Ancestry-specific gene expression profiles in TNBC tumors.

Abstract

1547 Background: Due to persistent disparities in breast cancer mortality, there has been a renewed focus on investigating tumor biology. Deeper exploration has exposed distinctions in tumor biology based upon self-reported race and ancestry. The disparities associated with Triple Negative Breast Cancer (TNBC) across the modern African Diaspora suggests that there is a genetic ancestry connection between its aggressive tumor biology and clinical outcomes. Understanding this connection could hold the key to improving clinical outcomes in this group. Methods: We investigated 75 TNBC primary tumors using Self-Reported Race (SRR) groups: African American (AA, n = 42) and European American (EA, n = 33). Using best practices established by TCGA, we analyzed bulk RNA sequencing to measure changes in genome-wide expression levels. We next quantified global ancestry in a novel manner using RNAseq variants using 1000 Genomes as the reference data. We then identified African and European ancestry-associated genes using a logistic regression (adjusted FDR p < 0.05) between quantified ancestry and gene expression levels. Results: We identified > 150 genes associated with quantified African ancestry. We also found using quantified ancestry was a more robust method to screen for differentially expressed genes than SRR. Using an updated TNBC subtyping method, we noted higher incidences of Basal-like 2 tumors in AAs. Pathway analyses indicated several canonical cancer pathways; including, TP53, NFKB1 and AKT, have altered functionality in patients of African descent. For example, TP53-associated genes were activated in TNBC tumors of AA versus EA. This upregulation, rather than loss of function, is suggestive of polymorphic and/or ancestry-specific expression regulation, likely driven by population-private genetic variants. Lastly, we used TCGA data to validate a subset of African ancestry-specific genes that were upregulated in AA patients in our cohort. Specifically, PIM3, ZBTB22 and PPP2R4 each retained significant upregulation, in our cohort, but also TNBC tumors from TCGA (p = 0.0018, 0.023 and 0.022, respectively). Conclusions: Our study has uncovered ancestry-specific gene expression profiles in TNBC tumors. The distinct distribution of TNBC subtypes and altered functional oncologic pathways are evidence that biological underpinnings in TNBC can be driven by shared genetic ancestry. These findings emphasize the need to investigate patient populations of various ancestral origins in order to fully appreciate the molecular diversity in tumor biology for precision of disease management.