1250 Gastric Intestinal Metaplasia Is a Common Finding Among Veterans Undergoing Endoscopy in South Texas

Abstract

INTRODUCTION: Gastric intestinal metaplasia (GIM) is a premalignant change of the mucosa of the stomach. Risk factors for intestinal metaplasia previously identified include chronic infection with Helicobacter pylori (HP), dietary factors, smoking, alcohol consumption, and chronic bile reflux. We aim to study the prevalence of GIM and its risk factors in a Texan Veteran population. METHODS: A retrospective chart review was performed among adults undergoing outpatient esophagogastroduodenoscopy (EGD) in 2013 and 2017 at a large Veterans Administration facility in Texas. Patient records were reviewed for demographic information, medication use, smoking and alcohol use, indication for EGD, presence of HP or GIM on gastric biopsies. Studies performed specifically to survey for known GIM were excluded. For patients with presence of GIM, data for type and location on biopsy were recorded. Student's t test was used for quantitative comparisons while the chi-square test or Fisher's exact test was used for qualitative comparisons using R. RESULTS: A total of 860 EGDs were reviewed, of which 340 (39.5%) underwent gastric biopsy. Of those undergoing gastric biopsy, 298 (87.7%) were male with a mean age of 58.0 years (±12.5) and 155 (45.6) were Non-Hispanic Caucasian, 92 (27.1) were Hispanic, and 39 (11.5%) were African-American. A total of 38 (11.2%) patients were found to have GIM. Among patients with GIM, 6 (15.8%) were found to have incomplete GIM. There was no significant difference in mean age among patients with GIM, 61.1 (±13.1) versus those without GIM, 57.6 (±12.4), P = 0.11. The rate of GIM among the non-Hispanic Caucasian population was significantly lower than patients of other ethnicities (7.7% vs. 16.9%, P = 0.02). We did not identify a statistically significant difference between GIM rates among those with HP and those without HP (18.4% vs. 10.3%, P = 0.17). PPI use was associated with lower rates of GIM (8.4% vs. 16.0%, P = 0.03). CONCLUSION: GIM was identified in 1 out of 9 patients in this Veterans cohort in Texas. As previously reported, GIM rates were higher in the Hispanic, African American and Asian patients. Patients taking PPIs were less like to have GIM. HP infection was not associated with GIM in this cohort. Longitudinal data is needed to better understand the risks of gastric cancer and how to risk stratify patients for ongoing surveillance in this diverse population.