Identification of Breast Cancer Molecular Subtypes

SINCE the advent of array-based technology and the sequencing of the human genome, scientistsattempted to bring new insights into breast cancer biology and prognosis. The identification of natural groups of tumors (called subtypes) from gene expression data was the subject of intense research this last decade. Such an identification usually involved two steps: 1. Identification of sets of co-expressed genes (gene clustering). 2. Identification of groups of similar tumors (sample clustering). Perou and Sorlie et al. were the first to conduct a gene expression profiling study of breast tumors in order to highlight the existence of subtypes, i.e. groups of tumors exhibiting similar ”genetic por-traits”. Their method consisted in (i) identifying the genes with low variance between tumor samples from the same patient but high variance between tumor samples from different patients (referred to as intrinsic gene list in the literature); (ii) performing a two-way hierarchical clustering to identify sets of co-expressed genes (Figure 1) and groups of similar tumors (Figure 2). In addition to highlight the importance of ER and HER2 phenotypes and proliferation, they also shown that the molecular subtypes exhibited different clinical outcome as illustrated by the survival curves in Figure 3. Fig. 1. Geneexpressionpatternsof 85experimental samples representing78 carcinomas, threebenign tumors, and fournormal tissues, analyzedbyhierarchical clustering using the 476 cDNA intrinsic clone set. (A) The tumor specimens were divided into five (or six) subtypes based on differences in gene expression. The cluster dendrogram showing the five (six) subtypes of tumors are colored as: luminal subtype A, dark blue; luminal subtype B, yellow; luminal subtype C, light blue; normal breast-like, green; basal-like, red; and ERBB2!, pink. (B) The full cluster diagram scaled down (the complete 456-clone cluster diagram is availabl