Background: Triple-negative breast cancer (TNBC) remains the most aggressive molecular subtype of breast cancer, with worse survival outcomes compared to other breast cancer subtypes. TNBC prevalence is highest among women of African descent worldwide, and through our previous work we have established a connection between West African ancestry (WAa) and higher rates of TNBC. Specifically, we have shown that prevalence rates of TNBC among West African and African American women are similar and higher than that of East African and White American women. We have also shown that quantified African ancestry is higher among TNBC cases compared to non-TNBC cases. To determine the influence African ancestry on the TNBC tumor environment, we sought to determine any differences in gene expression profiles of Ghanaian (West African) compared to Ethiopian (East African) women. Methods: RNA was extracted and sequenced from a pilot cohort of archival FFPE tumor tissue among Ghanaian (n = 19) and Ethiopian (n = 20) women. RNAseq reads were aligned, and quality of alignments were assessed, where de-duplicated samples with counts above 10M reads were included in the final analysis. Genetic ancestry was quantified by obtaining SNVs called from the RNAseq alignments, using GATK best practices. Differentially expressed genes lists were determined comparing Ghanaian vs. Ethiopian TNBC tumors, and also by identifying genes that were associated with increasing African ancestry. These gene lists, and log-fold change between comparison groups, were used as input for Ingenuity Pathway Analysis (IPA), to identify canonical pathways and de novo networks that are specific to Ghanaian or Ethiopian TNBC tumors. Results: Using 1KG populations as our reference to quantify genetic ancestry, we show that Ghanaian samples have >94% AFR ancestry, specifically matching population groups representative of WAa. The Ethiopian samples showed between 37-48% AFR ancestry, primarily represented by East African groups. Interestingly, there seems to be a significant proportion of EUR ancestry among the Ethiopians samples (30-49%), primarily represented by Italian ancestry. We have conducted the differential gene expression analysis in two ways. First, we have compared gene expression profiles between Ghanaian and Ethiopian tumors. In our preliminary analysis, we identified >600 genes (p < 0.01) that were differentially expressed between Ghanaian and Ethiopian TNBC tumors. Second, we used AFR ancestry as a continuous variable, where we conducted a linear regression analysis to identify genes associated with AFR ancestry. We identified >900 genes associated with AFR ancestry (p < 0.01), and this gene signature distinguished Ghanaian from Ethiopian tumors in an unsupervised hierarchical clustering. In comparing the differentially expressed gene lists from these two approaches, approximately 200 genes were shared, indicating the distinct value of both analyses. Using these gene lists as input for IPA analysis, we have begun to identify canonical pathways that have been altered by our differentially expressed genes, alongside de novo networks that differ between our Ghanaian and Ethiopian tumors. In our overlapping gene list, we see predicted differences in functions such as quantity of T lymphocytes, where genes downregulated in Ethiopian tumors may indicated reduced presence of these immune cells. Using CIBERSORT and xCell deconvolution methods, validation of these findings are ongoing. Conclusions and Ongoing work: This work highlights how ancestry-specific gene regulation can delineate differences in the tumor microenvironment among a cohort of African tumors. We are currently evaluating distribution of TNBC subtypes and estimation of immune cell populations in these tumors, to determine ancestry-specific differences in tumor heterogeneity and immune response.

Citation Format: Rachel Martini, Endale Gebregzabher, Princesca Dorsaint, Timothy Chu, Kanika Arora, Lee Gibbs, Zarko Manojlovic, Nicolas Robine, Andrea Sboner, Olivier Elemento, John Carpten, Lisa Newman, Melissa Davis. Gene expression profiles of Ghanaian and Ethiopian triple-negative breast tumors [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr SS1-07.