Genetic Determinants of Cardiovascular Disease Risk in Familial Hypercholesterolemia
http://www.100md.com
动脉硬化血栓血管生物学 2005年第7期
From the Departments of Vascular Medicine (A.C.M.J., E.S.v.A.-C., J.C.D., J.J.P.K.) and Clinical Epidemiology and Biostatistics (M.W.T.T.), Academic Medical Center, University of Amsterdam, the Netherlands; and the Department of Human Genetics (S.C., M.R.F., J.L.), Roche Molecular Systems, Inc, Alameda, Calif.
Correspondence to John J.P. Kastelein, Department of Vascular Medicine, Academic Medical Center, Meibergdreef 9, P.O.Box 22700, room F4-159.2, 1100 DE Amsterdam, the Netherlands. E-mail e.vandongen@amc.uva.nl
Abstract
Objective— To investigate the contribution of polymorphisms in multiple candidate genes to cardiovascular disease (CVD) risk in a large cohort of patients with heterozygous familial hypercholesterolemia (FH).
Methods and Results— We genotyped 1940 FH patients for 65 polymorphisms in 36 candidate genes. During 91.451 person-years, 643 (33.1%) patients had at least 1 cardiovascular event. Multifactorial Cox survival analysis revealed that the G20210A polymorphism in the prothrombin gene was strongly associated with a significantly increased CVD risk (GA versus GG; P<0.001).
Conclusions— In a large cohort of FH patients, we found that the G20210A polymorphism in the prothrombin gene is strongly associated with CVD risk. Our results constitute a step forward in the unraveling of the hereditary propensity toward CVD in FH and might lead to better risk stratification and hence to more tailored therapy for CVD prevention.
We investigated the contribution of 65 polymorphisms in 36 candidate genes to CVD risk in FH patients and found that the G20210A polymorphism in the prothrombin gene was associated with CVD risk. Our results constitute a step forward in the unraveling of the hereditary propensity toward CVD in FH.
Key Words: cardiovascular disease ? genetics ? hypercholesterolemia ? risk factors
Introduction
Familial hypercholesterolemia (FH) is a common hereditary disease, characterized by elevated levels of plasma low-density lipoprotein cholesterol (LDL-C) and premature cardiovascular disease (CVD).1 Characteristically, the mean age of onset of CVD is between 40 and 45 years in male FH patients and in female FH patients 10 years later.1,2 Nevertheless, the phenotypic expression of this disorder, in terms of onset and severity of atherosclerotic vascular disease, varies considerably.3
Unfortunately, a paucity of solid data exists on factors that contribute to these phenotypic differences. Previous studies have mostly focused on classical CVD risk factors and the functional variety among LDL receptor mutations.4–6 Although both influence the occurrence of CVD, they can only partially explain the observed large differences. We recently studied the contribution of classical risk factors to CVD in a large cohort of FH patients and demonstrated that <20% of the variation in CVD occurrence could be explained by these risk factors alone.7 Therefore, other still unknown and possibly genetic factors play an undeniable role in the development of CVD in these patients. Genetic differences affect susceptibility to disease and whereas premature atherosclerosis can be linked in rare cases to single-gene disorders, most individuals do not carry such DNA defects. The "common disorder, common variant" theory predicts that the majority of population-attributable variation in susceptibility to prevalent disease is caused by variants that occur in high frequency in multiple genes.8
Such genetic variation may also play an important role in the development of CVD in FH. This is substantiated by the fact that clustering of CVD occurs in FH kindred.9 Unfortunately, large-scale association studies involving multiple polymorphisms are lacking in FH. Our objective, therefore, was to investigate the contribution of polymorphisms in multiple candidate genes to CVD risk in a large cohort of patients with heterozygous FH.
Methods
Study Design and Study Population
The present investigation was a retrospective, multicenter, cohort study. The study design and study population have been described elsewhere.7 Briefly, lipid clinics in the Netherlands submit DNA samples from clinically suspected FH patients to a central laboratory for LDL receptor mutation analysis.10 We randomly selected hypercholesterolemic patients from this DNA bank database with the aid of a computer program (Microsoft Excel). These patients had been referred from 27 lipid clinics throughout the Netherlands (Figure I, available online at http://atvb.ahajournals.org).
Phenotypic data (including detailed information on CVD) were acquired by reviewing patient’s medical records by a trained team of data collectors.7 Guidelines for data collection from medical records were constructed for the purpose of the study and have been published.11 Written informed consent was obtained from all living patients. The Ethics Institutional Review Board of each participating hospital approved the protocol.
Power Calculation
To calculate CVD risk associated with genetic polymorphisms, at least 2000 FH patients were needed to reach a power of 80% to detect an odds ratio of 1.9 or more for carriers of the rare allele (please see http://atvb.ahajournals.org for the exact power calculation).
Selection of Patients and Inclusion Criteria
On reviewing 4000 medical records, a total of 2400 patients fulfilled the inclusion and exclusion criteria for participation and were included in the study (Figure I and Table I, available online at http://atvb.ahajournals.org). The FH diagnostic criteria were based on internationally established criteria.12–14
Selection of DNA Polymorphisms and Genetic Analyses
We genotyped 65 polymorphisms in 36 candidate genes that were previously implicated in CVD.15 Complete genotypes for all 65 polymorphisms could be obtained for 1940 (80.1%) of patients. In the remaining 460 patients, complete genotyping was impaired by technical difficulties. First, the amount of DNA was insufficient for adequate amplification in 235 patients. Among the remaining samples, complete genotyping for all polymorphisms was not possible. The missing data rate ranged from 0.51% to 5.3%, with an average rate of 1.7%. These patients did not differ clinically from the 1940 patients and were excluded from further analyses. Genomic DNA was extracted from peripheral blood leukocytes by standard procedures. Each sample was amplified using 2 pools of biotinylated polymerase chain reaction primers. Each polymerase chain reaction product pool was then hybridized to a linear array of sequence-specific oligonucleotide probes and alleles were detected using a hydrogen peroxidase-based chromogenic reaction, essentially as described previously.15 Samples were blinded for genotyping. The accuracy of genotyping in 500 randomly selected DNA samples was assessed by re-analysis of several polymorphisms in 3 genes (CETP, MTHFR, LPL). Less than 0.5% of discordant results were found.
