Abstract P2-05-02: Age-related genetic and epigenetic changes in estrogen-receptor positive breast cancer

Abstract

Abstract Background: Age is the main risk factor for developing breast cancer (BC) in women. The increase in incidence with age is dominated by estrogen receptor positive (ER+) BC despite the >90% reduction of estrogen levels during menopause and later life. Recent models suggest the lifetime risk of BC is correlated with the number of stem cell divisions with random mutations arising during each division and driving cancer development. However, only two genes are mutated in more than 20% of BC while several recurrent copy number alterations (CNAs) occur in more than 50%. New models of age-related risk factors need to be developed that incorporate mutations, CNAs, epigenetics, and other biological factors. As data of this type is limited in normal breast tissue, factors correlated with age of diagnosis in ER+ BC were identified to understand which variables changed with age and might impact risk of developing BC. Methods: In silico analysis of public databases was used to identify factors in ER+ primary BC and normal adjacent tissue (NT) that are correlated with age of diagnosis including DNA mutations, CNAs, DNA methylation and gene/protein expression (TCGA n=599ER+,113NT; METABRIC n=1435ER+,144NT). Results: DNA mutations accumulated with age in ER+ BC with the median mutation count below 28 in primary tumors diagnosis in patients <50yr and rising to more than 43 in patients >80yr. However, the two most frequently mutated genes (PIK3CA and TP53) showed no significant correlation with age. As previously reported, GATA3 mutation rate was nearly twice as high in younger patients (<50yr=0.25,>80yr=0.12). There was little evidence of a general accumulation of CNAs with age, but there were higher rates of gain in 16p (<50yr=0.42,>80yr=0.51) and loss of 1p (<50yr=0.18,>80yr=0.37) in older patients and lower rates of loss in 6q (<50yr=0.34,>80yr=0.24). The most significant correlations with age related to ESR1 including expression of ESR1 mRNA (rho=0.39,P=1.2e-23 Spearman), ERα protein (rho=0.35,P=1.7e-15 Spearman), and demethylation of ESR1 promoter (rho=-0.36,P=1.6e-13 Spearman). The levels of the activated form of ERα (pS118) also correlated with age but to a much lesser extent than total ERα (rho=0.12,P=0.01 Spearman). Demethylation and expression of ESR1 were highly correlated (rho=-0.50,P<1e-16 Spearman). Analysis of data from NT revealed correlation of age of diagnosis, expression of ESR1 (rho=0.34,P=0.001 Spearman) and demethylation of the ESR1 promoter (rho=-0.22,P=0.03 Spearman) but not with ERα expression (rho=0.32,P=0.19 Spearman) although only 19 NT had protein data. Conclusions: Analysis of DNA mutations and CNAs fit previous theoretical models with both the result of stochastic processes. In these models, the low impact on fitness from individual mutations allows the accumulation over time while the high fitness costs of CNAs prevent accumulation and are likely acquired in a single catastrophic event. The ability of ER+ BC to progress in the presence of very low postmenopausal estrogen levels may be partly explained by demethylation leading to higher ESR1 expression and maintenance of ERα activation. In theory, changes in expression and demethylation can be targeted to reduce the age-related increase in risk for developing BC. Citation Format: Schuster EF, Dowsett M. Age-related genetic and epigenetic changes in estrogen-receptor positive breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P2-05-02.