Breast cancer is a leading cause of cancer-death among women, where the clinicopathological features of tumors are used to prognosticate and guide therapy. DNA copy number alterations (CNAs), which occur frequently in breast cancer and define key pathogenetic events, are also potentially useful prognostic or predictive factors. Here, we report a genome-wide array-based comparative genomic hybridization (array CGH) survey of CNAs in 89 breast tumors from a patient cohort with locally advanced disease. Statistical analysis links distinct cytoband loci harboring CNAs to specific clinicopathological parameters, including tumor grade, estrogen receptor status, presence of TP53 mutation, and overall survival. Notably, distinct spectra of CNAs also underlie the different subtypes of breast cancer recently defined by expression-profiling, implying these subtypes develop along distinct genetic pathways. In addition, higher numbers of gains/losses are associated with the "basal-like" tumor subtype, while high-level DNA amplification is more frequent in "luminal-B" subtype tumors, suggesting also that distinct mechanisms of genomic instability might underlie their pathogenesis. The identified CNAs may provide a basis for improved patient prognostication, as well as a starting point to define important genes to further our understanding of the pathobiology of breast cancer. This article contains Supplementary Material available at http://www.interscience.wiley.com/jpages/1045-2257/suppmat

Prediction of the clinical outcome of breast cancer is multi-faceted and challenging. There is growing evidence that the complexity of the tumour micro-environment, consisting of several cell types and a complex mixture of proteins, plays an important role in development, progression, and response to therapy. In the current study, we investigated whether invasive breast tumours can be classified on the basis of the expression of extracellular matrix (ECM) components and whether such classification is representative of different clinical outcomes. We first examined the matrix composition of 28 primary breast carcinomas by morphology and gene expression profiling using 22K oligonucleotide Agilent microarrays. Hierarchical clustering of the gene expression profile of 278 ECM-related genes derived from the literature divided the tumours into four main groups (ECM1-4). A set of selected differentially expressed genes was validated by immunohistochemistry. The robustness of the ECM classification was confirmed by studying the four ECM groups in a previously published gene expression data set of 114 early-stage primary breast carcinomas profiled using cDNA arrays. Univariate survival analysis showed significant differences in clinical outcome among the various ECM subclasses. One set of tumours, designated ECM4, had a favourable outcome and was defined by the overexpression of a set of protease inhibitors belonging to the serpin family, while tumours with an ECM1 signature had a poorer prognosis and showed high expression of integrins and metallopeptidases, and low expression of several laminin chains. Furthermore, we identified three surrogate markers of ECM1 tumours: MARCO, PUNC, and SPARC, whose expression levels were associated with breast cancer survival and risk of recurrence. Our findings suggest that primary breast tumours can be classified based upon ECM composition and that this classification provides relevant information on the biology of breast carcinomas, further supporting the hypothesis that clinical outcome is strongly related to stromal characteristics.

Activating mutations of PIK3CA are the most frequent genomic alterations in estrogen receptor (ER)-positive breast tumors, and selective PI3Kα inhibitors are in clinical development. The activity of these agents, however, is not homogeneous, and only a fraction of patients bearing PIK3CA-mutant ER-positive tumors benefit from single agent administration. Searching for mechanisms of resistance, we observed that suppression of PI3K signaling results in induction of ER-dependent transcriptional activity, as demonstrated by changes in expression of genes containing ER binding sites and increased occupancy by the ER of promoter regions of upregulated genes. Furthermore, expression of ESR1 mRNA and ER protein were also increased upon PI3K inhibition. These changes in gene expression were confirmed in vivo in xenografts and patient-derived models and in tumors from patients undergoing treatment with the PI3Kα inhibitor BYL719. The observed effects on transcription were enhanced by the addition of estradiol and suppressed by the anti-ER therapies fulvestrant and tamoxifen. Fulvestrant markedly sensitized ER-positive tumors to PI3Kα inhibition, resulting in major tumor regressions in vivo. We propose that increased ER transcriptional activity may be a reactive mechanism that limits the activity of PI3K inhibitors, and that combined PI3K and ER inhibition is a rational approach to target these tumors.

In this paper, we propose a new method remMap - REgularized Multivariate regression for identifying MAster Predictors - for fitting multivariate response regression models under the high-dimension-low-sample-size setting. remMap is motivated by investigating the regulatory relationships among different biological molecules based on multiple types of high dimensional genomic data. Particularly, we are interested in studying the influence of DNA copy number alterations on RNA transcript levels. For this purpose, we model the dependence of the RNA expression levels on DNA copy numbers through multivariate linear regressions and utilize proper regularization to deal with the high dimensionality as well as to incorporate desired network structures. Criteria for selecting the tuning parameters are also discussed. The performance of the proposed method is illustrated through extensive simulation studies. Finally, remMap is applied to a breast cancer study, in which genome wide RNA transcript levels and DNA copy numbers were measured for 172 tumor samples. We identify a trans-hub region in cytoband 17q12-q21, whose amplification influences the RNA expression levels of more than 30 unlinked genes. These findings may lead to a better understanding of breast cancer pathology.