Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Sep 1.
Published in final edited form as: J Thromb Haemost. 2016 Aug 6;14(9):1798–1802. doi: 10.1111/jth.13392

Is there still room for additional common susceptibility alleles for venous thromboembolism?

D-A Trégouët *,†, A Delluc ‡, A Roche §, C Derbois , R Olaso , M Germain *,†, M de Andrade **, W Tang ††, D I Chasman ‡‡, A van Hylckama Vlieg §§, P Reitsma ¶¶, C Kabrhel ***, N Smith †††,‡‡‡,§§§, P-E Morange ¶¶¶,****
PMCID: PMC5152582  NIHMSID: NIHMS833031  PMID: 27326655

Abstract

Background

Through a meta-analysis of twelve Genome-wide Association Studies, the INVENT consortium identified two novel susceptibility loci for venous thromboembolism (VTE). This project has also generated other candidates that need to be further replicated.

Objectives

To assess the association with VTE of common SNPs that demonstrated strong statistical, but not genome-wide, significance in the INVENT cohorts.

Patients/Methods

Eleven SNPs were genotyped and tested for association with VTE in three case-control studies totaling 3,019 patients and 2,605 healthy individuals.

Results and Conclusions

None of the tested SNPs showed evidence for association with VTE. Different strategies are needed to decipher the whole spectrum of common and rare genetic variations associated with VTE risk.

Keywords: genome-wide association studies, genetics, polymorphisms, risk factor, venous thromboembolism

Introduction

In 2015, the International Network against VENous Thrombosis (INVENT) consortium reported the results of a meta-analysis of twelve Genome-Wide Association studies (GWAS) on venous thromboembolism (VTE) [1]. In the discovery phase of this study which was composed of 7,507 VTE cases and 52,632 controls nine loci reached the genome-wide threshold of 5×10−8 for declaring statistical significance. Three of these loci, SLC44A2, TSPAN15 and ZFPM2, had not previously been reported to associate with VTE risk. The first two replicated in three independent French case-control studies, MARTHA12, EDITH and FARIVE [1]. The association of SLC44A2 and TSPAN15 loci with VTE was also confirmed in a recent GWAS analysis of 6,135 self-reported VTE patients belonging to the 23andMe cohorts [2]. A meta-analysis conditioning on the well-established VTE-associated single nucleotide polymorphisms (SNPs) was also performed in the INVENT discovery cohorts but failed to identify any additional genome-wide significant associations [1]. However, at a lower statistical threshold of 1×10−5, 285 candidate SNP were identified in this conditional, 12-cohorts meta-analysis. After excluding associations at known VTE loci, and those at the ZFPM2 locus that did not replicate previously, 125 SNPs mapping to 11 loci remained. We reasoned that there might be true positive VTE-associations among these 11 loci and decided to further genotype the lead SNPs in the three French case-control studies that served as replication for the INVENT meta-analysis.

Material and Methods

Study populations

Three French case-control studies for VTE were used in this work: EDITH, FARIVE and MARTHA12. These are exactly those that had been employed for replicating the main findings of the INVENT meta-GWAS [1].

EDITH

The EDITH study is a case–control study that was designed to test interactions between genetic and environmental risk factors of VTE [3]. Between May 2000 and December 2009, all consecutive unselected in- and outpatients seen at Brest (West of France) University Hospital for symptomatic VTE (deep vein thrombosis (DVT) and/or pulmonary embolism (PE)) were asked to participate in the study. Diagnosis of DVT was confirmed by the absence of full compressibility of a proximal or distal vein of the deep lower limb on compression ultrasonography (CUS). PE Diagnosis was done by a segmental or larger artery filling defect on chest computed tomography, or the combination of high pre-test clinical probability of PE with high probability ventilation–perfusion lung scan according to the PIOPED criteria, or proximal DVT on CUS in a patient with suspected PE. Controls were selected from the roster of patients hospitalized in the same ward in the following 12 months after the case’s event date. Controls were not included if they had a past history of VTE or if they were receiving long-term anticoagulant therapy. DNA were available for 1,141 patients and 1,152 controls.

