CT differentiation of cholangiocarcinoma from periductal fibrosis in patients with hepatolithiasis

OBJECTIVE. The objective of our study was to determine useful CT findings for differentiating cholangiocarcinoma from periductal fibrosis in patients with hepatolithiasis. MATERIALS AND METHODS. CT images of 30 patients with hepatolithiasis and pathologically proven cholangiocarcinoma (n = 14) or periductal fibrosis (n = 16) were retrospectively reviewed. Helical CT scans were obtained before, 30 seconds after, and 65 seconds after the start of contrast material injection. Analysis of CT findings included evaluation for the presence of periductal soft-tissue density, bile duct wall thickening at the stricture site, ascites, portal vein obliteration, lymph node enlargement, and a duct stone; assessment of the degree of ductal dilatation; and evaluation of the enhancement pattern of periductal lesions, thickened ductal wall, and hepatic parenchyma. The CT attenuation coefficients of the thickened ductal wall and adjacent normal-looking bile duct were measured on images obtained during each phase. Among these findings, statistically significant variables were determined using the Fisher's exact test and Student's t test. Sensitivity and specificity values of the CT criteria were also calculated. RESULTS. The presence of periductal soft-tissue density (p = 0.002), higher enhancement of the duct than adjacent bile duct on portal venous phase images (p = 0.008), ductal wall thickening (p = 0.026), portal vein obliteration (p = 0.031), and lymph node enlargement (p = 0.031) were found to be the significant findings for differentiating cholangiocarcinoma from fibrosis in patients with hepatolithiasis. When any two or more of these five criteria were used in combination, we could identify 100% of the patients with cholangiocarcinoma but only 12.5% of the patients with fibrosis. CONCLUSION. Cholangiocarcinoma in patients with hepatolithiasis can be diagnosed using specific CT criteria.