Spatial transcriptomics reveals racial disparities in the transcriptional substructure of triple negative breast cancer.

Abstract Triple negative breast cancer (TNBC) is an aggressive and clinically challenging disease that disproportionately affects African-American (AA) women. Despite this well-known disparity, AA women have historically been underrepresented in large clinical and genomic studies of TNBC. Furthermore, while a genetic association of disease risk with African ancestry has been established, the tumor biology of AA TNBC has not been well characterized. We therefore sought to investigate the contribution of transcriptional substructure and tumor heterogeneity to TNBC disparities. We applied 10x Genomics Visium spatial transcriptomics (ST) technology to 28 samples representing primary TNBC tumors from 14 patients, 7 of whom were AA and 7 who were Caucasian. ST allows collection of whole transcriptome information from up to 5,000 spatially resolved coordinates within a tissue sample, providing remarkable insight into cellular organization within the larger framework of tissue architecture. Normalization of our ST expression data followed by clustering analysis confirmed extensive intra-tumoral subclonality, consistent with the genetic and molecular heterogeneity of TNBC. Most samples contained several transcriptionally distinct tumor clusters, characterized by distinctive activation of oncogenic and immunogenic signaling pathways revealed by gene set enrichment analysis (GSEA). For direct comparison of AA and Caucasian ST data, we performed an integration of all samples to consolidate the variation across genetic backgrounds. The integrated dataset was subject to clustering analysis to produce nine integrated clusters (ICs) representing shared cell populations that were present across all samples in the cohort. Interestingly, while all nine ICs were present in all individual samples, the representation of ICs varied by race. Molecular annotation of ICs using tools such as GSEA, ESTIMATE, and CIBERSORTx revealed that ICs enriched in Caucasian samples were immune-rich, while those enriched in AA samples were immunodeficient and distinctly hypoxic. These results support previous studies that have posited a unique tumor microenvironment in AA TNBC. Furthermore, investigation of the spatial organization of ICs using join count analysis revealed consistent non-random spatial incompatibility of hypoxic regions with immune infiltrate, which is consistent with established roles for hypoxia in immunosuppression. To examine the clinical significance of IC representation in TNBC, we used top marker genes of relevant ICs to perform survival analysis in publicly available breast cancer datasets (4929 samples). We found that high expression of IGKC, which was associated with Caucasian samples in our cohort, is significantly protective in basal-like breast cancer. Conversely, high expression of NDRG1, a hypoxia-related gene overrepresented in AA samples in our cohort, was significantly associated with poor survival. Thus, our study utilizes advanced molecular profiling to provide novel characterizations of AA TNBC with clinical implications in TNBC outcomes disparities. Citation Format: Rania Bassiouni, Michael O. Idowu, Lee D. Gibbs, Valentina Robila, Pamela J. Grizzard, Michelle G. Webb, Jiarong Song, Ashley Noriega, David W. Craig, John D. Carpten. Spatial transcriptomics reveals racial disparities in the transcriptional substructure of triple negative breast cancer [abstract]. In: Proceedings of the 15th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2022 Sep 16-19; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2022;31(1 Suppl):Abstract nr PR012.