Abstract
PURPOSE: Surgical resection remains the gold standard for hepatocellular carcinoma (HCC). Although various staging systems have been developed in recent years, the best tool for staging of HCC remains controversial. The aims of this study were to establish a new staging for patients with HCC undergoing surgical resection and to identify whether this staging is superior to other staging systems in predicting survival of resectable HCC. METHODS: The subjects of this retrospective study were 958 consecutive HCC patients who underwent surgical resection between 2000 and 2006. Predictors of survival were identified using the Kaplan-Meier method and the Cox model. The disease state was staged by our proposed Eastern staging system by integrating independent risk predictors, as well as six existing staging systems. The accuracy of prediction of 1-, 3-, and 5-year mortality for each system was evaluated by the area under the receiver operating characteristic curve (AUC). RESULTS: Macroscopic vascular invasion, multiple tumors, performance status 1-2, microscopic vascular invasion, extrahepatic spread, tumor size > 5 cm, albumin < 35 g/L, aspartate aminotransferase > 40 U/L, total bilirubin > 17 μmol/L, and presence of cirrhosis were identified as independent risk factors of survival after resection by multivariate analysis. The comparison of the results of the different staging systems showed that our Eastern staging had the best homogeneity (likelihood ratio χ(2) test 543.51, P < 0.001), monotonicity of gradients (linear trend χ(2) test 414.97, P < 0.001), and discriminatory ability (the highest AUCs for 1-, 3-, and 5-year mortality). CONCLUSIONS: Compared with other existing staging systems, our proposed Eastern staging system shows a superior predictive ability in a Chinese cohort of patients with resectable HCC, and it can give important prognostic information after surgery.