Multiomics in disparities research: Identifying African ancestry-specific mechanisms in the triple negative breast cancer tumor microenvironment

Abstract Triple negative breast cancer (TNBC) represents the most aggressive subtype of breast cancer (BC),often diagnosed as early-onset and late-stage disease with worse prognosis due to lack of targeted therapeutic options. In the US, approximately 10% of BC diagnoses annually are TNBC, however, frequency of TNBC diagnoses are significantly higher among African American (AA) women, driving a 40% higher mortality rate among AA women compared to White/European American (EA) women. Work from our International Center for the Study of Breast Cancer Subtypes (ICSBCS) cohort has linked west African ancestry to frequency of TNBC diagnoses, where we reported increased TNBC disease among women of west African ancestry (AA and Ghanaian) compared EA or east African (Ethiopian) women. Quantification of west African ancestry across the ICSBCS cohort has revealed significantly higher percentage of west African ancestry among TNBC patients, and this significant result remained when only considering AA women with TNBC. We hypothesize that African ancestry-specific factors drive differences in TNBC tumor etiology, and in combination with external factors yield a unique TNBC tumor microenvironment (TME). Approaching this work with a genetic ancestry (versus self-reported race) focus allows us the opportunity to define these unique factors that will ultimately translate into targeted therapeutic opportunities for this underserved population. To address this persisting disparity, we have leveraged multiomics approaches to: (1) identify unique African ancestry-specific gene expression profiles and enrichments using transcriptomic approaches, and (2) determine mutational profiles and signatures enriched among women of African ancestry with TNBC with whole genome sequencing (WGS). Utilizing a TNBC transcriptomics cohort enriched with women of African ancestry (n = 26), including AA, Ghanaian and Ethiopian TNBC patients, we have identified an African ancestry-associated gene signature comprised of 613 genes. After quantifying genetic ancestry across our patient cohort, we performed gene linear regression analyses to identify this gene signature. We also performed self-reported race (SRR) comparative analyses for differences in gene expression across population groups, and report that African ancestry associated gene are largely distinct from genes associated with SRR. Our African ancestry-associated gene signature revealed significant enrichment in immune cell trafficking pathways, where predicted activation of immune cell migration was observed among women with higher proportion of African ancestry. Deconvolution of the bulk RNAseq data revealed increases in tumor associated leukocytes proportions among AA and Ghanaian women, specifically B- and T- cell populations. Two separate validation cohorts looking at immunohistochemistry and multiplexed geospatial analysis also showed significant increases in T-cell infiltration into TNBC tumors of African ancestry women. Leveraging sub-continental ancestry estimations across the cohort, we additionally identified over 2000 genes associated with 5 African sub-continental reference populations, where significant enrichment in terms related to immune response was also observed. We hypothesize that these African ancestry-associated genes (1) will show enrichment with local ancestry estimates at the gene loci, and (2) that ancestral informative markers (AIMs) or population-private single nucleotide polymorphisms (SNPs) are expression quantitative trait loci (eQTLs) driving the observed African ancestry-specific gene expression observed. Workflow development is underway to apply local ancestry estimation methods to RNAseq data, to determine these associations in a loci-specific manner. WGS analysis was also completed on a cohort of patients with TNBC (n = 51) and hormone receptor positive (HR+) BC (n =23). Both cohorts are primarily comprised of AA patients and include Ethiopian patients with HR+ BC (n = 4) and TNBC (n = 3), and Ghanaian patients with TNBC (n = 3). Analysis of top mutated genes revealed significantly high frequency of TP53 mutations in TNBC (82%) compared to HR+ cases (35%) (p = 0.0001), with primarily missense and nonsense mutations in the P53 DNA-binding domain. All TNBC African cases harbored TP53 mutation, where only 1 Ethiopian HR+ case had a TP53 missense mutation. We compared COSMIC mutational signatures between TP53 mutation carriers and non-carriers and report a significant increase in proportion of Signature 3 (p = 0.0298), associated with double-strand break repair, and Signature 8 (p = 0.0026), which is associated with BC but has an unknown etiology. Of note, Signature 3 is often associated with germline BRCA1/2 mutations, and work is underway to determine BRCA1/2 status across our cohort. We have also reported the rs2363956 germline variant of the ANKLE1 gene as a potential TNBC protective allele among AA women, where ANKLE1 is reported to play a role in DNA repair pathways. We observe a higher fraction of Signature 3 across among the minor allele of ANKLE1 rs2363956 GG genotype, and a similar, lower fraction of Signature 3 among heterozygotes (GT) and major allele (TT) genotypes. DNAH5 and DNAH9 missense mutations were also detected among TNBC cases (16% and 14%, respectively), and not among HR+ cases (0%) (p = 0.05, p = 0.09, respectively). Mutations in DNA5 have been associated with chemosensitivity in ovarian cancer patients, and mutations in both DNAH5 and DNAH9 genes have been associated with chemosensitivity among patients with gastric cancer. GATA3 mutations were found exclusively among HR+ cases (26%, p = 0.0005), and PIK3CA mutations were also found at a significantly higher frequency among HR+ cases (30%) compared to TNBC cases (8%) (p = 0.03). Leveraging the diversity of our robust ICSBCS cohort, we have begun to define ancestry-associated phenotypes that are present in the TNBC TME of women with African ancestry. Our transcriptomic analyses have highlighted that ancestry-specific gene signatures are distinct from SRR-associated signatures, and we are currently working to define specific drivers of these ancestry-associated genes. We also highlight an increased immune response associated with higher fractions of African ancestry, that may provide opportunities to adapt current therapeutic options, such as immunotherapy, to this patient population. Our WGS work also highlights differences in mutational processes and mutation signatures among African ancestry women with TNBC and nonTNBC disease. We report an increased burden of TP53 mutations is present among TNBC patients and is highly correlated to increased Signature 3, which is associated with BRCAness and DNA repair deficiency. Work is currently underway to determine germline BRCA1/2 status of this patient population, as BRCA1/2 status may also be predictive of immunotherapy response. Taken together, these African ancestry specific phenotypes characterized among patients with TNBC highlight prospective biomarker or therapeutic opportunities for these historically underserved populations. Citation Format: Rachel Martini, Max Chao, Timothy Chu, Ishmael Kyei, Ernest Adjei, Mahteme Bekele, Paula Ginter, Kofi Gyan, Olivier Elemento, Nicolas Robine, John Carpten, Lisa Newman, Melissa Davis. Multiomics in disparities research: Identifying African ancestry-specific mechanisms in the triple negative breast cancer tumor microenvironment. [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 NG03.