Tumor Biology   2016 年 37 卷 4 期

2016.04.01

Prognosis of patients with advanced hepatocellular carcinoma (HCC) is under expectation. Life expectancy more than 3??months is one inclusion criteria for molecular targeted drugs in clinical trials. The main purpose of this research is to compare Model for End-Stage Liver Disease (MELD) and four MELD-based prognostic models in predicting the survival rate of advanced HCC patients. One hundred eighty-three patients with advanced HCC who were not amendable to standard anti-tumor therapy were retrospectively analyzed. Data were collected to classify patients according to MELD, Model for End-Stage Liver Disease with the incorporation of serum sodium (MELD-NA), Model for End-Stage Liver Disease to ascites and sodium (MELD-AS), integrated Model for End-Stage Liver Disease (iMELD), and Model for End-Stage Liver Disease to sodium (MESO) scores at diagnosis. 1-, 3-, and 6-month survivals were the end points used in the analysis. When predicting 1-month survival, MELD-AS, MELD, and MESO were the top 3 ranking staging systems. When predicting 3-month survival, area under the receiver operating characteristic curve (AUC) of MELD-AS is significantly higher than that of the other models (P < 0.05). When predicting 6-month survival, AUCs of MELD-AS and MELD-NA are significantly higher than those of the other models (P < 0.05). Cutoff point of MELD-AS is 23.11 with 40.5??% sensitivity and 93.8??% specificity at 1??month, 9.5 with 76.9??% sensitivity and 59.5??% specificity at 3??months, and 18.5 with 27.0??% sensitivity and 89.1??% specificity at 6??months. MELD-based scores of death group are significantly higher than those of survivors within 1 and 3??months (P < 0.001). Independent prognostic factors identified by multivariate analysis included persistent ascites, serum sodium, and thrombosis. MELD-AS is the best model in the prediction of short and intermediate survival among the five models for end-stage liver disease analyzed for Chinese advanced HCC patients.