Use of a SNP panel to refine risk estimates in women at high risk of breast cancer: Results from two randomized tamoxifen prevention trials.

1519 Background: An assessment of how panels of single-nucleotide polymorphism (SNP) breast cancer risk factors might improve risk assessment among high-risk women is needed for better targeting of preventive therapy and other risk reduction measures. Methods: A case-control study was designed using leukocytic or tissue DNA from two tamoxifen prevention trials (IBIS-I and Marsden). SNPs were assayed on the Sequenom platform and a relative risk score for 15 SNPs (SNP15) was formed using the most recent iCOGS SNP risk estimates. Baseline questionnaires were used to estimate 10-year relative risks from the Tyrer-Cuzick model (TC). Conditional logistic regression was used to assess performance. An updated panel using all validated risk SNPs (approx. 70) is also being evaluated and will be presented at the meeting. Results: 440 cases and 686 controls had median age at baseline of 49 (inter-quartile range 45 - 54). Hardy-Weinberg equilibrium was verified. SNP15 and TC appeared uncorrelated (Spearman 0.037, P=0....