Prospective evaluation of gadoxetic acid magnetic resonance for the diagnosis of hepatocellular carcinoma in newly detected nodules ≤2cm in cirrhosis.

Abstract

Most of the published studies about the diagnostic accuracy of gadoxetic acid-enhanced magnetic resonance (EOB-MR) for the non-invasive diagnosis of hepatocellular carcinoma (HCC) have had a retrospective design. Thus, we aimed to prospectively evaluate the diagnostic accuracy of EOB-MR for the non-invasive diagnosis of HCC in nodules ≤2cm detected by screening ultrasound (US) in patients with cirrhosis. Between July 2012 and October 2015, 62 consecutive asymptomatic Child-Pugh A-B cirrhotic patients with newly US-detected solitary nodules between 1 and 2cm were prospectively included in the study. Hepatic extracellular contrast-enhanced MR (ECCE-MR) followed by EOB-MR were obtained in less than 1-month interval. Two independent radiologists blindly reviewed the EOB-MR studies, and the diagnosis of HCC was assigned when the lesion showed arterial enhancement followed by portal venous phase washout and/or hypointensity on the hepatobiliary phase (HBP). The final HCC diagnosis was made by ECCE-MR according to the accepted non-invasive criteria, or by biopsy in lesions with atypical vascular profile. Final diagnoses were as follows: HCC (n=41), intrahepatic cholangiocarcinoma (n=2), colorectal metastases (n=1) and benign conditions (n=18). The sensitivity and specificity of EOB-MR for HCC diagnosis were 56.1% (95% CI: 39.7-71.5) and 90.5% (95% CI: 69.6-98.8), respectively, while sensitivity of ECCE-MR was 63.4% (95% CI: 46.9-77.9). The low rate of hypointense HCCs in the HBP and suboptimal liver uptake of contrast agent justify the low sensitivity of EOB-MR for HCC diagnosis. EOB-MR does not surpass the diagnostic accuracy of ECCE-MR for non-invasive diagnosis of HCC in nodules ≤2cm in cirrhotic patients.