Hepatocellular carcinoma in liver transplantation candidates: detection with gadobenate dimeglumine-enhanced MRI.

Abstract

The purpose of this study was to retrospectively evaluate the diagnostic performance of dynamic gadobenate dimeglumine-enhanced MRI with explant pathologic correlation in the detection of hepatocellular carcinoma (HCC) in patients undergoing liver transplantation. Forty-seven patients (28 men, 19 women; mean age, 49 years) underwent dynamic gadobenate dimeglumine-enhanced MRI within 3 months before primary liver transplantation. Dynamic imaging was performed before (unenhanced) and after (hepatic arterial, portal venous, equilibrium, and 1-hour delayed phases) IV bolus administration of gadobenate dimeglumine at 0.1 mmol/kg body weight. Retrospective image analysis to detect HCC nodules was performed independently by two abdominal radiologists who had no pathologic information. On a per-nodule basis, the sensitivity and positive predictive value were calculated for the two observers. Sensitivity and specificity in the diagnosis of HCC also were evaluated. Fisher's exact test was performed to determine whether there was a detection difference between HCC nodules 1 cm in diameter or larger and nodules smaller than 1 cm and to evaluate the differences in causes of false-positive MRI findings based on lesion size (>or= 1 cm vs < 1 cm). Twenty-seven patients had 41 HCCs. In HCC detection, gadobenate dimeglumine-enhanced MRI had a sensitivity of 85% (35 of 41 HCCs) and a positive predictive value of 66% (35 of 53 readings) for observer 1 and a sensitivity of 80% (33 of 41 HCCs) and a positive predictive value of 65% (34 of 52 readings) for observer 2. For both observers, sensitivity in the detection of HCCs 1 cm in diameter and larger (91-94%) was significantly different (p < 0.05) from that in detection of HCCs smaller than 1 cm (29-43%). Nonneoplastic arterial hypervascular lesions more often caused false-positive diagnoses of lesions smaller than 1 cm in diameter (80-86%) on MR images than of those 1 cm in diameter and larger (0-25%). The difference was statistically significant (p < 0.05) for both observers. In diagnosis, gadobenate dimeglumine-enhanced MRI had a sensitivity of 87% (20 of 23 patients) and a specificity of 79% (19 of 24 patients) for both observers. Dynamic gadobenate dimeglumine-enhanced MRI has a sensitivity of 80-85% and a positive predictive value of 65-66% in the detection of HCC. The technique, however, is of limited value for detecting and characterizing lesions smaller than 1 cm in diameter.