Abstract 6504: Germline genetically predicted body mass index is associated with endometrial cancer somatic transcriptomic, immune, and mutational signatures in The Cancer Genome Atlas

Abstract High body mass index (BMI) is a causal risk factor for endometrial cancer but the molecular mechanisms underlying this association remain elusive. Here we sought to characterize the tumor genomic landscape of endometrial cancers that have developed on a germline genetic background of predisposition to elevated BMI. We built a polygenic score (PGS) for adult BMI in women using effect size estimates and allele information on 242 independent (r2<0.05) variants associated with BMI at genome-wide significance (P<5x10-9) in 379,501 women of European ancestry. We performed sample and germline (blood) genotype quality control and imputation into the 1000 Genomes reference panel on data from 354 endometrial cancer cases of genetically inferred European ancestry from The Cancer Genome Atlas (TCGA). We assigned each woman in this TCGA cohort her genetically predicted life-course BMI based on the BMI PGS and found this to be modestly correlated with BMI at the point of diagnosis (r2=0.23; P=2.1x10-5). Multivariable linear (default) and quasi-Poisson (for zero-inflated counts) regression models were used to test for associations between the BMI germline PGS and endometrial cancer tumor genomic, transcriptomic, proteomic, and immune traits in TCGA. All analyses were adjusted for age, stage, microsatellite status and 10 genetic principal components. Mutational signature models were also adjusted for signature accuracy. We ranked 18,458 genes based on the association between their tumor expression and the BMI PGS and performed gene set enrichment analysis to identify associations between the BMI PGS and upregulation of genes in the IL6-JAK-STAT3 signaling (false discovery rate (FDR)=8.50x10-7), inflammatory response (FDR=7.03x10-6), interferon gamma response (FDR=5.49x10-5) and glycolysis (FDR=3.28x10-4) pathways. High BMI PGS had an inverse association with endometrial tumor EGFR (FDR=0.07) protein levels of the 131 tumor proteins profiled by reverse phase protein array. Endometrial tumors that had developed on a germline background predictive of high BMI were also associated with increased infiltration of activated mast cells (FDR=9.55x10-3) in our evaluation of 22 tumor immune cell infiltrates quantified by the CIBERSORT algorithm, as well as the mitotic and aging clock-like single base substitution (SBS) signatures 1 (FDR=0.01) and 5 (FDR=0.04). The two SBS signature associations and the activated mast cell association with the BMI PGS were substantially more pronounced in the subgroup of endometrial cancers with microsatellite instability. Thus, we combined germline and somatic data using a novel approach to identify endometrial cancer tumor molecular features associated with genetically predicted higher BMI, providing precision multi-omic portraits of endometrial cancers that develop on a background of adiposity. Citation Format: George Richenberg, Victoria Gray, Carina Owen, Tom Gaunt, Caroline Relton, Emma Vincent, Siddhartha Kar. Germline genetically predicted body mass index is associated with endometrial cancer somatic transcriptomic, immune, and mutational signatures in The Cancer Genome Atlas [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6504.