts (P = .184). Healed LTO was associated with superior final ROM across all planes. aTSA patients achieved higher final ROM, greater improvements in external rotation (22° vs. 16°, P = .010) and IR (2.4 vs. 0.8 points, P < .001), and higher final Single Assessment Numeric Evaluation and American Shoulder and Elbow Surgeons scores. On multivariable analysis, aTSA was associated with increased odds of IR improvement (odds ratio 4.0, 95% confidence interval 1.4-11.3, P = .005).

When subscapularis management was standardized using LTO, aTSA achieved superior IR compared to rTSA, independent of age, glenoid morphology, and LTO healing. These findings demonstrate that the reported ROM advantages of aTSA are intrinsic to implant biomechanics rather than subscapularis management, supporting its continued role for cuff-intact glenohumeral osteoarthritis.