Objective: At present, the differences in the efficacy between PET/CT and MRI for the detection of bone metastases in patients with cancer have not been clearly delineated. We performed a meta-analysis to compare the performance of FDG PET/CT with that of gadolinium-enhanced MRI for the detection of bone metastases in patients with cancer.
Methods: Studies about PET/CT and MRI for the detection of bone metastases were systematically searched in the MEDLINE, EMBASE, and EBM Review databases. We calculated sensitivities, specificities, diagnostic odds ratios, positive likelihood ratios, negative likelihood ratios (NLR), and constructed summary receiver operating characteristic curves using bivariate regression models for PET/CT and MRI, respectively.
Results: Across 9 studies (1116 patients), FDG PET/CT has similar patient-based sensitivity (0.803 vs 0.837), specificity (0.989 vs 0.977), diagnostic odds ratio (309.0 vs 221.9), positive likelihood ratio (61.7 vs 37.0), and negative likelihood ratio (0.200 vs 0.167) with gadolinium-enhanced MRI. Areas under the curve with 95% confidence interval for FDG PET/CT and gadolinium-enhanced MRI were 0.99 (0.98-0.99) and 0.98 (0.97-0.99), respectively.
Conclusions: FDG PET/CT and gadolinium-enhanced MRI have excellent diagnostic performance for the detection of bone metastases in patients with cancer.