UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
Familial hypercholesterolemia. The determination of phenotype
Jansen, A.C.M.
Publication date
2003
Link to publication
Citation for published version (APA):
Jansen, A. C. M. (2003). Familial hypercholesterolemia. The determination of phenotype.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.
Geneticc determinants of plasma
HDL-cholesteroll levels in familial
hypercholesterolemia a
AA study in 1000 FH heterozygotes
Emilyy S van Aalst-Cohen
1, Angelique CM Jansen
1, S Matthijs Boekholdt
2,
Michaell WT Tanck
3, Joep C Defesche
1, Jan Albert Kuivenhoven
1and
Johnn JP Kastelein
1Departmentss of Vascular Medicine', Cardiology2, and Clinical Epidemiology and
Biostatistics3,, Academic Medical Center, Amsterdam, the Netherlands
Abstract t
Background d
Familiall hypercholesterolemia (FH) is a common, hereditary disease, characterized by elevated levelss of plasma low-density lipoprotein cholesterol and premature cardiovascular disease (CVD).. Although the cause of FH is monogenic, there is wide variation in the onset and severityy of atherosclerotic disease. Additional environmental and genetic risk factors are presumedd to influence this clinical phenotype. In this respect, low high-density lipoprotein cholesteroll (HDL -C) has been shown to be a strong and independent risk factor. The current studyy focuses on the extent to which common genetic variants can explain the variation of
HDL-CC plasma levels in FH patients.
Methods s
AA well-characterized cohort of heterozygous FH patients was genotyped for common variants inn the genes encoding for ATP binding cassette transporter A1 (ABCA1), apolipoprotein (apo)) AIV, apoCIII, apoE, cholesteryl transfer ester protein (CETP), hepatic lipase (HL), lipoproteinn lipase (LPL), and two paraoxonases (PON1 and PON2).
Results s
Multiplee linear regression showed that, together, these variants explain 3.9% of the variation of HDL-CC plasma levels. When significant two-way interactions between the genetic variants were alsoo taken into account, the explained variation rose to 12.5%. In a regression mode! that also incorporatedd gender, smoking, alcohol use, body mass index and concomitant beta-blocker use ass covariates, the explained variation of HDL-C plasma levels even increased to 32.5%.
Conclusion n
Thiss study provides direct evidence that multiple, modestly penetrant, but highly prevalent geneticc variants can explain a substantial part of the variation of HDL-C plasma levels in a representativee cohort of heterozygous FH patients.
Introduction n
Familiall hypercholesterolemia (FH) is a common, hereditary disease, characterized by elevated levelss of plasma low-density lipoprotein cholesterol (LDL-C) and premature cardiovascular diseasee (CVD). FH is caused by genetic defects in the low-density-lipoprotein receptor (LDL-R)) gene, leading to an insufficient uptake of LDL-C from the circulation.' Characteristically, thee mean age of onset of CVD is between 40 and 45 years in male FH patients and in
femalee FH patients this occurs 10 years later.1-2 Although the cause of FH is monogenic,
theree is wide variation in the onset and severity of atherosclerotic disease in these ^atients 3
Thiss is often suggested to be related to environmental and additional genetic risk factors. Inn this respect, low high-density-lipoprotein cholesterol (HDL-C) has been shown an
independentt risk factor in FH patients.47 Therefore, early identification of FH patients with
loww HDL-C, in addition to the development of therapeutic strategies to specifically raise
HDL-CC in these high risk patients may be warranted.8 HDL-C levels, however, are affected
byy gender, obesity, smoking, diet, alcohol consumption and exercise in addition to numerous
geneticc factors9 and the relative contribution of these factors to HDL-C levels in FH patients
iss not known.10
Thee current study focuses on common single nucleotide polymorphisms (SNP's) in genes thatt could potentially affect HDL-C plasma levels in FH patients. Starting with apolipoprotein All (ApoAl), the molecule that is of the utmost importance to HDL metabolism, this introductionn will shortly describe the investigated genes, a number of genetic variants therein andd several reported effects on HDL-C. Special attention will be given to studies describing thesee genetic variants in FH.
ApoAll is the primary structural apolipoprotein of the HDL particle and crucial to overall reversee cholesterol transport (RCT). The ApoAl gene is compiexed with the genes encoding apolipoproteinss Clll, AIV and AV on chromosome 11, referred to as the ApoAl-C3-A4-A5 genee cluster. The impact of ApoAl genetic variants on HDL is variable. Mutations in the ApoAll gene mostly underlie familial hypoalphalipoproteinemia, in which patients exhibit
isolatedd low HDL-C levels and an increased incidence of premature CVD.11 In contrast, the
ApoAll Miiano mutant has been associated with a reduced incidence of coronary heart disease,
despitee very low HDL-C levels.12
Thee ATP binding cassette transporter 1 (ABCA1), located in cellular membranes, mediates cholesteroll efflux from macrophages to nascent HDL. Common variants in the ABCA1 gene havee recently been shown to significantly affect plasma HDL-C levels in the general
withh decreased triglycerides, increased HDL-C and a reduced risk of CAD in both the general
populationn as well as in FH patients.M
Thee cholesteryl ester transfer protein (CETP) is another key role player in RCT. This protein mediatess the transfer of cholesteryl esters from HDL-C to triglyceride-rich particles in exchange forr triglycerides. Variations in the CETP gene are determinants of CETP concentration, CETP
activityy and HDL-C, but their relation with CVD remains a controversial topic.8 Subjects with
hypercholesterolemiaa have been reported to have higher CETP activities, which could
contributee to the lower HDL-C levels and increased CVD risk seen in some of these patients.15
Inn line with these findings, a study of 101 FH patients demonstrated the association of the CETPP Taq1 B2 allele with a less atherogenic lipid profile, in terms of lower LDL-C and higher HDL-CC levels. Moreover, in this cohort, the B2 allele was associated with a lower prevalence
off arcus cornealis, xanthomas and CVD.16
Hepaticc lipase (HL) is an enzyme that catalyzes the hydrolysis of triglycerides and phospholipids.. Its catalytic activity contributes to the remodelling of LDL and HDL, resulting inn smaller, denser particles. In addition, HL promotes hepatic uptake of LDL and HDL.
