Chemoprevention of Gastric Carcinogenesis

Primary Objective \- The difference in cell DNA damage between patients treated with DFMO and patients treated with placebo at 6 months. The cell DNA damage is measured using the percent positive gastric epithelial cells assessed by IHC for gamma H2AX. The mean difference between the two groups at 6 months will be calculated, accounting for their baseline measurements. Secondary Objectives * The difference in cell DNA damage between patients treated with DFMO and patients treated with placebo for 18 months, and then followed for an additional 6 months. The cell DNA damage is measured using the percent positive gastric epithelial cells assessed by IHC for gamma H2AX. The mean difference between the two groups at 18 and 24 months will be calculated, accounting for their baseline measurements. * The differences in the gastritis histopathology score between patients treated with DFMO and patients treated with placebo for a total of 18 months, and followed for an additional 6 months. The gastritis histopathology score is measured with a quantitative scale 0.0-6.0, for atrophy, intestinal metaplasia, and dysplasia. The mean differences between the two groups at 6, 18, and 24 months will be calculated using mixed models, accounting for their baseline measurements. * Number of patients with quantitative toxicities. Toxicities will be assessed per CTCAE criteria, and each toxicity will be assigned an adverse event (AE) term according to CTCAE definitions (each AE term = unique representation of a specific event used for medical documentation and scientific analyses), and graded as defined by CTCAE (grade 1 = mild; grade 2 = moderate; grade 3 = severe or significant but not immediately life-threatening; grade 4 = life-threatening; grade 5 = death). * To evaluate whether candidate single nucleotide polymorphisms (SNPs) relevant to eflornithine (DFMO) efficacy.