Objective: Helical CT scanners allow multiple-phase sequential scans of the entire liver to be obtained during a single bolus injection of contrast material. The purpose of this study was to compare two injection protocols and to establish timing that would optimize detection of hepatomas less than 3 cm in diameter.
Subjects and methods: Triple-phase helical CT of the liver was evaluated in 217 patients who had liver cirrhosis and were referred for known or suspected hepatomas. Proof of individual neoplasms was based on biopsy results, surgical findings, or findings of other imaging studies. Sixty percent nonionic contrast material, infused at 2 or 4 ml/sec, was followed by sequential arterial-phase, portal-venous phase, and equilibrium-phase helical scans of the liver. Aortic and hepatic enhancement curves were constructed by measuring CT attenuation. The CT attenuation values of individual tumor lesions were also measured. We compared the degree of enhancement of normal structures and tumors obtained with four scan protocols (injection at 2 ml/sec with a 30-sec scan delay [n = 54], injection at 2 ml/sec with a 35-sec scan delay [n = 47], injection at 4 ml/sec with a 20-sec scan delay [n = 56], and injection at 4 ml/sec with a 25-sec scan delay [n = 60] and determined the optimal injection protocol and timing for CT acquisition.
Results: Peak aortic and hepatic enhancement was obtained earlier with the 4-ml/sec protocol (at 24 sec and 61 sec versus 36 sec and 90 sec for the 2-ml/sec protocol). The peak attenuation value of the aorta was higher with the 4-ml/sec protocol (330 H versus 186 H for the 2-ml/sec protocol). However, peak hepatic attenuation was similar for both protocols (98 H for the 4-ml/sec protocol versus 92 H for the 2-ml/sec protocol). Liver-tumor contrast was highest in the arterial phase with both protocols. The next highest contrast was obtained during the equilibrium phase. Liver-tumor contrast in the portal-venous phase was significantly lower than that in the other two phases. Tumor enhancement was significantly higher in scans obtained using the 4-ml/sec protocol with a delay time of 25 sec than those obtained with a delay time of 20 sec. In 109 hepatomas, 35 tumors were only seen or were most conspicuous during the arterial phase, four tumors were most conspicuous during the equilibrium phase, and one tumor was most conspicuous during the portal-venous phase.
Conclusion: Arterial-phase helical CT of the liver after 4-ml/sec injection of contrast material significantly improves detection of hepatomas less than 3 cm in diameter when performed in addition to delayed scanning. Portal-venous phase helical CT is of limited value in detecting small hepatomas.