Post on 22-May-2018
DOI: 10.1161/CIRCGENETICS.113.000400
1
Lipoprotein(a) Levels, Genotype and Incident Aortic Valve Stenosis: A
Prospective Mendelian Randomization Study and Replication in a Case-
Control Cohort
Running title: Arsenault et al.; Lipoprotein(a) and aortic valve stenosis
Benoit J. Arsenault, PhD1,2; S. Matthijs Boekholdt, MD, PhD3; Marie-Pierre Dubé, PhD1,2;
Éric Rhéaume, PhD1,2; Nicholas J. Wareham, MBBS, PhD4;
Kay-Tee Khaw, MBBChir5; Manjinder S. Sandhu, PhD5,6; Jean-Claude Tardif, MD1,2
1Montreal Heart Institute Research Center; 2Département de Médecine, Faculté de Médecine, Université de Montréal, Montréal, Québec, Canada; 3Department of Cardiology, Academic
Medical Center, Amsterdam, the Netherlands; 4MRC Epidemiology Unit, 5Department of Public Health and Primary Care, University of Cambridge, Cambridge; 6Genetic Epidemiology Group,
Wellcome Trust Sanger Institute, Hinxton, United Kingdom
Correspondence:
Jean-Claude Tardif, MD
Montreal Heart Institute
5000 Bélanger
Montréal (QC), H1T 1C8
Canada
Tel: 514-376-3330 ext. 3612
Fax: 514-593-2500
E-mail: jean-claude.tardif@icm-mhi.org
Journal Subject Codes: [8] Epidemiology
SSSS, PhPhPhPhDDDD4444;;;;
Cl dddd TTTT didididiffff MDMDMDMDf 1 2Tee Khaw, MBBChir ; Manjinder S. Sandhu, PhD ; Jean Claude Tardif, MDff
a ci me Pd G
Tee aw, C ; anj de S. Sa d u, D ; Jean Claude a di ,f
al HHHHeaeaeart Insnsnsstitititituuuteeee Resesessearcccch hh CeCeCeentnn eeere ;;; 2DéDéDéD papap rrtr emememmennnt t t t dededed MéMéMéM dedededeccinnnen , FaFaFaacuuuultltltltéééé dedede MMéddecité de MMMontréal,,, MoMM ntrééééalalal, Québec, Canaaadadd ; 3333Depapap rtment oof CaCCC rdddiiology, Academenter, Amsterdadadadam,m,m,m tttthehehehe NNNNetettheheheherlrlrrlanananandsdsdsd ;;;; 444MRMMRM CCC EpEpEpE idememememioioioiolololol gygygyy UUUUnininit,t,t, 5555DDDeD partment of Pddd PPPriririmamamaryryry CCCararareee, UUUnnnivivivererersisisitytyty ooofff CaCaCambmbmbririridddgegege, CaCaCambmbmbririridgdgdge;e;e; 6GeGeGenenenetititiccc EpEpEpidididemememioioiololologygygy GGG
WeWeWellllllllcocoomemme TTTTrurur sttstst SSSSananngegger rr InInInI stststititititututute,ee HHHHininininxtxttononon, UnUnUUnittitedededed KKKKinininingdgdgddomomom
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
2
Abstract:
Background - Although a previous study has suggested that a genetic variant in the LPA region
was associated with the presence of AVS, no prospective study has suggested a role for Lp(a)
levels in the pathophysiology of AVS. Our objective was to determine whether lipoprotein (a)
[Lp(a)] levels and a common genetic variant that is strongly associated with Lp(a) levels are
associated with an increased risk of developing aortic valve stenosis (AVS).
Methods and Results - Serum Lp(a) levels were measured in 17,553 participants of the EPIC-
Norfolk study. Among these study participants, 118 developed AVS during a mean follow-up of
11.7 years. The rs10455872 genetic variant in LPA was genotyped in 14,735 study participants
who simultaneously had Lp(a) levels measurements and in a replication study of 379 patients
with echocardiography-confirmed AVS and 404 controls. In EPIC-Norfolk, compared to
participants in the bottom Lp(a) tertile, those in the top Lp(a) tertile had a higher risk of AVS;
hazard ratio (HR) 1.57 [95% CI, 1.02-2.42] after adjusting for age, sex and smoking. Compared
to rs10455872 AA homozygotes, carriers of one or two G alleles were at increased risk of AVS;
HR=1.78 [1.11-2.87] and HR=4.83 [1.77-13.20], respectively. In the replication study, the
genetic variant rs10455872 also showed a positive association with AVS (odds ratio=1.57 [1.10-
2.26]).
Conclusions - Patients with high Lp(a) levels are at increased risk for AVS. The rs10455872
variant which is associated with higher Lp(a) levels is also associated with increased risk of
AVS, suggesting that this association may be causal.
