Abstract
OBJECTIVES: Glenohumeral osteoarthritis can be treated with anatomic (aTSA) or reverse shoulder arthroplasty (RSA) depending on the rotator cuff integrity and bone loss severity. The amount of glenohumeral motion as well as scapulothoracic compensation towards humerothoracic abduction after aTSA and RSA has not been well investigated but is thought to play an important role in determining total range of motion. Digitial Dynamic Radiography (DDR) provides a novel technique to measure the scapulohumeral rhythm (SHR) in response to surgical intervention by measuring the glenohumeral and scapulothoracic contributions to humerothoracic abduction. The purpose of this study is to compare SHR between aTSA and RSA for a primary diagnosis of glenohumeral osteoarthritis with the use of DDR. METHODS: DDR was performed on 74 shoulders, comprised of 44 RSA and 30 aTSA shoulders, after a primary diagnosis of glenohumeral osteoarthritis validated by the senior surgeon. All patients had glenoid bone loss and intact rotator cuffs. Shoulders with <6 months postoperative DDR, prior surgery, and patients <18 years old were excluded. Manual measurements of the angle between the humerus and the vertical midline and the lateral border of the scapula and horizontal midline were taken by two readers at 0°, 30°, 60°, 90°, and 120° of shoulder abduction, and the ratio of these measurements formed the SHR. SHR values were also compared to a cohort of 32 normal asymptomatic control shoulders. Data was compared using descriptive statistics, and inter-rater reliability of the manual measurements was assessed with intra-class correlations. RESULTS: The total number of measurements was 1,459 with an intraclass correlation coefficient of 0.99. The mean total range-of-motion (ROM) SHRs (0°-120°) for the RSA and aTSA cohorts were 2.00 and 1.97, respectively, compared to 2.51 for the normal shoulder controls. When comparing the top 10 highest total SHR in each cohort, the RSA cohort had a significantly higher (p<0.05) mean SHR of 2.67 compared to 2.49 for the aTSA cohort. The mean SHR for the RSA and aTSA cohorts from 0°-30°, 30°-60°, 60°-90°, and 90°-120° intervals were 2.73 and 1.88, 2.17 and 2.08, 2.54 and 2.70, and 2.64 and 3.30, respectively. In the previously stated intervals, the normal controls had mean SHRs of 2.76, 4.29. 3.89, and 3.38, respectively (Table 1). Final patient-reported outcomes including VAS, SSV, ASES for the RSA and TSA cohorts were 1.41 and 1.17, 83.8 and 84.2, and 91.4 and 93.9, respectively. CONCLUSIONS: Neither RSA nor aTSA restore in-vivo scapulothoracic biomechanics when performed for glenohumeral osteoarthritis, as shown by the markedly lower SHR to normal controls. While both RSA and aTSA produce similar mean total SHRs within the measured arc of motion, there remains variability in different intervals of shoulder abduction, likely due to the prosthesis biomechanics, specifically the polarity alteration. The clinical significance of scapula motion in shoulder arthroplasty remains poorly understood, but this work sheds light on its complexity and potential avenues to enhance range of motion in shoulder arthroplasty. Future studies should focus on biplane imaging and using paired pre and postoperative DDRs to increase study power.