RESUMEN
BACKGROUND: Sorafenib has demonstrated survival benefit in first-line treatment of advanced hepatocellular carcinoma (HCC); utility of sorafenib in patients with advanced HCC and Child-Pugh B (CP-B) liver function remains a subject of debate. METHODS: A systematic review identified studies using first-line sorafenib in patients with advanced HCC and CP-A/B liver function. Meta-regression analysis comprising linear regression was conducted to explore the association between the baseline factors and overall survival (OS). Differences between efficacy/safety and tolerability parameters were explored using meta-analysis. RESULTS: Thirty studies (12 Asian) comprising 8678 patients (August 2002 - September 2012) were included (four randomised controlled trials, 26 cohort studies). Median age was 61 years and 83% were men. Hepatitis B/C status was positive in 35%/22%, respectively. The CP status was available for 8577 patients (99%); CP-A, 79% and CP-B, 19%. Median OS on sorafenib for entire cohort was 7.2 months; 8.8 months in CP-A and 4.6 months in CP-B. Multivariable meta-regression analysis showed significant negative association between OS and proportion of patients with the Eastern Cooperative Oncology Group performance status 2 (P = 0.04) and CP-B liver function (P = 0.001). Among four studies reporting multivariable comparison of the CP status, CP-B was associated with significantly worse OS (P < 0.001). There were no differences in the response rate to sorafenib between patients with CP-A (4.6%) and CP-B (4.2%) liver function. Safety and tolerability were similar; 35% of patients with CP-A/B liver function developed grade III/IV adverse events (P = 0.7). Meta-regression analysis showed similar rates of treatment discontinuation without progression (P = 0.31) and treatment-related death (P = 0.94) in patients with CP-B liver function. CONCLUSION: CP-B liver function (versus CP-A) is associated with worse OS (but the similar response rate, safety and tolerability of first-line sorafenib, is unlikely to be clinically meaningful).