Key words: aortic valve stenosis, lipoprotein, Mendelian randomization, rs10455872
, y p ppp
ooonnnn ststststududududy y y y oooof f f f 373737379999 pppap titititiennnn
rfolk,,,, cocococompmpmpmpararararedededed ttttoooo
s in the bottom Lp(a) tertile, those in the top Lp(a) tertile had a higher risk of AV
o p
872 AA homozygotes, carriers of one or two G alleles were at increased risk of A
1
i [
s in n n thththt eee bobobob tttttttomomom LLLLp(a) tertile, those in theeee tttop Lp(a) tertile hhhadadadad a higher risk of AV
o ((HHHRH ) 1.57 [[[95959 % %% CICICII, 111.1 0202020 -2-222.4.4.42]2]2]] aaftff eeer adjjuustingng ffffororo agegge, seses x annd d d d smmmokokokokinininng.g.g. CCCComomomp
877772 222 AAAAA hommmozzyyygy ooteees, cccarrrriersrs ooof oneee oor ttwwo GGGG alleelesss wweree at iiinnncreasasasededed rrrisk k oof A
1.11-2.87] and HRHHR=4.8.88333 [1[1.7.7.77-133.20]0]0]],, rererespsppecectit veeelylyl . In ttheee rrepepeplicacacatit on study, the
iant rs1045587222 alllso shhhhowed a pppositiivii e association iiwi hthhh AAAAVSVSVSVS ((( ddoddds ratio=1.57 [
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
3
Introduction
Aortic valve stenosis (AVS) is the most common valvular disease in the Western world. Its
prevalence increases with age, reaching 2-3% of individuals aged over 65 years1. The burden of
AVS is high and is expected to double within the next 50 years2. To this date, the only effective
treatment for AVS is aortic valve replacement, for which costs have been estimated up to
$120,0003. Identification of the risk factors for AVS is likely to help the medical and scientific
communities develop novel and innovative treatment strategies. Up to now, male gender,
smoking, hypertension, dyslipidemia, metabolic syndrome and impaired glucose-insulin
homeostasis have been associated with AVS incidence and/or progression4, 5.
Genetic association studies have sought to determine whether genetic variants are
associated with AVS risk 6. Recently, Thanassoulis et al.7 performed a large-scale meta-analysis
of genome-wide scans for aortic valvular calcium in cohorts of the CHARGE consortium and
identified rs10455872 at the LPA locus as a susceptibility single nucleotide polymorphism (SNP)
for aortic valvular calcium. This association was replicated in two population-based studies of
AVS, namely the Copenhagen City Heart Study and the Malmö Diet and Cancer Study.
However, the association between Lp(a) levels and the risk of AVS was not investigated in these
AVS studies. The evidence that patients with AVS could be characterized by high Lp(a) levels is
scarce. Glader et al. showed that plasma levels of Lp(a) were almost 1.5-fold higher in 101
patients with AVS compared to matched controls, although this relationship did not reach
statistical significance8. Investigators of the Cardiovascular Health Study have shown that
patients with either aortic valve sclerosis or AVS had significantly higher Lp(a) levels compared
to controls but this association was documented cross-sectionally at baseline5. To this date, no
prospective study has suggested a role for Lp(a) levels in the pathophysiology of AVS.
ed glucose-insulinnnn
sion4444, 5555.
n
w a
-wide scans for aortic valvular calcium in cohorts of the CHARGE consortium a
s
al lar calci m This association as replicated in t o pop lation based st dies
netic c c asasassososociciciiataa iooonnn studies have sought to ddddeteteteermine whether gggenetic variants are
wwwwithhh AVS riskkk 666. RReR cccenttttlylylyl , ThThThThannnaassooullis eett al.7 pepepeperffforrrmmed aa largrgrgge-scccalalalale memm taaa-aaana
-widededd sssscacacacanns ffforor aaorortititicc valvlvlvulululular calalalciiciciumummm iiiinn cocohohohortrtrtrts offofof thehehee CCCCHAHAHAARGGGGEEEE consnsorortititiumum a
s10455872 at thehhh LPLLPAAA llllocus as a susceppp iitiibibibililililitytyty singlgllle nucllleotiiiidedd pppolymyy orphp ism
all lla ll ici ThThiis iia iti lili tedd iin t llatiio bba dd st ddiie by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
4
The objective of our study was to test the hypothesis that Lp(a) levels are associated with
an increased risk of developing AVS in a large population-based study of asymptomatic men and
women. We also tested the potential causality of this association by studying a common variant
in the LPA gene that is strongly associated with Lp(a) levels. We tested these hypotheses in the
EPIC-Norfolk prospective population study as well as in a case-control study of AVS.
