Gadoxetic acid-enhanced MRI for the characterization of hepatocellular carcinoma: A systematic review and meta-analysis

To establish the relative diagnostic accuracy of gadoxetic acid-enhanced magnetic resonance imaging (GA-MRI) compared with contrast-enhanced computed tomography (CE-CT), dynamic MRI (D-MRI), gadopentetic acid-enhanced MRI (GP-MRI), or gadobenic acid-enhanced MRI (GB-MRI) in the characterization of hepatocellular carcinoma (HCC).PubMed, EMBASE, the Cochrane Library, and the University of York CRD databases were searched to February 29 2016 for any studies that compared the diagnostic accuracy of GA-MRI to CE-CT, D-MRI, GP-MRI, or GB-MRI in patients with known or suspected HCC. Diagnostic accuracy outcomes (true positive, true negative, false positive, false negative) were extracted and analyzed using the bivariate model of Reitsma et al (2005).In studies comparing GA-MRI to CE-CT in patients with any-sized lesions, estimated sensitivities were 0.881 (95% confidence interval [CI] = 0.766, 0.944) and 0.713 (95% CI = 0.577, 0.819) respectively. Estimated specificities were 0.926 (95% CI = 0.829, 0.97) and 0.918 (95% CI = 0.829, 0.963), respectively. This difference was not statistically significant. In studies including patients with small lesions GA-MRI was superior to CE-CT, with estimated sensitivities of 0.919 (95% CI = 0.834, 0.962) and 0.637 (95% CI = 0.565, 0.704 and estimated specificities of 0.936 (95% CI = 0.882, 0.966) and 0.971 (95% CI = 0.937, 0.987), respectively. In studies comparing GA-MRI to D-MRI in patients with any-sized lesions estimated sensitivities were 0.907 (95% CI = 0.870, 0.934) and 0.820 (95% CI = 0.776, 0.857); estimated specificities were 0.929 (95% CI = 0.877, 0.961) and 0.934 (95% CI = 0.881, 0.964).GA-MRI has superior diagnostic ability to CE-CT in patients with small lesions. In patients with any-sized lesions there is no evidence that GA-MRI is superior to either CE-CT to D-MRI.3 J. Magn. Reson. Imaging 2017;45:281-290.