BACKGROUND AND AIM: The outcomes of sorafenib therapy in patients with advanced hepatocellular carcinoma (HCC) and impaired liver function remain unresolved. Although Child-Pugh (CP) classification is widely used for patient categorization, heterogeneity within a given CP class makes outcomes less predictable. The aim was to investigate the prognostic significance of CP score elements on the outcome of sorafenib in patients with advanced HCC and impaired liver function. METHODS: Of 1385 consecutive patients with advanced HCC in our center between January 2007 and December 2010, we reviewed the medical records of 325 patients who received sorafenib monotherapy. RESULTS: Median duration of sorafenib was 2.0 months (range 0.4-24.2) and median follow-up was 4.9 months (range 0.5-43.4). Disease control rates were significantly higher in CP class A (CPA) than in CP class B (CPB) patients. Median overall survival (OS) was 5.8 months. Subgroups with different CP scores showed significantly different OS (months): CPA5, 8.4; CPA6, 5.1; CPB7, 3.5; CPB8-9, 2.6 (P < 0.001). The presence of ascites was a significant prognostic factor in CPB7 patients (hazard ratio 2.262; P = 0.016). OS of CPB7 patients without ascites was similar to that of CPA6 patients (4.6 months) and was significantly longer than that of CPB7 patients with ascites (2.5 months; P = 0.027). OS of CPB7 patients with ascites was similar to that of CPB8-9 patients. CONCLUSIONS: CP score was more important than CP class in predicting the outcome of sorafenib therapy in patients with advanced HCC. Among the CP score components, presence of ascites was a significant prognostic factor, especially in CPB7 patients.
BACKGROUND AND AIM: The outcomes of sorafenib therapy in patients with advanced hepatocellular carcinoma (HCC) and impaired liver function remain unresolved. Although Child-Pugh (CP) classification is widely used for patient categorization, heterogeneity within a given CP class makes outcomes less predictable. The aim was to investigate the prognostic significance of CP score elements on the outcome of sorafenib in patients with advanced HCC and impaired liver function. METHODS: Of 1385 consecutive patients with advanced HCC in our center between January 2007 and December 2010, we reviewed the medical records of 325 patients who received sorafenib monotherapy. RESULTS: Median duration of sorafenib was 2.0 months (range 0.4-24.2) and median follow-up was 4.9 months (range 0.5-43.4). Disease control rates were significantly higher in CP class A (CPA) than in CP class B (CPB) patients. Median overall survival (OS) was 5.8 months. Subgroups with different CP scores showed significantly different OS (months): CPA5, 8.4; CPA6, 5.1; CPB7, 3.5; CPB8-9, 2.6 (P < 0.001). The presence of ascites was a significant prognostic factor in CPB7 patients (hazard ratio 2.262; P = 0.016). OS of CPB7 patients without ascites was similar to that of CPA6patients (4.6 months) and was significantly longer than that of CPB7 patients with ascites (2.5 months; P = 0.027). OS of CPB7 patients with ascites was similar to that of CPB8-9 patients. CONCLUSIONS: CP score was more important than CP class in predicting the outcome of sorafenib therapy in patients with advanced HCC. Among the CP score components, presence of ascites was a significant prognostic factor, especially in CPB7 patients.