A prospective observational case control study investigating the coronal plane scapular morphological differences in full-thickness posterosuperior cuff tears and primary glenohumeral osteoarthritis.

Abstract

The critical shoulder angle (CSA) is a surrogate marker of the coronal plane morphology of the scapula. CSA differences between scapulae could be due to differences in glenoid inclination (GI) or the location of the most lateral part of the acromion relative to the inferior glenoid, or both. An understanding of the hierarchy of the scapular morphological changes associated with glenohumeral osteoarthritis (GHOA) and rotator cuff (RC) tears would allow accurate biomechanical modeling. A prospective observational case control study was undertaken in which the GI, "nonglenoid"-CSA, acromial vertical offset index, acromial horizontal offset index, acromial horizontal-vertical offset index, and coronal plane angulation of the acromion (CPAA-m) were measured on high-quality radiographs to compare coronal plane scapular anatomy in: (1) patients with asymptomatic atraumatic full-thickness RC tears, (2) patients with symptomatic primary GHOA, and (3) a control group with no RC tear or GHOA treated for glenohumeral instability or symptomatic labral pathology. Intraobserver reliability of the measurements was performed. In the GHOA group, the GI was lower (less superiorly inclined) than the RC tear group (difference between the means: -4.8°, 95% confidence interval [CI] [-8.8°, -0.9°], P = .014) and the control group (difference between the means: -7.9°, 95% CI [-11.8°, -3.9°], P = .000); there was no difference in the acromial measurements. In the RC tear group, the nonglenoid-CSA was higher (difference between the means: 7.7°, 95% CI [3.0°, 12.3°], P = .001), the acromial vertical offset index was lower (difference between the means: -0.13, 95% CI [-0.24, -0.01], P = .026), and the acromial horizontal-vertical offset index was higher (difference between the means: 0.15, 95% CI [0.01, 0.28], P = .030) than the control group; there was no difference in the acromial horizontal offset index or the GI. The CPAA-m was lower (greater coronal plane downslope of the acromion) in both GHOA (difference between the means: -9.6°, 95% CI [-18.6°, -0.5°], P = .036) and RC tears (difference between the means: -9.9°, 95% CI [-19.0°, -0.9°], P = .029) compared with the control group. The intraclass correlation coefficients for intraobserver reliability demonstrated excellent reliability for the measurements (all >0.900). Scapulae associated with GHOA have lower GI, but no spatial differences in the location of the lateral acromion compared with a normal population. Scapulae associated with RC tears have a lower vertical offset of the lateral acromion, but no difference in horizontal offset or GI compared with a normal population. The downslope of the acromion in the coronal plane is greater (lower CPAA-m) in both RC tears and GHOA than the normal population.