Hepatocellular Carcinoma in Patients with Cirrhosis: Qualitative Comparison of Gadobenate Dimeglumine–enhanced MR Imaging and Multiphasic 64-Section CT

Abstract

To prospectively investigate whether combined interpretation of dynamic and hepatobiliary phase magnetic resonance (MR) images can improve the accuracy of gadobenate dimeglumine-enhanced MR imaging in the detection of hepatocellular carcinoma (HCC) compared with either dynamic MR or multiphasic multidetector computed tomographic (CT) images alone. Institutional review board approval and informed patient consent were obtained. Fifty-two patients (39 men, 13 women; mean age, 68 years; range, 38-81 years) suspected of having HCC underwent gadobenate dimeglumine-enhanced MR imaging and multiphasic 64-section multidetector CT. Images were qualitatively analyzed independently by three observers in three separate reading sessions. The alternating free-response receiver operating characteristic (AFROC) method was used to analyze the results. Differences in sensitivity and positive predictive values were calculated at a statistical significance of P < .05. A total of 67 HCCs were detected in 36 patients. The mean area under the AFROC curve (A(z)) was significantly higher for either the combined interpretation of dynamic and hepatobiliary phase MR images (A(z) = 0.95) or dynamic MR images alone (0.91) than for CT images (0.77) (P = .01 for both comparisons). The mean sensitivity of combined interpretation of MR images (0.72) was significantly higher than those of dynamic MR images alone (0.63) and multidetector CT images (0.61) (P = .008 and .001, respectively). The mean positive predictive value was not significantly different among the three imaging sets. The combined interpretation of dynamic and hepatobiliary phase MR images improves diagnostic accuracy of gadobenate dimeglumine-enhanced MR imaging for the detection of HCC compared with either dynamic MR or multiphasic multidetector CT images alone.