272 1.9999

Journal of Thrombosis and Haemostasis, 1: 1398–1402 Cardiovascular pharmacogenetics in the SNP era V . M O O S E R , D . M . W A T E R W O R T H , T . I S E N H O U R and L . M I D D L E T O NGenetics Research, GlaxoSmithKline, King of Prussia, Pennsylvania, USA To cite this article: Mooser V, Waterworth DM, Isenhour T, Middleton L. Cardiovascular pharmacogenetics in the SNP era. J Thromb Haemost across several haplotype blocks, thus potentially reducing the Summary. In the past pharmacological agents have contribu- number of SNPs for future whole-genome scans. Finally, a ted to a significant reduction in age-adjusted incidence of modest number of cases and controls appears to be sufficient to cardiovascular events. However, not all patients treated with detect genetic associations. There is little doubt that this type of these agents respond favorably, and some individuals may approach will have an impact on the way cardiovascular drugs develop side-effects. With aging of the population and the will be developed and prescribed in the future.
growing prevalence of cardiovascular risk factors worldwide,it is expected that the demand for cardiovascular drugs will Keywords: cardiovascular disease, genes, pharmacogenetics, increase in the future. Accordingly, there is a growing need to identify the ‘good’ responders as well as the persons at risk fordeveloping adverse events. Evidence is accumulating to indi- ‘The right drug for the right patient’, a new cate that responses to drugs are at least partly under genetic control. As such, pharmacogenetics – the study of variability indrug responses attributed to hereditary factors in different Age-adjusted cardiovascular mortality has markedly decreased populations – may significantly assist in providing answers in Western countries over the last few decades [1]. There is toward meeting this challenge. Pharmacogenetics mostly relies substantial evidence to indicate that this success is partly on associations between a specific genetic marker like single accounted for by the development of thrombolytic therapies nucleotide polymorphisms (SNPs), either alone or arranged in a and the prescription of antiplatelet and other agents in the specific linear order on a certain chromosomal region (haplo- secondary prevention of cardiovascular diseases, as well as types), and a particular response to drugs. Numerous associa- by a wider use of more effective pharmacological interventions tions have been reported between selected genotypes and to treat hypertension and dyslipidemia [2]. Despite these major specific responses to cardiovascular drugs. Recently, for in- accomplishments, cardiovascular diseases remain the major stance, associations have been reported between specific alleles cause of death in industrialized countries. This is due in part of the apoE gene and the lipid-lowering response to statins, or to the facts that cardiovascular risk factors remain highly the lipid-elevating effect of isotretinoin. Thus far, these types of prevalent, underdiagnosed and/or insufficiently treated [3]; studies have been mostly limited to a priori selected candidate not all patients respond equally well to pharmacological inter- genes due to restricted genotyping and analytical capacities.
ventions; and the population is aging. Moreover, there is ample Thanks to the large number of SNPs now available in the public evidence to indicate that, with the growing epidemics of obesity domain through the SNP Consortium and the newly developed [4,5] and Type 2 diabetes [6] progressively affecting Western technologies (high throughput genotyping, bioinformatics soft- countries, and the increasing prevalence of cardiovascular risk ware), it is now possible to interrogate more than 200 000 SNPs factors in developing countries [7,8], the incidence of cardio- distributed over the entire human genome. One pharmacoge- vascular diseases and the demand for safe and effective drugs to netic study using this approach has been launched by Glaxo prevent/treat these diseases will steadily increase worldwide.
SmithKline to identify the approximately 4% of patients who Accordingly, there is a growing need for ways to better identify are predisposed to developing a hypersensitivity reaction to people who have the highest chance to benefit from pharma- abacavir, an anti-HIV agent. Data collected thus far on the HLA cological interventions, and those who have the lowest risk of locus on chromosome 6 indicate that this approach is feasible.
developing side-effects when exposed to cardiovascular drugs.