Methods
EPIC-Norfolk study
The EPIC-Norfolk prospective population study is a population-based cohort of 25,639 men and
women aged between 39 and 79 years resident in Norfolk, UK. The design and methods of the
study have been described in details9. Participants were recruited from age-sex registers of
general practices in Norfolk as part of the 10-country collaborative EPIC study. The study cohort
was closely similar to UK population samples for many characteristics, including anthropometry,
blood pressure and lipids, but with a lower proportion of smokers. At the baseline survey
conducted between 1993 and 1997, participants completed a detailed health and lifestyle
questionnaire. Blood was taken by venipuncture into plain and citrate tubes. Blood samples were
processed for various assays at the Department of Clinical Biochemistry, University of
Cambridge, or stored at -80°C. Hospitalizations of study participants were identified through the
East Norfolk Health Authority database, which records all hospital contacts throughout England
and Wales for Norfolk residents. Vital status for all EPIC-Norfolk participants was obtained
through death certification at the Office for National Statistics. The underlying cause of death or
hospital admission was coded by trained nosologists according to the International Classification
of Diseases (ICD), Tenth Revision. Participants were identified as having incident aortic stenosis
if they were hospitalized with AVS (ICD code I35) as an underlying cause or if they died with
cohoooortrtrtrt oooof f f f 25252525,6,6,6,63939399 mmmmeeee
ed between 39 and 79 years resident in Norfolk, UK. The design and methods of
c c
y m
sure and lipids, but with a lower proportion of smokers. At the baseline survey
ed betweweweene 3933 aaand 79 years resident in Noorfr olk, UK. The ddesee ign and methods of
bbbeeeen describebebedd innn dddetetetaiaiaiilslssls999. PaPaPaP rtrticiiciciipannnts wwerre rerecrrruiuiuuited ddd frffromomo agege----sexx x rereregigigiistststererrs ss ofooo
ctiiicecececess s inininin NNNNorororo fofofolklklklk aaas pappp rttt oooff ff ththththee 1010100-c-c-c- ouououo ntntnttryryryry ccccolololo lalaaaboboboborararaatitiivevevee EEEEPIPIPIIC CCC ststststudududu y.y.y.y. TThehehehe ssstututuudydydy c
y similar to UK KKK poppopopupupupulalalatitititiononon ssssamamamamplplplp eseses fffforororo mmmmananannyyy y chararararacacaccteteteteriristststticicici s,s,s, iiincncncclulululuding anthropom
suuurere aandnd llipipididids,s, bbbutut wwitithhh aa lolowewerr prpropoporortititionon ooff smsmokokkerers.s. AAAtt thththee babbaseseliliinene ssururveveyy
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
5
AVS as an underlying cause. The Norwich District Health Authority Ethics Committee approved
the study, and all participants gave signed informed consent.
Montreal Heart Institute (MHI) Biobank
A total of 405 consecutive patients with mild to severe AVS were recruited. AVS was defined by
an aortic jet velocity >2.5 m/s. Patients with bicuspid AVS, AVS of rheumatic etiology or
patients who underwent aortic valve replacement for any other reason than AVS (aortic
insufficiency, infective endocarditis, congenital, etc.) were not included. Of these 405 patients,
176 underwent aortic valve replacement surgery. The control group included 415 patients
without AVS, with an anatomically normal tricuspid valve, and without aortic valve sclerosis or
stenosis as documented by echocardiography. All controls were characterized by an aortic jet
velocity <1.7 m/s. Cases were excluded if they had undergone radiotherapy due to any type of
cancer in the thoracic area (breast, trachea, bronchus or lung cancer) prior to the diagnostic of
AVS. Other exclusion criteria for both cases and controls included the presence of a bicuspid
aortic valve, renal insufficiency defined as serum creatinine level e
hyperparathyroidism, Paget’s disease of the bone or lupus erythematosus. The study protocol
was approved by the Montreal Heart Institute (MHI) Research Ethics Board and all participants
gave signed informed consent.
