Background: Hepatitis B virus (HBV) infection is reported to be associated with an increased risk of pancreatic cancer (PaC), but it remains controversial whether this is a causal relationship. In addition, it is unclear whether the status of HBV infection also affects PaC risk. Therefore, we conducted a meta-analysis to more closely examine the association between HBV infection and PaC.
Method: The studies included in the meta-analysis were identified and retrieved from PubMed and several other databases. The literature search was conducted up until August 2012. We adopted the Cochrane Collaboration's RevMan 5.1 in a combined analysis of pooled relative risk (RR) with their corresponding 95 % confidence intervals (CIs) using a random-effects and a fixed-effects model.
Results: Nine studies including 6 case-control and 3 cohort studies met eligibility criteria. The meta-analysis showed that the PaC risk was positively correlated with HBV infection when comparing with 'never exposed to HBV' subgroup, the pooled RR was 1.39 (95 % CI 1.22-1.59, p < 0.00001) in chronic HBV carriers, 1.41 (95 % CI 1.06-1.87, p = 0.02) in past exposure to HBV, and 3.83 (95 % CI 1.76-8.36, p = 0.0007) in active HBV infection. Using a stratified analysis, we also found that the risk of PaC was independent of smoking, alcohol drinking, and diabetes.
Conclusion: Findings from this meta-analysis strongly support that HBV infection is associated with an increased risk of PaC.