Sensitivity of superselective arteriography for small hepatocellular carcinoma compared with proximal arteriography and computed tomography during superselective arteriography.

PURPOSE To obtain successful arterial chemoembolization for hepatocellular carcinoma (HCC), we evaluated the sensitivity of proximal arteriography, superselective (subsegmental or more distal branch) arteriography and computed tomography (CT) during superselective arteriography and assessed the method for the injection of contrast medium. METHODS Thirty-two patients with 38 HCCs (< or = 5 cm) with a mean diameter of 2.2 cm underwent digital subtraction arteriography of proximal and superselective arteriography. In addition, they also had helical CT during superselective arteriography. The contrast medium was injected with a mechanical injector (n = 6 lesions) or by hand (n = 32) for superselective arteriography and CT during superselective arteriography. The amount of contrast medium used for superselective arteriography and CT during superselective arteriography with the mechanical injector was 3.5 times and 9 times that with manual injection, respectively. RESULTS Overall, 31 lesions (81.6%) were detected by proximal arteriography, 25 (65.8%) by superselective arteriography and 35 (92.1%) by CT during superselective arteriography. CT during superselective arteriography was significantly superior to superselective arteriography (P = 0.005). In both studies, manual injection of contrast medium had a significantly higher sensitivity than mechanical injection (P = 0.013). CONCLUSION To detect small HCC, CT during superselective arteriography showed significantly higher sensitivity than superselective arteriography. Manual injection of contrast medium was significantly superior to mechanical injection. Therefore, manual injection CT during superselective arteriography is recommended for accurately targeted, transarterial chemoembolization therapy.