Scapular Y: the forgotten radiograph in glenohumeral arthritis—novel measurement of posterior humeral head subluxation on scapular Y radiographs

Abstract

BackgroundAxillary radiographs, 2-dimensional (2D) axial computed tomography (CT) scans, and 3-dimensional (3D) software are described to quantify posterior humeral head subluxation (PHHS) in patients with glenohumeral arthritis (GHOA). This study’s purpose was to compare a novel method of PHHS calculated from upright scapular Y radiographs and compare it to legacy measures of PHHS on axillary radiographs, 2D CT, and 3D CT in patients with Walch type B glenoid morphology. MethodsPatients undergoing anatomic total shoulder arthroplasty for primary GHOA with Walch B glenoids and a full series of preoperative radiographs and 2D and 3D CT scans were retrospectively identified. Three fellowship-trained shoulder surgeons independently measured the percentages of PHHS on axillary and scapular Y radiographs and axial 2D CT. PHHS on 3D CT was measured volumetrically using surgical templating software. ResultsThirty patients were identified with GHOA undergoing anatomic total shoulder arthroplasty. There were 26 type-B2 (86.7%) glenoids, 3 type-B1 (10%) glenoids, and 1 type-B3 (3.3%) glenoid. The mean PHHS value for the entire group on the scapular Y radiographs was 77.0% ± 13.8%, axillary radiographs was 77.3% ± 8.6%, 2D CT was 74% ± 5.6%, and 3D CT was 83.3% ± 6.0%. PHHS on 3D CT was 6% greater than that on scapular Y (P = .04) and axillary (P < .01) radiographs and 9% greater than that on 2D CT (P < .0001). Correlation between scapular Y PHHS and all other imaging modalities was weak (axillary radiograph, r = 0.369; 2D CT, r = 0.097; 3D CT, r = 0.134). Correlation between axillary radiograph PHHS was moderate for 2D CT (r = 0.518) and weak for 3D CT (r = 0.398). Correlation of 2D CT PHHS to volumetric 3D CT was strong (r = 0.754). Excellent inter-rater reliability for a single measure between raters was observed on scapular Y radiographs (intraclass correlation coefficient; ICC = 0.922), and moderate inter-rater reliability was observed on axillary radiographs (ICC = 0.680) and 2D CT (ICC = 0.692). ConclusionScapular Y radiographs provide a novel, reliable method to measure PHHS. The weak-to-moderate correlation between scapular Y and other imaging modalities (axillary radiographs and 2D and 3D CTs) may be related to the dynamic position of the humeral head in the supine vs. upright position. Taken along with the axillary radiograph, the scapular Y radiograph provides a more comprehensive understanding of subluxation. With an increasing focus on reducing cost, utilizing the scapular Y method provides another objective measure of PHHS. Surgeons should be aware that PHHS on volumetric 3D CT was 9% greater than that on 2D CT and 6% greater than that on axillary or scapular Y radiographs.