Increasedd HL activity is often associated with low HDL-C levels17 while human HL deficiency
iss associated with hyperalphalipoproteinemia.18 However, the exact role of HL in RCT and
thee atherogenesis remains to be elucidated.
Lipoproteinn lipase (LPL) plays a key role in triglyceride metabolism. It hydrolyses triglycerides inn ApoB-containing lipoproteins and thereby generates constituents, i.e. phospholipids and apolipoproteinss that contribute to the HDL pool. Impaired LPL activity, due to LPL gene
defects,, is associated with low HDL-C levels.19 Two common LPL gene mutations, Asp9Asn
andd Asn291Ser, were shown to significantly modulate HDL-C and triglyceride levels20,
however,, no increased risk for coronary heart disease could be attributed to these mutations. Inn contrast, these mutations induced significant biochemical changes and, importantly, conferredd a significantly greater risk for CVD in FH.21'22
ApoEE is one of the major ligand proteins needed for clearance of triglyceride-rich remnant lipoproteinss by liver receptors. Three allelic variants of the apolipoprotein E gene, E2, E3 and E4,, encode for different ApoE isoforms. Patients with the E2/E2 genotype are characterized byy higher concentrations of triglyceride-rich lipoproteins and lower LDL-C levels, whereas oppositee effects are found in individuals with the E4/E4 genotype. In earlier studies, ApoE has nott been found to significantly contribute to the variation HDL-C levels nor the risk of CVD in
FHH patients.23 According to a recent study conducted in 450 children with FH, the apolipoprotein
otherr alleles, which might confer additional risk in those children.24
Paraoxonasee (PON) is an HDL-associated enzyme that is, in part, responsible for the protection off LDL against oxidation. This implicates members of the paraoxonase gene family in the developmentt of atherosclerosis in FH subjects. The first study on the role of PON1 in FH showedd an association between a PON1 gene variant and carotid wall thickness." This associationn could not be confirmed by a second study on the association between this
mutationn and CVD.26 In the latter study, however, another common polymorphism in the
PON22 gene was associated with CVD These allelic variants of PON1 and PON2 were not associatedd with HDL-C levels.
Itt is generally assumed that the combined effects of multiple common genetic variants
mightt explain a large part of HDL-C variation.27 However, surprisingly, the current literature
doess not provide direct evidence to support this hypothesis. Accordingly, it was our objective too determine the contribution of common genetic variants in a number of HDL-C related genee loci to the variation of HDL-C plasma levels in FH. For this purpose, we genotyped a cohortt of Caucasian heterozygous FH patients for common variants in the genes encoding forr ATP binding cassette transporter A1 (ABCA1), apolipoprotein (apo) AIV, apoCIII, apoE, cholesteryll transfer ester protein (CETP), hepatic lipase (HL), lipoprotein lipase (LPL), and twoo paraoxonases (PON1 and PON2). Subsequently, we quantified the extent to which thesee genetic variants explained the variation of HDL-C plasma levels in these patients.
Methods s
Studyy design a n d study population
Thee present investigation was a cross-sectional, multicenter, cohort study. Routinely, Lipid Clinicss in The Netherlands submit DNA samples from clinically diagnosed FH patients to a
centrall laboratory for LDL-R mutation analysis/3 This laboratory, located at the Academic
Medicall Center of the University of Amsterdam, manages a DNA-bank database, which currently containss over 9300 samples. We randomly selected 4000 hypercholesterolemic patients from thiss database who were subsequently screened for the FH diagnostic criteria as described below.. These patients had been referred from 27 Lipid Clinics throughout The Netherlands.
Dataa collection
Phenotypicall data were acquired by reviewing these patient's medical records by a well-trainedd team of 13 data collectors. Data collected from the medical record was supplemented byy information obtained from general practitioners, the patients themselves, and hospitals
thatt patients had visited formerly. In order to obtain consistent datasets, a number of qualityy measures were taken. All data collectors underwent extensive training; handbooks withh clinical definitions were used and inter-observer studies were carried out. Data collectors weree blinded for DNA genotyping results.
Clinicall data
FHH diagnostic criteria
Internationallyy uniform diagnostic criteria to identify FH patients are lacking and we therefore usedd a combination of established clinical diagnostic criteria from the US (the MedPed
criteria)25,, the UK (the Simon Broome Register criteria)30 and the Netherlands {the Dutch
Lipidd Clinic Network criteria).31 Male and female patients, 18 years and older, were included
inn the study when they met the following criteria: (I) the presence of a documented LDL-R
mutation,, or (II) an LDL-C level above the 95lh percentile for sex and age, in combination
withh (a) the presence of typical tendon xanthomas in the patient or in a first degree relative, orr (b) an LDL-C level above the 95th percentile for age and sex in a first degree relative, or (c) provenn CAD in the patient or in a first degree relative under the age of 60 years. Exclusion criteriaa were secondary hypercholesterolemia due to diabetes mellitus, renal, liver or thyroid disease,, and excessive alcohol consumption.
