Breast cancer affects one in eight women in the United States. Current treatment regimes tend to extend post-surgical adjuvant chemotherapy to lymph node negative patients with ever smaller original tumors, even though the majority of such patients do not appear to benefit from such treatment. The establishment of molecular signatures that define subgroups of these patients with different likelihoods of tumor recurrence would be useful allowing therapy to be tailored of the prognosis of the patient. We will also a novel global gene expression analysis procedure called RAGE to establish gene expression profiles for early node- negative breast tumors. A retrospective group of T1N0M0 and T2N0M0 breast tumors, 20 that have recurred within 5 years and 20 that have not, will be studied in Phase 1. The RAGE method will be adapted to a high throughput separation and analysis technology, multiple capillary electrophoresis, and throughput will be increased through various means to about 10,000 measurements per week. Expression levels of a set of approximately 3000 genes will be measured in each of the 40 retrospective tumors. These data will be combined with analogous data collected in Project 1 using SAGE on a subset of these tumors. The combined RAGE and SAGE data obtained in Phase I will be used to identify a set of approximately 700 candidate signature genes that show significant variance in expression between recurrent and non-recurrent tumors or that are expected to be important in breast tumorigenesis. The RAGE technique will then be used in Phase II to study expression of these candidate genes in a prospective set of 250 T1N0M0 and T2N0M0 breast tumors. Additional genes will be added to this set as new candidate genes are identified by further SAGE studies (Project 1) and as advances in our understanding of the biology of breast cancer are made. Ultimately, we expect to determine expression levels for about 1000 genes in the prospective tumor set. Novel genes whose expression may contribute to molecular signatures will be identified and cloned. Personnel in this Project will work closely with statisticians, information scientists and programmers in Project 3 and in the Data Management Core to analyze this dataset for molecular signatures characteristic of subsets of the tumor population.
Showing the most recent 10 out of 15 publications