Genotyping and laboratory measurements
In EPIC-Norfolk, the SNP was genotyped using Custom TaqMan® SNP Genotyping Assays
(Applied Biosystems, Warrington, UK). The genotyping assays were carried out using 10ng of
genomic DNA in a 2.5μl reaction volume on 384-well plates using a G-Storm GS4 Thermal
Cycler (GRI, Rayne, UK). The ABI PRISM® 7900HT Sequence Detection System (Applied
Biosystems, Warrington, UK) was used for end-point detection and allele calling. The SNP
cluded 415 patientntntnts s s s
ut ao tttrtiiiic v lllalve sclcllclererereros
documented by echocardiography. All controls were characterized by an aortic j
e
he thoracic area (breast, trachea, bronchus or lung cancer) prior to the diagnostic
r exclusion criteria for both cases and controls included the presence of a bicusp
e renal ins fficienc defined as ser m creatinine le el e
docucucumemementntntedededed bbby y y echocardiography. All cococontn rols were characacacterized by an aortic j
.77 mmm/s. Cases wwweree eeexcllclludududu edddd iiif ttthey hhhaad uunnderergogogogonenenn radaadiothhheeerapppyyy y dueee e tott annny tytytype
he thohohhorararaacicicic c arreaea ((((brbrb eaeasts , tracacacchhhhea, bbbbrororoncncncn hhuhuss oror lllunununng cacacancccerererer) ))) prprioioii r totototo ttthehehh dddiaiai gngnosostititic
r exclusion criteria fffof r bbbob thhhh cases andddd contr llols incllluddddeddd thhheh pppresence of a bicuspp
ll iin ffffiiciie ddefifi dd iti ini ll lel by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
6
passed the quality control criteria in the EPIC-Norfolk study (call rate > 95 %, blind duplicate
In the Montreal Heart Institute Biobank, genotyping was carried out using
both the Infinium HumanExome Beadchip and the Cardio-MetaboChip (Illumina, CA, USA).
Those genotyping panels have been described in details elsewhere10, 11. The rs10455872 was
genotyped using the Sequenom Mass Array. Genotyping was performed at the Beaulieu-Saucier
Université de Montréal Pharmacogenomics Centre. Seven duplicate controls genotyped on the
Infinium HumanExome Beadchip and 2 duplicates for the Cardio-MetaboChip had concordance
rates >0.9998. Quality checks for genotypes were performed to exclude completely failed SNPs,
uninformative SNPs, sample and genotyping call rates <98%, and SNPs with plate bias. SNPs
that failed Hardy-Weinberg equilibrium with P<10-7were excluded.
In EPIC-Norfolk, various laboratory measurements including a conventional lipid profile,
were performed at baseline as previously described9. When additional funding became available
in 2010, additional measurements were performed in a subset of the cohort with available stored
frozen blood samples. Lp(a) levels were measured with an immunoturbidimetric assay using
polyclonal antibodies directed against epitopes in apolipoprotein(a) (Denka Seiken, Coventry,
United Kingdom), as previously described 12. This assay has been shown to be apolipoprotein(a)
isoform-independent.
Statistical Analyses
For the prospective analysis in the EPIC-Norfolk cohort, study participants were excluded if
Lp(a) levels were missing. Baseline characteristics of study participants were compared between
participants who developed AVS during follow-up vs. those who did not using an unpaired
Student t-test for continuous variables with a normal distribution, a Mann-Whitney U test for
continuous variables with a non-normal distribution, or a chi-square test for categorical variables.
de completely failedededed S
Ps wiiiithththth plllal tttte bbbbiaiii s. SSSSNN
H 7
E p
r i
ditional measurements were performed in a subset of the cohort with available s
od samples Lp(a) le els ere meas red ith an imm not rbidimetric assa sin
Hardrdrdrdyy-WWW-Weieieiinbnnn erererggg equilibrium with P<1000-7777wwere excluded.
EPPIP CCCC-Norfolk, vvvariououus alalaabbob raaatotot ryyy mmeaassureemmentnts inininccluduuding a cooonnnventtitit ooonaaal lipppidddd p
rmedddd aaaattt bababab selililinene aass prpreviooussususlllly desesscrcrcribibbb dededd9999. WhWhWhW ennenen aaadddddddditttiooioionanalll fffufunddddininininggg bbbbecacameme aavavai
ditional measurements were pepp rfffformed ddd iiin a subset offff thehhh cohhhhort with available s
oddd lpl LLp((a)) lle lls ded ii hth iim t brbididiimet iri iin by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
7
Cox regression analysis was used to calculate hazard ratios (HR) and corresponding 95%
confidence intervals (95%CI) for the time to hospitalization or death due to AVS. Lp(a) levels or
rs10455872 genotype categories were used as independent variables. Regression models were
tested before and after adjustment for potential confounding risk factors. An instrumental
analysis was conducted to assess the per-allele increase of age- and sex-adjusted genetically
elevated Lp(a) levels using the following formula: Lp(a) = c + ß1(age) + ß2(sex) + ß3(LPA
genotype), as previously described 7. We used the additive model to test the association between
genetic variants in the IGF2R-SLC22A1-A2-A3-LPAL2-LPA-PLG region with a minor allele
frequency equal or above 5% and AVS risk using logistic regression after adjusting for age and
sex in the MHI Biobank. In EPIC-Norfolk, statistical analyses were performed using SPSS
software version 20 and we used PLINK for genetic associations studies in the Montreal Heart
Institute Biobank.
