Dynamic changes in gene expression in vivo predict prognosis of tamoxifen-treated patients with breast cancer

Abstract

Introduction: Tamoxifen is the most widely prescribed anti-estrogen treatment for patients with estrogen receptor (ER)-positive breast cancer. However, there is still a need for biomarkers that reliably predict endocrine sensitivity in breast cancers and these may well be expressed in a dynamic manner. Methods: In this study we assessed gene expression changes at multiple time points (days 1, 2, 4, 7, 14) after tamoxifen treatment in the ER-positive ZR-75-1 xenograft model that displays significant changes in apoptosis, proliferation and angiogenesis within 2 days of therapy. Results: Hierarchical clustering identified six time-related gene expression patterns, which separated into three groups: two with early/transient responses, two with continuous/late responses and two with variable response patterns. The early/transient response represented reductions in many genes that are involved in cell cycle and proliferation (e.g. BUB1B, CCNA2, CDKN3, MKI67, UBE2C), whereas the continuous/late changed genes represented the more classical estrogen response genes (e.g. TFF1, TFF3, IGFBP5). Genes and the proteins they encode were confirmed to have similar temporal patterns of expression in vitro and in vivo and correlated with reduction in tumour volume in primary breast cancer. The profiles of genes that were most differentially expressed on days 2, 4 and 7 following treatment were able to predict prognosis, whereas those most changed on days 1 and 14 were not, in four tamoxifen treated datasets representing a total of 404 patients. Conclusions: Both early/transient/proliferation response genes and continuous/late/estrogen-response genes are able to predict prognosis of primary breast tumours in a dynamic manner. Temporal expression of therapy-response genes is clearly an important factor in characterising the response to endocrine therapy in breast tumours which has significant implications for the timing of biopsies in neoadjuvant biomarker studies.