Outcome prediction of oestrogen receptor-positive breast cancer based on a panel of oestrogen receptor-regulated genes.

BACKGROUND: The response of oestrogen receptor-positive (ER+) breast cancers (BC) to endocrine therapy (ET) is variable. ER pathway-regulated genes have been proposed to play a role in response to ET. In this study, we investigated the prognostic and predictive impacts of the expression of key ER-regulated genes in BC. METHODS: The Cancer Genome Atlas data was used to identify differentially expressed genes (DEG) associated with ER-positivity. Of the DEGs (1329 upregulated and 1188 downregulated genes), 21 top genes showed biological and clinical relevance to ER functions and were further investigated. Publicly available transcriptomic datasets were utilised to evaluate the clinical significance of the expression of these 21 genes. The well-characterised Nottingham operable BC cohort was used to assess their protein expression. Genes that demonstrated prognostic significance on both levels were subsequently tested individually and in combination using multivariate Cox regression analysis. RESULTS: Of the 21 assessed ER-regulated genes, four genes (PR, GREB1, AR and BEX1) maintained their prognostic significance in ER+ BC at both the transcriptomic and proteomic levels. Multivariate Cox regression analyses showed that only PR and GREB1 are predictors of ET response independent of tumour grade, size or lymph node status. The combined PR-GREB1 expression was a strong predictor of ET response. CONCLUSIONS: This study showed that when several ER-related biomarkers were evaluated, only PR and GREB1 retained their independent prognostic significance and can be used, individually or in combination, to predict the response to ET in ER+ BC patients.