Bucket-handle meniscal tears of the knee: sensitivity and specificity of MRI signs

Objective: To determine the sensitivity and specificity of reported MRI signs in the evaluation of bucket-handle tears of the knee. Design and patients: A retrospective analysis of 71 knee MR examinations that were read as displaying evidence of a bucket-handle or "bucket-handle type" tear was performed. We evaluated for the presence or absence of the absent bow tie sign, the coronal truncation sign, the double posterior cruciate ligament (PCL) sign, the anterior flipped fragment sign, and a fragment displaced into the intercondylar notch. Sensitivity and specificity were calculated relative to the gold standard of arthroscopy. Results: Forty-three of 71 cases were surgically proven as bucket-handle tears. The absent bow tie sign demonstrated a sensitivity of 88.4%. The presence of at least one of the displaced fragment signs had a sensitivity of 90.7%. A finding of both the absent bow tie sign and one of the displaced fragment signs demonstrated a specificity of 85.7%. The double PCL sign demonstrated a specificity of 100%. The anterior flipped meniscus sign had a specificity of 89.7%. Conclusions: Bucket-handle tears of the menisci, reported in about 10% of most large series, have been described by several signs with MRI. This report gives the sensitivity and specificity of MRI for bucket-handle tears using each of these signs independently and in combination. MRI is shown to be very accurate for diagnosing bucket-handle tears when two or more of these signs coexist.