Dariush Nasrollahzadeh

Background:The mechanisms underlying pathways linking Helicobacter pylori infection, gastric atrophy, and esophageal adenocarcinoma have yet to be quantitatively investigated. Methods:Four hundred thirty-five patients with esophageal adenocarcinoma and 1,298 cancer-free subjects were included from the International Barrett’s and Esophageal Adenocarcinoma Consortium (BEACON). Samples were collected from 1993 to 2004. Gastric atrophy was determined by serum pepsinogens. H. pylori infection was assessed by antibodies against 15 antigens. Logistic regression models were fitted to estimate odds ratio (OR) and quantify the association between antibodies and gastric atrophy. A 4-way decomposition of the overall effect into natural direct and indirect effects was used to measure the mediating effects of mediators between H. pylori and esophageal adenocarcinoma while also considering their role as effect modifiers. Results:Seropositivity for 15 H. pylori antigens was associated with a lower risk of esophageal adenocarcinoma; the ORs ranged from 0.44 (95% CI, 0.40–0.49) to 0.91 (95% CI, 0.84–0.98). Seropositivity for multiple antigens was also associated with a reduced risk of esophageal adenocarcinoma. The effect of H. pylori infection was not mediated through gastric atrophy, body mass index (BMI), and gastroesophageal reflux disease (GERD), whether these factors were examined individually or combined, as the estimates of natural indirect effects were close to 0. Conclusions:This study confirmed the inverse association between H. pylori and esophageal adenocarcinoma, which was not mediated by gastric atrophy or the combination with BMI and GERD. Impact:It suggests refining esophageal adenocarcinoma risk prediction models and exploring other potential mediators and pathways between H. pylori and esophageal adenocarcinoma.

Supplementary Table S1. Distribution of esophageal adenocarcinoma (EAC) cases, controls

Supplementary Table S2. H. pylori multiplex serology antigens

Supplementary Table S3. Mediation analysis: each antibody was analyzed individually

Supplementary Table S4. Full output for Med4way with single mediator considered

Supplementary Table S5. Mediation effect of accumulated number of seropositive antigens with NSAID drug

Supplementary Table S6. The association between each antibody against individual H. pylori antigens and esophageal adenocarcinoma (EAC)

Supplementary Table S7. Association between total number of seropositive antigens and esophageal adenocarcinoma (EAC)

Background: The mechanism underlying pathways between H. pylori infection, gastric atrophy, and esophageal adenocarcinoma (EAC) is yet to be quantitatively investigated. Methods: 435 patients with EAC and 1298 cancer free subjects were included from the International Barrett’s and Esophageal Adenocarcinoma Consortium (BEACON). Samples collected from 1993 to 2004. Gastric atrophy was determined by serum pepsinogens. H. pylori infection was assessed by antibodies against 15 antigens. Logistic regression models were fitted to estimate odds ratio, to quantify the association between antibodies and gastric atrophy. The 4-way decomposition of the overall effect in natural direct and indirect effects was used to measure the intermediate effects of mediators between H. pylori and EAC, while also considering their role as effect modifiers. Results: Seropositivity for 15 H. pylori antigens was associated with a lower risk of EAC; the ORs ranged from 0.44 (95% CI; 0.40-0.49) to 0.91 (95% CI; 0.84-0.98). Seropositivity for multiple antigens was also associated with a reduced risk of EAC. The effect of H. pylori infection was not mediated through gastric atrophy, body mass index (BMI), and gastroesophageal reflux disease (GERD), whether these factors were examined individually or jointly, as the estimates of natural indirect effects were close to 0. Conclusions: This study confirmed the inverse association between H. pylori and EAC, which was not mediated by gastric atrophy nor the combination with BMI and GERD. Impact: It suggests refining EAC risk prediction models and exploring other potential mediators and pathways between H. pylori and EAC.