Additionall data on classical risk factors for CVD, family history of CVD, medication, physical examinations,, laboratory parameters, and extensive information on CVD were collected fromm the patient's medical records. All patients gave informed consent and the Ethics Institutionall Review Board of each participating hospital approved the protocol.
Classicall risk factors
Malee gender, age, smoking, body mass index (BMI) and the presence of hypertension and diabetess mellitus were considered classical risk factors. Smoking was defined as ever having
smokedd (yes/no). Body mass index was calculated from height and length (kg/m2).
Hypertensionn was defined when the diagnosis had been made by a physician and when antihypertensivee medication was prescribed, or if three consecutive measurements of blood pressuree were >140 mmHg systolic or > 90 mmHg diastolic. Diabetes mellitus was defined whenn the diagnosis had been made and medication (insulin or oral anti-diabetics) was prescribed,, or by a fasting plasma glucose of > 6.9 mmol/L.
Additionall factors known to influence plasma HDL-C levels
Inn addition to gender, smoking and BMI, data on alcohol use and concomitant beta-blocker usee were collected from the medical records. Alcohol use was defined by alcohol use at the timee of the first visit to the Lipid Clinic (yes/no). Concomitant beta-blocker use was defined
ass beta-blocker use at the time of determination of plasma HDL-C concentration (yes/no).
Cardiovascularr disease
CVDD was diagnosed by the presence of at least one of the following: (I) a myocardial infarctionn (Ml), proven by at least two of the following: (a) classical symptoms (>15 minutes), (b)) specific ECG abnormalities, (c) elevated cardiac enzymes {> 2 times local upper limit of normal);; (II) percutaneous coronary intervention or other invasive procedures; (III) coronary arteryy bypass grafting ; (IV) angina pectoris, diagnosed as classical symptoms in combination withh at least one unequivocal result of one of the following: (a) exercise test, (h) nurlpgr scintigram,, (c) dobutamine stress ultrasound, (d) a more than 70% stenosis on a coronary angiogram;; (V) ischemic stroke, focal neurological symptoms lasting for more than 24 hourss and demonstrated by CT- or MRI scan (VI) documented transient ischemic attack; (VII)) peripheral arterial bypass graft; (VIII) peripheral percutaneous transluminal angioplasty orr other percutaneous invasive intervention; (IX) (partial) amputation of an extremity due too peripheral arterial disease; (X) intermittent claudication defined as classical symptoms in combinationn with at least one unequivocal result of one of the following: (a) ankle/arm indexx <0.9 or (b) a stenosis <>50%) on angiogram or duplex scan.
Iff information on CVD did not strictly fulfil the above mentioned criteria, or if any suspect history,, symptoms or diagnostic evaluations were found in the record, the case was presented too an independent adjudication committee consisting of a cardiologist, a neurologist and a vascularr surgeon.
Laboratoryy analysis
Alll laboratory parameters were measured in fasting blood samples. Lipid levels, as stated in thee medical record, were determined after at least 6 weeks of withdrawal of any lipid-lowering medication.. Plasma total cholesterol, HDL-C, and triglycerides (TG) were measured by standard enzymaticc methods. LDL-C concentrations were calculated by means of the Friedewald
formula.322 Lipoprotein (a) concentrations were determined with immunonephelometric
methods.333 Homocysteine concentrations were measured using high-performance liquid
chromotography.344 Mutations in the LDL-R gene were assessed by methods described
previously.35 5
Selectionn of genetic variants
Basedd on the current literature, we selected 25 DNA polymorphisms in 10 genes that are knownn to affect HDL-C levels or that could potentially affect HDL-C levels. 23 bi-allelic SNP's inn 9 HDL-C related genes were used for the present analysis: ABCA1 ABCA1 ABCA1R2]gK,, ApoA4Thr3d7Ser, ApoA4G|n36QH|s, ApoC3C(.6dl)A, A p o C 3Q W ApoC3THl55jC,
A p o C 3c l ] M T,, ApoC3C 3 1 7 W, ApoC3T3206G, CETPM 2 9 ) A, CETPTgq]B, CETPC(.63m, CETPlte40Wai, HLc;_ 5ui T.. LPLT|.93)G, LPLflsp9A5n, LPLAw291Ser, LPLSciM7Sïop, P 0 N 1M e l W l e u, PON1G | n l 9 2 A t g ( and PON2Ser311Cys.
Inn addition, w e included the tri-allelic ApoE polymorphism that is determined by a r g i n i n e / cysteinee variations at codons 112 and 158 in our analysis.
Geneticc analyses
Genomicc DNA was extracted from peripheral blood leukocytes according to a standard protocol.. Genotyping was performed by Roche Molecular Systems, CA, USA. For all bi-allelicc sites, genotypes were generated using a PCR- and immobilized probe assay, as
describedd previously by Cheng et al.36 Researchers and laboratory personnel had no access
too neither identifiable nor clinical information.