FARIVE

The FARIVE study is a multicenter case-control study of 607 patients with a first episode of proximal DVT and/or PE [4]. DVT was diagnosed by venography and CUS while PE was diagnosed by spiral computed tomography,high probability ventilation-perfusion lung scan and pulmonary angiography. Patients younger than 18 years, with previous VT event, that had a diagnosis of active cancer or a history of malignancy less than 5 years previously, or have a short life expectancy because of other causes, were excluded. The control group consists of age- and sex-matched individuals free of venous and arterial thrombotic disease. Potential control subjects with cancer, liver or kidney failure, or a history of venous and/or arterial thrombotic disease are ineligible. DNA were available for 698 cases and 672 controls.

MARTHA12

The MARTHA12 study is composed of an independent sample of 1,245 VTE patients and 801 French healthy individuals. Patients have been recruited between 2010 and 2012 at the Thrombophilia center of La Timone hospital (Marseille, France) as part of the MARTHA study [5]. All patients had a history of a first VTE event documented by venography, Doppler ultrasound, angiography and/or ventilation/perfusion lung scan. They were all free of any chronic conditions and free of any well characterized genetic risk factors including anti-thrombin, protein C or protein S deficiency, homozygosity for FV Leiden or FII 20210A, and lupus anticoagulant. The control group was composed of two subsamples of healthy individuals: one consists of 475 healthy subjects recruited from the Marseille area, the second of 326 healthy heterozygous carriers of the FV Leiden or FII 20210A mutations selected from the national health examination centers of the French Social Security in collaboration with the Hemostasis and Thrombosis Study Group. DNA were available for 1,180 cases and 781 controls.

Single nucleotide polymorphisms genotyping and association testing

Eleven SNPs were selected for genotyping in the EDITH, FARIVE and MARTHA12 samples. These SNPs were identified as the lead SNP (or best proxy for it) at the 11 loci that demonstrated significant association with VTE at p < 10−5 in the INVENT meta-GWAS after adjusting for the known- and replicated- VTE associated SNPs [1]. The lead SNPs were USP34 rs10177303, CACNA2D3 rs72624844, FABP6 rs72812220, AKR1C1 rs114155856, WWP2 rs151186597, CPAMD8 rs10421800, KRI1 rs6511703, PPP5C rs3795043, LILRP2 rs8107460, ADRA1D rs78265041 and ICOSLG rs11701174 (Table 1). Genotyping was performed by taqman and sequenom technologies at the Centre National de Génotypage (CNG, Evry, France).

Table 1.

Selected single nucleotide polymorphisms that demonstrated suggestive statistical evidence at p < 10−5 for association with VTE in the INVENT meta-analysis.

Discovery Meta-Analysis INVENT consortium
Ncases = 7,507 Ncontrols=52,632
Chr Locus Lead SNP Alleles MAF OR(1) P(2) P(3)
2 USP34 rs10177303 G/A 0.37 0.889 [0.845 – 0.934] p = 4.44 10−6 p = 0.65
3 CACNA2D3 rs72624844 A/G 0.04 1.410 [1.239 – 1.605] p = 2.00 10−7 p = 0.21
5 FABP6 rs72812220 G/A 0.05 1.327 [1.173 – 1.500] p = 1.28 10−6 p = 0.86
10 AKR1C1 rs114155856 C/G 0.02 1.489 [1.258 – 1.763] p = 3.75 10−6 p = 0.10
16 WWP2 rs151186597 T/C 0.06 0.736 [0.643 – 0.841] p = 6.95 10−6 p = 0.98
19 CPAMD8 rs10421800 G/A 0.05 1.315 [1.182 – 1.465] p = 5.70 10−7 p = 0.07
19 KRI1 rs6511703 T/C 0.23 0.866 [0.816 – 0.919] p = 2.13 10−6 p = 0.71
19 PPP5C rs3795043 G/A 0.24 0.869 [0.817 – 0.920] p = 2.83 10−6 p = 0.52
19 LILRP2 rs8107460 C/T 0.13 1.206 [1.116 1.302] p = 1.98 10−6 p = 0.47
20 ADRA1D rs78265041 C/T 0.03 1.567 [1.297 – 1.893] p = 3.33 10−6 p = 0.11
21 ICOSLG rs11701174 C/T 0.28 0.877 [0.829 – 0.929] p = 7.98 10−6 p = 0.34
Open in a new tab
(1)