Results
Lipoprotein(a) levels and AVS risk in the EPIC-Norfolk study
In EPIC-Norfolk, Lp(a) levels were available in 17,553 study participants. Among these study
participants, 118 were identified with incident AVS during follow-up through March 31st 2010.
Baseline characteristics of the study participants are shown in Table 1 for cases and controls
separately. Of note, 250 study participants (1.4%) were on lipid-lowering therapy.
The association between Lp(a) levels and the risk of incident AVS is shown in Table 2.
Compared to participants in the bottom Lp(a) tertile, study participants in the top Lp(a) tertile
were at increased AVS risk, with an unadjusted HR=1.66 (1.08-2.56) and HR=1.57 (1.02-2.42)
when adjusting for age, sex and smoking. Further adjustment for LDL cholesterol levels had a
considerable impact on the relationship between Lp(a) levels and AVS risk. Given that the
ion with a minor aaaallllllllele
fter adjdjdjdjustitititing ffffor aaaagggge
M
ersion 20 and we used PLINK for genetic associations studies in the Montreal H
o
MHIHIHI BBBioioioiobabababanknnn . InInInI EPIC-Norfolk, statisticcccalalaa analyses were ppeeerfrrr ormed using SPSS
errsr iooon 20 and wweee usseeede PPPLILLL NKKKK fooor gennneetic aassocociaiaiattit oonsss sstudddieees inininn the MMMMooonttreaaal H
obankkk.
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
8
European Society of Cardiology guidelines have suggested that a “desirable” level of Lp(a) was
under 50 mg/dL, we assessed the relationship between AVS event rates in patients with Lp(a)
levels below versus equal or above 50 mg/dL. Our results suggest that even after adjusting for
age, sex, smoking and LDL cholesterol levels, patients with Lp(a) levels equal or above 50
mg/dL are at significantly increased AVS risk (Table 2).
LPA genetics, lipoprotein(a) levels and AVS risk in the EPIC-Norfolk study
The rs10455872 genetic variant was genotyped in 14,735 study participants in whom Lp(a)
levels were also available. Table 3 shows the association between Lp(a)-raising alleles and Lp(a)
levels, as well as with AVS risk using either an additive model or a dominant model, before and
after adjusting for Lp(a) levels, log-transformed Lp(a) levels and Lp(a) quintiles. There was a
positive association between the number of Lp(a)-raising alleles and lipoprotein(a) levels. There
was also a strong and positive association between rs10455872 and AVS risk, which was only
slightly affected after further adjustment for Lp(a) levels log-transformed Lp(a) levels or Lp(a)
quintiles.
To further address the potential causality of this genetic variant, we performed an
instrumental variable analysis, whereby the increment of Lp(a) levels per rs10455872 G allele
was tested for its association with AVS. Linear regression yielded a sex- and age-adjusted per-
allele increment of 31.1 mg/dL of plasma Lp(a) as shown in Figure 1. The HR for AVS risk was
1.95 (95% CI = 1.34-1.60) per each additional rs10455872 G allele while the HR for AVS risk
per 31.1 mg/dL Lp(a) increment was considerably lower (HR=1.31 [1.07-1.60]).
LPA locus and AVS risk in the Montreal Heart Institute Biobank
The clinical and echocardiographic characteristics of the MHI Biobank are shown in
Supplementary Table 1 and in Supplementary Table 2, respectively. The rs10455872 genetic
a))-raising alleles anananandd d
ominantttt moddddel,l,ll befefefefoooore
i
s T
strong and positive association between rs10455872 and AVS risk, which was o
e p
ing g g fofofof rrr LpLpLpL (a(a(a(a) leleleevevv ls, log-transformed Lppp(a(a(( )) levels and Lp(a)a)a)) quintiles. There wa
sooco iiiation betwweeeen tthehehehe numumumu beeeerrr offf LLp(aa))-raaissing g alalalllel llel s aand llil pppoprprproteiiiin(n(n(n(a))a) leveeelss. T
stronggg ananananddd poosisiititititiveve aasss ociaiaiatititition bbbbetetetweweweweeen rr 1s1s10404040 555555587887872222 ananandddd AVAVAVAVS SSS riiiisksksksk, hwhwhicici hhhh wawass o
ected after furthehhh r ddaddjujuj stment for LLp((p((a)a))) lllev llells loggg-tran fsffformeddd d LpLpLp(a( ) ) levels or Lp
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
9
variant along with 583 other SNPs in the IGF2R-SLC22A1-A2-A3-LPAL2-LPA-PLG region with
minor allele frequencies equal or above 5% were genotyped in this cohort. The association
between all SNPs in this region and AVS risk is shown in Figure 2. These results confirm that
the rs10455872 variant is indeed positively associated with AVS risk. However, several other
variants located upstream of the rs10455872 variant could also be predictive of AVS risk. Table
4 shows the results of the genetic association test for rs10455872 and for the other SNPs in this
region that were associated with AVS risk with a p-value below 1x10-3.
