Blood-Based Biomarkers of Early-Onset Breast Cancer
Abstract
Women with early-onset breast cancer are thought to have a higher contribution of inherited risk than those forming sporadic cancers at later ages. This inherited susceptibility to breast cancer might manifest as differences in gene expression patterns within key oncogenic pathways. While the normal breast is the ideal tissue in which to study this phenomenon, gene expression profiling of blood lymphocytes has been successfully used as a proxy in a variety of diseases including breast cancer. In the first phase of this project, we investigate the gene expression profile of untransformed blood lymphocytes in order to discover gene expression (mRNA and miRNA) signatures which can differentiate BRCA 1/2 negative women with a personal history of early-onset breast cancer and family history of breast cancer (n=51) from asymptomatic aged-matched women without a personal history of cancer or family history of breast cancer (n=50). Using adaboost methodology, we were able to differentiate cases from controls in our discovery cohort with 73% accuracy (sensitivity of 85% and specificity of 64%) using mRNA data alone. We are currently in the process of re-analyzing our data using the elastic net method to determine its robustness, prior to attempts at validation in an independent cohort.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 2014
- Accession Number
- ADA613793
Entities
People
- Allan Balmain
- David Quigley
- Laura Esserman
- Nasim Ahmadiyeh
Organizations
- University of California, San Francisco