Gene Expression and Pathway Activation Biomarkers of Breast Cancer Sensitivity to Taxanes.

Abstract

Taxanes are one of the most widely used classes of breast cancer (BC) therapeutics. Despite the long history of clinical usage, the molecular mechanisms of their action and cancer resistance are still not fully understood. Here we aimed to identify gene expression and molecular pathway activation biomarkers of BC sensitivity to taxane drugs paclitaxel and docetaxel. We used to our knowledge the biggest collection of clinically annotated publicly available literature BC gene expression data (12 datasets, n = 1250) and the experimental clinical BC cohort (n = 12). Seven literature datasets were used for biomarker discovery (n = 914), and the remaining five literature plus one experimental datasets (n = 336) - for the validation. We totally found 34 genes and 29 molecular pathways which could strongly discriminate good and poor responders to taxane treatments. The biomarker genes and pathways were associated with molecular processes related to formation of mitotic spindle and centromeres, and with the spindle assembly mitotic checkpoint. Furthermore, we created gene expression and pathway activation signatures predicting BC response to taxanes. These signatures were tested on the validation BC cohort and demonstrated strong biomarker potential reflected by mean AUC values of 0.76 and 0.77, respectively, which outperforms previously reported analogs. Taken together, these findings can deepen our understanding of mechanism of action of taxanes and potentially improve personalization of treatment in BC.