P. H. Liu1, M. E. Seaton2, J. P. Kronenfeld1, K. Khan1, A. L. Collier1, D. Franceschi1, A. S. Livingstone1, C. A. Hester1, N. Merchant1, J. Datta1 1University Of Miami, Surgical Oncology, Miami, FL, USA 2Summit Cancer Centers, Spokane Valley, WA, USA
Introduction: Gastric cancer (GC) in Hispanic patients is associated with a distinct epidemiology, as well as higher incidence of Helicobacter pylori infections. Prior H. pylori infection is associated with improved survival in US GC patients. However, the relationship between previous H. pylori infection, Hispanic ethnicity, and survival in GC has not been previously investigated.
Methods: All GC patients treated at our tertiary referral center were retrospectively reviewed (2010-2015). Any history of documented H. pylori infection was recorded. Overall survival was evaluated with Cox regression models, and interaction effects examined the relationship between Hispanic ethnicity and H. pylori infection. Kaplan Meier estimates assessed the survival impact of H. pylori-positivity in propensity-score matched Hispanic and non-Hispanic subgroups.
Results: Of 274 patients (45% localized GC, 56% Hispanic), the incidence of H. pylori-positivity was 16% in both Hispanic and non-Hispanic cohorts. Age, stage, tumor location, signet ring histology, and the proportion undergoing resection were similar between H. pylori-positive and -negative patients. In Cox regression analysis, ethnicity and H. pylori status were not independently associated with survival; however, there was a significant interaction between these variables (p=0.014) indicating that Hispanic ethnicity modified the effect of H. pylori on survival. In subgroup analyses, H. pylori-positivity was associated with improved survival among Hispanics (HR-adj 0.3, 95%CI 0.2-0.6, p=0.001), but not among non-Hispanics (HR-adj 0.9, p=0.8) (Figure). To corroborate these findings, we performed propensity-score matching of H. pylori-positive and -negative patients within Hispanic/non-Hispanic subgroups. In the matched Hispanic cohort, but not in the non-Hispanic cohort, H. pylori-positivity was significantly associated with improved survival compared with H. pylori-negativity (median not reached vs. 13 months, p=0.006).
Conclusion: This is the first study to identify a novel association between H. pylori-positivity and improved survival preferentially in Hispanic, but not non-Hispanic, patients. The environmental and molecular basis of this phenomenon warrants investigation.
