Breast cancer is the most frequently diagnosed cancer and the second leading cause of cancer-related death in women. Over 70% of breast tumors overexpress estrogen receptor (ER), which is a strong prognostic indicator and the most reliable predictor of response to endocrine therapy. ER status is routinely assessed in traditional clinical assays; however, these assays are highly sensitive and suppress the true biological range of ER expression. As a result, a number of clinically important features of ER expression have been overlooked in prior clinical and epidemiological studies. Genomic assays now allow for quantification of ER gene expression, offering greater resolution in measuring level of tumor ER, and digital image analysis can detect spatial heterogeneity of ER protein expression within tumor samples. Both of these features have shown associations with breast cancer survival, though prior studies have been limited by small, homogenous patient populations and limited measures of ER. Critical questions remain in understanding the role of ER in breast cancer outcomes. First, it is unknown whether the relationship between ER and survival varies between black and white women. This is important to understand because black women with ER positive breast cancer demonstrate significantly higher mortality rates than white women, even in randomized clinical trials. Second, we have previously shown that smoking is associated with significantly altered tumor ER gene expression, but the impact of other contemporaneous exposures on ER expression remains unexplored. Understanding how exposures modulate expression is critical to accurate interpretation of ER as a biomarker. We propose to investigate the role of ER heterogeneity in breast cancer outcomes using a population- based sample of 5,000 breast cancer patients, of which approximately half are black women. We hypothesize that increasing ER quantity and spatial homogeneity of expression will be associated with better outcomes, particularly in white women, and that exposure history will affect ER expression. We will test this hypothesis with two specific aims. First, we will estimate the association between quantitative measures of ER and survival, overall and by race. We will quantify ER using protein-based measures, including proportion of ER positive cells and intratumoral heterogeneity of expression, and Ribonucleic acid- (RNA) based measures, including Estrogen Receptor 1 (ESR1) expression and a multigene score reflecting estrogen signaling (luminal score). Second, we will evaluate how exposure history impacts quantitative ER and multigene biomarkers that include ER. We will estimate the association between body mass index, oral contraceptive use, hormone therapy use, alcohol use, and cigarette smoking and protein- and RNA- based measures of quantitative ER. These results will demonstrate how quantitative and spatial ER expression associate with survival outcomes and how contemporaneous exposures drive variability in ER expression in heterogenous breast cancer patient populations.
Estrogen receptor (ER) is arguably the most powerful prognostic and predictive biomarker in breast cancer, a disease which kills over 40,000 women in the United States annually. This project proposes to investigate the role of quantitative and spatial measures of ER in breast cancer outcomes in a diverse, population-based sample of breast cancer patients and evaluate the impact of contemporaneous exposures on tumor ER expression. The results from this project will help elucidate the underlying drivers of disparities in breast cancer outcomes and strengthen the interpretation of ER as a biomarker.
