Prediction of Collapse in Femoral Head Osteonecrosis: A Modified Kerboul Method with Use of Magnetic Resonance Images

Background: The hypothesis that the combined necrotic angle measurement from magnetic resonance imaging scans predicts the subsequent risk of collapse in hips with femoral head necrosis was tested. Methods: Thirty-seven hips with early stage osteonecrosis in thirty-three consecutive patients were investigated. With use of the modified method of Kerboul et al., we measured the arc of the femoral surface involved by necrosis on a midcoronal as well as a midsagittal magnetic resonance image (rather than an anteroposterior and a lateral radiograph) and then calculated the sum of the angles. On the basis of the magnitude of the resulting combined angle, hips were classified into four categories: grade 1 (<200°), grade 2 (200° to 249°), grade 3 (250° to 299°), and grade 4 (≥300°). After the initial evaluations, the hips were randomly assigned to a core decompression group or a nonoperative group. Patients underwent regular follow-up until femoral head collapse or for a minimum of five years. Results: Seven grade-4 hips and sixteen grade-3 hips had development of femoral head collapse by thirty-six months. Six of the nine grade-2 hips and none of the five grade-1 hips collapsed (log-rank test, p < 0.01). None of the four hips with a combined necrotic angle of ≤190° (the low-risk group) collapsed, all twenty-five hips with a combined necrotic angle of ≥240° (the high-risk group) collapsed, and four (50%) of the eight hips with a combined necrotic angle between 190° and 240° (the moderate-risk group) collapsed during the study period. Conclusions: The Kerboul combined necrotic angle, as ascertained with use of magnetic resonance imaging scans instead of radiographs, is a good method to assess future collapse in hips with femoral head osteonecrosis. Level of Evidence: Prognostic Level I. See Instructions to Authors on jbjs.org for a complete description of levels of evidence.