Abstract P3-06-02: Genetic polymorphisms from genome-wide association study associated with the metabolic and cell proliferation pathways affect the time to distant metastases of hormone receptor-positive and Her2-negative early breast cancer

Abstract

Abstract Background: Single nucleotide polymorphisms (SNPs) identified from genome-wide association study (GWAS) have been found to be associated with breast cancer risk. We hypothesized that candidate genes and genes derived from GWAS involving in the Estrone/Estradiol (E2)/Tamoxifen biosynthesis may influence the adjuvant hormonal therapy effect and the survival. In this study, we sought to investigate whether these SNPs are associated with prognosis of hormone receptor (HR)-positive breast cancer patients, especially in HER2-negative patients. Patients and methods: We selected breast cancer susceptibility SNPs identified by GWAS, SNPs in tamoxifen metabolizing related genes, and SNPs in estrogen receptor genes and estrogen metabolism genes, and genotyped for variations of above genes, including ALDH3A1, CYP2C19, COMT, CYP19, MAP3K1, FGFR2, TNRC9, HCN1, ERCC4, CYP3A5, UGT1A1, ER, ABCG2, CYP2B6, CYP2D6, 5p12 in 171 hormone receptor-positive, and Her2-negative early breast cancer patients (127 with negative lymph node [LN], and 44 with 1–3 positive LN). All patients received adjuvant hormonal therapy. The associations were examined between SNPs and distance disease-free survival (DDFS), and overall survival (OS) by using the log-rank test and Cox's proportional hazard model. Furthermore, we combined clinicopathologic features and SNPs into the risk score analysis to further validate above identified genetic markers. Results: We found that SNPs of CYP2B6 (rs3211371), FGFR2 (rs2981582), and MAP3K1 (rs889312) were significantly associated with DDFS and OS. Furthermore, in lymph node-negative patients, CYP2B6 (rs3211371), FGFR2 (rs2981582), MAP3K1 (rs889312) and 5p12 (rs10941679 and rs4415084) were significantly associated with DDFS and OS, while CYP3A5 (rs776746) was significantly associated with OS but not DDFS. We further assessed the associations of disease prognosis with the number of high-risk genotypes in CYP3A5, FGFR2, and MAP3K1, and showed significant dose-response relationships between the number of high-risk alleles at these 3 loci and DDFS (P = 0.005 for trend) and OS (P = 0.0008 for trend). When combining the clinicopathologic features and SNPs into the risk score analysis, patients were divided into 3 subgroups (subgroup 1, risk score<-1.438, n=43 [LN-positive, n=7]; subgroup 2, risk score between −1.438 and 1.708, n=85 [LN-positive, n=26]; subgroup 3, risk score>1.708, n=43 [LN-positive, n=11]). We found that around 20 % of subgroup 3 patients had early development of distant metastases (DM) in the upfront 3 years (the trend of DM appeared to persist at least 10 years), and subgroup 2 patients had higher risk of DM after 5 years, whereas subgroup 1 patients had no development of DM even after 12 years. Conclusion: Our results indicate that in addition to drug metabolic genes, genes related to cell proliferation, anti-apoptosis, and signaling transduction, for example, CYP3A5, CYP2B6, FGFR2, and MAP3K1 genes were associated with DDFS and OS in HR-positive/Her2-negative breast cancer patients. These findings provide additional insight that the genetic variants, or host factors, may affect the prognosis of breast cancer. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P3-06-02.