Abstract B63: ESR1 variants are not associated with breast cancer in African American women

Abstract

Abstract The estrogen receptor alpha (ESR1) gene is a candidate gene for breast cancer due to the ligand-activated transcription factor it encodes. The Breast Cancer Association Consortium (BCAC) identified an association of haplotype tagging SNPs in intron 4 of ESR1 and breast cancer risk in 21 European studies that include 53,994 subjects. We examined whether this association replicated in an African American sample from the Women's Health Initiative Single Nucleotide Polymorphism (SNP) Health Association Resource (WHI SHARe). Using this sample of 7800 postmenopausal women, of whom 316 have incident invasive breast cancer, we measured the genetic association of variants in ESR1 and breast cancer risk using Cox proportional hazards models adjusted for age, region, and principal components 1–4 taken from cluster analyses of genomic ancestry. All samples were genotyped on the Affymetrix 6.0 platform, and imputed with HapMap 2 populations (CEU and YRI) as reference. With regard to ER status, 68% of women in our sample (with known ER status) have ER-positive tumors. The most significant BCAC SNP, rs3020314, was not statistically significant in our study (HR: 1.09, 95% CI: 0.91–1.30) and we found no association between any of the other 130 SNPs tested in the ESR1 locus after Bonferroni correction. While our results for rs3020314 are in the same direction than the reported BCAC association it should be noted that the minor alleles differ by race; the C allele is more common in African Americans and the T allele is more common in BCAC (and the HapMap CEU sample). In our study, the minor (T) allele frequencies for rs3020314 are 0.295 for cases and 0.313 for controls. This lack of replication of the findings in intron 4 and the entire ESR1 locus suggests that ESR1 variants may not be associated with breast cancer in African American women. This may be due to a number of factors including linkage disequilibrium structure differences across races, small sample size and reduced power of the WHI sample, or may reflect disparity in tumor type, as African American women have a higher frequency of tumors that are estrogen receptor-negative. Citation Information: Cancer Epidemiol Biomarkers Prev 2011;20(10 Suppl):B63.