Statistical Analysis
Differences between subgroups were tested with 2 statistics or independent sample t test (for triglycerides [skewed distribution] on log-transformed data). To adjust for the effects of age and sex, we used multiple linear or logistic regression. Allele frequencies were calculated by genotype counting and for each (biallelic) polymorphism the deviation from Hardy–Weinberg equilibrium was tested by a 2 test with 1 degree of freedom. A likelihood ratio test was used to detect pair-wise linkage disequilibrium16 and the extent of disequilibrium was expressed in terms of D'=D/Dmax or D/Dmin.17 Cox proportional hazard regression with backward stepwise selection was used to model the association of multiple polymorphisms simultaneously with the occurrence of CVD. Polymorphisms were treated as categorical variables and patients homozygous for the common allele were used as the reference category. Follow-up started at birth and ended for each individual at the date of the first occurrence of established CVD. Patients without CVD were censored at the date of the last lipid clinic visit or at the date of death attributable to other causes. Because polymorphisms might express their untoward effects by way of, for example, hypertension, diabetes mellitus, obesity, or dyslipidemia, we did not introduce these factors as covariates in our models. Instead, we added those covariates that function independently from the polymorphisms: sex and smoking (time-dependent). For smoking, we implemented a linearly decreasing risk effect for the 3 years after cessation.18 Statistical analyses were performed using SPSS software (version 11.5; Chicago, Ill). Polymorphisms with P<0.05 for the likelihood ratio test were considered to be suggestively associated; those with P<0.001 were considered statistically significant.
Results
Clinical characteristics of the 1940 completely genotyped patients are described in Table 1. During 91.451 person-years, 643 (33.1%) patients had at least 1 cardiovascular event, including 29 individuals who died from documented CVD events. Mean age of onset of CVD was 48.2 years. Patients with CVD were older, more often males and smokers, and had a higher prevalence of hypertension and diabetes mellitus. More obesity and higher total cholesterol levels were also observed among CVD patients but were not significant after adjustment for age and sex. LDL-C levels did not differ between CVD and non-CVD patients. High-density lipoprotein cholesterol levels were lower (1.15±0.32 versus 1.24±0.36 mmol/L; P<0.001) and median triglyceride levels higher (1.76 versus 1.48 mmol/L; P<0.001) in patients with CVD.
TABLE 1. Clinical Characteristics of 1940 FH Patients With and Without Cardiovascular Disease
The 36 genes and 65 polymorphisms examined in the study and the frequencies of the least common alleles of the polymorphisms are presented in Table 2. All but 11 of the 65 polymorphisms studied were in Hardy–Weinberg equilibrium. The same polymorphisms showed deviation from Hardy–Weinberg equilibrium in patients with and without cardiovascular disease (data not shown). For 3 polymorphisms in the CETP gene (Asp442Gly, +1 G>A, and +3insT/in14), only homozygous wild-type individuals were found, which were excluded from further analyses.
TABLE 2. The 65 Polymorphisms in 36 Candidate Genes Examined in the Study
TABLE 2. Continued
Multifactorial Cox survival analysis, which included the 62 polymorphisms simultaneously, adjusted for sex and smoking, revealed that the G20210A polymorphism in the prothrombin gene exhibited the strongest association with an increased risk of CVD (GA versus GG; P<0.001) (Table 3). Heterozygous carriers of the G20210A polymorphism showed clearly reduced cardiovascular event-free survival rates (Figure). Subgroup analyses for gender, total cholesterol, LDL-C, high-density lipoprotein cholesterol, and triglyceride tertiles and year of birth (before 1930, 1930 to 1949, 1950 to 1969, or after 1970) were performed. For each subgroup, hazard ratios (HR) of 2.0 (not significant) were found, suggesting that the observed HR is not caused by a specific subgroup within the present population (data not shown). In addition, 4 other polymorphisms were identified to be suggestively associated with the risk of CVD (P<0.05). Heterozygous and homozygous carriers of the Met235Thr variant in the angiotensinogen gene and homozygous carriers of the Thr347Ser variant in the apolipoprotein (apo) A4 gene had an increased CVD risk (HR Met235Thrhet 1.25 (95% CI, 1.05 to 1.48) and HR Met235Thrhom 1.23 (95% CI, 0.98 to 1.54), respectively, and HR Thr347Serhom 1.37 (95% CI, 0.97 to 1.93). Conversely, homozygous carriers of the Ser311Cys variant in the paraoxonase-2 gene and homozygous carriers of the C1100T variant in the apoC3 gene had a decreased CVD risk (HR, Ser311Cyshom 0.69 [95% CI, 0.47 to 1.01] and HR, C1100Thom 0.65 [95% CI, 0.46 to 0.91], respectively). Strong linkage disequilibrium was observed between the polymorphisms C1100T in apoC3 and Thr347Ser in apoA4 (D'=–0.928). Therefore, multifactorial Cox survival analysis was performed using the genotype combination of the polymorphisms C1100T in apoC3 and Thr347Ser in apoA4 (Table 4). Compared with homozygotes for the common alleles, the SerSer+CC combination showed a significant effect on risk (HR 1.43 [95% CI, 1.01 to 2.03]), whereas ThrThr+TT combination showed a significant protective effect on risk (HR, 0.69 [95% CI, 0.49 to 0.97]).
TABLE 3. Multifactorial Cox Regression of Polymorphisms Associated With Cardiovascular Disease
Kaplan–Meier curves for cardiovascular event-free survival by G20210A genotype.
TABLE 4. HR of Apolipoprotein A4 Thr347Ser and Apolipoprotein C3 C1100T Genotype Combinations*
We investigated and confirmed the associations between specific polymorphisms and hypertension, diabetes mellitus, obesity, and dyslipidemia, which were not introduced as covariates in our models. However, to enable comparison of our results with those from earlier studies, we performed the same Cox regression model while adjusting not only for sex and smoking but also for hypertension (time-dependent), diabetes mellitus (time-dependent), body mass index, total cholesterol, LDL-C, high-density lipoprotein cholesterol, triglycerides, lipoprotein(a), and homocysteine levels. The detected HRs were similar (data not shown). The genotype distributions of polymorphisms associated with cardiovascular disease are depicted in Table 5.
TABLE 5. Distributions of Polymorphisms Associated With Cardiovascular Disease
Discussion
We investigated the contribution of a large number of polymorphisms in multiple candidate genes to CVD risk in FH patients. Strikingly, the G20210A polymorphism in the prothrombin gene was most strongly related to a significantly increased CVD risk. In addition, 4 other polymorphisms were suggestively associated with CVD (P<0.05). Two were associated with increased CVD risk, namely the Met235Thr variant in the angiotensinogen gene and the Thr347Ser variant in the apoA4 gene. In contrast, the Ser311Cys substitution in the paraoxonase-2 gene and the C1100T variant in the apoC3 gene were associated with decreased CVD risk. To our knowledge, this is the largest exploratory study on the association of genetic variants and CVD risk in FH.
The strengths of the present study lie in several areas. To begin with, by recruiting patients from all over the country, and by using our patient registration database, we minimized selection bias toward large families and genetically isolated communities. Moreover, by choosing a cohort design and deriving our "cases" and "controls" from a common population, we reduced the chance of population stratification and spurious associations.
Second, the large size of the cohort provided sufficient statistical power to detect relatively small relative risks or HRs. Our initial power statement was based on calculating an odds ratio. A power statement for HR was difficult to formulate, because other, less easy to predict, assumptions like median survival in control patients would have had to have been included in that power analysis. In retrospect, the average median survival of homozygous wild-type patients was 60 years, with a maximum follow-up of 85 years. Thus, in our cohort, HR of 1.6 (or 0.7 protective) could have been detected with a power of 80% (with a Bonferroni corrected 2-sided P=0.0008) using 1940 patients and assuming that 10% of the patients carried at least 1 rare allele. Because the carrier frequency was even higher than 10% for most polymorphisms, the actual power to detect a HR of 1.6 or higher exceeded 80%.