Allelic Odds Ratio (OR) [95%Confidence Interval] associated with the minor allele.

(2)

Association p-value resulting from the meta-analysis of INVENT GWAS conditioning on known VTE associated SNPs

(3)

p-value of the test for homogeneity of association (Cochran’s Q statistic) across the twelve discovery cohorts of the INVENT consortium.

Hardy-Weinberg equilibrium was checked by means of a Chi-square test in case and control subjects from each study separately. Association of genotyped SNPs with VTE risk was assessed in each of the three studies separately by use of a logistic regression model adjusted for age and sex. After having checked for the homogeneity of the associations across these studies using the Cochran’s Q statistic, results were then meta-analyzed using a fixed-effects model based on the inverse-variance weighting method. Because the aim of the project is to validate the candidate associations generated in the INVENT cohorts, one-sided association tests were conducted.

Results and Discussion

The genotype distributions of all SNPs but one, the KRI1 rs8113381, were compatible with the Hardy-Weinberg equilibirum (HWE). The distribution of the KRI1 rs8113381 that served as a proxy for the rs6511703 significantly (p <10−3) deviating from HWE in all studies, both in cases and controls, this SNP was then excluded from the association analysis. The minimum per-study genotyping success rate over the 10 remaining SNPs was 0.976. Results of the association tests were detailed in Table 2. None of the tested SNPs was statistically associated with VTE in the combined samples, with the smallest p-value being p = 0.027 for the AKR1C rs115641535. Odds ratios for all SNPs were close to unity or in the opposite direction compared to the INVENT findings.

Table 2.

Association of selected candidate SNPs with VTE in the MARTHA12, FARIVE and EDITH replication studies

USP34
rs2694642(1) 		CACNA2D3
rs72624844		 AKR1C1
rs115641535(1) 		FABP6
rs72812220		 WWP2
rs80215330(1) 		CPAMD8
rs10421800		 PPP5C
rs3795043		 LILRP2
rs7249176(1) 		ADRA1D
rs78265041		 ICOSLG
rs4818892(1)
MARTHA12
 Controls (N = 781) 0.385 0.064 0.028 0.042 0.052 0.056 0.245 0.121 0.023 0.341
 Cases (N = 1180) 0.344 0.064 0.022 0.030 0.069 0.063 0.254 0.111 0.016 0.317
 Allelic OR [95%CI](2) 0.84 [0.74–0.96] 1.02 [0.79–1.32] 0.78 [0.52–1.17] 0.70 [0.49–1.00] 1.32 [1.00–1.73] 1.13 [0.86–1.48] 1.05 [0.90–1.22] 0.91 [0.75–1.12] 0.71 [0.44–1.13] 0.89 [0.77–1.02]
FARIVE
 Controls (N = 672) 0.326 0.063 0.034 0.047 0.079 0.065 0.243 0.121 0.017 0.321
 Cases (N = 698) 0.355 0.062 0.022 0.035 0.065 0.052 0.253 0.140 0.018 0.315
 Allelic OR [95%CI](2) 1.14 [0.97–1.33] 0.98 [0.72–1.33] 0.63 [0.39–1.01] 0.72 [0.49–1.07] 0.80 [0.60–1.07] 0.78 [0.57–1.08] 1.06 [0.89–1.26] 1.19 [0.96–1.50] 1.05 [0.59–1.88] 0.97 [0.82–1.14]
EDITH
 Controls (N = 1152) 0.392 0.056 0.026 0.048 0.056 0.056 0.249 0.111 0.019 0.289
 Cases (N = 1141) 0.384 0.047 0.022 0.060 0.054 0.057 0.239 0.101 0.021 0.294
 Allelic OR [95%CI](2) 0.97 [0.86–1.09] 0.82 [0.63–1.07] 0.85 [0.58–1.25] 1.29 [0.99–1.69] 0.96 [0.75–1.25] 1.01 [0.79–1.29] 0.95 [0.83–1.09] 0.89 [0.74–1.08] 0.75[0.25–2.19] 1.03 [0.90–1.17]
All
 Allelic OR [95%CI](3) 0.96[0.89–1.04] 0.94[0.79–1.09] 0.77[0.60–0.97] 0.95[0.79–1.15] 1.01[0.87–1.19] 0.98[0.84–1.15] 1.00[0.93–1.10] 0.91[0.80–1.04] 0.82[0.58–1.16] 0.96[0.89–1.04]
 P (4) 0.292 0.405 0.027 0.626 0.866 0.849 0.824 0.178 0.267 0.351
 P(5) 0.0165 0.493 0.609 0.007 0.047 0.227 0.497 0.866 0.570 0.314
Post-Hoc power(6) 67% 91% 83% 76% 94% 80% 71% 82% 98% 70%
Open in a new tab
(1)

