Endoscopic Surveillance of Gastric Intestinal Metaplasia: A Retrospective Cohort Study
Ji Yoon Yoon,Erik Katcher,Ella Cohen,Stephen C Ward,Carol Rouphael,Steven H Itzkowitz,Christina P Wang,Michelle K Kim,Shailja C Shah
DOI: https://doi.org/10.1097/MCG.0000000000002039
2024-07-17
Abstract:Background: Gastric intestinal metaplasia (GIM) is a precancerous condition. Limited data exist on real-world clinical practice relative to guidelines. Aim: The aim of this study was to evaluate adherence to GIM risk stratification and identify factors associated with follow-up endoscopy. Materials and methods: We conducted manual chart review of patients with histologically confirmed GIM at an urban, tertiary medical center were identified retrospectively and details of their demographics, Helicobacter pylori, biopsy protocol, endoscopic/histologic findings, and postendoscopy follow-up were recorded. Multivariable logistic regression was used to identify factors independently associated with follow-up endoscopy. Results: Among 253 patients, 59% were female, 37% non-Hispanic White (NHW), 26% Hispanic, 16% non-Hispanic Black (NHB). The median age at index endoscopy was 63.4 years (IQR: 55.9 to 70.0), with median follow-up of 65.1 months (IQR: 44.0 to 72.3). H. pylori was detected in 21.6% patients at index EGD. GIM extent and subtype data were frequently missing (22.9% and 32.8%, respectively). Based on available data, 26% had corpus-extended GIM and 28% had incomplete/mixed-type GIM. Compared with NHW, Hispanic patients had higher odds of follow-up EGD (OR=2.48, 95% CI: 1.23-5.01), while NHB patients had 59% lower odds of follow-up EGD (OR=0.41, 95% CI: 0.18-0.96). Corpus-extended GIM versus limited GIM (OR=2.27, 95% CI: 1.13-4.59) was associated with follow-up EGD, but GIM subtype and family history of gastric cancer were not. Conclusions: We observed suboptimal risk stratification among patients with GIM and notable race and ethnic disparities with respect to endoscopic surveillance. Targeted interventions are needed to improve practice patterns and mitigate observed disparities.