Last, we applied more stringent criteria for statistical significance to avoid problems of "repetitive testing," which often occurs in studies of multiple polymorphisms. Furthermore, to examine possible false-positive associations caused by linkage disequilibrium between loci, we simultaneously assessed the effects of multiple polymorphisms in a multifactorial Cox model. Importantly, a cohort study design enabled us to determine HRs. Many association studies have had case-control designs with odds ratios as the outcome parameter. Whereas odds ratios mimic relative risks remarkably well when the outcome of interest is rare, this is not the case for CVD in FH, in which odds ratios might exaggerate the actual risk.
Several limitations of our study should also be mentioned. First, in this retrospective study, the primary source of data was the patient’s medical records. Medical records are primarily intended for patient care and not for research purposes. Second, no standardized information was available on lifestyle factors, such as dietary habits and physical activity. As a result, the interaction between these environmental factors and genetic variants could not be studied.
Third, our study included patients that were referred to a lipid clinic. In theory, patients with the most detrimental genetic profiles might have died before referral. Therefore, genetic polymorphisms associated with more severe CVD or early death could have been missed, leading to an underestimation of the risk. Fourth, 11 of the 65 polymorphisms studied were not in Hardy–Weinberg equilibrium. The exact reason for the deviation is not known and we can only speculate on this. Importantly, the deviation from Hardy–Weinberg equilibrium is not because of mixed ethnic groups, because the Dutch population is known to be a homogenous one. More than 99% of our patients were white and patients were randomly selected from all over the country. In addition, most deviations were caused by an excess of heterozygotes, which makes genotyping errors unlikely. Furthermore, the accuracy of genotyping in 500 randomly selected DNA samples was assessed by re-analysis of several polymorphisms in three genes, revealing that <0.5% of the results were discordant. However, it must be stressed that these 3 re-analyzed polymorphisms were not among the polymorphisms that were not in Hardy–Weinberg equilibrium. Finally, and most importantly, the 11 SNPs showed the same deviation from Hardy–Weinberg equilibrium in patients with CVD as well as in patients without CVD; therefore, we are of the opinion that this has not affected our results significantly. Fifth, it should be stressed that by the nature of this exploratory study the impact of the polymorphisms on intermediate end points such as coagulation factors (eg, prothrombin), blood pressure levels, angiotensinogen levels, and apolipoprotein levels, could not be assessed in the present study. Therefore, we could not determine the relationship of these intermediate end points with CVD, nor the subsequent pathological mechanisms involved in the development of CVD.
Our results are in line with earlier findings. The prothrombin mutation G20210A leads to increased plasma levels of prothrombin and an increased risk of venous thrombosis.19 Many studies have focused on the causative role of this mutation in arterial disease, often with negative results.20,21 However, in a comprehensive meta-analysis investigating the possible link between the G20210A prothrombin gene variant and different forms of premature ischemic heart disease in 12 043 patients, the G20210A polymorphism was shown to be a significant risk factor for myocardial infarction at a young age.22 In patients aged 55 years and younger, the odds ratio was 1.77 (95% CI, 1.16 to 3.42); in those aged 45 years and younger, the odds ratio was 2.30 (95% CI, 1.27 to 4.59). These odds ratios are similar to the HRs found in our study. In another meta-analysis, the association between the G20210A variant and arterial circulatory events was confirmed.23 In addition, 2 other large studies showed a pronounced effect of this polymorphism on CVD events in dyslipidemic patients.24,25 Although the relationship between this prothrombin mutation and CVD remains the subject of debate in non-FH populations, these findings support the strong association between G20210A carriership and premature CVD risk in our FH patients who have severe dyslipidemia from birth onwards.
Recently, the M235T polymorphism in the angiotensinogen gene was associated with increased levels of angiotensinogen and a corresponding increase in the risk of hypertension, as assessed in 45 267 individuals.26 However, a relationship between this variant and CVD risk has never been established in a large-scale study. Again, in our FH patients, the combined effect of elevated angiotensinogen and LDL-C levels might explain the increased risk of CVD.
We identified 2 common variants in the ApoA1-C3-A4-A5 gene cluster to be suggestively associated with CVD risk (Thr347Ser in apoA4 and C1100T in apoC3). Both appear to have (statistically) independent effects, because they remained significant in the multifactorial model and the estimated HRs of the univariate and multifactorial model were almost identical for both variants (data not shown). Apolipoproteins play a central role in lipid metabolism and this cluster of apolipoprotein genes has been identified as a locus with significant effects on triglyceride levels.27–29 Remarkably, in the first prospective study on genetic variants in this gene cluster and CVD risk in the general population, exactly the same 2 polymorphisms were associated with an increase and decrease in risk, respectively.30 However, in their multifactorial model, which also included other known risk factors such as body mass index and blood pressure, only the Thr347Ser variant remained significantly associated with CVD risk. They concluded that the apparent protective effect of C1100T in the apoC3 gene is most likely caused by strong linkage disequilibrium across the gene cluster. When body mass index, blood pressure, total cholesterol, and triglycerides were also added to our multifactorial model, the effect of the Thr347Ser variant was no longer significant (data not shown). However, this apparent discrepancy could be caused by differences in genotype frequencies between the 2 populations. Furthermore, in both populations the inference of functionality for these variants was hampered by the almost complete absence of patients homozygous for both the 347Ser and the 1100T allele.
Many studies support the antiatherogenic role of paraoxonase-1, an HDL-associated enzyme.31 The protective effect is generally considered to be caused by the ability of paraoxonase-1 to attenuate oxidative modification of lipoprotein particles. Paraoxonase-2, a family member of paraoxonase-1, is reported to have antioxidant properties and antiatherogenic capacities as well, although its exact physiological role is still unknown.31 Our results confirm a relationship between paraoxonase-2 and CVD, and are in concordance with the findings of an earlier study in FH patients, in which homozygotes for the Ser311Cys variant of paraoxonase-2 seem to be protected against CVD.32
Considering the limitations of association studies, we suggest our results should be replicated in addition to being performed in prospective studies of large and well-defined FH populations, in which genetic and environmental modifiers should be carefully monitored. The results of this hypothesis-generating study constitute a basis for further research and form a step forward in the unraveling of the underlying mechanisms of CVD in FH.
Acknowledgments
This study was supported by a grant of the Netherlands Heart Foundation (98/165). Kastelein is an established investigator of the Netherlands Heart Foundation (grant 2000D039). We are indebted to RMS personnel who supported this study. We acknowledge the members of the independent adjudication committee: R. J. G. Peters, MD, PhD, cardiologist; J. Stam, MD, PhD, neurologist; and D. Legemate, MD, PhD, vascular surgeon. We thank all the patients and the specialists of the participating lipid clinics.