Extended linkage disequilibrium can be detected readily, even To date, only a fraction of people treated with a particular drug fully benefit from such interventions. This is particularlyapparent in the case of antihypertensive therapies. It is esti- Correspondence: Dr Vincent Mooser, Director Medical Genetics (Cardio- vascular), Genetics Research, GlaxoSmithKline, 709 Swedeland Road, King À50% of patients who receive one type of therapy (for instance an ACE inhibitor or a calcium channel Tel.: þ1 610 270 7732; e-mail: vincent.2.mooser@gsk.com blocker) will have their blood pressure controlled by this # 2003 International Society on Thrombosis and Haemostasis treatment [9]. The remaining 50À75% will have received the these two disciplines have their own characteristics. Tradition- treatment, may experience side-effects, and yet their blood ally, the goal of disease genetics is to identify genetic variants pressure will not be normalized. This lack of consistency associated with a particular susceptibility to developing a [10] in terms of efficacy not only affects the individual patients, disease. The genetic contribution to diseases and the way such but also drug-producing companies. Indeed, many programmes diseases are inherited can be estimated from twin and family- have been terminated because the average response to one based studies. Moreover, because genetically determined dis- particular drug has been insufficient, yet a substantial propor- eases can be evaluated in multiple members within families, it is tion of patients may have adequately responded to this inter- possible to perform family-based molecular genetic studies, vention. Similarly, effective drugs have been removed from the such as linkage analyses. Linkage studies have been shown to be market because a fraction of patients have developed intolerable very powerful in identifying the molecular basis of Mendelian side-effects. It is obvious that a better way to identify the people disorders (for instance Liddle’s syndrome, a rare form of salt- who will adequately respond to the drug, and those who are sensitive hypertension due to gain-of-function mutation within prone to develop side-effects, would have a major impact on the the epithelial sodium transporter in the proximal tubule of the development and prescription of new cardiovascular drugs.
kidneys [18]). Most common diseases, however, are thought to Evidence is accumulating to indicate that pharmacogenetics be polygenic and multifactorial, due to interactions of environ- may significantly assist in meeting the challenge of ‘the right mental factors and a particular genetic make-up. As an example, drug for the right patient’ [11,12].
it has recently been demonstrated that smoking is very prevalentin early onset (usually considered highly genetically deter-mined) forms of coronary artery disease [19]. This observation further illustrates the need to include the environment as a factor The way individuals respond to a particular drug – in terms of in genetic studies on complex conditions.
both efficacy and side-effects – depends on a variety of para- Pharmacogenetics can rarely rely on family-based studies, meters, including compliance, bioavailability, drug–drug inter- because usually only one family member is treated with a actions, catabolization of the drug and its metabolites, the particular drug. Accordingly, the heritability and the mode of molecular mechanism responsible for the disease for which inheritance of such responses are hard to predict, and family- patients are treated, and what our ignorance prompts us to based molecular genetic studies would be difficult to perform.
designate as ‘idiosyncratic reactions’. Because a variety of At this stage pharmacogenetics is an experimental science that genes encoding enzymes, transporters or receptors involved mostly relies on association studies, i.e. studies comparing the in drug absorption, metabolism, excretion and mode of action distribution of sequence variants between cases who develop a are polymorphic (with some of these polymorphisms being particular response to the drug and appropriate controls who do functionally active), a genetic predisposition is likely to account not develop this type of response. The chance is that, in contrast for part of the interindividual variability in response to drugs to disease genetics, the environmental factor is limited (i.e. to [13,14]. The example of slow acetylators who carry a particular the drug), simplifying the analysis to some degree.
sequence variant within the N-acetyl-transferase 2 gene and A recent study illustrates how pharmacogenetics may assist thus poorly catabolize isoniazid or procainamide, illustrates in exploring the genetic basis of a complex disease, in this how one single gene variant can contribute to the occurrence of particular case hypertriglyceridemia. In this study, the authors severe side-effects, in this particular case hypersensitivity reac- postulated that hypertriglyceridemia, which occasionally ac- tions [15]. Another recent example of a particular response to companies the administration of isotretinoin (a vitamin A drugs that is mostly dependent on one single gene is provided by derivative used to treat acne), occurs preferentially in indivi- hyperbilirubinemia during administration of Tranilast. This duals who have a predisposition to lipid disorders, and that side-effect preferentially develops in carriers of one particular elucidation of the genetic basis for this side-effect may assist in variant of the UDG-glucuronosyltransferase 1 gene, the gene our understanding of hyperlipidemia. The data showed that responsible for Gilbert syndrome [16].
