Glenoid version: how to measure it? Validity of different methods in two-dimensional computed tomography scans

Recognition of the glenoid version is important for evaluation of different pathologies such as degenerative wear, shoulder instability, or congenital deformity. Surgical strategies can change significantly in the presence of major retroversion. There is no consensus on the method to use to evaluate version. This study compared different measurement strategies in 116 patients with shoulder computed tomography (CT) scans. We hypotheses that the methods will give different value for evolution.

Shoulder axial CT images were reviewed, and the image inferior to the base of the coracoid was selected. The glenoid version was measured according to the Friedman method and the scapula body method. Three orthopedic surgeons independently examined the images 2 times, and intraobserver and interobserver reliability was calculated using intraclass correlation (ICC).

Group 1 (n = 53): The average glenoid version was significantly different between the 2 measurement techniques for all 3 observers, with an average of -7.29° for the scapula body method and -10.43° for Friedman method. For group 2 (B2 glenoid group, n = 63): The most reliable method for measurement of B2 glenoid (glenoid with posterior erosion) version was the association of the Friedman line for the scapula axis and the intermediate glenoid line, with excellent intraobserver reliability (ICC > 0.957) and interobserver reliability (ICC = 0.954).

The glenoid version measurement is reliable on a 2D CT Scan. According to correlation found in our paper and those of the literature it seems that there is no advantage on 3D CT Scan to assess version in terms of reliability of measures.

Combining the Friedman method to determine the scapula axis with an intermediate glenoid line in B2 glenoid yield the most reliable measurements.