Objective: The distinction between benign (chemical) and tumor thromboses of the portal vein after treat,ent with percutaneous injection of ethanol for hepatocellular carcinoma is crucial for the proper management of the patient. The purpose of this study was to determine whether color Doppler sonography can be used to differentiate between the two types of thrombi.
Subjects and methods: Between October 1991 and April 1994, portal vein thrombosis was detected by color Doppler sonography in 19 patients (13 men and six women 59-77 years old; mean age, 67 years) who had hepatocellular carcinomas and who had received percutaneous ethanol injection (n = 11) or percutaneous ethanol injection after transcatheter arterial embolization (n = 8). The criterion for diagnosing tumor thrombosis by color Doppler sonography was the detection of pulsatile arterial flow in the thrombus. The benign or malignant nature of the thrombosis was subsequently established by percutaneous fine-needle biopsy of the thrombus; malignant thrombosis was seen in 13 patients, and chemical thrombosis was seen in six patients.
Results: Pulsatile arterial flow in the thrombus was observed by color Doppler sonography in 12 of the 13 malignant thrombi and in none of the bland thrombi. The flow was hepatopetal in seven cases and hepatofugal in five cases. The peak systolic frequency shift ranged from 0.59 to 2.65 kHz (mean, 1.35 kHz), and the resistive index ranged from 0.37 to 0.69 (mean, 0.55). The sensitivity and the specificity of color Doppler sonography for the detection of tumor thrombosis were 92% and 100%, respectively.
Conclusion: Our study shows that color Doppler sonography is a reliable way to differentiate between chemical and tumor thromboses of the portal vein in patients with hepatocellular carcinomas treated by ethanol injection. When the sonogram shows pulsatile arterial flow within the thrombus, percutaneous biopsy of the thrombus is unnecessary. The finding is always indicative of malignant thrombosis.