Publication: ASBrS 2026, Abstract Number: 1062

Authors:Nathalie Johnson, Mehran Habibi, Marcela Mazo, Ahmed Elkhanany, Fengting Yan, Nicole Sookhan, J. Jaime Alberty-Oller, Laila Samiian, Eric Brown, Linsey Gold, Alfredo Santillan, Eduardo Dias, Sasha Davis, Laura Lee, Constantine Godellas, Abigail Beard, Gordan Srkalovic, Beth-Ann Lesnikoski, Sahra Uygun, Christina Page, Nicole Stivers, Andrea Menicucci, William Audeh, Joyce O’Shaughnessy

Title:Genomically Basal-Type tumors demonstrate distinct immune profiles and chemosensitivity across self-reported race among patients enrolled in FLEX receiving neoadjuvant chemotherapy

Abstract

Background:

BluePrint, an 80-gene molecular subtyping signature, identifies intrinsic breast cancer (BC) biology beyond receptor-defined classifications. Prior studies have shown higher frequencies of genomically Basal tumors among Black and Latin American patients (pts) compared to White. However, limited data exist on chemotherapy response across race, and the biological basis of these differences remains unclear. Using whole transcriptome analysis (WTA) and clinical data from pts enrolled in the FLEX Study (NCT03053193), we investigated racial differences in pathological complete response (pCR) and associated molecular features. The objective was to evaluate pCR rates and characterize underlying biological and immune profiles among pts with BluePrint Basal early BC treated with neoadjuvant chemotherapy (NCT).

Methods:

A total of 451 BluePrint Basal tumors from pts self-identified as Black, Latin American, or White with early BC enrolled in FLEX who consented to WTA, clinical data collection and received NCT with available pCR data were included. Clinical characteristics across self-reported race/ethnicity were compared using chi-squared or Fisher’s exact tests. Univariate and multivariate logistic regression models assessed the association between clinical variables and achieving pCR. Differentially expressed genes (DEGs) were evaluated using limma, and pathway enrichment was performed using gene set enrichment analysis (GSEA) with Hallmark gene sets between pts who achieved a pCR vs no pCR, stratified by race. Immune cell abundances were characterized using xCell. Significance was defined as p < 0.05 for clinical variables and adjusted p < 0.05 for DEGs and xCell results.

Results:

The median age was 56 years, with the youngest among Latin American pts (n = 64; 47.5 years), followed by Black (n = 81; 56 years) and White (n = 306; 57 years; p < 0.001) pts. Overall, 67% of pts had TNBC, 28% HR+HER2- BC, 85% grade 3, 57% T2, and 83% lymph node-negative (LN-) tumors, with no significant differences observed across race. Nearly all tumors were MammaPrint (MP) High Risk 2 (H2) (93%) and 7% were High Risk 1 (H1), with the highest incidence of H2 (97%) among Latin American pts, followed by White (93%) and Black (89%, p=0.033) pts. Platinum-based regimens were most common overall (49%), followed by anthracycline + taxane (AC-T; 36%), and TC (7%; p = 0.077). 46% of pts achieved pCR, with the highest rates among Latin American (60%), followed by Black (47%), and White (43%; p=0.067) pts. In the multivariate analysis (Table), age (OR = 0.98, p = 0.048), MammaPrint Index (OR = 0.13, p = 0.0016), and platinum-based NCT (OR = 2.76, p = 0.0012) were significantly associated with pCR, while race was not independently associated with pCR.

Among all pts, GSEA for Hallmark gene sets revealed broad immune-related and inflammatory pathways in tumors that achieved pCR across all races. MYC targets, estrogen response, adipogenesis, and oxidative phosphorylation pathways were significantly upregulated among Latin American pts and downregulated among Black pts compared to White pts achieving pCR.

Immune XCell deconvolution revealed significant enrichment of plasma B cells (log₂FC = 0.70, adj. p = 0.048) and Treg cells (log₂FC = 0.43, adj. p = 0.039) among Black pts achieving pCR. In Latin American pts, pCR was associated with increases of M1 macrophages (log₂FC = 1.34, adj. p = 0.039), T cells, macrophages, and myeloid dendritic cells. White pts that did not achieve a pCR demonstrated significantly higher common myeloid progenitor cells (log₂FC = –1.03, adj. p = 0.039), stroma, and endothelial cells.

Conclusion:

BluePrint Basal-Type tumors demonstrate immune-active transcriptional profiles associated with NCT response across race, highlighting both shared and race-specific biological mechanisms of response. MammaPrint Index, age, and platinum-containing regimens remained significant variables associated with achieving pCR when accounting for clinical features, while race did not. Although self-reported race was not independently associated with pCR, the notably higher pCR rates observed among LA pts and the differences in WTA across racial groups highlight potential underlying biological or treatment response mechanisms that warrant further investigation in larger, racially diverse cohorts. Future analyses should also investigate WTA differences between pts achieving pCR and no pCR across different age groups to further understand the interaction between tumor biology, immune activation, and age-related factors influencing NCT response.