Abstract 7223: Evaluating treatment-related estrogen signaling and molecular subtype heterogeneity in a racially diverse HR-positive breast cancer cohort

Abstract African American women (AAW) with hormone receptor-positive (HR+) breast cancer have ∼50% higher mortality than European American women (EAW), and AAW are twice as likely to have weakly HR+ tumors (1-10% positivity). Although women with weakly HR+ tumors have 60% higher mortality and may still benefit from endocrine therapy (ET), they are less likely to receive ET. Clinical HR staining may not consistently reflect endocrine activity, and better characterization across HR positivity could help ensure optimal treatment. We performed targeted RNA sequencing on 118 tumors (54 EA, 64 AA) enrolled in the Detroit Research on Cancer Survivorship (ROCS) cohort and the Karmanos Cancer Institute Biobank to assess functional estrogen signaling with the Sensitivity to Endocrine Therapy (SET) index and heterogeneity in molecular subtypes using the Breast Cancer Consensus Subtype (BCCS) panel. Clinical HR staining was grouped as strongly (90%), moderately (11-90%), and weakly (1-10%) HR+. RNA-seq data were normalized and log2-transformed. SET values were grouped into low, intermediate, and high tertiles. BCCS subtypes were assigned using the BCCSclassifier R package, yielding five subtypes: ER-negative (ER-) non-basal, ER- basal, ER+ proliferative, ER+ stromal infiltration, and ER+ strong tumor signaling. Relationships between gene expression, SET score, HR categories, and race were evaluated using ANOVA and chi-square tests. Subtype distribution across HR categories and race was tested with multinomial logistic regression (α = 0.05).High SET scores were enriched among strongly HR+ tumors (χ2 = 11.9, p = 0.018). Further, 7% of weakly HR+ tumors had high SET scores, while 15% of strongly HR+ tumors had low SET scores, indicating discordance between clinical staining and treatment-related endocrine activity. Tumors from EAW vs. AAW had higher median SET values among those with intermediate (AA median= -0.21 vs. EA median= 0.017, p=0.028) and high (AA median=0.97 vs. EA median=1.88, p=0.16) SET scores. Considering BCCS, weakly HR+ tumors were more likely to have the ER-negative basal subtype compared to strongly HR+ tumors (OR=38.5, p=0.002) and to moderately HR+ tumors (OR=4.9, P = 0.044). Moderately HR+ tumors were also more likely to be classified as ER- basal (OR = 7.9, p = 0.024) compared to strongly HR+ tumors. The proportion of HR+ tumors assigned an ER-negative subtype differed significantly in AA vs EA patients overall (48.3% vs. 26.1%, p = 0.035), and similar trends were seen among the 1-10% (94% vs 67%, p=0.18) and 11-90% (52% vs 20%, p=0.11) HR groups. We observed meaningful differences in hormone signaling phenotypes between AAW and EAW, suggesting that clinical HR staining may not fully capture the underlying endocrine activity. This emphasizes the need for deeper molecular characterization to guide targeted treatment strategies and address disparities. Citation Format: Abigail M. Fielder, Gregory Dyson, Julie Ruterbusch, David Carr, Julie Boerner, Ann G. Schwartz, Kristen S. Purrington, . Evaluating treatment-related estrogen signaling and molecular subtype heterogeneity in a racially diverse HR-positive breast cancer cohort abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7223.