S6-6: Expression of Key Estrogen-Regulated Genes (ERGs) Differ Substantially across the Menstrual Cycle in ER+ Breast Tumours.

Abstract

Abstract Aim: To determine whether there are substantial changes in the expression of ERGs in estrogen receptor positive (ER+) breast cancer through the menstrual cycle. Background: Plasma levels of estradiol (E2) vary from c.100pM to c.1000pM and progesterone levels from <3nM to >50nM through the menstrual cycle. The changes in E2 are proportionally similar to those that occur in postmenopausal women treated with an aromatase inhibitor which lead to profound changes in the expression of ERGs. However, there are only inconsistent data on whether the cyclical changes in hormone levels during the menstrual cycle affect gene expression in ER+ tumours. Methods; 173 paraffin-embedded ER+ breast carcinomas were analysed from premenopausal patients in which day of menstrual cycle and hormonal data were recorded at the time of surgery1 enabling accurate definition of the timing of the ovarian cycle. The patients were ascribed to one of 3 pre-defined time windows of the menstrual cycle: window 1: days 27–35 + 1–6 (low circulating E2 and progesterone); window 2: days 7–16 (high E2 and low progesterone); window 3: days 17–26 (moderate E2 and high progesterone). RNA was extracted (RecoverAll; Ambion) and QPCR used to measure expression of ESR1, 4 ERGs (TFF1, PGR, GREB1 and PDZK1)2 and 3 housekeeping genes. Results: ESR1 expression did not differ significantly across the menstrual cycle but there was strong evidence of differences in the ERGs (Kruskal-Wallis; p=0.0015 to 0.093). Gene expression levels of the 4 ERGs were 51–109% higher in window 2 than window 1 (see table) and this was significant for PGR, GREB1 and TFF1 and approached significance for PDZK1. The expression of all the ERGs was lower (14-59%) in window 3 compared to window 2 (significant for TFF1). PGR and GREB1 expression was significantly higher (30-40%) in window 3 compared to window 1. The average expression of these 4 ERGs gives the AvERG, a previously defined index of estrogen responsiveness2. This showed similar changes to the individual ERGs but the reduced variability of the AvERG led to even greater statistical significance. A control set of ER− tumours (n=83) showed very low expression of the ERGs and no cyclical changes. Discussion: These data reveal significant changes in breast tumour biology across the menstrual cycle. As these changes are likely to occur only in tumours that are estrogen-dependent their measurement may potentially be developed as a test of a tumour's likely response to estrogen deprivation. Several ERGs are present in currently used molecular profiling tests; variable expression of these markers through the menstrual cycle may affect the interpretation of the tests. Proliferation-related genes are currently being assessed to determine if their expression varies in a similar way. 1. Zurrida et al., Crit Rev Oncol Hem 2001, 38; 223–30 2. Dunbier et al., JCO 2010, 28; 1161–7 Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr S6-6.