Abstract
BACKGROUND: There is little information regarding whether the combination of Helicobacter pylori (H. pylori) antibody and serum pepsinogen (sPG), which is a marker of the degree of atrophic gastritis, has a discriminatory ability for detecting incident gastric cancer. We examined this issue in a long-term prospective cohort study of a Japanese population. METHODS: A total of 2446 Japanese community-dwelling individuals aged ≥40 years were stratified into four groups according to baseline H. pylori serological status and sPG: Group A (H. pylori[-], sPG[-]), Group B (H. pylori[+], sPG[-]), Group C (H. pylori[+], sPG[+]), and Group D (H. pylori[-], sPG[+]), and participants were followed up prospectively for 20 years. RESULTS: During the follow-up, 123 subjects developed gastric cancer. Compared with that in Group A, the cumulative incidence of gastric cancer was significantly increased in Groups B, C, and D, whereas no significant difference was found between Groups C and D. The multivariable-adjusted risk of gastric cancer was significantly increased in Group B (hazard ratio [HR], 4.08; 95% confidence interval [CI], 1.62-10.28) and in Groups C and D combined (HR 11.1; 95% CI, 4.45-27.46). When the multivariable model with H. pylori antibody was changed into that with the combination of H. pylori antibody and sPG, the C statistics for developing gastric cancer increased significantly (0.773 vs 0.732, P = 0.005), and the continuous net reclassification improvement value was 0.591 (P < 0.001). CONCLUSIONS: Our findings suggest that the combination of H. pylori antibody and sPG is a useful tool for predicting the development of gastric cancer.