Statisticall analysis
Linearr regression was used to estimate the effect of individual genotypes on HDL-C concentrationn without lipid-lowering medication. Multiple linear regression was used to estimatee the total amount of HDL-C variation explained by multiple genotypes, and by multiplee genotypes plus a set of five covariates, i.e.: gender, smoking, alcohol use, BMI and concomitantt beta-blocker use. Genotypes were entered into the model as categorical variables. .
Twoo different multiple regression models were tested: (1) a backward regression model thatt included the main effects only of all genotypes in the initial model; and (2) a combined-effectss model that included all main effects and a selection of two-way interactions {gene-gene,, gene-environment and environment-environment interactions). All two-way interaction termss were assessed, and those that were significant after adjustment for the main effects off all genotypes were included in the initial model. Backward selection was then performed onn the main effects of the genotypes not included in the selected interactions. A p-value of << 0.05 was considered significant. For both models, the percentage variation in baseline
HDL-CC concentration explained by the final model (R2) was calculated. Statistical analyses
weree performed using SPSS software (version 11.0, Chicago, Illinois).
Results s
Afterr thorough review of 4000 medical records and application of the diagnostic criteria, 24000 patients were defined as having FH. From these, a total of 1002 individuals could be completelyy genotyped for 25 selected polymorphisms. These 1002 subjects did not differ significantlyy from the original cohort of 2400 patients with regards to any of the clinical characteristics.. These characteristics of the study population are summarized in table 1. The
Tablee 1 Clinical characteristics of the genotyped FH patients Numberr of subjects
Demographics s
Malee gender (%)
Agee at first visit to Lipid Clinic (year) Agee at last visit to Lipid Clinic (year) Totall Lipid Clinic follow-up (years)
Riskk factors
Smoking,, ever (%)
Alcoholl use at first visit to Lipid Clinic (%) Hypertension n
Diabetess mellitus (%)
First-degreee family history of premature CVD (%)
Physicall examination
BMII (kg/m2)
Systolicc blood pressure (mmHg) Diastolicc blood pressure (mmHg) Tendonn xanthomas (%)
Laboratoryy parameters
Totall cholesterol (mmol/L) LDLL cholesterol (mmol/L) HDLL cholesterol (mmol/L) Triglyceridess (mmol/L) Glucosee (mmol/L) HbA1C(%) ) Lp(a)) (mg/L) Homocysteinee (micromol/L) Cardiovascularr disease
Totall CVD (including mortality) by last visit to Lipid CI nicc (%) Totall coronary artery disease by last visit to Lipid Clinic (%) Totall cerebrovascular disease by last visit to Lipid Clinic (%) Totall peripheral ischemic disease by last visit to Lipid Clinic (% ) Valuess are given as mean levels standard deviation
glucose,, HbAIC, Lp{a), and homocysteine are giver brackets.. CVD indicates cardiovascular disease; BMI,
exceptt wh< ass median bodyy mass 1002 2 49.0 0 44.11 ( 12.3) 48.77 ( 12.7) 3.77 [1.5-6.6] 73.9 9 75.7 7 9.4 4 4.8 8 54.3 3 25.11 ) 1344 ( 18) 822 ( 10) 44% % 9.366 ( 1.94) 7.333 ( 1.95} 1.200 ) 1.577 [1,11-2.23] 5.00 [4.6-5.4] 5.55 [5.1-6.1] 1666 [67-460] 11.11 [9.0-13.4] 28.0 0 23.9 9 3.0 0 4.4 4
=ree given as percentages. Triglycerides, withh the interquartile range between ndex;; Lp{a), lipoprotein(a).
averagee HDL-C concentration without lipid-lowering medication was 1.20 + 0.35 mmol/l. Genotypee counts and mean HDL-C plasma levels according to genotype are presented for thee 23 bi-allelic SNP's in table 2.
Forr the tri-atlelic ApoE polymorphism, the E3E3, E3E4, E2E3, E2E4, E4E4, and E2E2 genotypes weree defined in 574, 286, 64, 30, 46, and 2 subjects, respectively. HDL-C plasma levels for E3E3,, E3E4, E2E3, E2E4, E4E4, and E2E2 genotypes were 1.20 , 1.20 , 1.21
Tablee 2 Genotype counts and HDL-C plasma levels for alt investigated SNP's
Commonn allele homozygotes Heterozygotes
