Value of anterosuperior rotator cuff and labral tears at MRI for predicting long head of biceps tearing at arthroscopy

Abstract

To determine whether magnetic resonance imaging (MRI) signs of subscapularis (SBT), supraspinatus (SST), or superior labral tearing predict long head of biceps (LHB) tendon tearing at arthroscopy. Subjects with MRI studies followed by arthroscopy were collected. Radiologists graded rotator cuff (RC) tendons and the superior labrum using three grade classifications and blinded to arthroscopy. Correlation between imaging variables and surgical outcome was expressed in terms of odds ratios and determined using a stepwise logistic regression model. Selection criteria identified 89 participants. Statistically significant increase in odds of finding a partial LHB tendon tear at arthroscopy were noted for both MRI readers with any SBT tear (OR=4.1-5.6, p<0.0001 to 0.002), full-thickness SST (OR=8-20.4, p=0.002 to 0.006), and combined SST-SBT tears (OR=5.1-7.6, p<0.0001 to 0.002) and relative to grade 0 MRI scores for those categories. Statistically significant increase in the odds of finding any LHB tendon tear at arthroscopy were noted for both MRI readers with any SBT tear (OR=9.6 to 14.6, p<0.0001), full-thickness SST (OR=9.0 to 52.0, p<0.0001 to 0.0004) and combined SST-SBT tears (OR=8.2 to 15.1, p<0.0001) at MRI and relative to grade 0 MRI scores for these categories. No significant predictive effect was found for the labral categories. LHB tendons should be closely scrutinised if anterosuperior rotator cuff tears, and SBT tears in particular, are found on MRI.