References
Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver CR, et al, eds. The metabolic basis of inherited disease. New York, NY: McGraw-Hill; 2001: 2863–2913.
Risk of fatal coronary heart disease in familial hypercholesterolaemia. Scientific Steering Committee on behalf of the Simon Broome Register Group. BMJ. 1991; 303: 893–896.
Sijbrands EJ, Westendorp RG, Defesche JC, de Meier PH, Smelt AH, Kastelein JJ. Mortality over two centuries in large pedigree with familial hypercholesterolaemia: family tree mortality study. BMJ. 2001; 322: 1019–1023.
Ferrieres J, Lambert J, Lussier-Cacan S, Davignon J. Coronary artery disease in heterozygous familial hypercholesterolemia patients with the same LDL receptor gene mutation. Circulation. 1995; 92: 290–295.
Sauvage Nolting PR, Defesche JC, Buirma RJ, Hutten BA, Lansberg PJ, Kastelein JJ. Prevalence and significance of cardiovascular risk factors in a large cohort of patients with familial hypercholesterolaemia. J Intern Med. 2003; 253: 161–168.
Umans-Eckenhausen MA, Sijbrands EJ, Kastelein JJ, Defesche JC. Low-density lipoprotein receptor gene mutations and cardiovascular risk in a large genetic cascade screening population. Circulation. 2002; 106: 3031–3036.
Jansen AC, Aalst-Cohen ES, Tanck MW, Trip MD, Lansberg PJ, Liem AH, van Lennep HW, Sijbrands EJ, Kastelein JJ. The contribution of classical risk factors to cardiovascular disease in familial hypercholesterolaemia: data in 2400 patients. J Intern Med. 2004; 256: 482–490.
Risch N, Merikangas K. The future of genetic studies of complex human diseases. Science. 1996; 273: 1516–1517.
Sijbrands EJ, Westendorp RG, Paola LM, Havekes LM, Frants RR, Kastelein JJ, Smelt AH. Additional risk factors influence excess mortality in heterozygous familial hypercholesterolaemia. Atherosclerosis. 2000; 149: 421–425.
Umans-Eckenhausen MA, Defesche JC, Sijbrands EJ, Scheerder RL, Kastelein JJ. Review of first 5 years of screening for familial hypercholesterolaemia in the Netherlands. Lancet. 2001; 357: 165–168.
Jansen AC, Aalst-Cohen ES, Hutten BA, Buller HR, Kastelein JJ, Prins MH. Guidelines were developed for data collection from medical records for use in retrospective analyses. J Clin Epidemiol. 2005; 58: 269–274.
Mortality in treated heterozygous familial hypercholesterolaemia: implications for clinical management. Scientific Steering Committee on behalf of the Simon Broome Register Group. Atherosclerosis. 1999; 142: 105–112.
US MedPed program. http:/www.medped.org. 2005.
Defesche J.C. Familial Hypercholesterolemia. In: Betteridge DJ, ed. Lipids and Vascular Disease. London, UK: Martin Dunitz Ltd; 2000: 65–76.
Cheng S, Grow MA, Pallaud C, Klitz W, Erlich HA, Visvikis S, Chen JJ, Pullinger CR, Malloy MJ, Siest G, Kane JP. A multilocus genotyping assay for candidate markers of cardiovascular disease risk. Genome Res. 1999; 9: 936–949.
Slatkin M, Excoffier L. Testing for linkage disequilibrium in genotypic data using the Expectation-Maximization algorithm. Heredity. 1996; 76: 377–383.
Lewontin RC. The interaction of selection and linkage. II. Optimum models. Genetics. 1964; 50: 757–782.
Rea TD, Heckbert SR, Kaplan RC, Smith NL, Lemaitre RN, Psaty BM. Smoking status and risk for recurrent coronary events after myocardial infarction. Ann Intern Med. 2002; 137: 494–500.
Poort SR, Rosendaal FR, Reitsma PH, Bertina RM. A common genetic variation in the 3'-untranslated region of the prothrombin gene is associated with elevated plasma prothrombin levels and an increase in venous thrombosis. Blood. 1996; 88: 3698–3703.
No evidence of association between prothrombotic gene polymorphisms and the development of acute myocardial infarction at a young age. Circulation. 2003; 107: 1117–1122.
Yamada Y, Izawa H, Ichihara S, Takatsu F, Ishihara H, Hirayama H, Sone T, Tanaka M, Yokota M. Prediction of the risk of myocardial infarction from polymorphisms in candidate genes. N Engl J Med. 2002; 347: 1916–1923.
Burzotta F, Paciaroni K, De S V, Crea F, Maseri A, Leone G, Andreotti F. G20210A prothrombin gene polymorphism and coronary ischaemic syndromes: a phenotype-specific meta-analysis of 12 034 subjects. Heart. 2004; 90: 82–86.
Kim RJ, Becker RC. Association between factor V Leiden, prothrombin G20210A, and methylenetetrahydrofolate reductase C677T mutations and events of the arterial circulatory system: a meta-analysis of published studies. Am Heart J. 2003; 146: 948–957.
Doggen CJ, Cats VM, Bertina RM, Rosendaal FR. Interaction of coagulation defects and cardiovascular risk factors: increased risk of myocardial infarction associated with factor V Leiden or prothrombin 20210A. Circulation. 1998; 97: 1037–1041.
Gardemann A, Arsic T, Katz N, Tillmanns H, Hehrlein FW, Haberbosch W. The factor II G20210A and factor V G1691A gene transitions and coronary heart disease. Thromb Haemost. 1999; 81: 208–213.
Sethi AA, Nordestgaard BG, Tybjaerg-Hansen A. Angiotensinogen gene polymorphism, plasma angiotensinogen, and risk of hypertension and ischemic heart disease: a meta-analysis. Arterioscler Thromb Vasc Biol. 2003; 23: 1269–1275.
Groenendijk M, Cantor RM, de Bruin TW, Dallinga-Thie GM. The apoAI-CIII-AIV gene cluster. Atherosclerosis. 2001; 157: 1–11.
Talmud PJ, Hawe E, Martin S, Olivier M, Miller GJ, Rubin EM, Pennacchio LA, Humphries SE. Relative contribution of variation within the APOC3/A4/A5 gene cluster in determining plasma triglycerides. Hum Mol Genet. 2002; 11: 3039–3046.
Olivier M, Wang X, Cole R, Gau B, Kim J, Rubin EM, Pennacchio LA. Haplotype analysis of the apolipoprotein gene cluster on human chromosome 11. Genomics. 2004; 83: 912–923.
Wong WM, Hawe E, Li LK, Miller GJ, Nicaud V, Pennacchio LA, Humphries SE, Talmud PJ. Apolipoprotein AIV gene variant S347 is associated with increased risk of coronary heart disease and lower plasma apolipoprotein AIV levels. Circ Res. 2003; 92: 969–975.
Ng CJ, Shih DM, Hama SY, Villa N, Navab M, Reddy ST. The paraoxonase gene family and atherosclerosis. Free Radic Biol Med. 2005; 38: 153–163.
