Identifying gastric cancer risk before cancer suspicion: Helicobacter pylori prevalence and virulence heterogeneity in a multicenter endoscopy cohort.

e22534 Background: Gastric cancer remains a leading cause of cancer mortality worldwide. Helicobacter pylori is a WHO class I carcinogen, yet its detection and characterization in real-world, non-oncologic endoscopy populations remain limited. Identifying modifiable gastric cancer risk before cancer suspicion represents a critical prevention opportunity. Methods: This was a multicenter, cross-sectional study of patients undergoing upper endoscopy for non-oncologic indications at a tertiary center in Mexico. Clinical, endoscopic, histopathologic and lifestyle variables were collected. Gastric biopsies were evaluated by H&E and molecular assays for H. pylori , five virulence-associated genes were identified. Prevalence was estimated with exact 95% Confidence Intervals (CI). Exploratory group comparisons used contingency tables with chi 2 of Fisher exact tests, as appropriate, bilateral p-values were interpreted descriptively. Results: Among 92 biopsies tested molecularly, H. pylori was detected in 28.3% (n = 26, CI 19.4–38.6). In the paired analytic cohort with clinical and questionnaire data (n = 89), H. pylori was identified histologically in 11 samples and differed significantly across gastritis severity categories (p = 0.007), with higher inflammatory activity among positive samples. Premalignant lesions were present in 10% of patients. Molecular profiling revealed substantial virulence heterogeneity, vacA was detected in 85% (n = 22) of H. pylori- positive biopsies, with a wide quantitative range (max. > 17,000 copies), cagA was not detected, while other virulence associated genes showed variable presence. Exploratory analyses suggested associations between H. pylori infection and select socioeconomic and lifestyle factors, including consuming non-purified water. Despite evidence of carcinogenic infection, referred prior eradication therapy was rare. Conclusions: In real-world endoscopy cohort without cancer suspicion, H. pylori infection and inflammation-associated risk states are common and frequently untreated. Results of the characterization and count of virulence-associated genes, suggests risk is not necessarily binary (infected vs not), rather graded, biologically plausible, and measurable at the point of routine care. Treating risk as graded supports pragmatic prevention strategies and endoscopy-based pathways to identify and treat high-risk states.