Discussion
Results of the present study show that individuals with high Lp(a) levels are at increased risk of
AVS. To the best of our knowledge, this study is the first study with a longitudinal design to
suggest that Lp(a) levels are a strong risk factor for AVS. Because reverse causality cannot be
fully addressed in this study design, we have genotyped a common variant in the LPA region
(rs10455872) that is closely associated with Lp(a) levels. Our results also show that the number
of the rs10455872 G allele associated with elevated Lp(a) levels is strongly associated with AVS
risk, which suggests that the relationship between Lp(a) and AVS is likely to be causal. In
addition, several other variants located upstream of the LPA gene were also shown to be
associated with AVS risk.
With the exception of the recent report of the CHARGE consortium7, the literature on
Lp(a) levels and AVS risk is scarce. The association between Lp(a) levels and presence of aortic
valve sclerosis and stenosis was first demonstrated in a cross-sectional analysis of the
Cardiovascular Health Study5. A previous case-control study has suggested that Lp(a) levels may
be elevated in patients with AVS8 and one other study observed that patients with aortic valve
calcification had significantly higher Lp(a) levels compared to patients without aortic valve
the present study show that individuals with high Lp(a) levels are at increased ris
he best of our knowledge, this stud is the first study with a lo itudinal design t
t t
s o
2) that is closely associated with Lp(a) levels. m
the preesesesent ssstuudydydyy show that individuals witith h high Lp(a) levvelee s are at increased ri
hee bbbeb st of our r knknknowwleleleedgdgdgge,e,e, ttthihihiis ststtstududuu y isiii thee ffirsttst stututtudydydy wwwititithhh h a a lllongnggitititi udududdininininalalall ddddesessigigigign t
t LLLp(p(p(p(a)a)a)) leleelevevevevelslsls arererere aaa sssstrtrttronnnng gg g ririririsksksks facacacactototoor r r r fofofor r r AVAVAVAVSSS.S BBBBececece auauaua sesesee rrrreveveve ererererseese ccccauauaua sasasasaliitytytyty cccananannnononot
ssed in this studdddy yyy dedededesiss gngngngn, wewewewe hhhavavavavee e gegegegenonononotytytyt pepepeeddd d a cocococommmmmmmmononn vvvvarararriaiaiai ntntntt iiiin nnn the LPA regio
2)2) tthahatt isis cclolloseselylly aassssocociaiiateteddd wiwiththth LLp((p(a)a) lllevevelells.s. OuOuO rr reresusultltltss alallsoso shhshowow ttthahatt ththee nunumm
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
10
calcification13. Also, despite an early finding suggesting that hyperlipidemic transgenic rabbits
expressing the human apolipoprotein(a) gene have increased vascular calcification, studies that
have sought to establish a link between Lp(a) levels and coronary calcification in humans have
yielded conflicting results14-16. Considerable evidence supports a role for Lp(a) in the
pathophysiology of AVS. First, Lp(a) is a risk factor for cardiovascular disease (CVD) and AVS
shares a number of common risk factors with CVD. Second, at the cellular level, Lp(a) and
isolated apo(a) have been shown to trigger smooth muscle cell proliferation via activation of
transforming growth factor (TGF)- 17. Given that there is now compelling evidence for a role of
TGF- signaling in the pathophysiology of AVS (as recently reviewed by Xu et al.18), Lp(a)
could induce valvular calcification via the TGF- pathway, although experimental studies are
required to confirm this association. Lp(a) has also been shown to activate macrophages and
increase interleukin-8 secretion in vitro19. In addition, oxidized, but not native LDL particles
could activate macrophages and promote valvular calcification in humans20. Since Lp(a) can act
as an important carrier of oxidized phospholipids, which have been identified as significant
predictors of the onset of CVD in a nested case-control dataset of the EPIC-Norfolk study, Lp(a)
may promote calcification via its pro-oxidant properties.