Leus FR, Zwart M, Kastelein JJ, Voorbij HA. PON2 gene variants are associated with clinical manifestations of cardiovascular disease in familial hypercholesterolemia patients. Atherosclerosis. 2001; 154: 641–649.(Angelique C.M. Jansen; Em)
Correspondence to John J.P. Kastelein, Department of Vascular Medicine, Academic Medical Center, Meibergdreef 9, P.O.Box 22700, room F4-159.2, 1100 DE Amsterdam, the Netherlands. E-mail e.vandongen@amc.uva.nl
Abstract
Objective— To investigate the contribution of polymorphisms in multiple candidate genes to cardiovascular disease (CVD) risk in a large cohort of patients with heterozygous familial hypercholesterolemia (FH).
Methods and Results— We genotyped 1940 FH patients for 65 polymorphisms in 36 candidate genes. During 91.451 person-years, 643 (33.1%) patients had at least 1 cardiovascular event. Multifactorial Cox survival analysis revealed that the G20210A polymorphism in the prothrombin gene was strongly associated with a significantly increased CVD risk (GA versus GG; P<0.001).
Conclusions— In a large cohort of FH patients, we found that the G20210A polymorphism in the prothrombin gene is strongly associated with CVD risk. Our results constitute a step forward in the unraveling of the hereditary propensity toward CVD in FH and might lead to better risk stratification and hence to more tailored therapy for CVD prevention.
We investigated the contribution of 65 polymorphisms in 36 candidate genes to CVD risk in FH patients and found that the G20210A polymorphism in the prothrombin gene was associated with CVD risk. Our results constitute a step forward in the unraveling of the hereditary propensity toward CVD in FH.
Key Words: cardiovascular disease ? genetics ? hypercholesterolemia ? risk factors
Introduction
Familial hypercholesterolemia (FH) is a common hereditary disease, characterized by elevated levels of plasma low-density lipoprotein cholesterol (LDL-C) and premature cardiovascular disease (CVD).1 Characteristically, the mean age of onset of CVD is between 40 and 45 years in male FH patients and in female FH patients 10 years later.1,2 Nevertheless, the phenotypic expression of this disorder, in terms of onset and severity of atherosclerotic vascular disease, varies considerably.3
Unfortunately, a paucity of solid data exists on factors that contribute to these phenotypic differences. Previous studies have mostly focused on classical CVD risk factors and the functional variety among LDL receptor mutations.4–6 Although both influence the occurrence of CVD, they can only partially explain the observed large differences. We recently studied the contribution of classical risk factors to CVD in a large cohort of FH patients and demonstrated that <20% of the variation in CVD occurrence could be explained by these risk factors alone.7 Therefore, other still unknown and possibly genetic factors play an undeniable role in the development of CVD in these patients. Genetic differences affect susceptibility to disease and whereas premature atherosclerosis can be linked in rare cases to single-gene disorders, most individuals do not carry such DNA defects. The "common disorder, common variant" theory predicts that the majority of population-attributable variation in susceptibility to prevalent disease is caused by variants that occur in high frequency in multiple genes.8
Such genetic variation may also play an important role in the development of CVD in FH. This is substantiated by the fact that clustering of CVD occurs in FH kindred.9 Unfortunately, large-scale association studies involving multiple polymorphisms are lacking in FH. Our objective, therefore, was to investigate the contribution of polymorphisms in multiple candidate genes to CVD risk in a large cohort of patients with heterozygous FH.
Methods
Study Design and Study Population
The present investigation was a retrospective, multicenter, cohort study. The study design and study population have been described elsewhere.7 Briefly, lipid clinics in the Netherlands submit DNA samples from clinically suspected FH patients to a central laboratory for LDL receptor mutation analysis.10 We randomly selected hypercholesterolemic patients from this DNA bank database with the aid of a computer program (Microsoft Excel). These patients had been referred from 27 lipid clinics throughout the Netherlands (Figure I, available online at http://atvb.ahajournals.org).
Phenotypic data (including detailed information on CVD) were acquired by reviewing patient’s medical records by a trained team of data collectors.7 Guidelines for data collection from medical records were constructed for the purpose of the study and have been published.11 Written informed consent was obtained from all living patients. The Ethics Institutional Review Board of each participating hospital approved the protocol.
Power Calculation
To calculate CVD risk associated with genetic polymorphisms, at least 2000 FH patients were needed to reach a power of 80% to detect an odds ratio of 1.9 or more for carriers of the rare allele (please see http://atvb.ahajournals.org for the exact power calculation).
Selection of Patients and Inclusion Criteria
On reviewing 4000 medical records, a total of 2400 patients fulfilled the inclusion and exclusion criteria for participation and were included in the study (Figure I and Table I, available online at http://atvb.ahajournals.org). The FH diagnostic criteria were based on internationally established criteria.12–14
Selection of DNA Polymorphisms and Genetic Analyses
We genotyped 65 polymorphisms in 36 candidate genes that were previously implicated in CVD.15 Complete genotypes for all 65 polymorphisms could be obtained for 1940 (80.1%) of patients. In the remaining 460 patients, complete genotyping was impaired by technical difficulties. First, the amount of DNA was insufficient for adequate amplification in 235 patients. Among the remaining samples, complete genotyping for all polymorphisms was not possible. The missing data rate ranged from 0.51% to 5.3%, with an average rate of 1.7%. These patients did not differ clinically from the 1940 patients and were excluded from further analyses. Genomic DNA was extracted from peripheral blood leukocytes by standard procedures. Each sample was amplified using 2 pools of biotinylated polymerase chain reaction primers. Each polymerase chain reaction product pool was then hybridized to a linear array of sequence-specific oligonucleotide probes and alleles were detected using a hydrogen peroxidase-based chromogenic reaction, essentially as described previously.15 Samples were blinded for genotyping. The accuracy of genotyping in 500 randomly selected DNA samples was assessed by re-analysis of several polymorphisms in 3 genes (CETP, MTHFR, LPL). Less than 0.5% of discordant results were found.
Statistical Analysis
Differences between subgroups were tested with 2 statistics or independent sample t test (for triglycerides [skewed distribution] on log-transformed data). To adjust for the effects of age and sex, we used multiple linear or logistic regression. Allele frequencies were calculated by genotype counting and for each (biallelic) polymorphism the deviation from Hardy–Weinberg equilibrium was tested by a 2 test with 1 degree of freedom. A likelihood ratio test was used to detect pair-wise linkage disequilibrium16 and the extent of disequilibrium was expressed in terms of D'=D/Dmax or D/Dmin.17 Cox proportional hazard regression with backward stepwise selection was used to model the association of multiple polymorphisms simultaneously with the occurrence of CVD. Polymorphisms were treated as categorical variables and patients homozygous for the common allele were used as the reference category. Follow-up started at birth and ended for each individual at the date of the first occurrence of established CVD. Patients without CVD were censored at the date of the last lipid clinic visit or at the date of death attributable to other causes. Because polymorphisms might express their untoward effects by way of, for example, hypertension, diabetes mellitus, obesity, or dyslipidemia, we did not introduce these factors as covariates in our models. Instead, we added those covariates that function independently from the polymorphisms: sex and smoking (time-dependent). For smoking, we implemented a linearly decreasing risk effect for the 3 years after cessation.18 Statistical analyses were performed using SPSS software (version 11.5; Chicago, Ill). Polymorphisms with P<0.05 for the likelihood ratio test were considered to be suggestively associated; those with P<0.001 were considered statistically significant.
