META: an MRI-based scoring system differentiating metastatic from osteoporotic vertebral fractures

BACKGROUND CONTEXT: Magnetic resonance imaging (MRI) is useful for the differential diagnosis of osteoporotic vertebral fractures (OVFs) and metastatic vertebral fractures (MVFs), but no single finding is absolutely conclusive. PURPOSE: The purpose of the present study was to create a scoring system to facilitate the correct diagnosis of MVFs by integrating several MRI findings. STUDY DESIGN: This is a retrospective and single-center observational study that attempts to create a diagnostic scoring system by discriminant analysis. PATIENTS SAMPLE: We included 100 OVFs and 100 MVFs in thoracolumbar vertebrae of which MR images were obtained within 60 days from the suspected time of fractures. OUTCOME MEASURES: The sensitivity and specificity of known important MRI findings were assessed, and the classification accuracy of the scoring system was investigated. METHODS: Seven MRI findings of these fractures were analyzed to evaluate their sensitivity and specificity. Using these findings as variables, discriminant analysis was performed in 140 fractures as a training set, and the classification accuracy was calculated in the remaining 60 fractures as a test set. Additionally, the images of these 60 fractures were reviewed by another blinded reviewer to investigate the interobserver reliability of each finding. RESULTS: All findings had high specificity with low-to-moderate sensitivity. Eight variables were selected in the final discriminant function. A simpler scoring system (MRI Evaluation Totalizing Assessment [META]) was created by approximating the coefficients and the constant term by integral numbers. The classification accuracy was calculated to be 96.6% in the test set. The interobserver reliability of the key findings varied, but the final discrimination conducted by META had the high agreement between the two reviewers (k=0.93). CONCLUSIONS: This novel scoring system, META, could prove to be a useful tool for the differential diagnosis of OVFs and MVFs. It is simple and physician friendly, yet highly accurate. (C) 2015 Elsevier Inc. All rights reserved.