It has been suggested that more than 90% of the variance in serum Lp(a) concentrations
is explained by variations in and around the LPA locus on chromosome 6q27, which makes Lp(a)
an ideal candidate for Mendelian randomization studies21. The purpose of an analysis of this kind
is to test whether lifelong exposure to a biomarker, in our case elevated Lp(a) levels, would
increase AVS risk independently of potential confounders by other phenotypes or clinical
characteristics that could influence serum Lp(a) levels. We have chosen to study rs10455872 for
three reasons: 1) it is the common variant that has shown the strongest association with Lp(a)
lling evidence for a a a a ro
d by XuXX et al.18181818),),),) LLLLppp(a
c a
confirm this association. Lp(a) has also been shown to activate macrophages an
t e
a a
ce vvvvalalala vulalallarr r cccalclclcciiifi ication via the TGF- pata hhway, althouuugh eexxxperimental studies a
cococonnnfn irm this aassssocciaaationnnn.. . Lp(a) hahhas aaalsso bbeeen shshhshowooown totto acttivvvattte macrcrrroopo hahahages an
terleukin-8 secrereretittt on in vvvititro19. In additioooon,n, oxidiiiizezez d,,,, but notot nnnative LDL particle
aaatetete mmmacacacrororophphphp agagageseses aaandndnd ppprororomomomotetete vvvalalalvuvuvulalalarrr cacacalclclcififificicicatatatioioionnn ininin hhhumumumananansss20. SiSiSincncnceee LpLpLp(a(a(a))) cacaca
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
11
levels in genome-wide association studies22, 2) this genetic variant is frequently used as a proxy
of the number of Kringle-IV-type 2 repeats (KIV-2) in the LPA gene23 and 3) it was the SNP
with the strongest association with aortic valvular calcium in a recent GWAS7. Additionally,
several other genetic variants located upstream of the LPA locus were identified as being
associated with aortic valvular calcification in this GWAS. Interestingly, we found that several
genetic variants located upstream of the LPA locus were associated more strongly with AVS than
rs10455872. Although these were not in linkage disequilibrium with rs10455872, they could
nevertheless be associated with different apo(a) isoforms, apo(a) mRNA expression and Lp(a)
levels22.
Although neither statins, nor fibric acid derivatives influence Lp(a) levels, several other
lipid-lowering therapies have been shown to lower plasma Lp(a) levels to varying extent. These
include proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors24, antisense
oligonucleotides targeting apolipoprotein B mRNA 25, thyroid hormone analogues26 and niacin27.
A recent report from the European Atherosclerosis Society consensus panel suggested at the time
screening for Lp(a) levels in patients at intermediate or high CVD risk28. The panel also
suggested the use of niacin for CVD risk reduction in patients with plasma Lp(a) levels above the
80th percentile. This recommendation, together with the findings of the current study, raises the
question of whether or not niacin [or any other Lp(a)-lowering agent] could be beneficial for the
prevention of AVS or the improvement of aortic valve function in patients with AVS and high
Lp(a) levels. Because the association between the rs10455872 variant and AVS risk was partly
independent of plasma Lp(a) levels and because genetic alleles associated with high elevated
Lp(a) was more strongly associated with AVS risk than a similar increment of actual plasma
Lp(a) levels, it is possible that the association between this genetic variant and AVS risk may be
NA exppression and d d d LpLLL
hough neither statins, nor fibric acid derivatives influence Lp(a) levels, several o
i T
p
o
port from the E ropean Atherosclerosis Societ consens s panel s ggested at th
hougugugh h h neneneititithehhh r sttststatins, nor fibric acid derrrivivivivata ives influence LLLLp(ppp a) levels, several o
innng therapies hahahave bbbeen n n n sshs owwwnn ttto lowwweer pplaasmmaa LpLpLp(a))) llevelsss toooo vvvaryiyiyiyingngng eexxxtennnttt. T
proteteeinninin ccccoonveve trtrtasasee susubtbb illlisisisinininin////kexininin ttttypypypy e 9999 (P(P(PPCSCSCSSK9KK9K9)))) innhiihihibibbibittotorsrs24444, anaanantitititisensnsee
otides targegg tinggg apopp lilililipopp prpp otein B BB RmRRNANANAA 2525255, ,, hhthyryy oiiidd d hhhoh rmone an lalllogggues26 and ni
t ffr thhe EE AAthhe lle isi SSo iciet ll tedd at thh by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
12
mediated through mechanisms other than the difference in plasma Lp(a) levels. Whether this is
due to the lower number of KIV-2 repeats per apo(a) associated with this variant or another
segregating phenotype is unknown. This finding also questions whether all individuals with high
Lp(a) or only those with genetic profiles associated with elevated Lp(a) should be targeted with
Lp(a)-lowering agents for the prevention and/or regression of AVS. It should also be considered
that this aspect contrasts with the results reported by the PROCARDIS consortium who
investigated the association between two genetic variants associated with Lp(a) levels (including
rs10455872) and CVD risk23. In that study, adjustment for Lp(a) levels completely eliminated
the relationship between the genotype score and CVD risk. Another potential explanation for the
stronger association between genetic profiles associated with elevated Lp(a) compared to plasma
measured Lp(a) levels with AVS risk is the fact that these may represent lifelong exposures and
are less susceptible to random measurement error, as opposed to a single plasma Lp(a)
concentration measured on one occasion.
