Posteroinferior relevant scapular neck offset in reverse shoulder arthroplasty: key player for motion and friction-type impingement in a computer model.

Abstract

Range of motion (ROM) and prevention of notching remain a challenge for reverse shoulder arthroplasty (RSA). Both may be affected by the morphology of the scapula. The purpose of this study was to define anteroinferior (a) and posteroinferior (p) relevant scapular neck offset (RSNO) and to examine the hypothesis that pRSNO is significantly smaller than aRSNO, and influences rigid body motion (RBM). Adapting glenosphere implantation strategies may therefore be of value. In this computer model study, we used deidentified computed tomographic scans of 22 patients (11 male and 11 female; mean age: 72.9 years) with massive cuff tears without joint space narrowing. Eight RSA glenoid configurations were tested with a constant neck-shaft angle (145°). Two baseplate types (25 mm; 25 + 3 mm lateralized) and 4 glenospheres (GS) (36 mm; 36 +2 mm of eccentricity; 39 mm; 39 + 3 mm) were used. RSNO was defined as the standardized measurement of the horizontal distance from the inferior extent of the GS to the bony margin of the scapula after baseplate positioning (flush to inferior glenoid extent; neutral position: 0° inclination and 0° version-both software computed). There was a highly significant difference between pRSNO and aRSNO for both genders (P<.001). pRSNO was always smaller than aRSNO. pRSNO was strongly correlated with external rotation (ERO: 0.84) and extension (EXT: 0.74) and moderately correlated with global ROM (GROM: 0.68). There was a moderately strong correlation between aRSNO and internal rotation (IRO: 0.69). pRSNO was strongly correlated with aRSNO, EXT, ERO, IRO, adduction (ADD) and GROM (0.82, 0.72, 0,8, 0.71, 0.82, 0.76) in female patients and with EXT and ERO (0.82, 0.89) in male patients. The median pRSNO allowing for at least 45° ERO and 40° EXT was 14.2 mm for men and 13.8 mm for women. For all patients and models, pRSNO ≥14 mm increased EXT, ERO, and GROM significantly compared with pRSNO <14 mm (P<.001). The combination of lateralization and inferior overhang (eccentricity) led to the most significant increase of pRSNO for each GS size (P<.001). This is one of the first RSA modeling studies evaluating nonarthritic glenoids of both genders. The lateral scapular extent to glenoid relationship is asymmetric. pRSNO is always smaller than aRSNO for both genders and was a critical variable for EXT and ERO, demonstrating additional strong correlation with aRSNO, IRO, ADD, and GROM in female patients. pRSNO ≥14 mm was a safe value to prevent friction-type impingement. Combining increased glenosphere size, lateralization, and inferior overhang gives the best results in this computer-simulated setting.