Magnetic Resonance Imaging Quantitative Indices Predict Collapse in Osteonecrosis for Lesions Occupying More Than the Medial Two-Thirds of the Weight-Bearing Portion of the Femoral Head.

Abstract

Accurate prediction of femoral head collapse remains challenging in osteonecrosis of the femoral head (ONFH), particularly in hips classified as Japanese Investigation Committee (JIC) Types C1 or C2. Conventional two-dimensional radiographic classifications provide insights into the three-dimensional morphology of necrotic lesions. This study evaluated the association between three-dimensional magnetic resonance imaging (MRI)-based quantitative indices and femoral head collapse. This retrospective longitudinal study included 95 hips in 69 patients who had nontraumatic ONFH classified as JIC Stages 1 or 2 and Types C1 or C2. Quantitative MRI-based parameters, including the necrotic volume ratio, necrotic depth ratio, and central depth ratio, were measured using coronal and sagittal MRI images and three-dimensional reconstruction. Hips progressing to JIC Stages 3A or higher were classified as the collapse group, whereas those remaining at Stages 2 or lower for at least 36 months were classified as the non-collapse group. There were 46 hips that progressed to collapse, whereas 49 did not. Multivariate logistic regressions and receiver operating characteristic curve analyses were performed. Multivariate analyses identified three independent predictors of collapse: central depth ratio (odds ratio [OR], 1.08; P < 0.05); necrotic depth ratio (OR, 0.91; P < 0.05), reflecting an inverse association after multivariate adjustment; and necrotic volume ratio (OR, 1.08; P < 0.05). The JIC Stage remained significantly associated with collapse, whereas JIC Type was not significant. The areas under the curve (AUC) were 0.73 for the central depth ratio, 0.69 for the necrotic depth ratio, and 0.79 for the necrotic volume ratio. A combined model demonstrated discriminative performance (AUC, 0.84). The MRI-based quantitative indices, including volumetric and slice-based parameters, were independently associated with femoral head collapse in ONFH. Combined assessment of volumetric and depth-related parameters may improve risk stratification beyond conventional radiographic classification and support clinical decision-making.