Breast cancer patients whose tumours are oestrogen receptor (ER)-negative are more likely to have poorer outcomes than those with ER-positive tumours. These patients are often treated with standard chemotherapy regimes, rather than the hormonal and targeted treatments used for ER-positive patients.

Chen et al. obtained large published datasets of drug response and gene expression profiles (breast cancer patient cohorts, plus in vitro studies of cancer cell lines). They then used these to develop genomic biomarker signatures for the agents paclitaxel, 5-fluorouracil, cyclophosphamide and doxorubicin, which are commonly used as adjuvant chemotherapies for ER-negative patients. These models were subsequently validated in independent patient datasets.

The authors hope that their proposed biomarker panels could be used to predict which patients are most likely to respond to a particular agent. This would facilitate personalised treatment decisions, rather than agents being used in a ‘trial and error’ manner.

Read the full article in Endocrine-Related Cancer 25 595–605.