Gene expression profiles of estrogen receptor-positive and estrogen receptor-negative breast cancers are detectable in histologically normal breast epithelium.

PURPOSE Previously, we found that gene expression in histologically normal breast epithelium (NlEpi) from women at high breast cancer risk can resemble gene expression in NlEpi from cancer-containing breasts. Therefore, we hypothesized that gene expression characteristic of a cancer subtype might be seen in NlEpi of breasts containing that subtype. EXPERIMENTAL DESIGN We examined gene expression in 46 cases of microdissected NlEpi from untreated women undergoing breast cancer surgery. From 30 age-matched cases [15 estrogen receptor (ER)+, 15 ER-] we used Affymetryix U133A arrays. From 16 independent cases (9 ER+, 7 ER-), we validated selected genes using quantitative real-time PCR (qPCR). We then compared gene expression between NlEpi and invasive breast cancer using four publicly available data sets. RESULTS We identified 198 genes that are differentially expressed between NlEpi from breasts with ER+ (NlEpiER+) compared with ER- cancers (NlEpiER-). These include genes characteristic of ER+ and ER- cancers (e.g., ESR1, GATA3, and CX3CL1, FABP7). qPCR validated the microarray results in both the 30 original cases and the 16 independent cases. Gene expression in NlEpiER+ and NlEpiER- resembled gene expression in ER+ and ER- cancers, respectively: 25% to 53% of the genes or probes examined in four external data sets overlapped between NlEpi and the corresponding cancer subtype. CONCLUSIONS Gene expression differs in NlEpi of breasts containing ER+ compared with ER- breast cancers. These differences echo differences in ER+ and ER- invasive cancers. NlEpi gene expression may help elucidate subtype-specific risk signatures, identify early genomic events in cancer development, and locate targets for prevention and therapy.