Objectives: To analyse the association between serum alpha-foetoprotein (AFP) levels and sustained virological response (SVR) in treated patients.
Methods: One-hundred patients with chronic hepatitis C were treated with pegylated interferon alpha-2a plus ribavirin for 48 weeks. The primary endpoint was SVR. Linear regression analysis was performed to identify clinical, biological, and histological factors affecting baseline AFP levels. The association between pretreatment serum AFP and SVR was assessed by multivariate logistic regression analysis.
Results: Of 100 patients, 95 were infected with genotype 4, one with genotype 1, and four with undetermined genotype. The median serum AFP level was 4.5 ng/ml and AFP values ranged from 1.2 to 49.8 ng/ml. In multivariate analysis, higher fibrosis stage and higher steatosis score were independently associated with higher serum AFP levels. SVR rate was 61.0% (61/100), and was lower for patients with AFP levels above rather than below the median value (40.8% versus 80.4%, respectively, P < 0.001). In multivariate analysis, including adjustment for age, gender, body mass index, steatosis score, fibrosis stage, ALT level, haemoglobin level, clotting time, HCV RNA viral load, and treatment dose received, a baseline serum AFP level above the median value was associated with a lower SVR rate (OR [95% CI]=0.10 [0.03-0.42], P < 0.001). None of the seven patients with increased (above 15 ng/ml) pretreatment AFP achieved SVR.
Conclusions: In this study, higher baseline serum AFP levels independently predicted a lower SVR rate among patients with chronic hepatitis C. If confirmed with genotypes other than 4, these findings would suggest adding serum AFP to the list of factors predictive of treatment response.