rs2694642, rs115641535, rs80215330, rs7249176 and rs4818892 were used as proxies for rs10177303, rs114155856, rs151186597, rs8107460 and rs11701174, respectively. Data for the KRI1 rs8113381 that served as a proxy for the lead rs6511703 were not shown as its distribution did not follow Hardy-Weinberg equilibrium.

(2)

Allelic Odds Ratio (OR) [95%Confidence Interval] associated with the minor allele derived from a logistic regression analysis adjusted for age and sex.

(3)

Meta-analyzed adjusted allelic OR derived from a standard fixed-effect meta-analysis of the results observed in the three replication studies.

(4)

P-value of the combined allelic OR derived from a standard fixed-effect meta-analysis of the results observed in the three replication studies.

(5)

P-value of the test for homogeneity of associations (Cochran’s Q statistic) across the 3 studies.

(6)

Power of the combined replication studies to replicate, at the 0.05/10 significance level, the allelic effect observed in the INVENT discovery cohorts (one-sided test).

We also investigated whether these associations could be enhanced in specific subgroups of patients, and for this end patients were stratified according to the clinical manifestations of VTE (deep vein thrombosis or pulmonary embolism) or to the presence of strong genetic risk factors (F2 rs1799963 or F5 rs6025 mutations). Also in these analyses we did not observe evidence for preferential association of the tested SNPs in specific strata.

Despite an average power of ~80% to detect in the three case-control populations studied an odds ratio equal to or greater than the INVENT findings (Table 2), we were not able to replicate any of these associations. The lack of validation of the 11 SNPs does not preclude the existence of common variants with weaker effects or with similar effect size but in no linkage disequilibrium with SNPs assessed by imputation techniques, of common variants exercing haplotype or more complex effects, or of rare variants associated with much stronger genetics effects.

Our findings suggest that none of the 11 common SNPs we selected and that demonstrate suggestive statistical evidence at p < 10−5 (which corresponds to allelic odds ratio greater than 1.12) in the INVENT GWAS tag for VTE-associated SNPs. Additional statistical power and possibly denser genotyping are needed to identify new associations with incident VTE.

Essentials.

  • Genetic architecture of venous thromboembolism (VTE) remains to be fully disentangled.

