The number of attendees of this subcommittee meeting was approximately 200.

Dr. L. Iacoviello presented the planned activities of the SubCommittee for the next year. Major issues were:

• Preparation of Guidelines for planning studies of genetic epidemiology
• Information on genetic polymorphisms available in "candidate" genes for haemostasis and thrombosis, on their functionality, on the protocols for their determination and on the validation of genotyping
• Standardisation of genetic polymorphism nomenclature
• Registry of on-going studies on the association between genetic and biochemical haemostatic variables and cardiovascular disease
• Development of a WEB page on our Subcommittee, in connection with the ISTH Web-page (http://www.negrisud.it/ssc )
• Systematic meta-analysis of the studies on predictive genetic and biochemical haemostatic variables in cardiovascular disease
• Preparation of Guidelines for blood management for homocysteine evaluation.

Ongoing meta-analysis on biochemical predictive Haemostatic Variables in Cardiovascular Disease.

Dr. G Lowe underlined the necessity for meta-analysis to include a minimum of 1000 CHD cases in prospective studies, to consider the effect of confounders and the nature of samples to be analysed for antigen levels of haemostatic variables. In the framework of the British Regional Heart Study, they found a strong correlation between D-dimer, von Willebrand factor (vWF), and t-PA antigen levels measured in paired citrated plasma and serum samples.  However, carefullness is still necessary in the interpretation of these results, until validation by further studies will be available.  In contrast, their self-association over 5 years was lower and needs evaluation in further studies. Results of meta-analyses for fibrin D-dimer, vWF, and t-PA as markers of coronary risk were presented. They all included more than 1,000 CHD cases and showed a significant association with the risk of CHD. Adjustment for confounders did not modify the association for D-Dimer and vWF, while the association with t-PA decreased. A Fibrinogen Trialist Collaboration has been started to perform meta-analyses including more than 10,000 cases.

Methodological approaches for association population-based studies in genetics. Dr. A. Di Castelnuovo presented the use of association population-based studies in genetics. Association studies are based on linkage disequilibrium (LD), a population-based concept identified by noting than certain alleles of two genes occur together in the population more often than by chance.

A major limitation of the approach is confounding by population stratification that can occur in ethnically mixed populations. A way to measure and correct for stratification has been recently proposed, by genotyping a moderate number of genetic markers unlinked with the disease and using the average of association across the unlinked markers as a direct measure of stratification.

Correct sample size estimation is of fundamental importance, especially in testing for gene-environment interactions. Methods to calculate sample size in the presence of interactions have been reviewed. Interaction must be supported by statistical tests, and not solely suggested by data. Synergy index (S) is a useful measure of interaction in case-control studies.

Relevant references to the topic are:

1. MJ Khoury et al. Fundamentals of Genetic Epidemiology, Oxford University Press, 1993
2. KJ Rothman et al. Modern Epidemiology, Lippincot-Raven, 1998
3. J Ott et al. Am. J. Hum. Genet. 67:289-294;2000
4. JK Pritchard et al. Am. J. Hum. Genet. 65:220-228,1999
5. ER Reich et al. Genetic Epidemiology 20:4-16,2001
6. WD Dupont et al. Controlled Clin. Trials 11:116-128,1990
7. SJ Hwang et al. Am. J. Epidem. 140:1029-1037,1994
8. I Foppa et al. Am. J. Epidem. 146:596-604,1997
9. M Lundberg et al. Epidemiology 7:655-656,1996
10. A Di Castelnuovo et al. Statistica 2001, in press
11. J Hoh et al. PNAS; 97:9615-9617,2000

The Utility of Family Studies for Understanding the Genetics of Thrombosis. Dr. J. Blangero provided an overview of a method for linkage analysis in extended pedigrees known as Variance Component Method (VCM). VCM is a powerful strategy to map genes related with quantitative trait loci (QTL) involved in pathogenesis of complex disease. VCM used data from large pedigrees containing more than 20 individuals in more than 3 generations. The method is based on a decomposition of the variability of the QTL in genetic and environmental causes. He showed that for a disease with large prevalence (35%), the VCM method is more powerful than other linkage-based strategies.

The author presented also some results of the GAIT project, a project aimed at identification of QTL involved in thrombosis, that made use of the VCM strategy in analysing data from Spanish families.

