To cite this article: Morange P-E, Saut N, Antoni G, Emmerich J, Tré gouë t D-A. Impact on venous thrombosis risk of newly discovered gene variants associated with FVIII and VWF plasma levels. J Thromb Haemost 2011; 9: 229-31.Through a genome-wide association study (GWAS), seven new single nucleotide polymorphisms (SNPs) were recently found to be associated with plasma levels of factor (F) VIII and von Willebrand Factor (VWF) by the CHARGE consortium [1]. These SNPs were located in five genes that have not previously been suspected to be associated with these protein levels. More specifically, VWF plasma levels were found to be influenced by SNPs located in TC2N (rs10133762), CLEC4M (rs868875), SCARA5 (rs2726953), STAB 2 (rs4981022) and STXBP5 (rs9390459). The last three genes were also found to be associated with FVIII plasma levels, either through the same SNP (rs9390459 for STXBP5) or via different SNPs (rs9644133 for SCARA5 and rs12229292 for STAB 2). Because high plasma levels of these two proteins are considered risk factors for venous thrombosis (VT) [2-4], we investigated the influence of these SNPs on VT risk using the results of our previously published GWAS on VT (in silico GWAS) and two candidate gene case-control studies, MARTHA and FARIVE, all three based on independent French populations [5].Two of these seven SNPs, rs9390459 and rs9644133, were available in the DNA chip array used in our previous GWAS, but none of them showed any trend of an association with VT risk (Table 1). The five other SNPs identified in CHARGE were not typed in our GWAS but according to the HapMap database, three of them, rs12229292, rs2726953 and rs10133762, were in strong linkage disequilibrium with SNPs available in our GWAS. Their best proxies were rs4981021 (pairwise r 2 = 0.879), rs4276643 (r 2 = 0.884) and rs1884841 (r 2 = 0.961), respectively (Table 1). Among those, only the rs1884841 showed a promising significant association with VT (Table 1) as the rs1884841-T allele was more frequent in cases than in controls (0.498 vs. 0.432, P = 6.45 · 10 )4). The latest two hit SNPs identified in CHARGE, rs4981022 and rs868875, were not in strong LD with any of the GWAS typed SNPs, with best proxies being rs608773 (r 2 = 0.124) and rs4964629 (r 2 = 0.266), respectively. As a consequence, we decided to genotype the rs1884841 in the MARTHA study composed of 1150 VT patients and 801 healthy individuals [5] for further support of association with VT as well as both SNPs that did not have good proxies in our GWAS, rs4981022 and rs868875. While the rs4981022 did not show any evidence of an association with VT risk in MARTHA, the rs868875-G allele was found to be less frequent in cases than in controls (0.292 vs. 0.338, P = 0.0025) and the rs1884841-T more frequent in cases than in controls (0.487 vs. 0.437, P = 0.0024) (Table 1). Consequently, these last two SNPs were further examined for support of an association with VT in a sample of 594 VT patients and 588 controls part of the FARIVE study [5]. We were not able to confirm th...