those individuals have a higher risk of developing metabolic It is anticipated, however, that atypical drug responses are syndrome (an aggregation of lipid disorders, hypertension and rarely due to one gene only. For instance, it is generally accept- glucose intolerance) than individuals for whom plasma trigly- ed that African-Americans respond better to diuretics (and less ceride levels remained unchanged during this therapy. As a well to ACE inhibitors [17]) than Caucasians, due to a higher proof of concept, it was shown that the lipid response to iso- susceptibility to salt retention (and thus a lesser activated renin- tretinoin was closely associated with the apoE E2 allele (which angiotensin system), and that this susceptibility does not seem is associated with Fredrickson Type III hyperlipoproteinemia) to be dependent on one single gene. Elucidation of the genetic and apoE E4 allele (which is associated with higher lipid levels basis for this particular response to antihypertensive agents may in the plasma than the wild-type apoE E3 allele) [20].
not only allow us to accurately predict who will most likely benefitfrom one particular type of drug, it would also significantly Pharmacogenetics: the candidate–gene association studies contribute to our understanding of the hypertensive disease.
This latter example illustrates how discoveries in pharmaco- Thus far, restricted genotyping and analytical capacities have genetics may impact on disease genetics, and vice versa, even if limited pharmacogenetics to association studies of a priori # 2003 International Society on Thrombosis and Haemostasis Table 1 Pharmacogenetic studies: candidate-gene vs. whole-genome SNP association studies selected candidate genes. In this approach, genes to be tested are identify genes of susceptibility to a particular drug response usually selected based on a previous understanding of the way [16] (Table 1). One particular advantage of such an approach is drugs are metabolized, or based on the biological pathway that that, beyond finding tests to better predict the response to the drug, this unbiased approach may reveal totally unexpected A variety of conditions need to be met simultaneously for genetic associations. As such, this type of pharmacogenetic genetic associations to be detected, both for medical genetics approach may yield major benefits to disease genetics, as and pharmacogenetics. The example of apoE-associated lipid response to isotretinoin may help illustrate this point. For such The concept of whole-genome-based association studies is an association to be detected, genetic variants obviously need to relatively simple and takes advantage of the accumulating be identified and accurately analyzed (using methods like knowledge on SNPs. SNPs represent changes in nucleotides restriction fragment length polymorphisms, allele-specific am- that are present in a substantial proportion of the population.
plification or direct sequencing). These variants should by Most SNPs are bi-allelic, making binary technologies applic- themselves be functionally active (for instance by modifying able to identify and analyze them. It is estimated that there is the affinity of apoE to the LDL-receptor), or should be in one SNP on average for every 1000 base pairs (hence the linkage disequilibrium (see below) with other variants located concept that humans share 99.9% of the genome in common), in exons, in intron-exon junctions or regulatory sequences that and that there are approximately 3 million SNPs in the human are functionally active. In addition, the functional impact of genome (which contains $3 billion base pairs). Accordingly, these variants should not be fully compensated by other me- the basic idea of a whole-genome SNP association study is to chanisms (for instance by upregulation of the LDL-receptor) compare the frequency of these variants between cases and that would abolish the phenotypic expression of this particular controls. To be successful, this type of approach relies on seve- genotype. Moreover, the distribution of the variant should be ral factors: the availability of SNPs in sufficient numbers to sufficiently different in cases and controls, and the number of cover the entire genome; very powerful and accurate genotyp- cases and controls examined should be large enough (usually ing capabilities to examine large collections of SNPs; appro- several hundred individuals) for associations to be detected.
priate analytical approaches to detect significant associations; Finally, appropriate environmental factors may need to be evalu- and the availability of genomic DNA from large cohorts of well- ated and included in the analysis to reveal genetic associations.
phenotyped cases and controls. Moreover, because analysis of 3 A large number of associations between predefined candi- million SNPs for each case and control is not presently feasible, date-genes and specific responses to drugs have been described one must rely on the fact that several SNPs will provide the so far [for reviews, see, among others refs 14, 15, 21 and 22]. In same information as many SNPs in a particular region, due to particular, associations have been reported between sequence the phenomenon of linkage disequilibrium (Fig. 1).
