Abstract C130: A genetic ancestry-informed single-cell atlas of the endometrial cancer tumor microenvironment reveals ancestry-associated immune features with potential clinical implications

Abstract

Genetic ancestry-associated differences in the tumor immune profile have been identified in multiple cancer types but remain largely uncharacterized in endometrial cancer (EC). These tumor microenvironment (TME) features, such as cytotoxic immune cell infiltration and functionality, influence disease progression, tumor immune escape, and response to immunotherapy. Single-cell transcriptomic studies provide essential insight into TME features but have notably lacked racial/ethnic diversity. This is especially problematic in EC, in which mortality rates for Black women are double those of any other racial/ethnic group. Thus, the inadequate diversity of single-cell data limits insight into disease behavior and therapeutic potential in populations disproportionately affected by the disease. To address this gap, we developed a genetic ancestry-aware single-cell atlas of the EC TME to identify ancestry-associated shifts in immune population proportions, functionality, and communication networks, as well as the implications of these features for patient survival. Our single-cell atlas encompasses 69 unique samples from eight datasets, including scRNA-seq and snRNA-seq data from normal (n = 36), hyperplastic (n = 5), and cancerous (n = 28) endometrium. Of the 69 total samples, only 13 reported patient race/ethnicity, which included 8 Black and 5 White patients. We inferred genetic ancestry from single-cell data by extracting genotypes at ancestry-informative SNP sites to estimate the fractional contribution of sub-continental ancestry groups. We profiled ancestry-associated shifts in cellular composition and functionality, comparing cell type proportions and differential expression across ancestry groups. We also profiled cellular communication networks using CellChat. Finally, we leveraged this single cell atlas as a reference for deconvolution of bulk RNA-seq in The Cancer Genome Atlas (TCGA) uterine corpus endometrial carcinoma (UCEC) and uterine carcinosarcoma (UCS) cohorts (n = 602) to assess impacts of differential immune infiltration on patient survival. We identified 9 broad cell populations present in the single cell atlas (ciliated and unciliated epithelium, stromal fibroblasts, endothelium, lymphocytes, myeloid cells, B cells, mast cells, and cancer stem cells), as well as functionally distinct sub-populations within the lymphocyte and myeloid clusters. We identified distinct ancestry-associated shifts in the expression profiles and communication networks of immune cell subsets, particularly of immunoregulatory processes. The ability of tumor-infiltrating immune cells to overcome tumor-induced immunosuppression has significant implications for disease progression, and this atlas enables deconvolution-based survival analysis in the TCGA cohort to assess prognostic implications of these ancestry-associated immune shifts. Overall, our findings suggest the presence of genetic ancestry-influenced immunoregulatory shifts in the endometrial cancer TME which require further investigation for equitable immunotherapy and biomarker development. Emma Saart, Kassidy Lundy-Perez, Nyasha Chambwe. A genetic ancestry-informed single-cell atlas of the endometrial cancer tumor microenvironment reveals ancestry-associated immune features with potential clinical implications [abstract]. In: Proceedings of the 19th AACR Conference on the Science of Cancer Health Disparities; 2026 Oct 18-21; Atlanta, GA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2026;35(10 Suppl):Abstract nr C130.