BACKGROUND: The long-term prognosis of patients with hepatocellular carcinomas is still disappointing primarily because of intrahepatic recurrence. Among many factors, portal vein invasion is considered to be the most important risk factor leading to recurrence, and it is thought to be the direct evidence of tumor invasiveness. In clinical aspects, it seems to be more practical to determine the factors associated with portal vein invasion. METHODS: Three hundred and seventy-one patients who underwent a curative hepatic resection for hepatocellular carcinomas between 1991 and 1995 at Seoul National University Hospital were included in this study. Portal vein invasion was identified histopathologically. The prognostic factors of hepatocellular carcinoma and the risk factors linked to portal vein invasion were analyzed by both univariate and multivariate analyses. RESULTS: Portal vein invasion was detected in thirty seven patients (10%), and the 5-year overall and disease-free survival rates in this group were 39.9% and 6.5%, respectively. The 5-year overall and disease-free survival rates of patients without portal vein invasion were 60.1% and 36.8%, respectively. In multivariate analysis using Cox's proportional hazards model, portal vein invasion was proven to be the most important risk factor in both overall and disease-free survival. In a multiple stepwise logistic regression analysis, the size and the number of the tumors were strong independent predictors of portal vein invasion by a hepatocellular carcinoma. CONCLUSIONS: Intrahepatic recurrence and patient survival in cases of hepatocellular carcinomas were closely related to portal vein invasion. Patients with tumor sizes larger than 4 cm or with multiple tumors should be monitored closely for early recurrence.
Carcinoma, Hepatocellular* ; Disease-Free Survival ; Humans ; Logistic Models ; Multivariate Analysis ; Portal Vein* ; Prognosis ; Proportional Hazards Models ; Recurrence ; Risk Factors* ; Seoul