Virtual assessment of internal rotation in reverse shoulder arthroplasty based on statistical shape models of scapular size

Background: The purpose of this study was to assess impingement-free internal rotation (IR) in a virtual reverse shoulder arthroplasty simulation using a Statistical Shape Model based on scapula size. Methods: A database of over 10,000 scapulae utilized for preoperative planning for shoulder arthroplasty was analyzed with a Statistical Shape Model to obtain 5 scapula sizes including the mean and 2 standard deviations. For each scapula model, one glenosphere size (33-42 mm) was selected as the best fit based on consensus among 3 shoulder surgeons. Virtual implantation variables included 1) lateral offset (0-12 mm in 2-mm increments), 2) inferior eccentricity (0, 2.5, 5, and 7.5 mm), and 3) posterior eccentricity (0, 2.5, and 5 mm). The neck shaft angle was fixed at 135° with an inlay design humeral prosthesis. IR at the side (IR0) and in abduction (IRABD) were then simulated. Results: Maximum impingement-free IR0 was reached with increasing inferior offset in combination with increasing lateralization. Lateralization was the most important variable in increasing impingement-free IRABD. Maximum IRABD was reached at 4-6 mm of lateralization with smaller scapula (−2 to 0 standard deviation). Increasing lateralization up to 12 mm continues to increase IRABD for larger-sized scapula (+1 to +2 standard deviation). Optimal inferior offset and lateralization to maximize IR did have a small loss of external rotation in neutral abduction. There was no loss of external rotation in 60° of abduction. Conclusion: In a virtual model, the glenosphere position required to maximize IR varied by scapula size. For smaller scapulae, maximum IR0 was reached with a combination of 2.5-mm inferior offset and 0- 4 mm of lateralization. For larger scapulae, maximum IR0 was reached with a combination of 2.5 mm of inferior offset and 4 mm of lateralization. The amount of lateralization required to maximize IRABD also varies by scapula size. Maximum IRABD was reached in smaller scapula with 4-6 mm of lateralization and at least 12 mm of lateralization in larger scapula. These findings may be applied in the clinical decision-making process knowing that impingement-free IR and IRABD can be maximized with combinations of inferior offset and lateralization based on scapula size with minimal effect on external rotation and external rotation in 60° of abduction.