It is important for both the patient and the surgeon to determine whether a meniscal lesion can be repaired before undertaking surgery. The purpose of this study was to examine the pertinence of clinical signs and determine the value of imaging findings for the preoperative diagnosis of bucket-handle meniscal tears. This preliminary study was conducted before undertaking an analysis of preoperative criteria of reparability in a homogeneous group of meniscal lesions.
This retrospective series included 33 arthroscopically-proven bucket-handle meniscal tears in patients who underwent arthrography and/or arthroscan and/or MRI preoperatively. The images were reviewed by two senior radiologists who established a consensus diagnosis. Clinically, the type of blockage and the presence of permanent flexion before surgery were noted. The following items were noted on the imaging results: fragment displacement (fragment in the notch on the coronal slice) anterior megahorn, double PCL, and serpent sign on the sagittal slice. Longitudinal, transversal extension and position of the bucket-handle were noted. We searched for correlations with the intraoperative findings.
Fourteen patients had a history of knee blocking and 15 had permanent flexion before surgery. Only 10 patients had the typical association of blocking and flexion. Certain diagnosis of bucket-handle meniscal tear was provided by MRI (13/15), arthroscan (6/7), and arthrography (10/24) giving an equivalent sensitivity for the two slice imaging techniques. The sign of a fragment in the notch on the coronal slice was a constant finding. The double PCL sign was sensitive for medial meniscal tears and for lateral meniscal tears with associated ACL tears. The diagnosis was successfully established in all 9 patients who underwent several explorations (2 or 3). Buckle-handle meniscal tear was not identified in 9 patients (arthrography 7, MRI 2).
Our findings demonstrate that the preoperative diagnosis of bucket-handle meniscal tears cannot be properly established on clinical criteria of typical blocking and/or permanent flexion. They confirm that arthrography is not contributive to diagnosis and that the absence of a slice image is detrimental to diagnosis. The sensitivity of the two slice imaging methods was similar. The key sign was the presence of a fragment in the notch on the coronal slice; in the three cases where this sign was absent, the reason was found to be the small size of the displaced fragment (resolution limit) and time between imaging and arthroscopy. The characteristic features of the bucket-handle lesions observed in this series are exactly the same as reported in earlier reports but to our knowledge provide the first data on comparative performance of arthroscan and MRI.
The noninvasive nature of MRI and the possibility of assessing the meniscal wall and the quality of the meniscal tissue make MRI the exploration of choice for preoperative assessment of meniscal tears.