variants within genes encoding metabolizing enzymes (like The SNP Consortium (http://www.snp.cshl.org/), a joint CYP2D6) and increased response to warfarin (and subsequent academia–industry initiative, was commissioned to identify risk of bleeding) or higher incidence of side-effects when and release sequences of SNPs. The success of this initiative exposed to b-blockers. Similarly, severe arrhythmias have been has been tremendous, in the sense that the Consortium has re- associated with sequence variants within genes encoding po- cently made available in the public domain more than 2 million tassium channels and exposure to antiarrhythmics. In the same SNP sequences distributed over the entire human genome (for way, the lipid-lowering effect of statins has been associated with examples see http://www.ncbi.nlm.nih.gov/). The major ques- specific alleles of the apoE or the CETP genes.
tions that now arise are (i) how to genotype these SNPs at areasonable cost, (ii) how many of these SNPs are needed toprovide adequate sensitivity and specificity for pharmacoge- Pharmacogenetics: moving to whole-genome SNP netic studies, and (iii) what is the required sample size.
At present, the only reliable way to answer these questions is Rapid technological improvements in high-throughput geno- to perform the experiments and generate experimental data.
typing, and developments in bioinformatics are now opening GlaxoSmithKline has recently launched such an initiative. The the way for an unprejudiced exploration of the entire genome to goal of this project is to identify genetic variants that are # 2003 International Society on Thrombosis and Haemostasis Fig. 1. Schematic representation of the linkagedisequilibrium block structure of genomic DNA.
Haplotype blocks and recombination hotspots areshown in green and red, respectively.
Identification of recombination hotspots is basedon observed recombinations in the commonhaplotypes, shown as gaps or crossovers in thehaplotypes. Once this structure is known, SNPsthat contain all the mapping information can bechosen. Using this type of strategy, the numbersof informative SNPs is expected to be reduced to3–5 per haplotype block.
associated with hypersensitivity reaction to abacavir, an anti- the genome is required to establish the degree of generality of HIV agent [23]. Data accumulated thus far on the HLA locus on chromosome 6 are very encouraging. They indicate that such an Whatever the mechanism, the observation of haplotype blocks experiment is feasible. Moreover, large regions of linkage has clear implications for linkage disequilibrium mapping, be it disequilibrium have been detected, forming haplotype blocks.
for whole SNP genome scans or for specific gene regions (Fig. 1).
These data are important, as they suggest that the number of One advantage is that fewer SNPs will be required to map SNPs to be examined can be reduced in other similar future associations, as a relatively smaller number will be required to initiatives (see below), and that cluster-analysis algorithms may provide sufficient representative information within the blocks.
be developed and standardized to generate individual SNP Furthermore we will know exactly where a greater saturation of profiles (SNP Printssm) that define genetic-susceptibility re- markers is required, in the regions of low linkage disequilibrium.
sponses to drugs. Finally, the data indicate that a modest number However, the presence of large haplotype blocks will also limit of cases and controls may be sufficient to achieve adequate the resolution of association methods to fine-map a susceptibility sensitivity and specificity [23,24].
gene, if it falls within one of these blocks, as additional markerswill not necessarily provide any more information.
Determining these haplotype blocks may therefore be pivotal Linkage disequilibrium and haplotype blocks: potential in the success of whole-genome association studies, and to this end a major worldwide effort is underway to create a haplotype Haplotypes are ancestral segments of chromosomes that have map of the human genome, which is expected to be completed been inherited as a unit throughout the generations with little in 2–3 years. The question then arises, how different are the genetic shuffling or mutation. They can be directly observed by patterns of linkage disequilibrium among different ethnic typing individuals within families for genetic markers and groups, and will it be necessary to have separate maps for each following the coinheritance of alleles from neighbouring mar- group? A recent study found that Caucasian and Asian haplo- kers through the generations. An alternative way to determine types are very similar, but haplotypes of African origin are quite haplotypes is to perform allele-specific sequencing, a technol- different [27]. The authors estimate that approximately half the ogy that is just starting to be utilized for that very purpose [25].
human genome exists in blocks of 44 kb or larger in Caucasian Alleles that occur together in this fashion are said to show allelic and Asian samples, and 22 kb or larger in African and African- association, and ‘linkage disequilibrium’ is the extent of this co- American individuals. Within each block 3–5 haplotypes typi- cally capture about 90% of all chromosomes in each population.
