Abstract Background: Clinical assessment of breast cancer progression and therapeutic response typically relies on qualitative immunohistochemistry results for ER, PR, and HER2 proteins, which may be improved by adherence to ASCO/CAP Guidelines. To identify more robust, clinically relevant molecular markers, we conducted in silico analyses of a comprehensive, de-identified database that integrates quantified protein biomarker levels, qPCR, and microarray gene expression data, as well as long-term clinical outcomes. Because TNBC represents an aggressive disease that is difficult to treat, our multifarious investigation focuses on differentiating biomarkers for this subtype that improve clinical outcomes. Methods: Biomarker Triads were created from ER and PR proteins quantified by radio-ligand binding (n = 13,966) and enzyme immunoassays (n = 4,408), expressed as fmol/mg cytosol protein. HER2 protein was quantified in 504 and 901 specimens using either ELISA or EIA, respectively, and expressed as HNU/mg membrane protein. Clinical data were available for 1,194 ER/PR cases, and complete ER/PR/HER2 protein Triads were available for 189 of these. However, survival outcomes (PFS and OS) were only available for a subset of these patients. Expression levels of these genes and 82 other cancer-associated genes were validated by RTqPCR of 580 specimens after RNA extraction and purification. Microarray analysis (∼22,000 genes) was conducted on RNA from each of 247 LCM-isolated carcinoma samples. Biomarker Triads were defined as ER+/PR+/HER2+ (TPBC) or ER-/PR-/HER2- (TNBC), using either protein levels or cognate gene expression (ESR1, PGR, ERBB2). Differential gene expression was assessed using nonparametric tests and Spearman correlations in R (version 4.4.2). For the subset of patients with survival data, exploratory Kaplan-Meier and Cox proportional hazards models were used to assess associations between gene expression and progression-free survival (PFS) or overall survival (OS). Results: Neither ER/PR assay platform influenced ER+/PR+ or ER-/PR- status, and neither ER nor PR levels were associated with HER2 expression. TPBC biopsies showed elevated expression of BL2 (CADM1), CAXII, ERBB2, and ERBB4 compared to TNBC, supporting their use in subtype definition. TPBC also exhibited increased IL10 and VEGFATV2, while TGFB1 was downregulated in both TPBC and ER+ cases. BRCA1/2 expression was reduced in ER-/PR- tumors, while NAT1/2 and COMT were elevated in ER+, ER+/PR+, and TPBC samples. FOXA1 and ART3 emerged as potential targets from gene-based Triad definitions using microarray data. In a subset of cases with available clinical outcomes, exploratory survival analysis suggested that TPBC triads were associated with more favorable PFS and OS compared to TNBC, and elevated expression of certain genes (e.g., ESR1, FOXA1) trended with improved survival outcomes. Conclusions: This integrative dataset, which links rigorously collected clinical biopsies, precise biomarker protein quantification, and gene expression profiles from LCM-procured carcinoma cells, enabled the discovery of genes differentially expressed across biomarker-defined subtypes. Biomarker Triads based on protein and gene expression levels in primary breast tissue biopsies revealed a landscape of genomic diversity in carcinomas, which improved molecular subtype classification. Select genes, including the transcription factor FOXA1 and ART3, reported to be involved in cell migration, represent probable targets for drug development and/or companion diagnostics in ER+ and triple-negative breast carcinomas. Citation Format: M. W. Daniels, J. L. Wittliff. Genomic Diversity of Breast Carcinomas Exhibiting Biomarker Triads Derived from Quantified Levels of ER, PR and HER2 Proteins or from Expression of Their Cognate Genes 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 PS4-05-20.
Daniels et al. (Tue,) studied this question.