Combined Effects of Thrombosis Pathway Gene Variants Predict Cardiovascular Events

Abstract

The genetic background of complex diseases is proposed to consist of several low-penetrance risk loci. Addressing this complexity likely requires both large sample size and simultaneous analysis of different predisposing variants. We investigated the role of four thrombosis genes: coagulation factor V (F5), intercellular adhesion molecule 1 (ICAM1), protein C (PROC), and thrombomodulin (THBD) in cardiovascular diseases. Single allelic gene variants and their pair-wise combinations were analyzed in two independently sampled population cohorts from Finland. From among 14,140 FINRISK participants (FINRISK-92, n = 5,999 and FINRISK-97, n = 8,141), we selected for genotyping a sample of 2,222, including 528 incident cardiovascular disease (CVD) cases and random subcohorts totaling 786. To cover all known common haplotypes (>10%), 54 single nucleotide polymorphisms (SNPs) were genotyped. Classification-tree analysis identified 11 SNPs that were further analyzed in Cox's proportional hazard model as single variants and pair-wise combinations. Multiple testing was controlled by use of two independent cohorts and with false-discovery rate. Several CVD risk variants were identified: In women, the combination of F5 rs7542281 × THBD rs1042580, together with three single F5 SNPs, was associated with CVD events. Among men, PROC rs1041296, when combined with either ICAM1 rs5030341 or F5 rs2269648, was associated with total mortality. As a single variant, PROC rs1401296, together with the F5 Leiden mutation, was associated with ischemic stroke events. Our strategy to combine the classification-tree analysis with more traditional genetic models was successful in identifying SNPs—acting either in combination or as single variants—predisposing to CVD, and produced consistent results in two independent cohorts. These results suggest that variants in these four thrombosis genes contribute to arterial cardiovascular events at population level.

Single Nucleotide Variants in the Protein C Pathway and Mortality in Dialysis Patients

“…In addition, the combination of THBD rs1042580 with different Factor V SNVs was associated with an increased risk of cardiovascular events [12]. Other THBD SNVs have also been associated with cardiovascular outcomes when combined with PROC SNVs or factor V Leiden in the general population [12]. However, we did not find an association between THBD rs1042580 AG/GG genotypes and mortality in dialysis patients.…”

Factor V Leiden mutation is not a predisposing factor for acute coronary syndromes

“…2,4,7 However, its prevalence in India is highly variable in different populations. A study in the North Indian population showed clear absence of FVL in North Indian population.…”

“…10,11 Another risk factor in Factor V gene is FV H1299R. 4,9 In a study in northern part of India FVL and prothrombin G20210A mutations were observed with an allele frequency 2%, 12 therefore, in the present study the prevalence of FVL (G1691A) mutation an important genetic cause of thromboembolism was studied in subjects of Rayalaseema region of India.…”

Prediction of individual genetic risk of complex disease

“…The central problem of selecting a set of predictors when the number of measured variables is very large and greater than the number of observed phenotypic values is a common statistical challenge -the large p small n paradigm -arising in a wide range of diverse fields (including gene expression data, for example [19]) and a number of these methods have been applied to SNP selection [20][21][22][23][24][25]26 ]. Just as the individual SNP associations discovered in GWAS need to be validated in independent datasets, prediction equations estimated from one dataset need to be validated in other independent data.…”