Analysis of Prognostic and Predictive Genomic Signatures Using Archival Paraffin-embedded Tumor Specimens in Breast Cancer

Major Aims of study: 1. To create a gene expression-based prognostic device that complements or exceeds the prognostic utility of conventional biomarkers of breast cancer outcome. 2. To identify one or more clinical subgroups of patients for which the prognostic device outperforms, or substantially adds to, the prognostic performance of conventional markers that currently determine therapeutic strategies. Sub-Aims of study: 1. Assess the prognostic value of the multiple gene expression signatures, alone and in combination, using a large cohort of breast cancer patients for which pathology, treatment and outcome is available. A "training" and "testing" design is proposed. 2. Evaluate the utility of a prognostic device that measures gene expression levels from formalin-fixed paraffin-embedded specimens (FFPEs) of primary resected tumors. The investigators will utilize the Affymetrix Quantigene 2.0 Assay and/or the Illumina BeadXpress VeraCode DASL Gene Expression Assay (FDA-approved IVDMIA.) 3. For specific clinical subgroups of patients/tumors, the investigators will mathematically identify additive or synergistic prognostic relationships between genes and gene signatures that, in combination, will yield maximal risk prediction (distant metastases-free survival) for patients. 4. Compare the prognostic utility of the investigators device to that of the conventional prognostic variables that are currently used to determine therapeutic strategy. 5. Incorporate the prognostic signatures into a practical prognosis algorithm that seeks to include conventional measures of outcome such as tumor size, histologic grade, nodal status, patient age, or Nottingham index, etc. The investigators hypothesize that adequate quality and quantity of tumor RNA may be extracted from archival paraffin-embedded tumor specimens for gene expression profiling, and that archival tumor-derived genomic signatures may be used as prognosticators or predictors in breast cancer.