  • 11 newly discovered candidate polymorphisms were genotyped in 3,019 VTE cases and 2,605 controls

  • None of the 11 polymorphism were significantly associated with VTE risk

  • Additional major efforts are needed to identify VTE-associated genetic variants

Acknowledgments

This work was conducted as part of the INVENT consortium whose contributors list is given in the Appendix. This study recieved grant support for NHLBI R01HL116854 and NHLBI 1R01HL116854. This work was partially supported by the GenMed LABEX (ANR-10-LABX-0013), the French Clinical Research Infrastructure Network on Venous Thrombo-Embolism (F-CRIN INNOVTE) and the ICAN Institute for Cardiometabolism and Nutrition (ANR-10-IAHU-05). We thank Cécile Dulary and Fernando Dos Santos for their technical support in genotyping MARTHA12, FARIVE and EDITH samples.

Footnotes

Addendum

A. Delluc, A. Roche, M. De Andrade, W. Tang, D. Chasman, A. van Hylckama Vlieg, P. Reitsma, C. Kabrhel, N. Smith, D-A. Trégouët and P-E. Morange participated to data collection and study design. C. Derbois and R. Olaso organized the genotyping analyses. D-A Trégouët and M. Germain performed statistical data analysis. D-A. Trégouët drafted the document that were further reviewed by A. Delluc, M. De Andrade, W. Tang, D. Chasman, P. Reitsma, C. Kabrhel, A van Hylckama Vlieg, N. Smith and P-E Morange.

References

  • 1.Germain M, Chasman DI, de Haan H, Tang W, Lindström S, Weng LC, de Andrade M, de Visser MC, Wiggins KL, Suchon P, Saut N, Smadja DM, Le Gal G, van Hylckama Vlieg A, Di Narzo A, Hao K, Nelson CP, Rocanin-Arjo A, Folkersen L, Monajemi R, Rose LM, et al. Meta-analysis of 65,734 individuals identifies TSPAN15 and SLC44A2 as two susceptibility loci for venous thromboembolism. Am J Hum Genet. 2015;96:532–42. doi: 10.1016/j.ajhg.2015.01.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Hinds DA, Buil A, Ziemek D, Martinez-Perez A, Malik R, Folkersen L, Germain M, Mälarstig A, Brown A, Soria JM, Dichgans M, Bing N, Franco-Cereceda A, Souto JC, Dermitzakis ET, Hamsten A, Worrall BB, Tung JY, METASTROKE Consortium, INVENT Consortium. Sabater-Lleal M. Genome-wide association analysis of self-reported events in 6135 individuals and 252 827 controls identifies 8 loci associated with thrombosis. Hum Mol Genet. 2016 Feb 9; doi: 10.1093/hmg/ddw037. pii: ddw037 [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Oger E, Lacut k, Le Gal G, Couturaud F, Guénet D, Abalain JH, Roguedas AM, Mottier D, EDITH COLLABORATIVE STUDY GROUP Hyperhomocysteinemia and low B vitamin levels are independently associated with venous thromboembolism: results from the EDITH study: a hospital-based case-control study. J Thromb Haemost. 2016;4:793–799. doi: 10.1111/j.1538-7836.2006.01856.x. [DOI] [PubMed] [Google Scholar]
  • 4.Trégouët DA, Heath S, Saut N, Biron-Andreani C, Schved JF, Pernod G, Galan P, Drouet L, Zelenika D, Juhan-Vague I, Alessi MC, Tiret L, Lathrop M, Emmerich J, Morange PE. Common susceptibility alleles are unlikely to contribute as strongly as the FV and ABO loci to VTE risk: results from a GWAS approach. Blood. 2009;113:5298–5303. doi: 10.1182/blood-2008-11-190389. [DOI] [PubMed] [Google Scholar]
  • 5.Oudot-Mellakh T, Cohen W, Germain M, Saut N, Kallel C, Zelenika D, Lathrop M, Trégouët DA, Morange PE. Genome wide association study for plasma levels of natural anticoagulant inhibitors and protein C anticoagulant pathway: the MARTHA project. Br J Haematol. 2012;157:230–239. doi: 10.1111/j.1365-2141.2011.09025.x. [DOI] [PubMed] [Google Scholar]

RESOURCES