It was originally thought that linkage disequilibrium would They propose that to perform a fully powered association study mostly reflect genetic distance and would decay in a fairly linear will require as many as 300 000 SNPs in non-Africans and manner over increasing distances. However, closer examination 1 million SNPs in Africans. Theoretically, this would suggest of specific chromosomal regions has revealed instead irregular that populations who have undergone population bottlenecks linkage disequilibrium patterns that are mostly position depen- and hence have more extensive linkage disequilibrium would be dent. These patterns are composed of large stretches of DNA more useful for initial localizations, and that populations with ($5–100 kb) where recombinations are not observed and link- shorter range linkage disequilibrium would be more useful for age disequilibrium is high (haplotype blocks), and short inter- fine mapping. At this stage, it is still debated whether the vening regions ($1–5 kb) where linkage disequilibrium is low.
optimal strategy will be that simple.
Jeffreys et al. showed that regions of linkage disequilibrium A haplotype map of the human genome may therefore be breakdown in the class II region of the major histocompatibility critical in streamlining the process involved in association complex correspond precisely with meiotic recombination hot- mapping and may provide much information on the structures spots observed by typing sperm [26], suggesting a possible and histories of human populations. However, assuming that the explanation for the position-dependent nature of linkage dis- ‘common variant–common disease’ theory is correct, it is equilibrium breakdown. However, replication in other areas of possible that rare variants will easily be missed with this # 2003 International Society on Thrombosis and Haemostasis strategy and more than one mutation in one disease gene will changing demography and disease prevalence in the US. Diabetes Care split the association signal, rendering it undetectable. Thus, 7 Yusuf S, Reddy S, Ounpuu S, Anand S. Global burden of cardiovascular there is a great deal of speculation and projection at the current diseases: Part II. Variations in cardiovascular disease by specific ethnic time, which will only be resolved by the elucidation of some groups and geographic regions and prevention strategies. Circulation common disease or drug-response genes that will allow proof of 8 Reddy KS. Cardiovascular diseases in the developing countries: dimen- sions, determinants, dynamics and directions for public health action.
Public Health Nutr 2002; 5: 231–7.
Cardiovascular pharmacogenetics: hopes, hurdles and 9 Sever P. The heterogeneity of hypertension: why doesn’t every patient respond to every antihypertensive drug? J Hum Hypertens 1995; 9(Suppl. 2): S33–S36.
The data accumulated thus far are very encouraging, and 10 Waeber B, Brunner HR. The multifactorial nature of hypertension: the suggest that the concept of ‘the right drug for the right patient’ greatest challenge for its treatment? J Hypertens Suppl 2001; 19 (Suppl.
3): S9–16.
is becoming a closer reality. There are still major hurdles to 11 Marshall A. Getting the right drug into the right patient. Nat Biotechnol overcome. Large clinical trials need to be performed (in various ethnic groups) to consolidate the validity of this approach; such 12 Roses AD. Pharmacogenetics. Hum Mol Genet 2001; 10: 2261–7.
initiatives are underway in industry and academia (ex GenHat) 13 Evans WE, Relling MV. Pharmacogenomics: translating functional [21], and the results of these experiments should be available genomics into rational therapeutics. Science 1999; 286: 487–91.
14 Roden DM, George AL Jr. The genetic basis of variability in drug within the next few years. Next, additional technological responses. Nat Rev Drug Discov 2002; 1: 37–44.
improvements are necessary to bring the cost of genotyping 15 Meyer UA. Pharmacogenetics and adverse drug reactions. Lancet 2000; down to a level that allows pharmacogenetics to become economically attractive; here again, DNA chip technologies 16 Roses AD. Genome-based pharmacogenetics and the pharmaceutical are presently being developed that should generate tests that are industry. Nat Rev Drug Discov 2002; 1: 541–9.
17 Gibbs CR, Beevers DG, Lip GY. The management of hypertensive affordable. In addition, regulators, payers, physicians and disease in black patients. QJM 1999; 92: 187–92.
patients should agree with such projects. In this respect, phar- 18 Rossier BC, Pradervand S, Schild L, Hummler E. Epithelial sodium macogenetics may raise ethical issues that are similar to the channel and the control of sodium balance: interaction between genetic ones raised by disease genetics. A major effort will be required and environmental factors. Annu Rev Physiol 2002; 64: 877–97.
from the patients, their physicians, academia and industry to 19 Jomini V, Oppliger-Pasquali S, Wietlisbach V, Rodondi N, Jotterand V, Paccaud F, Darioli R, Nicod P, Mooser V. Contribution of major overcome these hurdles. In our opinion, the potential beneficial cardiovascular risk factors to familial premature coronary artery dis- impact of pharmacogenetics for the patients and society in ease: the GENECARD project. J Am Coll Cardiol 2002; 40: 676–84.
