Computed tomography (CT) has been the gold standard for measuring glenoid version in patients with glenohumeral osteoarthritis or instability. However, few studies have compared measurements of glenoid version on magnetic resonance imaging (MRI) versus CT. Clinicians should feel confident in assessing glenoid version on MRI in the absence of CT imaging. In fact, MRI could be comparable to CT even in cases where the entire scapula is not visible on MRI. A 5 difference in glenoid version between imaging modalities represents a clinically significant difference.
See related article on page 99G lenoid version plays an important role in the normal biomechanics of the shoulder and can be adversely affected by glenohumeral arthritis or instability. 1,2 The normal range is considered to be 2 of anteversion and 9 of retroversion. The Friedman et al. 3 study was the first to demonstrate that the glenoid was more retroverted on computed tomography (CT) in patients with severe glenohumeral osteoarthritis compared with controls. This landmark study helped to establish CT as the gold standard measurement for glenoid version, but the question remains whether magnetic resonance imaging (MRI) can be used for measuring glenoid version. 1,3 In the study, "Magnetic Resonance Imaging Correlates With Computed Tomography for the Glenoid Version Calculation Despite Lack of Visibility of Medial Scapula," Parada, Shaw, Antosh, Eichinger, Li, Curry, and Provencher set out to answer this question by examining the reliability of measuring glenoid version on MRI versus CT imaging. 4 They studied 32 patients who were treated operatively for glenohumeral instability in the absence of glenohumeral arthritis. Each patient underwent both CT and MRI scans, and 2 fellowship-trained orthopaedic surgeons independently measured scapular width and glenoid version for each imaging series on 2 separate occasions. Intrarater reliability was excellent for scapular width and glenoid version, and interrater reliability was excellent for scapular width but moderate for glenoid version. Scapular width was 24.7 mm longer on CT versus MRI, and 78% of the scapular width could be seen on MRI versus CT. Glenoid version did not differ between MRI and CT.Similar to Parada et al., previous studies have found no difference in glenoid version on MRI versus CT in patients with shoulder dislocation, primary shoulder osteoarthritis, and those planning to undergo shoulder arthroplasty. 2,4-6 Beaulieu-Jones et al. 7 found that posterior glenoid bone loss was associated with increased posterior glenoid version on CT or MRI in patients with recurrent posterior shoulder instability and bone injury; however, there was no comparison between imaging modalities. Two studies have also demonstrated that posterior instability was associated with glenoid version measured on MRI. 8,9 Based on these results, clinicians should feel confident in assessing glenoid version on MRI in the absence of CT imaging.Parada et al. should be commended for assessing intra-and interrater reliability, alt...