Interferon and prevention of hepatocellular carcinoma in viral cirrhosis: an evidence-based approach

Background/Aims: To evaluate by meta-analysis of available literature wheth er interferon (IFN) reduces the incidence of hepatocellular carcinoma (HCC) in patients with hepatitis B virus (HBV) or hepatitis C virus (HCV)-relate d Child A cirrhosis. Methods: Three randomized controlled trials and 15 nonrandomized controlled trials, including 4614 patients and comparing IFN to no treatment, were se lected. Data on the incidence of HCC in IFN treated and untreated patients were extracted from each study. Meta-analysis by the DerSimonian and Laird risk difference (RD) method was used to pool observations. Results: A different incidence of HCC between treated and untreated cirrhot ic patients was observed for HCV (overall RD -12.8%; 95% CI -8.3 to -17.2%, P < 0.0001) and HBV (overall RD -6.4%; 95% CI - 2.8 to -10%, P < 0.001). I n HCV-related cirrhosis, the rate of HCC development was lower in sustained responders to IFN than in untreated patients (overall RD -19.1%; 95% CI - 13.1 to similar to 25.2, P < 0.00001). With low heterogeneity among trials (P = 0.053), and also in nonresponders vs, untreated patients (overall RD - 11.8%; 95% CI -6.4 to -19.1%, P < 0.0001), although with significant hetero geneity. Inconsistency among the studies was a major problem, both for HCV (chi (2) = 58.16 with 13 DF; P < 0.0001) and HBV (<chi>(2) = 26.4 with 6 DF ; P = 0.0001) related cirrhosis, and also when follow-up was shorter than 6 0 months. Consistent results were only observed when assessing data from Eu ropean reports: in this subgroup no preventive effect of HCC was shown for HBV (overall RD - 4.8%; 95% CI - 11.1-1.5%, P, not significant), and only a weak effect for HCV (overall RD -10%; 95% CI -5.9 to -14.2%; P < 0.0001). Conclusions: Literature data pooling suggests a slight preventive effect of IFN on HCC development in patients with HCV-related cirrhosis. The magnitu de of this effect is low and the observed benefit might be due to spurious associations. The preventive effect is more evident among sustained respond ers to IFN. IFN does not seem to affect the rate of HCC in HBV-related cirr hosis. (C) 2001 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved.