FOR THE BRESCIA HCC STUDY risk (RR) estimates for these factors. The relationships beWe performed a case-control study to assess the association tween HBV and HCV infections, and between these infections of hepatocellular carcinoma (HCC) with hepatitis B virus and alcohol drinking, may be of relevance and may contrib-(HBV) and hepatitis C virus (HCV) infection and alcohol ute to the discrepancies observed. Mutual negative confounddrinking. We recruited as cases 172 subjects with an initial ing interactions between HBV and HCV infections have been diagnosis of HCC, who were admitted to the two major hospifound in most studies, 3-6 whereas positive interactions have tals in the province of Brescia, northern Italy, and 332 subbeen observed in a few. 3,4 Positive interaction between HBV jects, sex-, age-, and hospital-matched, who were admitted to infection and alcohol intake as a cause of liver injury has the Departments of Ophthalmology, Dermatology, Urology, been suggested in some studies but has not been confirmed Cardiology, and Internal Medicine, as controls. Of the HCC by others. 7,8 Few studies have yet documented the joint effect cases, 23.8% were positive for HBsAg and 37.8% for HCV of HCV infection and alcohol drinking: some data suggest RNA; among the controls, 5.4% were positive for HBsAg and synergism between the two, but further research is needed 4.8% for HCV RNA. History of heavy alcohol intake (ú80 g to clarify the relationship. 9 of ethanol per day for at least 5 years) was found amongIn this study, we address the role of HBV, HCV infection 58.1% of the cases and among 36.4% of the controls. The and its genotypes, and alcohol drinking, and their interacrelative risks (RRs) for HBsAg, HCV RNA positivity, and tions in HCC causation, through a case-control study perheavy alcohol intake were, respectively: 11.4 (95% confidence formed in Brescia, North Italy. This locality showed a high interval: 5.7-22.8), 23.2 (95% confidence interval: 11.8-45.7), incidence of liver cancer as compared with the nine Italian and 4.6 (95% confidence interval: 2.7-7.8). Positive interacareas covered by a Cancer Registry, especially in men: the tions (synergisms) between both HBsAg positivity and HCV incidence rate, age-standardized on the Italian population, RNA positivity and heavy alcohol intake were found, sugwas 32.2 per 100,000 in Brescia versus 15 per 100,000 in gesting more than additive effects of viral infections and alcothe provinces of Turin and Varese, North Italy. 10 hol drinking on the risk of HCC. Infection with HCV genotype 1b showed a higher risk than type 2 (RR Å 2.9; 95% confi-PATIENTS AND METHODS dence interval: 0.9-10), suggesting a major role for the former type in causing HCC. On the basis of population attributable Design and Subjects risks (AR), heavy alcohol intake seems to be the single mostThe study base was the population resident in the province of relevant cause of HCC in this area (AR: 45%), followed by Brescia, North Italy. The province includes the city of Brescia (about HCV (AR: 36%), a...