INTRODUCTION: Imaging evaluation of the acetabular labrum is essential before the hip arthroscopy for cases with femoroacetabular impingement (FAI) or acetabular labrum tear. Magnetic Resonance arthrography has been utilized in this regard, recent development of 3.0 Tesla MRI and novel imaging sequences make it possible to visualize the joint structures including acetabular labrum without arthrography. In this study, we evaluated acetabular labrum tear using radial sequence 3D 3.0 Tesla Multiple Echo Recombined Gradient Echo Magnetic Resonance Imaging (MERGE MRI) based on modified Czerny's classification, compared with hip arthroscopic findings.

METHODS: We retrospectively reviewed MERGE MRI findings performed as a preoperative evaluation from December 2013 to September 2018. A total of 71 hips including 29 hips of femoroacetabular impingement (FAI), 26 hips of borderline development dysplasia of the hip (BDDH) and 16 hips of early-stage osteoarthritis (OA) under Tönnis grade 2 were enrolled this study. MERGE MRI findings were evaluated by radial imaging in anterior region (2-3 o'clock), anterolateral region (1-2 o'clock), and lateral region (12-1 o'clock). Occasional cases have strongly degenerated that could not be categorized by original Czerny's classification is defined as stage 4, which is defined as new stage in this study. Specifically, stage 4A is defined as the low signal change area suggesting ossification. Stage 4B is defined as the mixture of high and low signal change area, the acetabular labrums are unclear and thickened. The hip arthroscopic findings were evaluated based on the Multicenter Arthroscopic Hip Outcomes Research Network (MAHORN) classification of the acetabular labrum morphology in the anterior, anterolateral and lateral region respectively. The acetabular labrum tear was diagnosed in case with partial, complete, flap tear and complex/ degenerative by MAHORN classification. In MERGE MRI, acetabular labrum tear was defined in cases with Czerny’s stage 1 or upper grade. We compared MERGE MRI findings with the hip arthroscopic findings and calculated sensitivity, specificity and positive/negative predictive value (PPV/NPV) in terms of the existence of acetabular labrum tear.

RESULTS: MERGE MRI findings revealed acetabular labrum tear more frequently in anterolateral than lateral region (p<0.01). Especially in FAI and BDDH cases, acetabular labrum tear was more frequently observed by MERGE MRI in anterolateral than lateral region (p = 0.02). In OA cases, a relatively higher rate of stage 4 was confirmed. In hip arthroscopic findings, 94% showed acetabular labrum tear in anterior, 96% in anterolateral and 47% in lateral region. In FAI and BDDH cases, acetabular labrum tear was more frequently recognized in anterior and anterolateral than lateral region (p<0.01). In comparison with MERGE MRI findings and the hip arthroscopic findings, the sensitivity, specificity, PPV, NPV of MERGE MRI was 85%, 56%, 89%, 46%, respectively. In anterior region, the sensitivity and specificity was 79% and 50%. In anterolateral region, those were 96% and 50%. In lateral region, those were 70% and 57%.

DISCUSSION: The Czerny's classification is an evaluation of acetabular labrum tear by MR arthrography (1). MERGE MRI is a non-invasive examination without arthrography which constructs a wideband high resolution images with less influence of artifacts. This is the first study evaluating 3.0 Tesla 3D MERGE MRI findings for evaluating acetabular labrum tear comparing with arthroscopic findings. One of the superiority of 3D images is that we can evaluate radial images by reconstruction. Thus, it is possible to identify the anatomical location by clock display position. In the current study, acetabular labrum tear was frequently observed in 0 to 2 o’clock as same as previously reported (2). MERGE MRI revealed sufficient sensitivity regarding acetabular labrum tear. Especially in anterolateral regions, it revealed excellent sensitivity. Previous similar study examined acetabular labrum tear using 3.0 Tesla MRI without arthrography, reporting sensitivity of 98% and specificity of 100% (3). Thus non-arthrographic MRI with 3.0 Tesla will be alternative modality for acetabular labrum tear imaging.
 
SIGNIFICANCE: We applied 3D 3.0 Tesla MERGE MRI for the evaluation of acetabular labrum tear comparing with arthroscopic findings. MERGE MRI revealed excellent sensitivity for a diagnosis of acetabular labrum tear, especially in anterolateral region.

REFERENCES: 
*1 Czerny C et al. American Journal of Roentgenology. 1999; 173: 345-349
*2 Yoon LS et al. Skeletal Radiol. 2007; 36(11):1029-33
*3 Crespo AM et al. Eur J Radiol. 2017; 88: 109-116

1909 - The Diagnostic Performance of Multiple Echo Recombined Gradient Echo MRI in Acetabular Labrum Tear -Comparison with arthroscopic findings-