Abstract P4-07-01: Correlation of Breast Cancer Susceptibility Loci with Patient Characteristics, Metastasis-Free Survival and Expression of the Nearest Genes

Abstract

Abstract Background: Genome-wide association studies has identified single nucleotide polymorphisms (SNPs) in several loci being associated with breast cancer risk: fibroblast growth factor receptor 2 (FGFR2), trinucleotide repeat containing 9 (TNRC9 or TOX3), mitogen-activated protein kinase kinase kinase 1 (MAP3K1), lymphocyte-specific protein 1 (LSP1) and the imprinted maternally expressed H19, and loci in gene deserts at 8q24 and 2q35. However, the mechanism by which these loci confer breast cancer risk in patients is still largely unknown. Here, we addressed this question by associating these SNPs with the mRNA expression of the genes located at or nearest to the SNP in the same linkage disequilibrium region. In addition, we associated the SNPs with clinical, pathological and patient characteristics and prognosis. Material and Methods : SNPs tagging breast cancer loci were genotyped in genomic tumor DNA samples of 2,480 breast cancer patients. All samples were collected between 1978-2004. The mean age was 55.6 years and median follow-up up was 106 months. The SNPs were correlated with patients and tumor characteristics. Of the 1,262 patients with lymph-node negative disease and who did not receive any adjuvant systemic therapy, SNP status was associated with distant metastasis-free survival (MFS) using Cox regression analysis. Finally, in a subset of 1,400 of the 2,480 patients, the mRNA expression of FGFR2, TNRC9, MAP3K1, LSP1 and H19 genes was determined by quantitative RT-PCR and correlated with SNP genotypes. Results: The SNP rs2981582 in FGFR2 was significantly associated with ER and PgR status of the tumors (both p=0.001). Besides weak associations with tumor grade (FGFR2, p=0.01), ER (MAP3K1, p=0.03; LSP1, p=0.03), and PgR (LSP1, p=0.05), no other association with any clinical or pathological variable for any of the SNPs was observed. Of the SNPs analyzed, only rs2107425 near H19 was significantly linked in uni-and multivariable analysis with MFS (Hazard ratio [HR]=1.44, 95% Confidence interval [CI]: 1.04-1.98; p=0.026; HR=1.53; 95% CI: 1.09-2.14; p=0.013, respectively) with the more aggressive minor allele displaying a recessive trade. Interestingly, the minor allele of SNP rs3803662 was significantly associated with lower mRNA expression of the 8 kb downstream TNRC9 gene (p=0.0019). However, none of the other risk alleles, including the one in FGFR2, had an association with mRNA expression of the nearest located gene. Conclusions: In agreement with previous studies, a clear correlation of the SNPs in FGFR2 with ER and PR status of tumors was observed. The lower level of TNRC9 mRNA in tumors having the minor allele genotype suggests that TNRC9 may act as a tumor suppressor gene and its expression might have a protective effect. A significant association of the SNP near H19 with poor outcome without apparent effect on H19 mRNA expression suggests that this prognostic SNP in a well-known imprinted region is not linked to prognosis through altering H19 gene expression. This study indicates that most of the studied SNPs that confer breast cancer risk are not linked to prognosis nor do they effect, in primary tumors mRNA expression of the nearest gene. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P4-07-01.