Search

The clinical utility of color Doppler sonography in the differentiation of hepatocellular carcinoma from metastases and hemangioma was investigated in 72 hepatocellular carcinomas (80 lesions), 30 metastases (82 lesions), and 39 hemangiomas (54 lesions). Overlaps of color patterns were found among hepatocellular carcinoma, metastases and hemangioma. Pulsatile waves from lesions with the basket, vessels within tumor, or spot patterns, or lesions measuring less than 3 cm with detectable signals, did not favor the diagnosis of hemangioma. In conclusion, color Doppler sonography can aid in the differentiation of hepatocellular carcinoma from hemangioma but may be unreliable in the differentiation of hepatocellular carcinoma from hypervascular metastases.

Color Doppler sonography was used to investigate the changes in intrahepatic portal hemodynamics in early stage hepatic abscesses (21 lesions). At time of patients' admission, 10 lesions measuring over 6 cm showed reverse flow from at least one of their corresponding segmental portal branches, but 11 lesions measuring 6 cm or less did not show this phenomenon. All portal branches with reverse flow eventually returned to a normal direction of flow after successful medication. In conclusion, a large early stage hepatic abscess may show temporary reverse portal flow on color imaging. Thus, a differential diagnosis of hepatic tumors with reverse portal flow should include early stage hepatic abscesses.

Percutaneous intravascular ethanol injection of the supplying vessel of the tumor using color Doppler imaging was performed in three patients (three lesions) with hepatocellular carcinoma measuring more than 3 cm. Intravascular injection was achieved in six out ot seven attempts. The total amount of ethanol injected intravascularly was small (10, 15,0 and 26 ml for each patient). One attempt failed and the ethanol was injected perivascularly. No serious complication was noted. After treatment, two patients showed remarkable decrease in tumor size and one showed massive tumor necrosis. In conclusion, ethanol injection using the intravascular approach may be of value in the palliative management of large hepatocellular carcinomas.

Ninety-seven patients with hepatocellular carcinoma had duplex scanning from the patent or residual portal lumen and comparative angiography. Duplex ultrasonography identified five of nine patients with angiographic evidence of arterioportal shunting. Furthermore, Doppler results indicated the presence of the shunt in one patient without angiographic evidence. Duplex scanning from the portal lumen may be useful in the detection of arterioportal shunting. This procedure can be used to evaluate the hemodynamic change of the recipient portal vein under natural physiologic conditions. However, small shunts without hepatofugal portal blood cannot be detected by this procedure.

Thirty-one patients with hepatocellular carcinoma underwent duplex scanning of their lesions with comparative angiography and serum alpha-fetoprotein determinations before and after transcatheter arterial embolization. Duplex ultrasound correctly identified 26 patients with patent tumor vessels after embolization. The Doppler signals became undetectable after embolization in the other five patients. Comparative angiography demonstrated good devascularization of their lesions. However, tiny tumor blush or small satellite nodules around the periphery of the devascularized masses were found. Therefore, the presence of signals after therapy indubitably needs further embolization. The timing of the second embolization for the patient with undetectable Doppler signals after the first embolization needs to be justified by further evaluations.

Sixteen patients with hepatocellular carcinoma had duplex pulsed Doppler scanning of the portal thrombus and comparative angiography. Fifteen patients demonstrated detectable Doppler signals from the thrombus. The Doppler signals were corroborated by the "thread-and-streaks" sign on angiography in eight patients. This suggests that signals come from tumor vessels within the thrombus. Therefore, the presence of Doppler signals in the thrombotic portal venous system is not only due to partial obstruction of the lumen. Tumor thrombus from hepatocellular carcinoma must be considered, especially when the detectable signals are pulsatile waves.