Understanding the biology of disparity-associated genomically stable gastric tumors

PROJECT SUMMARY Gastric cancer (GC) has a high mortality rate in Latinos. This is likely the result of multi-level social determinants of health and population-specific factors. Our recent cancer registry-based study showed that Latinos have high rates of poor prognosis diffuse GCs. Compared to intestinal-type tumors, Latinos with diffuse tumors had ~50%

lower 5-year survival. In addition to the histological classification, TCGA described four GC molecular subtypes, including the genomically stable (GS) subtype, characterized by diffuse histology, low mutation and copy number alteration rates and a high frequency of “undruggable” CDH1 and RHOA mutations. GS tumors are

chemotherapy-resistant, immunologically “cold,” and have the worst prognosis. Interestingly, we recently showed that ~50% of all Latino patients have GS tumors, a fraction that is >3-fold higher than in Asians and Whites. Hence, our studies suggest that Latinos are enriched with poor prognosis GC histological and molecular

subtypes, partly explaining GC disparities. Given the high GS prevalence in Latinos, we obtained paired tumor- normal reduced representation bisulfite sequencing (RRBS) data from 15 Latino patients and re-analyzed TCGA methylation data to carry a preliminary GS epigenome analysis. We found that GS tumors, compared to other

subtypes, are enriched with differentially methylated genes (DMGs) in key pathways such as WNT, TGF-beta and PI3K/AKT signaling, which may explain their aggressiveness and chemoresistance. Our analyses also highlighted key limitations of existing publicly available GS epigenome data. For example, TCGA lacked Latino

samples and profiled primarily tumor-only samples using sub-optimal methods as we found that RRBS identified many methylated regions that were missing in TGCA methylation arrays and which appeared to be GS-tumor, and possibly Latino-specific. Our preliminary studies therefore highlighted several research gaps that will be

addressed in the present application. The objectives of the present application are to advance our understanding of GS tumor biology using epigenomic and functional genomic approaches. To this end, we will leverage our expertise in cancer biology, health disparities, patient-derived modeling, preclinical studies, and

genome engineering. Our hypothesis is that Latino GS tumorigenesis follows unique biological pathways, some which may be amenable to therapeutic development. In Aim 1, we will characterize the epigenome of 100 Latino GS tumors using RRBS. Aim 2 will investigate GS tumorigenesis using ethnic-appropriate Latino gastric

organoids and isogenic modeling for CDH1 and RHOA. In Aim 3, we will perform a loss of function screen that will be used to identify synthetically lethal associations in CDH1- and RHOA-mutant GS cell lines, some of that maybe amenable to molecularly guided therapies. Our study will significantly advance our understanding of GS

tumors' biology and therapeutic vulnerabilities. As GS tumors are likely drivers of higher mortality rates in Latinos, our study will also significantly advance cancer health equity in the country.