ABCA1 1 ABCA1 1 ABCA1 1 ApoA4 4 ApoA4 4 ApoC3 3 ApoC3 3 ApoC3 3 ApoC3 3 ApoC3 3 ApoC3 3 CETP P CETP P CETP P CETP P HL L LPL L LPL L LPL L LPL L PON1 1 PON1 1 PON; ; C69T T ins319 9 R219K K Thr347Ser r Gln360His s C(-641)A A C(-482)T T T(-455)C C C1100T T T3206G G C3175G G C(-629)A A C(-631)A A He405Val l TaqlB B C(-514)T T T(-93)G G Asp9Asn n Asn291Ser r Ser447X X Met55Leu u Gin192Arg g Ser311Cys s mean n 1.20 0 1.21 1 1.18 8 1.21 1 1.20 0 1.21 1 1.20 0 1.20 0 1.22 2 1.19 9 1.20 0 1.16 6 1.20 0 1.20 0 1.15 5 1.18 8 1.20 0 1.20 0 1.20 0 1.20 0 1.20 0 1.20 0 1.19 9 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + SD D 0.37 7 0.34 4 0.32 2 0.35 5 0.35 5 0.33 3 0.33 3 0.33 3 0.35 5 0.32 2 0.34 4 0.32 2 0.35 5 0.35 5 0.32 2 0.34 4 0.35 5 0.35 5 0.35 5 0.35 5 0.35 5 0.34 4 0.35 5 n= = 451 1 778 8 544 4 679 9 847 7 373 3 524 4 388 8 535 5 396 6 782 2 262 2 865 5 454 4 320 0 602 2 959 9 965 5 934 4 828 8 427 7 470 0 590 0 mean n 1.20 0 1.17 7 1.24 4 1.18 8 1.19 9 1.20 0 1.20 0 1.21 1 1.18 8 1.21 1 1.18 8 1.20 0 1.17 7 1.20 0 1.21 1 1.23 3 1.09 9 1.10 0 1.08 8 1.23 3 1.19 9 1.20 0 1.22 2 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + SD D 0.33 3 0.35 5 0.38 8 0.35 5 0.32 2 0.36 6 0.36 6 0.36 6 0.36 6 0.37 7 0.36 6 0.34 4 0.34 4 0.35 5 0.34 4 0.36 6 0.26 6 0.26 6 0.29 9 0.33 3 0.34 4 0.35 5 0.34 4 n= = 429 9 210 0 391 1 271 1 152 2 468 8 395 5 460 0 378 8 428 8 206 6 513 3 131 1 436 6 504 4 339 9 42 2 36 6 67 7 160 0 449 9 432 2 348 8
HDL-CC levels are presented as mean + SD. ABCA1 indicates ATP-binding cassette A1; ApoA4, a pol ipo protein A4;; ApoC3, apolipoprotein C3; CETP, cholesteryl ester transfer protein; HL, hepatic lipase; LPL, lipoprotein lipase;; PON, paraoxonase.
Tablee 3 Explained variation of HDL-C plasma concentration
Mainn effects model Combined effects model
Variabless Main effects only Main effects + selectedd 2-way interactions
Genotypess 3.9% (10) 12.5% (13) Genotypess plus covariates 20.2% (42) 32.5% (72)
Revaluess for multivariate linear regression models incorporating all genetic variants only, and genetic variantss plus covariates (gender, smoking, alcohol, BMI and concomitant beta-blocker use). Values are presentedd for models that incorporate main effects only, and main effects plus 2-way interactions that contributedd significantly to the explained HDL-C variation. The number of variables per model is given betweenn parentheses.
Combinedd effect of genetic variants on HDL-C plasma levels
Withoutt covariates, the main-effects model consisted of five polymorphisms that together accountedd for 3.9% of the population variability in HDL-C plasma levels (Table 3). These five polymorphismss were CETPTaq|B, ABCA1R215K, LPLA=n29!^r, ApoC3cl00T and ApoA4Th,M,Ser.
Raree allele mean n 1.20 0 1.17 7 1.15 5 1,11 1 1.00 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 17 7 17 7 17 7 17 7 18 8 15 5 25 5 1.35 5 1.19 9 1.26 6 1.23 3 1.21 1 1.21 1 0.96 6 1.17 7 1.19 9 1.18 8 1.17 7 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
--+ --+ + + + + + + homozygotes s SD D 0.33 3 0.36 6 0.29 9 0.32 2 0.43 3 0.34 4 0.36 6 0.35 5 0.32 2 0.34 4 0.23 3 0.39 9 0.40 0 0.33 3 0.41 1 0.34 4
--0.32 2 0.32 2 0.33 3 0.36 6 n= = 122 2 14 4 67 7 52 2 3 3 161 1 83 3 154 4 89 9 178 8 14 4 227 7 6 6 112 2 178 8 61 1 1 1 1 1 1 1 14 4 126 6 100 0 64 4 p-value e 0.981 1 0.337 7 0.011 1 0.073 3 0.569 9 0.455 5 0.749 9 0.575 5 0.165 5 0.507 7 0.640 0 0.010 0 0.413 3 0.931 1 0.005 5 0.094 4 0.109 9 0.208 8 0.011 1 0.554 4 0.762 2 0.820 0 0.432 2 RR2 2 0 . 0 1 % % 0.2% % 0.9% % 0.5% % 0 . 1 % % 0.2% % 0 . 1 % % 0 . 1 % % 0.4% % 0 . 1 % % 0 . 1 % % 0.9% % 0.2% % 0 . 0 1 % % 1.1% % 0.5% % 0.4% % 0.3% % 0.9% % 0 . 1 % % 0 . 1 % % 0 . 0 1 % % 0.2% %
Thee combined-effects model explained 12.5% of HDL-C variation. In total, 12 main effects andd 5 gene-gene interaction effects remained in the combined-effects model. The interactions weree (ABCA1C69T x ApoA4Thr347Ser), (ABCA1C69T x ApoC3T3205G), (ApoC3CM82)T x CETPTsqie), (ApoC3T[. 455)cc x LPLSer447stop), and (ApoE x CETPTanJ with individual Revalues of 1.0,1.9,1.8, 2.0 and 1.6%,
respectively.. In addition to the main effects of the polymorphisms in the interaction terms, the mainn effects of ABCA1D11Q„ ApoC3riinnT, HL,tl„1T and LPL ^Q1C remained significant.
