Genome-wide Determinants of Prognosis for Early-Onset Breast Cancer This application addresses broad Challenge Area (08) Genomics and specific Challenge Topic, 08-CA-101: Augmenting Genome-Wide Association Studies Prognosis for breast cancer is disproportionately worse for early-onset cases, who are more likely to be negative for estrogen and progesterone receptors and present with higher grade tumors and advanced stage disease. To date, there have been several publications on genome-wide association (GWA) study of breast cancer risk, but there has been no such publication on breast cancer prognosis. The proposed study will allow us to conduct GWA study analyses of prognosis of early-onset breast cancer with good statistical power, using existing research resources from ongoing study of breast cancer risk. We have access to data on relevant prognostic and treatment data for a large number of breast cancer patients with comprehensive repository of germline and tumor DNA and RNA. Our application has four components. First, we plan to use existing genetic data from a genome-wide association (GWA) study of early-onset breast cancer to conduct a GWA scan for prognostic outcomes. We will examine associations for 610,000 SNPs and CNVs (Illumina Platform) in the 3,000 cases (diagnosed at age 50 or before) in relation to several prognostic outcomes (e.g. overall mortality, breast cancer-specific mortality) using ~10 years of follow-up data. In a second stage, we will genotype ~2,700 markers showing the strongest associations in stage 1 of this study in an independent sample of breast cancer patients diagnosed at age 50 or prior (n=1,000) and drawn from the same source populations as the cases studied in stage 1. Genotyping will be done using Illumina's Golden Gate platform. Data on these ~2,700 SNPs from both stages will then be analyzed jointly to identify SNPs showing significant associations with breast cancer prognosis, accounting for population stratification and multiple testing. In a third stage, we will obtain sequence data for the regions identified in the GWA analysis from the 1,000 genomes project (or HapMap if data is not available at the time of analysis). The LD patterns observed in this data will be used to identify candidate causal prognostic alleles and impute their genotypes in the sample used in the stage 1 analysis (i.e., samples with genome wide data). These markers will then be evaluated for associations with prognosis and interaction with clinical factors and other prognostic alleles. In a final stage, we will characterize and validate the function and molecular basis of the effects of the prognostic alleles, by running whole genome gene expression, microRNA expression and methylation profiles in tumor tissues for 200 breast cancer cases. These cases will consist of 100 """"""""carriers"""""""" and 100 """"""""non-carriers"""""""" for each of the candidate causal prognostic alleles (in each group, 50 with poor outcome, 50 with good outcome). These analyses will be conducted using Illumina's genome-wide gene expression, methylation and microRNA panels. Overall, we believe that our proposed comprehensive investigations of genome-wide analysis of early- onset breast cancer prognosis will address a critical research question in the highly cost-efficient and timely manner, providing important data for the first time on genotype-prognosis associations for pre-menopausal, early-onset breast cancer, while elucidating the molecular basis of these prognostic effects. The proposed research will generate knowledge that can be applied for future pharmacological, translational, epidemiological, and basic research.
In this project, we propose to search for and characterize inherited genetic factors that influence breast cancer prognosis (i.e., survival, cancer recurrence). We plan to comprehensively test over 600,000 genetic markers for associations with prognosis using data on ~4,000 breast cancer patients who have been under observation for ~10 years. We will characterize the biological effects of the genetic factors we identify using various techniques comprehensively characterize breast tumors based on molecular features related to tumor DNA and RNA.
