Comparison of gadoxetic acid disodium-enhanced MRI and biphasic spiral CT in detection of hepatocellular carcinoma in patients meeting the Milan criteria

Objective: To prospectively compare gadoxetate disodium-enhanced MRI with biphasic spiral CT in the detection of HCC in patients within the Milan criteria. Methods: The study was designed as a prospective, open-label, within-patient with a corresponding blinded reading. AUCs were also applied to compare for the comparison of combined unenhanced and gadoxetic acid disodium-enhanced MRI versus biphasic spiral CT using the Hanleyand McNeil method. Results: For all liver lesions meeting Milan criteria, the diagnostic accuracy across the three readers was significantly greater with Gadoxetic acid disodium-enhanced MR imaging than with biphasic spiral CT. The AUC of GAD enhanced MRI were 0.873 with 95% CI = 0.801-0.925, 0.928 with 95% CI = 0.869-0.967 and 0.854 with 95% CI = 0.780-0.911 respectively in Reader 1, 2, and 3. The AUC of biphasic spiral CT were 0.724 with 95% CI = 0.638-0.800, 0.705 with 95% CI = 0.617-0.783 and 0.669 with 95% CI = 0.580-0.751 respectively in Reader 1, 2, and 3. The sensitivity across the three readers was significantly greater with Gadoxetic acid disodium-enhanced MR imaging than with biphasic spiral CT. Average across Readers were 0.851 with 95% CI = 0.745-0.962 vs. 0.692 with 95% CI = 0.596-0.793, respectively. Conclusion: Compared with biphasic spiral CT, Gadoxetic acid disodium-enhanced MRI yields significantly higher diagnostic accuracy and sensitivity in the detection of HCC in patients within Milan criteria.