20 Rodondi N, Darioli R, Ramelet AA, Hohl D, Lenain V, Perdrix J, Wiet- lisbach V, Riesen WF, Walther T, Medinger L, Nicod P, Desvergne B,Mooser V. High risk for hyperlipidemia and the metabolic syndrome after an episode of hypertriglyceridemia during 13-cis retinoic acid therapy foracne: a pharmacogenetic study. Ann Intern Med 2002; 136: 582–9.
The authors are employees of GlaxoSmithKline. They thank 21 Arnett DK, Boerwinkle E, Davis BR, Eckfeldt J, Ford CE, Black H.
Allen Roses, Virginia Schmith, Wayne Anderson and Sanjay Pharmacogenetic approaches to hypertension therapy: design and rationale for the Genetics of Hypertension Associated Treatment(GenHAT) study. Pharmacogenomics J 2002; 2: 309–17.
22 Ordovas JM, Mooser V. The APOE locus and the pharmacogenetics of lipid response. Curr Opin Lipidol 2002; 13: 113–7.
23 Hetherington S, Hughes AR, Mosteller M, Shortino D, Baker KL, Spreen 1 Tunstall-Pedoe H, Vanuzzo D, Hobbs M, Mahonen M, Cepaitis Z, W, Lai E, Davies K, Handley A, Dow DJ, Fling ME, Stocum M, Bowman Kuulasmaa K, Keil U. Estimation of contribution of changes in coronary C, Thurmond LM, Roses AD. Genetic variations in HLA-B region and care to improving survival, event rates, and coronary heart disease hypersensitivity reactions to abacavir. Lancet 2002; 359: 1121–2.
mortality across the WHO MONICA Project populations. Lancet 2000; 24 Lai E, Bowman C, Bansal A, Hughes A, Mosteller M, Roses AD.
Medical applications of haplotype-based SNP maps: learning to walk 2 Kuulasmaa K, Tunstall-Pedoe H, Dobson A, Fortmann S, Sans S, Tolonen before we run. Nat Genet 2002; 32: 353.
H, Evans A, Ferrario M, Tuomilehto J. Estimation of contribution of 25 Patil N, Berno AJ, Hinds DA, Barrett WA, Doshi JM, Hacker CR, changes in classic risk factors to trends in coronary-event rates across Kautzer CR, Lee DH, Marjoribanks C, McDonough DP, Nguyen BT, the WHO MONICA Project populations. Lancet 2000; 355: 675–87.
Norris MC, Sheehan JB, Shen N, Stern D, Stokowski RP, Thomas DJ, 3 Wong MD, Shapiro MF, Boscardin WJ, Shapiro MF, Boscardin WJ, Trulson MO, Vyas KR, Frazer KA, Fodor SP, Cox DR. Blocks of limited Ettner SL. Contribution of major diseases to disparities in mortality.
haplotype diversity revealed by high-resolution scanning of human chromosome 21. Science 2001; 294: 1719–23.
4 Freedman DS, Khan LK, Serdula MK, Galuska DA, Dietz WH. Trends 26 Jeffreys AJ, Kauppi L, Neumann R. Intensely punctate meiotic recom- and correlates of class 3 obesity in the United States from 1990 through bination in the class II region of the major histocompatibility complex.
5 Flegal KM, Carroll MD, Ogden CL, Carroll MD, Ogden CL, Johnson 27 Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, CL. Prevalence and trends in obesity among US adults, 1999–2000.
Higgins J, DeFelice M, Lochner A, Faggart M, Liu-Cordero SN, Rotimi C, Adeyemo A, Cooper R, Ward R, Lander ES, Daly MJ, Altshuler D.
6 Boyle JP, Honeycutt AA, Narayan KM, Hoerger TJ, Geiss LS, Chen H, The structure of haplotype blocks in the human genome. Science 2002; Thompson TJ. Projection of diabetes burden through 2050: impact of # 2003 International Society on Thrombosis and Haemostasis

Source: http://www.biotec.com.uy/centro/cardiologia/272.pdf