Objective: The aim of this study was to determine the accuracy of CT arterioportography and hepatic digital subtraction angiography, separately and combined, for the detection of hepatocellular carcinoma in the cirrhotic liver by using thin-section liver explant histopathologic findings.
Subjects and methods: Fifty-nine patients with liver cirrhosis were examined with CT arterioportography and digital subtraction angiography as a part of preoperative diagnostic workup for liver transplantation. Before liver explantation, CT arterioportograms and digital subtraction angiograms were prospectively evaluated in a blinded manner, separately by two CT radiologists and two angiographers, respectively, and combined by two reviewer teams, each including a CT radiologist and an angiographer. In addition, each examination was retrospectively evaluated using direct comparison with the corresponding thin-section liver explant specimens
Results: There were 39 histologically confirmed hepatocellular carcinomas. In both prospective and retrospective assessments, the reviewers achieved the best performance with CT arterioportography and digital subtraction angiography combined (area under the curve [A(z)] 0.82). The diagnostic confidence in the detection of hepatocellular carcinoma was higher with digital subtraction angiography (A(z), 0.81) than that with CT arterioportography (A(z), 0.68). Prospectively, sensitivity and specificity were 75% and 60% for CT arterioportography, 77% and 80% for digital subtraction angiography, and 84% and 81% for CT arterioportography and digital subtraction angiography combined, respectively. Retrospectively, sensitivity and specificity were 80% and 62% for CT arterioportography; 82% and 79% for digital subtraction angiography; 87% and 81% for CT arterioportography and digital subtraction angiography combined, respectively. Five hepatocellular carcinomas, one poorly and four well differentiated, with a mean size of 1.4 cm were not detectable on the CT arterioportography and digital subtraction angiography combination. False-positive findings were 20, 11, and 10 on CT arterioportography, digital subtraction angiography, and the CT arterioportography and digital subtraction angiography combination.
Conclusion: Combining CT arterioportography with digital subtraction angiography enabled reliable detectability of moderately and poorly differentiated hepatocellular carcinomas in cirrhotic livers but was less sensitive for the detection of well-differentiated hepatocellular carcinomas and resulted in a relatively high rate of false-positive findings.