Objective: The purpose of this study was to assess the diagnostic performance of in-phase and opposed-phase chemical-shift imaging (CSI) for differentiating benign and malignant vertebral bone marrow lesions (BMLs).
Materials and methods: The PubMed and EMBASE databases were searched for diagnostic accuracy studies comparing conventional gradient-echo CSI or the Dixon method for differentiating benign and malignant vertebral BMLs with histopathologic or best-value comparator results. Methodologic quality was assessed with the Quality Assessment of Diagnostic Accuracy Studies 2 tool. Bivariate modeling and hierarchical summary ROC (HSROC) modeling were performed to evaluate the diagnostic performance of CSI. For heterogeneity exploration, we performed meta-regression analyses.
Results: Twelve studies including 663 lesions of 591 patients were included. CSI showed a pooled sensitivity of 0.92 (95% CI, 0.84-0.96), pooled specificity of 0.89 (95% CI, 0.81-0.93), and HSROC AUC of 0.95 (95% CI, 0.93-0.97) for differentiating benign from malignant vertebral BMLs. The corresponding values for differentiating benign from malignant compression fractures were 0.96 (95% CI, 0.81-0.99), 0.89 (95% CI, 0.83-0.93), and 0.93 (95% CI, 0.91-0.95). In meta-regression analysis, minimum TR (< 100 ms), flip angle (< 50°), and Dixon method tended to have higher specificity. Study population, slice thickness (< 5 mm), minimum TE (< 2.3 ms), flip angle (< 50°), and blinding also significantly affected heterogeneity (p < 0.05).
Conclusion: In-phase and opposed-phase CSI has excellent diagnostic performance for differentiating benign and malignant vertebral BMLs. CSI with a short TR, small flip angle, and Dixon method is recommended for more accurate diagnosis as specificity increases.
Keywords: bone marrow neoplasm; chemical-shift imaging; compression fracture; meta-analysis; spine.