gms | German Medical Science

Objectives: Several classification systems have been proposed to categorize or quantify osteonecrosis of the femoral head (ONFH). However, little is known regarding the relationship among necrotic volume, necrotic angle and location of the necrotic lesion. This study aimed to investigate the relationship among these classification systems and to evaluate their reliability in terms of prognosis.

Methods: This study included 101 hips of 74 consecutive patients with non-collapsed ONFH, which were classified as stage 1 or 2 according to the Association Research Circulation Osseous classification. A three-dimensional spoiled gradient-echo sequence (3D-SPGR) MRI was performed for all hips; the slice interval was 1 mm without inter-slice gaps. First, we investigated the relationship among the Steinberg classification, modified Kerboul method, and Japanese Investigating Committee (JIC) classification by comparing the necrotic volume, combined necrotic angle, and size and location of the necrotic lesion on the mid-coronal image. Necrotic volume was measured according to the Steinberg grade (A, B, and C) using 3D-SPGR MRI. We also measured the combined necrotic angle on the basis of the modified Kerboul method and divided it into four grades (1, 2, 3, and 4). We categorized the size and location of the necrotic lesion at the weight-bearing portion using the JIC classification, which consists of four types (A, B, C1, and C2) and is based on mid-coronal images (Fig 1). We also measured the distance between the mid-coronal and coronal images that had the largest necrotic area. Second, we investigated the occurrence of collapse at a minimum 2-years follow-up (mean; 9 years, range; 2-16 years) and evaluated the reliability in these classification systems to predict the prevalence of collapse.

Results and Conclusion: A positive correlation was observed between the Steinberg grade and modified Kerboul grade (r = 0.81), as well as between the Steinberg grade and the JIC type was also observed (r = 0.69). The mean distance between the mid-coronal and coronal images that had the largest necrotic area was 9.0 mm anteriorly. The collapse rates in the Steinberg classification were 15% for grade A, 50% for grade B, and 80% for grade C; those in the modified Kerboul method were 9% for grade 1, 50% for grade 2, 70% for grade 3, and 88% for grade 4; and those in the JIC classification were 0% for type A, 6% for type B, 61% for type C1, and 78% for type C2.

These three classification systems were equally reliable to predict the prevalence of collapse. While necrotic volume can effectively prognosticate ONFH, its measurement is time-consuming. In contrast, using the modified Kerboul method and the JIC classification is relatively easy. In particular, the JIC classification is the simplest because it is categorized with respect to the mid-coronal image only. We conclude that the JIC classification is as reliable as the other classification systems and while it is practically the easiest for classifying necrotic lesions.