[Gene signatures for breast cancer, clinical utility and therapeutic applications].

Gene signatures quantify hormone receptors and proliferation genes, combining multivariate prediction models. Hormone-negative tumors have greater proliferation and the prognostic value is limited. The first generation of prognostic signatures (Oncotype DX, MammaPrint, Genomic Degree Index) predict recurrence at 5 years. Subsequent tests (Prosigna, EndoPredict, Breast Cancer Index) have better prognostic value for recurrence and are predictive of early relapse. There are no useful prognostic genetic tests for hormone-negative tumors, or predictors of response to treatment. The recent expansion of high-performance technology platforms including the low-cost sequencing of tumor-derived DNA and circulating RNA and the reliable rapid quantification of microRNAs offer new opportunities to build prediction models.