Abstract P1-06-07: Enhanced differentiation of clinical behavior by breast carcinomas utilizing quantification of estrogen and progestin receptor proteins and gene expression subsets

Abstract

Abstract Background: IHC analyses of clinical biomarkers for breast cancer management yield semi-quantitative results potentially complicated by variation in methods and interpretation. We are mining a comprehensive database established over four decades that contains deidentified results of quantified protein and RNA biomarkers with clinical outcomes to detect candidates for companion diagnostics and molecular targets. Assays were performed in a Reference Laboratory holding both CLIA and Commonwealth of Kentucky licenses. Methods: This retrospective investigation utilized ER/PR protein levels, expressed as fmol/mg cytosol protein, that were quantified for each of 2339 carcinoma biopsies with FDA-approved kits and cut-off values by radio-ligand binding assay (LBA, NEN/DuPont) and/or enzyme immunoassay (EIA, Abbott Labs). Of these, 454 specimens had both LBA and EIA results recorded. Biomarker results and de-identified clinical outcomes were examined using REMARK criteria. Gene expression, quantified biomarkers, features of primary breast cancers and clinical patient outcomes were analyzed by univariable and multivariable Cox regressions, Fisher’s Exact Test, Kaplan Meier plots and with R software v4.0.0. Microarray results of ~22,000 genes were obtained from RNA of LCM-procured breast carcinoma cells of 247 primary breast biopsies and expression levels of 85 cancer-associated genes were determined by RTqPCR of 278 specimens after RNA extraction and purification with established procedures. Results: Quantified levels of both ER and PR protein were expressed to a greater extent in primary carcinomas from post-menopausal patients compared to cancers from premenopausal women regardless of assay type (p = 0.005). ESR1 expression by microarray was only greater in LCM-procured carcinoma cells of post-menopausal patients while both ESR1 and PGR relative expression levels measured by qPCR were greater in tissue biopsies of post-menopausal women (p = 0.05). Box plots of untransformed EIA/LBA results of either ER (n=454) or PR (n=452) using specimens on which both quantifying assays were performed revealed a significant positive ER bias by EIA (median ratio=1.54) compared to that of PR (median ration=1.13). Transformed data of ratios confirmed this observation. Kaplan Meier plots of patient outcomes with ER or PR or both assessed by each assay type suggested LBA was a slightly better predictor. Microarray analyses of LCM-procured cells revealed 2994 genes distinguishing ER+ from ER- breast cancers of which ~200 genes were identified that differentiated 4 distinct molecular subtypes (2 ER+ and 2 ER-) exhibiting different outcomes upon Kaplan-Meier analyses. Expression levels of the 85 genes (including ERBB2), validated by qPCR with RNA from biopsy tissue sections, were evaluated relative to quantified levels of either ER or PR protein or both in primary breast cancer biopsies. Collectively, use of quantified ER/PR protein or ESR1/PGR expression with clinical outcomes enhanced identification of small molecular signatures and drug development candidates. Conclusions: Measurements of ER, PR and HER2 proteins with tests that quantify their levels as well as by microarray and qPCR assessment of ESR1, PGR and ERBB2 relative gene expression were correlated with prediction of clinical outcomes of breast carcinomas exhibiting various features (e.g., pathology classification, TNBC). Results reinforced importance of assessing levels of the clinically relevant biomarkers in a quantified fashion to enhance their use in breast cancer management and prediction of risk of recurrence. These investigations also revealed numerous over-expressed genes in carcinomas with poor clinical outcomes suggesting candidates for development of novel therapeutics. Citation Format: James L Wittliff, Michael W. Daniels. Enhanced differentiation of clinical behavior by breast carcinomas utilizing quantification of estrogen and progestin receptor proteins and gene expression subsets [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-06-07.