ABSTRACT
This study was carried out in order to assess the accuracy of ultrasound in a) screening for hepatocellular carcinoma in patients hospitalized for cirrhosis and b) determining the possibility of hepatic resection. From January 1983 to January 1987, 492 patients with cirrhosis were investigated for hepatocellular carcinoma using clinical examination, serum alphafetoprotein measurement and ultrasound study. Four hundred and thirty-seven patients had alcohol-related cirrhosis (88.8 p. 100); ascites was found in 280 cases (56.9 p.100). Ultrasonography-guided aspiration cytology was performed whenever a tumor was found. Four hundred and nineteen patients underwent ultrasonography and alpha-fetoprotein chemistry. The diagnosis of hepatocellular carcinoma was confirmed in 66 of 419 patients (15.8 p. 100). Sixty of the 88 patients diagnosed as having tumor on ultrasound were found to have hepatocellular carcinoma. In 6 out of 66 patients (9.1 p. 100), the tumor was not identified by ultrasonography but the level of alphafetoprotein was high (greater than 1,000 ng/ml). Sixteen cancers (24.2 p. 100) were diagnosed by ultrasonography only. All patients were men with alcohol-related cirrhosis, their ages ranging from 52 to 82 years. The results of ultrasound investigation were: 8 solitary tumors, 5 multicentric tumors, 3 diffuse tumors. Only 2 of the 16 tumors were resectable. We conclude that ultrasound is an accurate procedure in the diagnosis of hepatocellular carcinoma associated with cirrhosis. However, in hospitalized patients, the value of screening for hepatocarcinoma is small because liver resection is often impossible due to the wide spread tumor involvement or advanced liver cirrhosis at the time of diagnosis. On the other hand, screening should be performed in patients with compensated cirrhosis.
