Abstract PO-002: The effect of genetic ancestry and molecular signatures on cancer survival disparities: A pan-cancer analysis

Abstract

Abstract Cancer incidence and mortality rates in the United States (US) are not equitably distributed among all population groups, especially for racial and ethnic minorities. While non-genetic factors, including socioeconomic and environmental differences, have long been cited as the main contributors to cancer health disparities (CHD), new evidence points to a possible role of genetic and biological differences in CHD but with limited understanding and remain largely undiscovered. In this study, disparities in cancer survival outcome among US population groups were characterized based on self-identified race and ethnicity (SIRE) and genetic ancestry (GA), using univariate and multivariate survival analyses. Moreover, group-specific gene expression and mutation signatures that contribute to cancer survivorship disparities were systematically analyzed using feature selection and multiple and generalized linear regression models. Kaplan-Meier analysis of overall survival of The Cancer Genome Atlas (TCGA) cohort showed significantly different survival probability for the following GA groups: East Asian (EA) vs. European (EU) in head and neck squamous cell carcinoma (HNSC), African (AF), EA and EU vs. Admixed American (AA) in kidney renal clear cell carcinoma (KIRC), and EA, AA vs. EU and EA vs. AA in skin cutaneous carcinoma (SKCM). Multivariate Cox proportional hazard modeling using GA showed significantly greater risk of mortality for AF compared to EU (hazard ratio(HR)= 1.838, p= 0.021) with breast invasive carcinoma (BRCA) and HNSC (HR= 2.01, p= 0.005), Native Americans (NA) compared to EU (HR= 1.36, p=<0.001) with HNSC, and EA compared to EU (HR= 5.502, p= <0.001) with SKCM. In contrast, significantly lower risk of mortality was experienced by NA compared to EU (HR= 0.625, p= 0.043) with KIRC, and AA compared to EU (HR= 0.15, p= 0.005) with SKCM. Only three out of the six significant GA cancer survivorship differences were confirmed by using SIRE, underscoring the power of GA analysis. Eight differentially expressed genes (DEG), 5 for BRCA, 2 for HNSC and 1 for SKCM, in GA groups showed significant interactions between gene expression and ancestry in affecting survivorship and contributing to CHD. No differentially mutated genes (DMG) between GA groups showed significant interaction with ancestry in association to survival for four cancers. Characterization of CHD using GA outperformed SIRE in data precision, missingness, model stability, and detection of significant associations. Differences in molecular signatures between GA groups and its impact on differential survival outcomes supports the role of biological differences contributing to cancer survival disparities. Significant interactions between gene expression and ancestry on survivorship suggests that gene regulatory differences may play a greater role on cancer outcome than structural differences due to gene variants, and also indicates that directionality and effect size of genes on cancer outcome may vary based on individual’s GA. Citation Format: Kara Keun Lee, John F. McDonald, King I. Jordan. The effect of genetic ancestry and molecular signatures on cancer survival disparities: A pan-cancer analysis [abstract]. In: Proceedings of the AACR Virtual Conference: Thirteenth AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2020 Oct 2-4. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2020;29(12 Suppl):Abstract nr PO-002.