Results
Clinical characteristics of the 1940 completely genotyped patients are described in Table 1. During 91.451 person-years, 643 (33.1%) patients had at least 1 cardiovascular event, including 29 individuals who died from documented CVD events. Mean age of onset of CVD was 48.2 years. Patients with CVD were older, more often males and smokers, and had a higher prevalence of hypertension and diabetes mellitus. More obesity and higher total cholesterol levels were also observed among CVD patients but were not significant after adjustment for age and sex. LDL-C levels did not differ between CVD and non-CVD patients. High-density lipoprotein cholesterol levels were lower (1.15±0.32 versus 1.24±0.36 mmol/L; P<0.001) and median triglyceride levels higher (1.76 versus 1.48 mmol/L; P<0.001) in patients with CVD.
TABLE 1. Clinical Characteristics of 1940 FH Patients With and Without Cardiovascular Disease
The 36 genes and 65 polymorphisms examined in the study and the frequencies of the least common alleles of the polymorphisms are presented in Table 2. All but 11 of the 65 polymorphisms studied were in Hardy–Weinberg equilibrium. The same polymorphisms showed deviation from Hardy–Weinberg equilibrium in patients with and without cardiovascular disease (data not shown). For 3 polymorphisms in the CETP gene (Asp442Gly, +1 G>A, and +3insT/in14), only homozygous wild-type individuals were found, which were excluded from further analyses.
TABLE 2. The 65 Polymorphisms in 36 Candidate Genes Examined in the Study
TABLE 2. Continued
Multifactorial Cox survival analysis, which included the 62 polymorphisms simultaneously, adjusted for sex and smoking, revealed that the G20210A polymorphism in the prothrombin gene exhibited the strongest association with an increased risk of CVD (GA versus GG; P<0.001) (Table 3). Heterozygous carriers of the G20210A polymorphism showed clearly reduced cardiovascular event-free survival rates (Figure). Subgroup analyses for gender, total cholesterol, LDL-C, high-density lipoprotein cholesterol, and triglyceride tertiles and year of birth (before 1930, 1930 to 1949, 1950 to 1969, or after 1970) were performed. For each subgroup, hazard ratios (HR) of 2.0 (not significant) were found, suggesting that the observed HR is not caused by a specific subgroup within the present population (data not shown). In addition, 4 other polymorphisms were identified to be suggestively associated with the risk of CVD (P<0.05). Heterozygous and homozygous carriers of the Met235Thr variant in the angiotensinogen gene and homozygous carriers of the Thr347Ser variant in the apolipoprotein (apo) A4 gene had an increased CVD risk (HR Met235Thrhet 1.25 (95% CI, 1.05 to 1.48) and HR Met235Thrhom 1.23 (95% CI, 0.98 to 1.54), respectively, and HR Thr347Serhom 1.37 (95% CI, 0.97 to 1.93). Conversely, homozygous carriers of the Ser311Cys variant in the paraoxonase-2 gene and homozygous carriers of the C1100T variant in the apoC3 gene had a decreased CVD risk (HR, Ser311Cyshom 0.69 [95% CI, 0.47 to 1.01] and HR, C1100Thom 0.65 [95% CI, 0.46 to 0.91], respectively). Strong linkage disequilibrium was observed between the polymorphisms C1100T in apoC3 and Thr347Ser in apoA4 (D'=–0.928). Therefore, multifactorial Cox survival analysis was performed using the genotype combination of the polymorphisms C1100T in apoC3 and Thr347Ser in apoA4 (Table 4). Compared with homozygotes for the common alleles, the SerSer+CC combination showed a significant effect on risk (HR 1.43 [95% CI, 1.01 to 2.03]), whereas ThrThr+TT combination showed a significant protective effect on risk (HR, 0.69 [95% CI, 0.49 to 0.97]).
TABLE 3. Multifactorial Cox Regression of Polymorphisms Associated With Cardiovascular Disease
Kaplan–Meier curves for cardiovascular event-free survival by G20210A genotype.
TABLE 4. HR of Apolipoprotein A4 Thr347Ser and Apolipoprotein C3 C1100T Genotype Combinations*
We investigated and confirmed the associations between specific polymorphisms and hypertension, diabetes mellitus, obesity, and dyslipidemia, which were not introduced as covariates in our models. However, to enable comparison of our results with those from earlier studies, we performed the same Cox regression model while adjusting not only for sex and smoking but also for hypertension (time-dependent), diabetes mellitus (time-dependent), body mass index, total cholesterol, LDL-C, high-density lipoprotein cholesterol, triglycerides, lipoprotein(a), and homocysteine levels. The detected HRs were similar (data not shown). The genotype distributions of polymorphisms associated with cardiovascular disease are depicted in Table 5.
TABLE 5. Distributions of Polymorphisms Associated With Cardiovascular Disease
Discussion
We investigated the contribution of a large number of polymorphisms in multiple candidate genes to CVD risk in FH patients. Strikingly, the G20210A polymorphism in the prothrombin gene was most strongly related to a significantly increased CVD risk. In addition, 4 other polymorphisms were suggestively associated with CVD (P<0.05). Two were associated with increased CVD risk, namely the Met235Thr variant in the angiotensinogen gene and the Thr347Ser variant in the apoA4 gene. In contrast, the Ser311Cys substitution in the paraoxonase-2 gene and the C1100T variant in the apoC3 gene were associated with decreased CVD risk. To our knowledge, this is the largest exploratory study on the association of genetic variants and CVD risk in FH.
The strengths of the present study lie in several areas. To begin with, by recruiting patients from all over the country, and by using our patient registration database, we minimized selection bias toward large families and genetically isolated communities. Moreover, by choosing a cohort design and deriving our "cases" and "controls" from a common population, we reduced the chance of population stratification and spurious associations.
Second, the large size of the cohort provided sufficient statistical power to detect relatively small relative risks or HRs. Our initial power statement was based on calculating an odds ratio. A power statement for HR was difficult to formulate, because other, less easy to predict, assumptions like median survival in control patients would have had to have been included in that power analysis. In retrospect, the average median survival of homozygous wild-type patients was 60 years, with a maximum follow-up of 85 years. Thus, in our cohort, HR of 1.6 (or 0.7 protective) could have been detected with a power of 80% (with a Bonferroni corrected 2-sided P=0.0008) using 1940 patients and assuming that 10% of the patients carried at least 1 rare allele. Because the carrier frequency was even higher than 10% for most polymorphisms, the actual power to detect a HR of 1.6 or higher exceeded 80%.
