Abstract PS3-06-16: Mapping neutrophil architecture with integrative spatial analysis to reveal clinically relevant functional heterogeneity

Background: Neutrophils are increasingly recognized as a key component of the tumor microenvironment. However, their spatial organization and interactions with other cells remain poorly characterized. The clinical and biological significance of these spatial patterns requires further exploration. Our study aims to identify clinically relevant neutrophil spatial architectures in breast cancer and elucidate their biological underpinnings. Methods: We established a large-scale breast cancer cohort covering all standard therapeutic regimens (adjuvant setting, n = 629; neoadjuvant setting, n = 381), with matched hematoxylin and eosin (H&E)-stained whole slide images (WSIs). We refined an artificial intelligence-based computational pathology framework we had proposed1 to identify neutrophils in WSIs and characterize their spatial patterns (i.e., their spatial relationships with neighboring cell types). Then we studied the correlation between these patterns and patient outcomes. To investigate the biology basis of distinct neutrophil spatial patterns, we performed single-cell spatial transcriptomics profiling (n = 104) and integrative analysis, with functional validation using patient-derived organoids (PDOs). Results: We found neutrophil spatial architectures demonstrated significant prognostic and predictive value, particularly in triple-negative breast cancer (TNBC). Specifically, we identified two clinically relevant patterns in TNBC: (1) Neutrophils in closer proximity to tumor cells predicted improved prognosis, which was validated in independent cohorts. Mechanistically, neutrophils upregulated MHC class I on tumor cells via secreted factors (e.g., TNFα), thereby enhancing CD8+ T cell-mediated recognition and cytotoxicity. This process was further corroborated by evidence from a neutrophil-PDO-CD8+ T cell co-culture system. (2) Neutrophils neighboring immune cells correlated with superior response to neoadjuvant immunotherapy. These neutrophils developed potent antigen-presenting capabilities through concurrent upregulation of MHC class I and II molecules. They primed CD4+ naive T cells and expanded CD8+ effector/memory T cells, thus fostering an immunotherapy-responsive microenvironment. Conclusion: Our study delineates the spatial heterogeneity of neutrophils, mapping their functional plasticity to spatial patterns with clinical relevance. These findings may inform the development of clinically actionable biomarkers for prognosis stratification and treatment response prediction in breast cancer patients. References 1Zhao, S. et al. Single-cell morphological and topological atlas reveals the ecosystem diversity of human breast cancer. Nat Commun 14, 6796, doi:10.1038/s41467-023-42504-y (2023). Citation Format: Y. Xu, D. Ma, X. Wu, Z. Qiu, L. Dai, H. Zhang, S. Zhao, Y. Xiao, Y. Jiang, Z. Shao. Mapping neutrophil architecture with integrative spatial analysis to reveal clinically relevant functional heterogeneity [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS3-06-16.