Abstract 2607: A GWAS From The Vitamin Intervention For Stroke Prevention (VISP) Clinical Trial Identifies Variants That Contribute To Recurrent Stroke And Combined Vascular Endpoints

Keith L Keene,Fang Chen,Wei-Min Chen,Fang-Chi Hsu,Josyf C Mychaleckyj,Kimberly F Doheny,Marcia Adams,Cathy C Laurie,Stephanie M Gogarten,Patrick J Heagerty,Ebony B Bookman,Bradford B Worrall,Michele M Sale,

DOI: https://doi.org/10.1161/str.43.suppl_1.a2607

IF: 10.17

2012-02-01

Stroke

Abstract:We have performed a genome-wide association study (GWAS), using the Illumina Omni 1M SNP platform, in 2,100 stroke patients from the Vitamin Intervention for Stroke Prevention (VISP) trial. VISP was a controlled, double blinded clinical trial, designed to determine whether the daily intake of high dose folic acid, vitamins B6 and B12 reduce recurrent cerebral infarction. Extensive quality control (QC) measures were performed, resulting in a total of 737,081 SNPs for analysis. Principal Component Analysis (PCA), implemented in KING, was utilized to address confounders due to population substructure. We performed genome wide association (survival) analyses for recurrent stroke and combined vascular endpoints (stroke, MI, and CHD death), using stratified Cox model. Clinic sites were treated as strata and age, sex, treatment arm, and the top 10 principal components were used as covariates for these analyses. There is no inflation in our GWAS scan results, with GC lambda values of 1.008 and 0.995 for recurrent stroke and combined vascular endpoints, respectively. For recurrent stroke, we observed association with two SNPs that reached genome wide significance (≤10 -8 ). Both SNPs were located in a gene sparse region of chromosome 1p31 (P=1.73×10 -8 ; beta=-0.64; P=3.43×10 -8 ; beta=-0.62). Two additional SNPs approached genome wide significance, one located in the same 1p31 region (P=7.60×10 -7 ) and the second located in the GMPR gene (P=8.69×10 -7 ). Variants in the GMPR gene region have been associated with LDL-cholesterol measures in previous GWAS. While no SNPs reached genome wide significance for combined vascular endpoints, six SNPs approached genome wide significance (P values ranging from 9.51×10 -6 to 2.56×10 -7 ). These variants are located in genes that have implications in vascular disease. In particular, GPR37L1 has been identified as a potential target for heart failure in mouse expression studies while NSMAF may play a role in regulating TNF-induced cellular responses (e.g. inflammation). Additionally, ILDR2 is located in a region that has shown previous linkage for hypertension, suggesting potential biological implications for these variants in relation to stroke and vascular disease. Our GWAS scans represent the first large scale genetic association study of recurrent stroke and combined vascular events. Our findings have identified novel associations for genes that contribute to recurrent stroke and composite vascular endpoints.

peripheral vascular disease,clinical neurology