Last, we applied more stringent criteria for statistical significance to avoid problems of "repetitive testing," which often occurs in studies of multiple polymorphisms. Furthermore, to examine possible false-positive associations caused by linkage disequilibrium between loci, we simultaneously assessed the effects of multiple polymorphisms in a multifactorial Cox model. Importantly, a cohort study design enabled us to determine HRs. Many association studies have had case-control designs with odds ratios as the outcome parameter. Whereas odds ratios mimic relative risks remarkably well when the outcome of interest is rare, this is not the case for CVD in FH, in which odds ratios might exaggerate the actual risk.
Several limitations of our study should also be mentioned. First, in this retrospective study, the primary source of data was the patient’s medical records. Medical records are primarily intended for patient care and not for research purposes. Second, no standardized information was available on lifestyle factors, such as dietary habits and physical activity. As a result, the interaction between these environmental factors and genetic variants could not be studied.
Third, our study included patients that were referred to a lipid clinic. In theory, patients with the most detrimental genetic profiles might have died before referral. Therefore, genetic polymorphisms associated with more severe CVD or early death could have been missed, leading to an underestimation of the risk. Fourth, 11 of the 65 polymorphisms studied were not in Hardy–Weinberg equilibrium. The exact reason for the deviation is not known and we can only speculate on this. Importantly, the deviation from Hardy–Weinberg equilibrium is not because of mixed ethnic groups, because the Dutch population is known to be a homogenous one. More than 99% of our patients were white and patients were randomly selected from all over the country. In addition, most deviations were caused by an excess of heterozygotes, which makes genotyping errors unlikely. Furthermore, the accuracy of genotyping in 500 randomly selected DNA samples was assessed by re-analysis of several polymorphisms in three genes, revealing that <0.5% of the results were discordant. However, it must be stressed that these 3 re-analyzed polymorphisms were not among the polymorphisms that were not in Hardy–Weinberg equilibrium. Finally, and most importantly, the 11 SNPs showed the same deviation from Hardy–Weinberg equilibrium in patients with CVD as well as in patients without CVD; therefore, we are of the opinion that this has not affected our results significantly. Fifth, it should be stressed that by the nature of this exploratory study the impact of the polymorphisms on intermediate end points such as coagulation factors (eg, prothrombin), blood pressure levels, angiotensinogen levels, and apolipoprotein levels, could not be assessed in the present study. Therefore, we could not determine the relationship of these intermediate end points with CVD, nor the subsequent pathological mechanisms involved in the development of CVD.
Our results are in line with earlier findings. The prothrombin mutation G20210A leads to increased plasma levels of prothrombin and an increased risk of venous thrombosis.19 Many studies have focused on the causative role of this mutation in arterial disease, often with negative results.20,21 However, in a comprehensive meta-analysis investigating the possible link between the G20210A prothrombin gene variant and different forms of premature ischemic heart disease in 12 043 patients, the G20210A polymorphism was shown to be a significant risk factor for myocardial infarction at a young age.22 In patients aged 55 years and younger, the odds ratio was 1.77 (95% CI, 1.16 to 3.42); in those aged 45 years and younger, the odds ratio was 2.30 (95% CI, 1.27 to 4.59). These odds ratios are similar to the HRs found in our study. In another meta-analysis, the association between the G20210A variant and arterial circulatory events was confirmed.23 In addition, 2 other large studies showed a pronounced effect of this polymorphism on CVD events in dyslipidemic patients.24,25 Although the relationship between this prothrombin mutation and CVD remains the subject of debate in non-FH populations, these findings support the strong association between G20210A carriership and premature CVD risk in our FH patients who have severe dyslipidemia from birth onwards.
Recently, the M235T polymorphism in the angiotensinogen gene was associated with increased levels of angiotensinogen and a corresponding increase in the risk of hypertension, as assessed in 45 267 individuals.26 However, a relationship between this variant and CVD risk has never been established in a large-scale study. Again, in our FH patients, the combined effect of elevated angiotensinogen and LDL-C levels might explain the increased risk of CVD.
We identified 2 common variants in the ApoA1-C3-A4-A5 gene cluster to be suggestively associated with CVD risk (Thr347Ser in apoA4 and C1100T in apoC3). Both appear to have (statistically) independent effects, because they remained significant in the multifactorial model and the estimated HRs of the univariate and multifactorial model were almost identical for both variants (data not shown). Apolipoproteins play a central role in lipid metabolism and this cluster of apolipoprotein genes has been identified as a locus with significant effects on triglyceride levels.27–29 Remarkably, in the first prospective study on genetic variants in this gene cluster and CVD risk in the general population, exactly the same 2 polymorphisms were associated with an increase and decrease in risk, respectively.30 However, in their multifactorial model, which also included other known risk factors such as body mass index and blood pressure, only the Thr347Ser variant remained significantly associated with CVD risk. They concluded that the apparent protective effect of C1100T in the apoC3 gene is most likely caused by strong linkage disequilibrium across the gene cluster. When body mass index, blood pressure, total cholesterol, and triglycerides were also added to our multifactorial model, the effect of the Thr347Ser variant was no longer significant (data not shown). However, this apparent discrepancy could be caused by differences in genotype frequencies between the 2 populations. Furthermore, in both populations the inference of functionality for these variants was hampered by the almost complete absence of patients homozygous for both the 347Ser and the 1100T allele.
Many studies support the antiatherogenic role of paraoxonase-1, an HDL-associated enzyme.31 The protective effect is generally considered to be caused by the ability of paraoxonase-1 to attenuate oxidative modification of lipoprotein particles. Paraoxonase-2, a family member of paraoxonase-1, is reported to have antioxidant properties and antiatherogenic capacities as well, although its exact physiological role is still unknown.31 Our results confirm a relationship between paraoxonase-2 and CVD, and are in concordance with the findings of an earlier study in FH patients, in which homozygotes for the Ser311Cys variant of paraoxonase-2 seem to be protected against CVD.32
Considering the limitations of association studies, we suggest our results should be replicated in addition to being performed in prospective studies of large and well-defined FH populations, in which genetic and environmental modifiers should be carefully monitored. The results of this hypothesis-generating study constitute a basis for further research and form a step forward in the unraveling of the underlying mechanisms of CVD in FH.
Acknowledgments
This study was supported by a grant of the Netherlands Heart Foundation (98/165). Kastelein is an established investigator of the Netherlands Heart Foundation (grant 2000D039). We are indebted to RMS personnel who supported this study. We acknowledge the members of the independent adjudication committee: R. J. G. Peters, MD, PhD, cardiologist; J. Stam, MD, PhD, neurologist; and D. Legemate, MD, PhD, vascular surgeon. We thank all the patients and the specialists of the participating lipid clinics.
References
Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver CR, et al, eds. The metabolic basis of inherited disease. New York, NY: McGraw-Hill; 2001: 2863–2913.
Risk of fatal coronary heart disease in familial hypercholesterolaemia. Scientific Steering Committee on behalf of the Simon Broome Register Group. BMJ. 1991; 303: 893–896.
