Does the Critical Shoulder Angle Correlate With Rotator Cuff Tear Progression?

BACKGROUND

The critical shoulder angle (CSA) has been reported to be associated with rotator cuff disease and has been suggested as an etiology for cuff tears. However, it is unclear whether acromial morphologic characteristics such as CSA are a cause or effect because all studies to date have been retrospective.

QUESTIONS/PURPOSES: (1) How often can the CSA be reliably measured? (2) Is the CSA associated with rotator cuff disease? (3) Is the CSA correlated with baseline tear size or tear enlargement? (4) Does the CSA change with time?

METHODS

In this retrospective comparison of longitudinally collected data, patients with asymptomatic rotator cuff tears underwent ultrasonography and standardized AP radiographs at enrollment and yearly thereafter during a median of 4 years. Three hundred ninety-five patients were included, of whom 14 were excluded as they were not yet eligible for 2-year followup and 68 (18%) were lost to followup, leaving 313 study patients who were evaluated with 1433 radiographs. Patients with adhesive capsulitis with normal rotator cuffs and radiographically normal scapulae were included as control subjects (119 subjects). Two observers (PNC, DS) measured the CSA in a blinded fashion. Radiographs that met Suter-Henninger criteria for CSA measurement reliability were included. For the study group, 179 of the 313 (57%) patients with radiographs that met Suter-Henninger criteria were further analyzed; the remainder were excluded from this study. For the control group, 50 of 119 (42%) subjects met criteria and were further analyzed. Tear enlargement was found in 94 patients, and the CSA was compared in patients with tears and control subjects, and in tears with or without enlargement, and was correlated with tear size. In a subgroup of the study group in which 59 of 179 patients had a minimum of 3 years between initial and followup radiographs, two CSA measurements were performed to measure change.

RESULTS

In total, of the 1552 radiographs evaluated, only 326 (21%) were of sufficient quality to measure the CSA. The CSA was higher among patients with cuff tears than control subjects (34° ± 4° versus 32° ± 4°; mean difference, 2.0°; 95% CI, 0.7°-3.2°; p = 0.003). The CSA did not correlate with baseline tear length (ρ = 0.22, p = 0.090) or width (ρ = 0.16, p = 0.229). The CSA was not different between tears that enlarged and those that were stable (34° ± 3° versus 34° ± 4°; mean difference, 0.2°; 95% CI, -0.9° to -1.4°; p = 0.683). The CSA did not change over time (CSA Time 1: mean 33° ± 4° SD; CSA Time 2: mean 33° ± 4° SD; mean difference, -0.2°; 95% CI, -0.6° to 0.1°; p = 0.253).

CONCLUSIONS

Even with a longitudinal protocol, most radiographs are of insufficient quality for CSA measurement. Although patients with a history of degenerative cuff disease have higher CSA values than control subjects, the difference is small enough that it could be influenced by measurement error in practice; in any case, a difference of the magnitude we observed is likely to be clinically unimportant. The CSA is not correlated with tear size or tear progression, and does not seem to change with time. These results suggest that the CSA is unlikely to be related to rotator cuff disease.

LEVEL OF EVIDENCE

Level II, prognostic study.