The Extension-Thickness-Damage (ETD) score: a pre-operative hip MR arthrography-based classification to predict type of labrum surgery.

INTRODUCTION: Hip magnetic resonance arthrography (MRA) is the pre-operative imaging modality of choice in patients with labral damage, with several classifications of labral tears having been reported based on MRA findings. Nevertheless, none of the available classification systems allows the surgeon to predict before surgery how a labral tear could be treated. Our purpose was to develop a new MRA-based scoring system of labral tears to predict before surgery the treatment option more suitable for labral tears. MATERIALS AND METHODS: Forty-seven patients (29 males and 18 females; mean age: 35.9 ± 12.4) performed hip MRA for suspicious of femoroacetabular impingement and were afterwards subjected to arthroscopic treatment. Two musculoskeletal radiologists reviewed all pre-operative examinations and provided the Extension-Thickness-Damage score for each patient, based on Extension of tear, Thickness of labrum, and type of Damage. Chondral lesions grading was based on the arthroscopic findings according to Konan classification. For statistical purposes, patients were divided into two groups, depending on the type of treatment: labral repair or debridement. Mann-Whitney U, Chi-square, receiver operator curves, and Cohen kappa statistics were used. RESULTS: 35/47 underwent repair, while 12/47 were debrided. In both groups, the median chondral damage was grade III, with no significant differences (p = 0.439). The median Extension-Thickness-Damage score in the repair group (6) was significantly lower (p < 0.001) than that in the debridement group (8). The highest diagnostic performance (area under the curve) of Extension-Thickness-Damage was 0.819. The inter-observer agreement was substantial in the evaluation of Extension (k = 0.626) and Thickness (k = 0.771), and almost perfect for Damage (k = 0.827). Higher scores of Extension and Thickness were more frequently associated with debridement (p < 0.001; p = 0.0016, respectively), with no significant differences on the basis of Damage parameter (p = 0.284). CONCLUSIONS: The MRA-based Extension-Thickness-Damage score could represent a helpful pre-operative tool, expressing the extent of the damage and its reparability before arthroscopy.