Abstract
Breast cancer immune phenotypes influence treatment response and clinical outcomes, yet their ancestry-specific variations remain underexplored. Here, we analyzed transcriptomic data from over 13,000 breast tumors across six ancestry groups to characterize immune-stromal profiles and their association with ancestry, biological features, treatment response, and survival outcomes. Expression patterns were validated by spatial proteomics and immunohistochemistry. K-means clustering consistently identified three immune phenotypes (Hot, Moderate, or Cold) that varied significantly by ancestry, age, molecular subtype, and prognosis. Logistic regression and ancestry-associated analyses revealed that while immune phenotypes were primarily driven by PAM50 subtype, age, and disease stage, notable ancestry-related differences persisted, with European ancestry generally exhibiting higher immune and stromal activity across breast cancer subtypes. Hot tumors, enriched in the Basal-like and HER2 subtypes, were associated with younger age, higher immune infiltration, and improved overall survival. African ancestry was linked to elevated immune scores and upregulation of BTLA-mediated T cell co-inhibition, suggesting sensitivity to immunotherapy. European and East Asian tumors showed stromal enrichment, particularly inflammatory and myofibroblastic cancer-associated fibroblasts, associated with poor prognosis. Core immune activation genes (e.g., CD3, CD2, and CXCL10) were conserved, while ancestry-specific signatures and chemokine signaling were identified. This study uncovers both shared and ancestry-specific immunogenomic features of breast cancer, highlighting the role of ancestry and other biological features in shaping the tumor immune microenvironment. These findings re-emphasize the need for population-informed approaches in breast cancer immunotherapy and biomarker development, to ensure equitable precision oncology strategies across global populations.