Does the method of Helicobacter pylori detection influence the association with gastric cancer risk?

Background: The exact role of Helicobacter pylori as a causative agent of gastric cancer is still under debate. The aim of this study was to determine how the use of different diagnostic methods for detection of H. pylori influences the measures of prevalence of the infection and thus the association with risk of gastric adenocarcinoma. Methods: We included 72 cases and 324 controls in an endoscopy clinic-based matched case-control study. Culture of H. pylori and immunohistochemical staining were performed on gastric biopsies. Serum samples were tested for H. pylori IgG by conventional ELISA and by immunoblotting. Results: The overall prevalence of H. pylori was 68% based on all 4 diagnostic methods, 79% in the cases and 66% in the controls. Highest agreement, 91%, was observed between culture and immunohistochemistry with a Kappa value of 0.81. Immunoblotting detected the highest number of H. pylori-positive subjects in both cases and controls. The association of H. pylori positivity with gastric cancer was generally weaker and statistically non-significant using culture and immunohistochemistry compared with the serological tests, of which IgG ELISA yielded the higher odds ratio (OR 2.5, 95% confidence interval 1.4-4.4). Conclusion: The study shows that relative risk estimates for the association between H. pylori and gastric cancer risk are to some extent determined by the diagnostic method used to detect H. pylori infection.