Whenn the five covariates were incorporated into the model, the main-effects model consisted off four polymorphisms (ABCA1 R219j<j H Lq w, LPLAsn291Ser and CETPTaqlB) explaining 20.2% of
HDL-CC variation. The combined-effects model could even explain 32.5% of HDL-C variation while containingg 18 main effects and eight gene-gene, two gene-environment and one
environment-environmentt interactions. The most important gene-gene interactions were (ABCA1C59Tx
ApoC3T3;06G),, ( A p o C 3w x CETPTaqlB), (LPLSer447Slop x PON2Ser311Cy5) and ( P O M ^ , 1Cys x CETPIW0Wal)
withh individual Revalues of 2.1, 1.8, 1.1 and 1.1%, respectively. The most significant
gene-environmentt interactions were (CETP|le40BVa| x alcohol use) and (PON1MetS5LDu x gender) with
Discussion n
Inn accordance with previous reports of FH patients4,6, low HDL-C in our cohort was a strong
andd independent risk factor for the development of CVD (data not shown). This clearly indicatess that an in-depth investigation of the determinants of these HDL-C levels in FH is warranted.. The present investigation provides evidence that the combined effects of common geneticc variants involved in HDL metabolism can explain 12.5% of the variation of HDL-C plasmaa levels. Moreover, when gender and several environmental factors were taken into
accountt a strikinn 32.5% of the variation in HDL-C levels could be exn!ained. Therefore
modulationn of these genes could be anticipated to have substantial effects on HDL-C in thesee individuals. This is an important finding in view of recent evidence that a 1 % increase
inn HDL-C can yield a 3% risk reduction of CVD.37 It should be noted, however, that
gene-genee and gene-environment interactions account for a substantial proportion of the overall 32.5%% variation in HDL-C. This indicates that targeting specific genes may only be efficacious inn the context of a specific metabolic background.
Severall reports have been published regarding such gene-environment interactions. In one
case-controll study conducted with 1474 patients enrolled in the ECTIM study38, it appeared
thatt LPLA5p9Asn mutation carriers solely develop an atherogenic lipid profile in the presence of a
metabolicc factor, such as obesity.39 Another case-control ECTIM sub-study, conducted with
6088 patients38, demonstrated an intriguing interaction between alcohol consumption and the
CETPTaa 1B polymorphism. 40
The effect of the B2 allele on plasma HDL-C levels was absent in patientss who drank little alcohol, but increased significantly with increasing alcohol consumption. Inn addition, the odds ratio for myocardial infarction was markedly decreased in heavy drinkers, butt only in the presence of a 82 allele. The interactions in the combined-effects model presented inn our study will be the focus of future research in this large FH cohort.
Severall aspects must be taken into account when interpreting the results of the present study.. To begin with, the present study focuses on variation in HDL-C levels and the contributionn thereof to the clinical expression in FH. Although the inverse relationship betweenn HDL-C levels and CVD risk is well founded, alterations in HDL regulating genes thatt result in high HDL-C levels have not consistently been associated with cardiovascular protection.. Several genetic variants have been reported to be associated with a lower risk
off CVD, but independent of HDL-C levels.41 A2 Therefore, studying the contribution of common
geneticc variants to HDL-C variation should ultimately be translated into their contribution to CVDD risk. Moreover, it has been suggested that not only the level of HDL-C, but also the
compositionn contributes to CVD risk.43 Unfortunately, HDL-C subclasses were not determined
variantss with CVD risk, as prospectively assessed over time.
Anotherr aspect that deserves further attention is the study design. Due to the retrospective designn and reliance on documentation in the medical records, no standardized information wass available regarding dietary habits and physical activity. Therefore we cannot estimate the contributionn of these environmental factors that are known to modulate HDL-C plasma levels. Finally,, statistical aspects of our study may require some explanation. Statistical analysis of geneticc population studies is still in development. A major issue in this field is the interpretation off datasets with a large number of genetic variables. These analyses have a tendency to be
' r i u o r ff i t + o W i o t h o n i i m h n r n f o v n j a | r i | r i n w a n a h joe ^ n f i f Q g i ' l l g S o r e v e n c
' i r n a c c o t ; t h e n u m b e r
off observations. In the current analysis, we used 'conventional' multiple linear regression. Wee incorporated main effects of genotypes and, subsequently, we selected only those interactionn terms that contributed significantly to the explained HDL-C variation after adjustmentt for all other variables in the model. Our dataset comprised 1002 individuals, andd the total number of variables in the final combined-effects model was 72. Fortunately,
thiss is substantially lower than the 1:10 ratio that is generally considered acceptable.44
Conclusion n
inn hypercholesterolemia cohorts, such as FH patients, the variation in HDL-C levels is thought too contribute significantly to the overall risk in these individuals. However, little is known aboutt the actual contribution of genes and environment to HDL in FH. Understanding HDL metabolismm and RCT may lead to improved cardiovascular risk assessment of FH patients andd patients in the general population at high risk for CVD. In addition, HDL-associated genee products may be promising targets for raising HDL-C levels in these high-risk patients inn order to protect against atherosclerosis.
Acknowledgments s
Thiss study was supported by a grant of the Netherlands Heart Foundation (98/165). J.J.P. Kasteleinn is an established investigator of the Netherlands Heart Foundation (grant D039/ 66510).. We are indebted to Suzanne Cheng and Jia Li of Roche Molecular Systems, Alameda, California,, USA, for performing the genotyping and generously providing research reagents. Wee thank the members of the independent adjudication committee for their expert advice: Dr.. R.J.G. Peters, cardiologist, Prof. Dr. J. Stam, neurologist and Prof. Dr. D. Legemate, vascularr surgeon. We thank all patients who participated in the study and the specialists of thee participating Lipid Clinics throughout the Netherlands.
