Hepatocellular Carcinoma

Abstract

To correlate the appearance of hepatocellular carcinoma on delayed (60 minutes) postcontrast T1-weighted gradient echo images with the mode of action of gadobenate dimeglumine (Gd-BOPTA) and the anatomic and pathologic characteristics of the lesions. A total of 34 patients with hepatocellular carcinoma and varying degrees of diffuse liver disease were studied. T2-weighted spin echo and T1-weighted spin echo and gradient echo images were acquired before and 60 minutes after the intravenous administration of 0.1 mmol/kg Gd-BOPTA. Qualitative and quantitative evaluations of the images were performed and correlated with histologic findings. The quantitative evaluation, performed on T1-weighted gradient echo images, looked at the percentage increase of liver enhancement after Gd-BOPTA administration, the lesion-to-liver contrast/noise (C/N) ratio before and after Gd-BOPTA administration, and the C/N variation after Gd-BOPTA administration. Qualitative assessment considered the morphologic features of the lesions as well as the visual variation of contrast before and after Gd-BOPTA administration. Finally, a histologic evaluation was made of the degree of differentiation of the lesions and of the presence of fatty metaplasia, necrosis, bile, or intratumoral peliosis. Among the parameters affecting lesion identification were the extent of liver function, degree of vascularization, residual functionality of the tumor cells, and characteristics of the neoplastic tissue. Positive correlations (Spearman coefficients = 0.359 and 0.393, respectively) were observed precontrast between the degree of liver failure and the amount of contrast noise, and postcontrast between the amount of intralesional fatty metaplasia and the extent to which lesion conspicuity worsened after Gd-BOPTA administration. An inverse correlation (Spearman coefficient = -0.330) was observed between the degree of lesion differentiation and the visible appearance after Gd-BOPTA administration, with well-differentiated lesions tending toward worsened conspicuity postcontrast. A statistically significant difference (P = 0.001) was observed in the mean precontrast C/N ratio for lesions later showing unchanged conspicuity and worse conspicuity on postcontrast images, respectively. Marked variation (P = 0.019) was also observed between Child A and B cirrhotic patients for the degree of hepatic enhancement on postcontrast images. The results suggest that liver parenchyma signal intensity is influenced by the extent to which liver function is compromised, that residual hepatocytic functionality permits Gd-BOPTA uptake by certain lesions and that this uptake might subsequently impair the observed C/N ratio on delayed images, and that the worsening of lesion conspicuity on postcontrast images is influenced also by high quantities of intralesional fatty metaplasia.