Background: We aimed to evaluate the safety, efficacy and prognostic impact of baseline and early clinical markers in both Child-Pugh A and B patients with advanced hepatocellular carcinoma (HCC).
Methods: We prospectively studied 89 Japanese patients with HCC (Child-Pugh A, n = 59; Child-Pugh B, n = 30) who were started with sorafenib between May 2010 and July 2013.
Results: Frequency of sorafenib-related adverse events was almost similar between Child-Pugh score 5, 6, and 7 patients. The rate of liver dysfunction, including any grade encephalopathy, ≥ grade 3 ascites, or ≥ grade 3 bilirubin increased, in Child-Pugh score ≥8 group was significantly higher than that in the other groups. The median overall survival of Child-Pugh score 5, 6, 7 and ≥8 patients were 14.5, 11.1, 8.7 and 4.6 months, respectively. Patients in Child-Pugh score 6 had significantly longer OS than those in Child-Pugh score 7 (P = 0.049). Multivariate analysis identified macrovascular invasion (MVI), alpha-fetoprotein (AFP), Child-Pugh score and aspartate aminotransferase (AST) as baseline predictors of survival. However, extrahepatic metastasis (EHM) was not a significant prognostic factor. In addition, decrease in AFP level and development of hand-foot skin reaction within 4 weeks after sorafenib initiation were closely associated with favorable survival.
Conclusion: It is possible that not only Child-Pugh score 5 and 6 but also 7 patients are eligible for future clinical trials with sorafenib or similar drugs. Various survival predictors identified in this study might be considered as stratification factor. Although both MVI and EHM is a phenotype of advanced HCC, MVI should be discriminated from EHM because of the prognostic impact on survival in sorafenib-treated advanced HCC patients.