The three-dimensional glenoid vault model can estimate normal glenoid version in osteoarthritis

Glenohumeral arthroplasty can involve correcting pathologic glenoid tilt or version. Predicting the physiologic glenoid version for a particular individual can be difficult. We propose using a previously validated, 3-dimensional, glenoid vault model as a template to predict normal glenoid version. Computed tomography scans of both shoulders were obtained in 14 subjects with unilateral glenohumeral osteoarthritis. Custom-developed graphic software was used to create a 3-D reconstruction of each scapula. Within the software, the vault model was placed in a best-fit orientation into each glenoid vault independently by 3 observers who were blinded to the contralateral scapula. Measurement differences between the glenoid and vault model were analyzed by repeated-measures analysis of variance. Standard errors of measurement (SEM) were calculated. Interobserver and intraobserver reliabilities were assessed. The healthy glenoid version averaged –7.0° (SEM, 0.7°; range, 0° to –14°). The arthritic glenoid version averaged –15.6° (SEM, 0.7°; range, 1° to –33°; P < .0001). The version of the implanted vault model measured –7.1° (SEM, 0.7°; range, –1° to –15°) on the healthy side and –7.2° (SEM, 0.7°; range –2° to –11°) on the arthritic side. Measurements between observers were not significantly different (P = .98). Interobserver and intraobserver correlation coefficients were 0.79 (P < .001) and 0.80 (P < .001). In the arthritic glenoid, the vault model reproducibly closely approximated the version of the normal contralateral glenoid, –7.2° vs –7.0° (P = .99) and is a novel and accurate method of estimating the normal glenoid version. This technique may be valuable in correcting pathologic glenoid version due to arthritis.