Taking Prognostic Signatures to the Clinic

In this chapter, the authors focus on genomic predictors related to prognosis and the way these predictors are expected to impact patient management in the future and accelerate the transition from the “empirical” to “tailored” oncology. Clinicians have long recognized that the diagnosis of breast cancer includes tumor types with different natural histories and responses to various treatments. They proposed a molecular classification of breast cancer with at least five subgroups on the basis the expression patterns of 500 “intrinsic genes.” The normal breast-like subgroup showed genes expressed by adipose and nonepithelial tissues and also revealed a strong expression of basal epithelial genes. By combining the hypothesis-driven approach with comparative cross-species functional genomics, G. V. Glinsky et al. identified an 11-gene “death-from-cancer” signature. This signature can identify a lethal subset of human cancers of various origins (including breast cancer) with a high propensity to metastasize.