Posterior sentinel sign of the biceps tendon as an arthroscopic indicator of supraspinatus tears

Introduction: 

Magnetic resonance imaging (MRI) is highly accurate for diagnosing full-thickness rotator cuff tears but inaccurate in detecting small anterosuperior cuff tears near the rotator interval. Expanding on the concepts of long head of biceps tendon (LHBT) instability and the anterior sentinel sign, the present study evaluated the posterior sentinel sign, a novel arthroscopic indicator of supraspinatus tears. 

Methods: 

This monocentric prospective cohort study included patients undergoing arthroscopic repair of rotator cuff tears at a single institution. The integrity and stability of the LHBT were assessed intra-articularly for abrasion, delamination, or partial tears. Sentinel signs were classified as anterior or posterior based on their location on the LHBT. Rotator cuff tears were evaluated, and their correlation with sentinel signs was analyzed using the chi–squared test. To evaluate the diagnostic value of the sentinel signs, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the curve (AUC) were calculated. 

Results: 

This study included 69 patients. The posterior sentinel sign was observed in 21/39 patients with supraspinatus tears (53.8%) and in 1/30 patient without (3.3%), showing a significant association (X² = 19.9, p < 0.001, Phi = 0.54). For detecting supraspinatus tears, the posterior sentinel sign had a sensitivity of 53.8% (95%CI: 38.6–68.4), specificity of 96.7% (83.3%-99.4), PPV of 95.5% (78.2–99.2), NPV of 61.7% (47.4%-74.2) and AUC of 0.79 (95%CI: 0.68–0.89). The anterior sentinel sign was present in 17/24 patients with subscapularis tears (70.8%) and absent in those without (0/45), showing a significant association (X² = 42.3, p < 0.001, Phi = 0.78). For detecting subscapularis tears, it had a sensitivity of 70.8% (95%CI: 50.8–85.1), specificity of 100% (92.1–100), PPV of 100% (81.6–100), NPV of 86.5% (74.7–93.3) and AUC of 0.97 (95%CI: 0.87–0.99). 

Conclusions: 

Expanding our evaluation of the previously reported arthroscopic LHBT sentinel sign with a more detailed description of location, two main types can be discriminated. The anterior sentinel sign, which is associated with subscapularis tears, and the posterior sentinel sign, which is associated with supraspinatus tears. The presence of a sentinel sign should prompt the clinician tosearch for (occult) anterosuperior rotator cuff tears. Level of evidence: Level III.