Sijbrands EJ, Westendorp RG, Defesche JC, de Meier PH, Smelt AH, Kastelein JJ. Mortality over two centuries in large pedigree with familial hypercholesterolaemia: family tree mortality study. BMJ. 2001; 322: 1019–1023.
Ferrieres J, Lambert J, Lussier-Cacan S, Davignon J. Coronary artery disease in heterozygous familial hypercholesterolemia patients with the same LDL receptor gene mutation. Circulation. 1995; 92: 290–295.
Sauvage Nolting PR, Defesche JC, Buirma RJ, Hutten BA, Lansberg PJ, Kastelein JJ. Prevalence and significance of cardiovascular risk factors in a large cohort of patients with familial hypercholesterolaemia. J Intern Med. 2003; 253: 161–168.
Umans-Eckenhausen MA, Sijbrands EJ, Kastelein JJ, Defesche JC. Low-density lipoprotein receptor gene mutations and cardiovascular risk in a large genetic cascade screening population. Circulation. 2002; 106: 3031–3036.
Jansen AC, Aalst-Cohen ES, Tanck MW, Trip MD, Lansberg PJ, Liem AH, van Lennep HW, Sijbrands EJ, Kastelein JJ. The contribution of classical risk factors to cardiovascular disease in familial hypercholesterolaemia: data in 2400 patients. J Intern Med. 2004; 256: 482–490.
Risch N, Merikangas K. The future of genetic studies of complex human diseases. Science. 1996; 273: 1516–1517.
Sijbrands EJ, Westendorp RG, Paola LM, Havekes LM, Frants RR, Kastelein JJ, Smelt AH. Additional risk factors influence excess mortality in heterozygous familial hypercholesterolaemia. Atherosclerosis. 2000; 149: 421–425.
Umans-Eckenhausen MA, Defesche JC, Sijbrands EJ, Scheerder RL, Kastelein JJ. Review of first 5 years of screening for familial hypercholesterolaemia in the Netherlands. Lancet. 2001; 357: 165–168.
Jansen AC, Aalst-Cohen ES, Hutten BA, Buller HR, Kastelein JJ, Prins MH. Guidelines were developed for data collection from medical records for use in retrospective analyses. J Clin Epidemiol. 2005; 58: 269–274.
Mortality in treated heterozygous familial hypercholesterolaemia: implications for clinical management. Scientific Steering Committee on behalf of the Simon Broome Register Group. Atherosclerosis. 1999; 142: 105–112.
US MedPed program. http:/www.medped.org. 2005.
Defesche J.C. Familial Hypercholesterolemia. In: Betteridge DJ, ed. Lipids and Vascular Disease. London, UK: Martin Dunitz Ltd; 2000: 65–76.
Cheng S, Grow MA, Pallaud C, Klitz W, Erlich HA, Visvikis S, Chen JJ, Pullinger CR, Malloy MJ, Siest G, Kane JP. A multilocus genotyping assay for candidate markers of cardiovascular disease risk. Genome Res. 1999; 9: 936–949.
Slatkin M, Excoffier L. Testing for linkage disequilibrium in genotypic data using the Expectation-Maximization algorithm. Heredity. 1996; 76: 377–383.
Lewontin RC. The interaction of selection and linkage. II. Optimum models. Genetics. 1964; 50: 757–782.
Rea TD, Heckbert SR, Kaplan RC, Smith NL, Lemaitre RN, Psaty BM. Smoking status and risk for recurrent coronary events after myocardial infarction. Ann Intern Med. 2002; 137: 494–500.
Poort SR, Rosendaal FR, Reitsma PH, Bertina RM. A common genetic variation in the 3'-untranslated region of the prothrombin gene is associated with elevated plasma prothrombin levels and an increase in venous thrombosis. Blood. 1996; 88: 3698–3703.
No evidence of association between prothrombotic gene polymorphisms and the development of acute myocardial infarction at a young age. Circulation. 2003; 107: 1117–1122.
Yamada Y, Izawa H, Ichihara S, Takatsu F, Ishihara H, Hirayama H, Sone T, Tanaka M, Yokota M. Prediction of the risk of myocardial infarction from polymorphisms in candidate genes. N Engl J Med. 2002; 347: 1916–1923.
Burzotta F, Paciaroni K, De S V, Crea F, Maseri A, Leone G, Andreotti F. G20210A prothrombin gene polymorphism and coronary ischaemic syndromes: a phenotype-specific meta-analysis of 12 034 subjects. Heart. 2004; 90: 82–86.
Kim RJ, Becker RC. Association between factor V Leiden, prothrombin G20210A, and methylenetetrahydrofolate reductase C677T mutations and events of the arterial circulatory system: a meta-analysis of published studies. Am Heart J. 2003; 146: 948–957.
Doggen CJ, Cats VM, Bertina RM, Rosendaal FR. Interaction of coagulation defects and cardiovascular risk factors: increased risk of myocardial infarction associated with factor V Leiden or prothrombin 20210A. Circulation. 1998; 97: 1037–1041.
Gardemann A, Arsic T, Katz N, Tillmanns H, Hehrlein FW, Haberbosch W. The factor II G20210A and factor V G1691A gene transitions and coronary heart disease. Thromb Haemost. 1999; 81: 208–213.
Sethi AA, Nordestgaard BG, Tybjaerg-Hansen A. Angiotensinogen gene polymorphism, plasma angiotensinogen, and risk of hypertension and ischemic heart disease: a meta-analysis. Arterioscler Thromb Vasc Biol. 2003; 23: 1269–1275.
Groenendijk M, Cantor RM, de Bruin TW, Dallinga-Thie GM. The apoAI-CIII-AIV gene cluster. Atherosclerosis. 2001; 157: 1–11.
Talmud PJ, Hawe E, Martin S, Olivier M, Miller GJ, Rubin EM, Pennacchio LA, Humphries SE. Relative contribution of variation within the APOC3/A4/A5 gene cluster in determining plasma triglycerides. Hum Mol Genet. 2002; 11: 3039–3046.
Olivier M, Wang X, Cole R, Gau B, Kim J, Rubin EM, Pennacchio LA. Haplotype analysis of the apolipoprotein gene cluster on human chromosome 11. Genomics. 2004; 83: 912–923.
Wong WM, Hawe E, Li LK, Miller GJ, Nicaud V, Pennacchio LA, Humphries SE, Talmud PJ. Apolipoprotein AIV gene variant S347 is associated with increased risk of coronary heart disease and lower plasma apolipoprotein AIV levels. Circ Res. 2003; 92: 969–975.
Ng CJ, Shih DM, Hama SY, Villa N, Navab M, Reddy ST. The paraoxonase gene family and atherosclerosis. Free Radic Biol Med. 2005; 38: 153–163.
Leus FR, Zwart M, Kastelein JJ, Voorbij HA. PON2 gene variants are associated with clinical manifestations of cardiovascular disease in familial hypercholesterolemia patients. Atherosclerosis. 2001; 154: 641–649.(Angelique C.M. Jansen; Em)