Frequency of glenoid chondral lesions on MR arthrography in patients with anterior shoulder instability.

PURPOSE: To describe the frequency of glenoid chondral abnormalities in relation to Hill Sachs (HS) lesions in MR arthrograms of patients with anterior shoulder instability versus controls. Such glenoid lesions can directly impact surgical decision-making and approach, and potentially negatively impact outcome if missed.

MATERIALS AND

METHODS

Retrospective analysis of direct MR shoulder arthrograms in 165 subjects, (101 with anterior instability/64 controls) was performed independently by 2 blinded musculoskeletal radiologists. Outcome measures were the presence of a HS, anterior labral pathology and glenoid chondral injury. Kappa statistic, Pearson Chi-square and Mann-Whitney analysis were employed for analysis.

RESULTS

Inter-observer variability for the presence of HS, labral and chondral lesions was 0.964, 0.965 and 0.858 respectively, with intra-observer variability of 1.0, 0.985 and 0.861 for the principle reader. 58% of patients and 8% of controls had HS (p<0.001). 72% of patients and 25% of controls had anterior labral injury (p<0.001). 36% of instability patients and 10% controls had glenoid chondral lesions (p<0.001). 46% of anterior instability patients with HS defects had chondral injury as opposed to 21% of patients without HS defects (p=0.009). Depth of the HS lesion did not increase the likelihood of a glenoid chondral lesion (p=0.7335).

CONCLUSION

In the clinical anterior instability cohort, we demonstrated a statistically significant higher number of HS and glenoid chondral lesions than in controls. In anterior instability patients, the presence of a HS lesion confers a statistically significant greater likelihood of having a glenoid chondral lesion when compared to patients with instability and no HS.