Twin studies to evaluate the genetics of haemostatic variables. Dr. P Grant presented examples on how to define the genetic components of single factors related to a syndrome, such as the insulin resistance syndrome. The inheritability of these factors has been suggested by studies on offsprings of probands with diabetes, which showed increased levels of factors related to the metabolic syndrome (including haemostatic factors) as compared with those of the normal population. He presented results on 501 twin pairs that showed a high inheritability of F VII, F XII, F VIII, fibrinogen, vWF, PAI-1, F XIII levels. Moreover, also levels of activation markers of coagulation and fibrinolysis, as such as D-Dimer, prothrombin fragments, thrombin-antithrombin complex, showed a significant inheritability. These results were confirmed by the Leeds Family Study on 537 subjects from 89 families. However, when the most studied polymorphisms in haemostatic genes were tested in these models, they accounted only for a very small percentage of the total inheritability.

The CANVAS project and the candidate gene approach . Dr. F Cambien presented the "candidate gene approach" to understand the genetic component of ischaemic cardiovascular disease. The strategy consists in identifying all the possible candidate genes and all the possible relevant polymorphisms in functional parts of the genes. 300-600,000 single nucleotide polymorphisms in regulatory portions of the genes could be expected. It is necessary to genotype all the polymorphisms of a certain gene, since the functional one could be not in linkage disequilibrium with the others (the example of Apo A1 has been given). Dr Cambien then presented the GENE-CANVAS project. The project started in 1998 with the aim of screening for polymorphisms in candidate genes for ischaemic cardiovascular disease, providing assay conditions, allele and genotype frequencies and linkage disequilibrium of identified polymorphisms, and testing associations with ischaemic cardiovascular disease. At 2001, 29 studies, 102 gene and 454 polymorphisms have been analysed and are available at the Web site http://genecanvas.idf.inserm.fr .

The second session of the meeting was focused on differences in the association between levels of haemostatic and inflammatory variables and the risk of ischaemic vascular disease.

Homocysteine evaluation: problems and predictive value. Dr. M Cattaneo could not be present at the meeting.

Inflammation variables in the prediction of cardiovascular disease. Dr. Kluft reported on the ‘Leiden Meeting on Inflammation and Cardiovascular Disease’. Information is available at the web site www.haemost.nl and published in the Italian Heart Journal 2001;2:155-199. Three inflammatory markers have been discussed: CRP, SAA and SFLA2. CRP seems to perform better, since it presents very limited preanalytical problems, has a reliable assay, is widely available and can be considered as both a marker and a functional parameter. It is a systemic marker, but is also localised at the site of atherosclerotic plaques.

The intraindividual variation of CRP levels is about 30%. An upper quartile of PCR level >3mg/L has been established as a cut-off for high CAD risk.

Finally, treatment options have been discussed. The DALI study showed that atorvastatin treatment is able to strongly reduce CRP levels at the dose of 80 mg/day. The dose response for effects on CRP levels was not parallel with those of cholesterol lowering, suggesting that the reduction of CRP was not dependent on the effect of atorvastatin on cholesterol. Evaluation of cost/effect benefit of statins in primary prevention showed an acceptable cost/benefit for people over 45 years.

Inflammatory markers and sexual hormones.Dr. Cushman presented an overview of the studies on the effect of hormone replacement therapy (HRT) on CRP levels. HRT doubles the levels of CRP and this effect could explain the results of the two trials that showed an increased risk of CAD after HRT. On the contrary, tamoxifen therapy reduced CRP levels. In the HERS trial, the increased risk of CAD in the first months of follow-up was associated to an increase in the levels of CRP.

The VITA project: haemostatic variables and prediction of deep vein thrombosis and arteriosclerosis progression.Dr. Tosetto presented the VITA study (1993 to 1997), a cross-sectional randomised study of 15,109 subjects in the area of Vicenza, Italy. The principal aim of the study was to define the prevalence and the impact of haemostatic variables on the thrombotic risk. For each subject recruited into the study information on family structure, history of thrombosis and DNA samples are available. The main results of the study confirmed the association of APC resistance with the risk of VTE and did not support an association of prothrombin polymorphism with VTE.

Secondary aims were the prospective evaluation of the risk and a cross-sectional evaluation of the impact of haemostatic variables on arteriosclerosis (evaluated as IMT of the common carotid artery) in a subsample of 2,000 subjects. The results showed an association between high fibrinogen levels and the risk of both high IMT and incident myocardial infarction. The VITA study is also part of a European collaborative study on the genetics of thrombosis, the GENERALE project.

The final discussion pointed out that the different methodological approaches to genetic studies are complementary; that genetic studies may be relevant in the understanding of the physiopathology of haemostasis and in defining the genetic components of haemostatic variables levels, rather than the risk of CVD; and that it is important to determine genetic variables together with the circulating levels of the corresponding proteins.