Abstract 4934: Promoter methylation biomarkers for breast cancer risk stratification using an unbiased whole genome approach

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DC Background: Silencing of tumor suppressor gene expression by promoter methylation occurs frequently in breast cancer and can affect the expression of >100 genes per tumor. Since tumorigenesis is a multistep process and occurs over a large period of time, promoter methylation of some genes may occur very early in this process and can be used for risk stratification or occur very late and be used for early detection. The objective of our study is to identify genes that are frequently methylated in breast cancers and can be used for individualized breast cancer risk stratification. Methods and Results: We have used an unbiased whole genome approach to identify breast cancer methylation markers in addition to 18 candidate markers obtained from breast cancer literature. Six breast cancer cell lines and 6 short term benign breast primary cultures were exposed to 0.5 µM 5-aza-2′-deoxycytidine or DMSO for 5 days; total RNA was isolated and expression differences were determined by gene expression profiling using Illumina arrays. After normalization and bioinformatic analysis, 288 genes of interest were identified. The methylation status of each promoter was analyzed by methylation specific PCR (MSP) using an independent set of 10 breast cancer cell lines, 6 benign breast primary cultures, and 4 lymphocyte samples. Of the 288 genes tested, 204 genes were excluded because they were: not methylated in breast cancer cell lines (109), methylated in lymphocytes (91), or methylated in all benign breast cultures at the same intensity as the cancer samples (4). Of the remaining 84 genes, 58 genes with the highest methylation frequency in the breast cancer cell lines were analyzed by MSP with a panel of 15 primary breast cancers and 15 random periareolar fine needle aspirate (RP-FNA) samples from benign breast tissue (5 breast cancer patients, 5 unaffected Gail high-risk patients, and 5 unaffected Gail average-risk patients). Thirty seven genes were methylated in at least 4 breast cancer samples and had a cancer/RP-FNA ratio of >2. Unsupervised hierarchical clustering separated the genes into 3 groups based on the primary breast tumor data. One group containing 10 genes was highly methylated in all primary tumors with an average of 7.6/10 methylated genes per tumor (range 3-10). 21 genes representing all groups have been selected for quantitative MSP analysis. We are currently testing these genes in a panel of 200 RP-FNA samples including primary breast tumors and benign breast tissue from cancer and unaffected patients to identify biomarkers for breast cancer risk stratification. These data will be presented at the meeting. Conclusions: We have identified 37 candidate genes that are highly methylated in breast cancer and less frequently in benign breast tissue. Quantitative MSP analysis is underway using clinical samples to determine the best biomarkers for individualized breast cancer risk assessment. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 4934.