P-25 Questionnaire-assisted surveillance strategy for patients with precancerous conditions of gastric cancer

Surveillance decisions for gastric cancer [GC] mostly focused on advanced-stage lesions and were made based baseline severity/extension diagnosis of non-neoplastic according to current guidelines. However, misclassification may occur due sampling error or endoscopy quality in practice. We aimed develop a simple risk-stratification strategy aid the decision-making surveillance. In multicenter community-based cohort, 75,051 participants receiving enrolled during 2015-2017 followed up until 2021. Cumulative incidence rates [CIRs] precancerous conditions calculated by Kaplan-Meier method compared Log-rank tests. Mixed-effects cox regression models used detect predictors build risk scores. Another cohort including 26,586 was external validation. During median follow-up 6.25 years, CIRs 0.302%, 0.436%, 4.756% normal group, lesions, neoplastic respectively. The one-year GC higher than group (1.462% vs. 0.047%, p < 0.001), but no difference found lesions. Four selected calculate scores, male (hazard ratios [HR]=4.17, 95% confidence interval: 1.32-12.50), ≥60 years (HR=3.48, 1.57-7.72), smoking (HR=3.07, 1.31-7.20), limited vegetable consumption (HR=3.34, 1.45-7.21). High-risk lesion patients had (HR=2.71, 1.46-5.02), four-year CIR reached first-year follow-up. Similar trends observed validation cohort. summary, we provided evidence Chinese population progressively-increasing with advance mucosa significance management GC. advocate additional risk-based surveillance exposure risks. Specifically, our findings suggest that should be within one year after initial high 3-5 years. also encourage future studies validate effectiveness consider more factors.