References s
11 Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia.. In: Scriver CR, Beaudett AL, Sly WS, Alle D, editors. The metabolicc basis of inherited disease. New York:: McGraw-Hill, 2001: 2863-913. 2.. Risk of fatal coronary heart disease in
familiall hypercholesterolaemia. Scientific Steeringg Committee on behalf of the Simon Broomee Register Group. BMJ.
1991;303:893-896. .
3.. Sijbrands EJ, Westendorp RG, Defesche JC ett al. Mortality over two centuries in large pedigreee with familial
hypercholesterolaemia:: family tree mortality study.. 8MJ, 2001;322:1019-1023.
4.. Hill JS, Hayden MR, Frohlich J, Pritchard PH. Geneticc and environmental factors affecting thee incidence of coronary artery disease in heterozygouss familial hypercholesterolemia. Arteriosclerr Thromb. 1991; 11:290-297. 5.. Tato F, Keller C, Schuster H et al. Relation
off lipoprotein(a) to coronary heart disease andd duplexsonographic findings of the carotidd arteries in heterozygous familial hypercholesterolemia.. Atherosclerosis.
1993;101:69-77. .
6.. Ferrieres J, Lambert J, Lussier-Cacan S, Davignonn J. Coronary artery disease in heterozygouss familial hypercholesterolemia patientss with the same LDL receptor gene mutation.. Circulation. 1995;92:290-295 7.. Real JT, Chaves FJ, Martinez-Uso I et al.
Importancee of HDL cholesterol levels and thee total/ HDL cholesterol ratio as a risk factorr for coronary heart disease in molecularlyy defined heterozygous familial hypercholesterolaemia.. Eur Heart J. 2001;22:465-471. .
8.. de Grooth GJ, Kuivenhoven JA, Stalenhoef AFF et al. Efficacy and safety of a novel cholesteryll ester transfer protein inhibitor, JTT-705,, in humans: a randomized phase II dose-responsee study. Circulation.
2002;105:2159-2165. .
9.. Rader DJ. High-Density Lipoproteins and Atherosclerosis.. Am J Cardiol.
2002;90(Suppl):62i-70i. .
10.. Ordovas JM. HDL Genetics: candidate genes,, genome wide scans and gene-environmentt interactions. Cardiovasc Drugs Ther.. 2002;16:273-281.
11.. Schaefer EJ. Familial lipoprotein disorders andd premature coronary artery disease. Med Clinn North Am. 1994;78:21-39.
12.. Franceschini G, Sirtori CR, Bosisio E et al. Relationshipp of the phenotypic expression of thee A-IMilano apoprotein with plasma lipid andd lipoprotein patterns. Atherosclerosis.
1985;58:159-174. .
13.. Clee SM, Zwinderman AH, Engert JC et al. Commonn genetic variation in ABCA1 is associatedd with altered lipoprotein levels andd a modified risk for coronary artery disease.. Circulation. 2001;103:1198-1205. 14.. Cenarro A, Artieda M, Castillo S et al A
commonn variant in the ABCA1 gene is associatedd with a lower risk for premature coronaryy heart disease in familial
hypercholesterolaemia.. J Med Genet. 2003;40:163-168. .
155 Tato F, Vega GL, Tall AR, Grundy SM. Relationn between cholesterol ester transfer proteinn activities and lipoprotein cholesterol inn patients with hypercholesterolemia and combinedd hyperlipidemia. Arterioscler Thrombb Vase Biol. 1995;15:112-120. 16.. Carmena-Ramon R, Ascaso JF, Real JT et al.
Associationn between the TaqlB polymorphismm in the cholesteryl ester transferr protein gene locus and plasma lipoproteinn levels in familial
hypercholesterolemia.. Metabolism. 2001;50:651-656. .
17.. Zambon A, Deeb SS, Pauletto P et al. Hepaticc lipase: a marker for cardiovascular diseasee risk and response to therapy. Curr Opinn Lipidol. 2003;14:179-189.
18.. Hegele RA, Little JA, Connelly PW Compoundd heterozygosity for mutant hepaticc lipase in familial hepatic lipase deficiency.. Biochem Biophys Res Commun.
1991;179:78-84. .
19.. Pimstone SN, Gagne E, Gagne C et al. Mutationss in the gene for lipoprotein lipase: aa cause for low HDL cholesterol levels in individualss heterozygous for familial hypercholesterolemia.. Arterioscler Thromb Vasee Biol. 1995;15:1704-1712.
20.. Kastelein JJ, Ordovas JM, Wittekoekk ME et al.. Two common mutations {D9N, N291S) in lipoproteinn lipase: a cumulative analysis of theirr influence on plasma lipids and
lipoproteinss in men and women. Clin Genet. 1999;56:297-305. .
21.. Wittekoek ME, Pimstone SN, Reymer PW et al.. A common mutation in the lipoprotein lipasee gene (N291S) alters the lipoprotein phenotypee and risk for cardiovascular diseasee in patients with familial hypercholesterolemia.. Circulation. 1998;97:729-735. .
