1Department of Surgery, National Cancer Center Hospital, 2Department of Diagnostic Radiology, National Cancer Center Hospital
Due to recent advance of diagnostic modalities, many small (≤ 5 cm) hepatocellular carcinomas has been found in Japan. Before 1980, α-fetoprotein was the main clue to find out a small hepatocellular carcinoma. However, since 1980, ultrasound and computed tomography (CT) has become the main diagnostic methods in detecting a small hepatocellular carcinoma. Serum α-fetoprotein level was lower than 400 ng/ml in 60% of these patients with a small hepatocellular carcinoma. We laparotomized 72 patients with 93 hepatocellular carcinomas less than 5 cm in diameter and resected 86 tumors in 68 patients. Sensitivity of small hepatocellular carcinoma was ultrasound 90.1%, angiography 86.8% CT 91.3%. Ultrasound was the most excellent diagnostic modality in diagnosis of the location of the tumor, because ultrasound can demonstrate the portal and the hepatic venous branches simultaneously. The characteristics of the small hepatocellular carcinoma were mosaic pattern of internal echoes, posterior echo enhancement, thin halo and lateral shadows. These findings were recognized 68%, 61%, 37% and 18% respectively. However, tumors less than 2 cm were demonstrated as a simple low echo area. The liver tumor with mosaic pattern and posterior echo enhancement was seen only in hepatoma. Predictivity of hepatocellular carcinoma with these two findings was 60% in patients with hepatocellular carcinoma from 2.1 to 5 cm in diameter. Prediction of hystology of the liver tumor less than 2 cm in diameter was difficult with any diagnostic modalities. For this purpose, ultrasonically guided percutaneous aspiration cytology is generally accepted as a reliable procedure. However, aspiration cytology made correct diagnosis in 44% of the tumor which could not be predicted with any diagnostic modalities. It is due to low grade malignancy of hepatocellular carcinoma which is very common (33%) among tumors less than 2cm in diameter.