Objectives: Risk of pancreatic cancer between Helicobacter pylori infected and noninfected persons is controversial, and therefore a meta-analysis was performed.
Methods: PubMed was searched up to September 2014. Only population-based nested case-control studies comparing the serological prevalence of Helicobacter pylori between pancreatic cancer cases and cancer-free controls were eligible. Pooled odds ratios (ORs) and 95% confidence intervals (CIs) for pancreatic cancer risk between Helicobacter pylori infected and noninfected persons were estimated.
Results: Five eligible nested case-control studies were included, with 1446 pancreatic cancer cases and 2235 cancer-free controls. On the whole, the proportion of pancreatic cancer cases among those infected with Helicobacter pylori was not significant different from those noninfected (OR, 0.99; 95% CI, 0.65-1.50; P = 0.96). Likewise, seropositivity of cytotoxin-associated gene A (CagA) showed nonsignificant association with pancreatic cancer (OR, 0.92; 95% CI, 0.65-1.30; P = 0.63). The CagA-positive virulent strains of Helicobacter pylori did not increase the risk of pancreatic cancer (OR, 0.97; 95% CI, 0.50-1.89; P = 0.93). However, CagA-negative nonvirulent strains of Helicobacter pylori had a significant increased risk for pancreatic cancer (OR, 1.47; 95% CI, 1.11-1.96; P = 0.008).
Conclusions: The CagA-negative non-virulent strains of Helicobacter pylori may be a potential risk factor of pancreatic cancer. High-quality prospective large-scaled studies are required for more conclusive results.