DNA copy number stratification of breast cancer samples identifies distinct biological mechanisms

B30 A number of published studies have suggested that DNA copy number and gene expression can be an important diagnostic and therapy determinant for cancer. In this study, a set of 58 FFPE breast cancer samples were hybridized on high-density oligonucleotide aCGH arrays. Using a novel hierarchical clustering algorithm the samples were stratified into three distinct clusters. One cluster was predominantly samples that had mostly a 1q gain event, the second cluster had 8p loss and 8q gain, and the third had both the 1q gain as well as the 8p loss and 8q gain. Performing survival analysis, we identified one cluster to have significantly better outcome than the other two. We have then integrated a set of gene expression profiles previously reported in the literature/databases as prognostic markers and identify distinct hot spots where changes in gene expression correlate with copy number change uniquely in each cluster. Using this analysis we were able to identify genomic structure and corresponding gene expression with pathological biological variables and outcome.