A New Stage and Classification of Osteonecrosis of the Femoral Head Based on MRI With Pathophysiological...

To explore the influence of the new stage and classification system based on the correlation between MRI and pathological features on the collapse, symptoms, and function of femoral head necrosis. Existing radiographic staging systems primarily evaluate structural changes and lesion morphology but provide limited information regarding intraosseous biological activity and pathological heterogeneity. Clinically, patients within the same ARCO stage often demonstrate markedly different symptoms, collapse patterns, and treatment responses, highlighting the need for an MRI-based classification reflecting underlying pathological status.

This retrospective study included 100 patients (200 hips) who were diagnosed with bilateral ONFH. According to the degree of fibrosis in the necrotic area in a mid-coronal section on hip MRI, the MRI stage of ONFH can be divided into three types. F1, borderline fibrosis (banded low signal); F2, partial fibrosis in the necrotic area (partial low signal); F3, global fibrosis in the necrotic area (global low signal). Based on the level of inflammatory response, the MRI classification of ONFH can be divided into four. E0, pathological stable type (no bone marrow edema and inflammatory); E1, partial edema of the femoral head and neck; E2, global edema of the femoral head and neck; E3, extensive edema of the proximal femur. The radiological evaluations were based on plain X-rays and MRI.

There was a significant difference in the collapse rate between F1 (7.1%), F2 (66.3%), and F3 (92.1%) (p < 0.001). The collapse rate in E0 (11.1%), E1 (66.7%), E2 (85.3%), and E3 (77.8%) was significantly different (p < 0.001). There was also a significant difference in pain level and hip function between MRI stage and classification (p < 0.001). The later the stage, the higher the pain level, and similar results in hip functions. With the aggravation of bone marrow edema, the patient's pain level increases and hip function decreases. ROC analysis demonstrated that MRI stage based on fibrosis showed the highest discriminative ability for femoral head collapse (AUC = 0.819), compared with ARCO stage (AUC = 0.756) and MRI classification based on inflammatory response (AUC = 0.714). However, given the cross-sectional design and the potential for incorporation bias, these findings should be interpreted with caution.

MRI-based assessment of fibrosis extent and inflammatory activity provides additional information on intraosseous pathological status that is not captured by conventional radiographic stage systems and is significantly associated with femoral head collapse, pain, and hip function. This classification may have potential value for disease stratification and clinical assessment. However, its role in guiding treatment decisions and predicting longitudinal outcomes requires further validation in prospective studies.