Proteomic Prediction of Breast Cancer Risk: A Cohort Study

摘要

Our objective is to develop and test proteomic methods for the prediction of breast cancer risk, an approach that has not been attempted previously. Our underlying hypothesis is that proteomic analysis of serum will identify proteins differentially expressed in women who do versus those who do not develop invasive breast cancer, and that these differences will be identifiable prior to the clinical presentation of breast cancer. Our work is being conducted in two phases, a training phase and a test phase. Both phases will be conducted as case-control studies nested in a population-based cohort of women who were members of Kaiser Permanente. These serum specimens were collected between 1986 and 1992. We have finalized our cohort definition, finalized the definition of cases and controls; finalized the criteria for matching controls to cases; selected the cases and controls; pulled and aliquotted the serum specimens. For the proteomic analysis, we have developed a detailed protocol for analysis of the serum samples. Briefly, the serum sample is loaded onto an immunoaffinity column to deplete twelve abundant proteins, and the flow-through fraction is collected and subjected to tryptic digestion. Subsequently, the peptides are labeled with iTRAQ reagents and fractionated by cation exchange chromatography (SCX). Six pooled SCX fractions are separately loaded onto a reverse phase column and followed by MALDI-TOF/TOF (4800 Proteomic Analyzer) analyses. The data collected are automatically processed, combined, and searched against a human protein database. This procedure has been thoroughly tested for quantitation and complexity (dynamic range) and a reproducibility study is underway. By applying this high-resolution proteomic approach to a prospective setting, this ongoing project should enhance our ability to identify those women at increased risk of breast cancer and intervene before they progress to cancer.