Validate Gene Expression and Proteomic Signatures Predictive of Treatment for Response for Breast Cancer Patient

Many chemotherapeutic agents are active in breast cancer, although response rate to any individual drug is only 30-50%. The choice of chemotherapy is empirical, and development of a chemosensitivity assay is desirable, to reduce costs, unnecessary toxicity, and loss of window of opportunity to cure. Single molecular markers to predict sensitivity are not highly accurate, as chemotherapy resistance mechanisms likely involve complex pathways. High-throughput technologies such as gene expression microarray and Proteinchip array allow simultaneous analysis of thousands of genes, and hundreds of proteins, and may be more informative. We previously conducted a study on patients with measurable primary breast tumor who received primary chemotherapy with an alternating regimen of adriamycin and docetaxel, and generated tumor genomic and tumor and plasma proteomic signatures that predicted for clinical and pathological response using high throughput discovery platforms. This protocol aims to recruit 20 patients as an independent test set to validate the genomic and proteomic signatures generated previously. Half the patients will be randomized to receive 4 cycles of pre-operative adriamycin (Arm A) allowing validation of the adriamycin-specific signatures, while the other half will be randomized to receive 4 cycles of pre-operative docetaxel (Arm B) allowing validation of the docetaxel-specific signatures. Subjects will then undergo resection of the primary breast tumor, followed by 4 cycles of adjuvant therapy with the alternative drug (docetaxel in Arm A, adriamycin in Arm B). Serial tumor and plasma samples will be obtained for genomic and proteomic analysis. The previously generated genomic and proteomic signatures will be applied to this independent dataset to categorize patients into good and poor responders, and the prediction correlated with actual treatment responses. Secondary goals include the correlation of patient genotype with drug pharmacokinetics, and the correlation of chemotherapy-induced peripheral blood mononuclear cell gene expression changes with treatment response and toxicities