22.. Wittekoek ME, Moll E, Pimstone SN et a!. A frequentt mutation in the lipoprotein lipase genee {D9N) deteriorates the biochemical andd clinical phenotype of familial hypercholesterolemia.. Arterioscler Thromb
\/^i-^\/^i-^ D;^I 1 nnn.1 n o i n o T i n
23.. Berglund L, Wiklund O, Eggertsen G et al. Apolipoproteinn E phenotypes in familial hypercholesteroiaemia:: importance for expressionn of disease and response to therapy.. J Intern Med. 1993;233:173-178. 24.. Wiegman A, Sijbrands EJ, Rodenburg J et al.
Thee Apolipoprotein {epsilon }4 Allele Conferss Additional Risk in Children with Familiall Hypercholesterolemia. Pediatr Res. 2003;53:1008-1012. .
25.. Leus FR, Wittekoek ME, Prins J et ai. Paraoxonasee gene polymorphisms are associatedd with carotid arterial wall thicknesss in subjects with familial hypercholesterolemia.. Atherosclerosis. 2000;149:371-377. .
26.. Leus FR, Zwart M, Kastelein JJ, Voorbij HA. PON22 gene variants are associated with clinicall manifestations of cardiovascular diseasee in familial hypercholesterolemia patients.. Atherosclerosis 2001;1 54:641-649. .
27.. Heller DH, Faire De U., Pedersen NL et al. Geneticc and environmental influences on serumm lipid levels in twins. N Engl J Med. 1993;328:1150-1156. .
28.. Umans-Eckenhausen MA, Defesche JC, Sijbrandss EJ et al. Review of first 5 years of screeningg for familial hypercholesteroiaemia inn the Netherlands. Lancet. 2001;357:165-168. .
29.. US MedPed program, http:/ www.medped.orgg . 2003.
30.. Mortality in treated heterozygous familial hypercholesteroiaemia:: implications for clinicall management. Scientific Steering Committeee on behalf of the Simon Broome Registerr Group. Atherosclerosis.
1999;142:105-112. .
31.. Defesche J. Familial Hypercholesterolemia, In:: Betteridge DJ, editor. Lipids and Vascular Disease.. London, UK: Martin Dunitz Ltd, 2000:65-76. .
32.. Friedewald WT, Levy Rlr Fredrickson DS.
Estimationn of the concentration of low-densityy lipoprotein cholesterol in plasma, withoutt use of the preparative
ultracentrifuge.. Clin Chem. 1972;18:499-502. .
33.. Marcovina S, Albers JJ, Scanu AM et al. Use off e reference material proposed by the Internationall Federation of Clinical Chemistryy and Laboratory Medicine to evaluatee analytical methods for the determinationn of plasma lipoprotein (a). Clin
34.. Spaapen, LJM, Waterval, WAH, and Vies, J. H.. S. Detection of hyperhomocystetnemia at prematuree cerebrovascular disease. Dutch Associationn of Clinical Chemistry 17, 194-199.. 1992.
35.. Fouchier SW, Defesche JC, Umans-Eckenhausenn MW, Kastelein JP. The molecularr basis of familial
hypercholesterolemiaa in The Netherlands. Humm Genet. 2001; 109:602-61 5. 36.. Cheng S, Grow MA, Pallaud C et al. A
multilocuss genotyping assay for candidate markerss of cardiovascular disease risk. Genomee Res. 1999;9:936-949. 37.. Rubins HB, Robins SJ, Collins D et al.
Gemfibrozill for the secondary prevention of coronaryy heart disease in men with low levelss of high-density lipoprotein cholesterol. Veteranss Affairs High-Density Lipoprotein Cholesteroll Intervention Trial Study Group. NN Engl J Med. 1999;341:410418. 38.. Parra HJ, Arveiler D, Evans AE et al. A
case-controll study of lipoprotein particles in two populationss at contrasting risk for coronary heartt disease. The ECTIM study. Arterioscler Thromb.. 1992,12:701-707.
39.. Mailly F, Fisher RM, Nicaud V et al. Associationn between the LPL-D9N mutation inn the lipoprotein lipase gene and plasma lipidd traits in myocardial infarction survivors fromm the ECTIM Study. Atherosclerosis.
1996;122:21-28. .
40.. Fumeron F, Betoulle D, Luc G et al. Alcohol intakee modulates the effect of a
polymorphismm of the cholesteryl ester transferr protein gene on plasma high densityy lipoprotein and the risk of myocardiall infarction. J Clin Invest. 1995;96:1664-1671. .
41.. Franceschini G, Sirtori CR, Bosisio E et al. Relationshipp of the phenotypic expression of thee A-l Miiano apoprotein with plasma lipid andd lipoprotein patterns. Atherosclerosis. 1985;58:159-174. .
42.. Zwarts KY, Clee SM, Zwinderman AH et al. ABCA11 regulatory variants influence coronaryy artery disease independent of effectss on plasma lipid levels. Clin Genet. 2002;61:115-125. .
43.. Cheung MC, Brown BG, Wolf AC et al. Alteredd particle size distribution of apolipoproteinn A-l cholesterol containing lipoproteinss in subjects with coronary artery disease.. J Lipid Res. 1991;32:383-394. 44.. Steyerberg EW, Harrell FE, Jr., Borsboom GJ
ett al. Internal validation of predictive models:: efficiency of some procedures for logisticc regression analysis. J Clin Epidemiol. 2001;54:774-781. .