Several aspects of the study design need to be taken into account when interpreting the
results of the present study. The longitudinal design of this study is a major advantage to follow
the natural occurrence of AVS in an apparently healthy population, because it avoids the inherent
biases associated with cross-sectional study designs. However, despite the fact that our cohort
was large, the low incidence of AVS in general population samples still resulted in relatively few
events. In addition, given the fact that we used AVS-related hospitalizations and mortality as
outcome, this outcome definition is likely to be restricted to the most severe and symptomatic
cases, rather than including just the mild and asymptomatic ones. Another limitation is the fact
that in EPIC-Norfolk, the AVS study outcome was based on ICD-coded hospitalizations and
mortality, and not validated against echocardiography reports. In addition, AVS diagnosis was
s completelyy elimimimiinannn t
otenttttiiiialll l ex lllplanatioioioionnnn f
s p
L levels with AVS risk is the fact that these may represent lifelong exposures
s
o
eral aspects of the st d design need to be taken into acco nt hen interpreting
sociaiaiaiatititiononon bbbbetee weweweene genetic profiles assoccciaiaiaiatett d with elevated d d LpLLL (a) compared to p
Lpp(aaa) levels wiiithhh AAAVVVSV risisisiskk k is tthehehe ffacttt thhat thhesese mmmmayy rrreppressennnt llliffffelonnnnggg eexppposussures
sceptititiblblblbleeee ttto rranandddodomm mem asurururreeeementttt eererrororor r, aass ooppopopooseeedddd too aaaa ssinini gllglgle plplplplasasaasma LLL (p(p(a)a)))
on measured on one occasioiii n.
lal ts ff hth st dd dde isi dd to bb takke iinto nt hhe iinte iti by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
13
indeed confirmed by echocardiography in the MHI replication study, which included a sufficient
number of cases.
In conclusion, the results of the present study suggest that patients with high Lp(a) levels
are at increased risk for AVS. The fact that a common genetic variant in LPA is simultaneously
associated with both serum Lp(a) levels and risk of AVS further suggests that this association is
likely to be causal. Our report also shows that several other genetic variants in the expanded LPA
region could also be associated with AVS. Whether pharmacological interventions that influence
Lp(a) levels will reduce the risk of AVS in patients with high Lp(a) levels should be tested in
randomized clinical trials.
Acknowledgments: We would like to thank the study participants as well as the clinical staff of
the EPIC-Norfolk study and of the Montreal Heart Institute’s Biobank.
Funding Sources: BJA is supported by a post-doctoral fellowship from the Canadian Institutes
of Health Research (CIHR). JCT holds the Canada Research Chair in translational and
personalized medicine and the University of Montreal endowed research chair in atherosclerosis.
EPIC-Norfolk is supported by research programme grants from the Medical Research Council
UK and Cancer Research UK.
Conflict of Interest Disclosures: None.
References:
1. Passik CS, Ackermann DM, Pluth JR, Edwards WD. Temporal changes in the causes of aortic stenosis: A surgical pathologic study of 646 cases. Mayo Clin Proc. 1987;62:119-123.
2. Lung B, Vahanian A. Degenerative calcific aortic stenosis: A natural history. Heart. 2012;98 Suppl 4:iv7-13.
vels should be testtttededede
d
N
o t
Research (CIHR). JCT holds the Canada Research Chair in translationa
dgggmmments: Wee wwwouuldldldd lllikikikke e e totto ttthahahannknkn thehehee sstudydy parartiiticicicic ppantntn s asasa wwwellllll aaas tthehehehe cclililil nininin cacacal l l st
Norrrfofofofolk studydydydy aannnd oof ff the MMoM nttrer aaala Heeearrt Insnstitutut tte’s Biobbab nkk.
ources: BJA is suppppopp rt dded bbbby yy a popp st-ddddoctoralll l ffef llowshhhhipipi ffffrom thhhhe CaC nadian Insti
ReReeseseararchchch (((CICICICIHRHRHRHR).).).) JJJCTCTCTCT hhhholololdsdsds ttthehehe CCCCananadadadaa ReReReR seseararchchchh CCCChahahah iririr iiiinn trtrt ananslslsllatatioioionanalll aandndnd
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
14
3. Rosamond W, Flegal K, Friday G, Furie K, Go A, Greenlund K, et al. Heart disease and stroke statistics--2007 update: A report from the american heart association statistics committee and stroke statistics subcommittee. Circulation. 2007;115:e69-171.
4. Rosenhek R, Binder T, Porenta G, Lang I, Christ G, Schemper M, et al. Predictors of outcome in severe, asymptomatic aortic stenosis. N Engl J Med. 2000;343:611-617.
5. Stewart BF, Siscovick D, Lind BK, Gardin JM, Gottdiener JS, Smith VE, et al. Clinical factors associated with calcific aortic valve disease. Cardiovascular health study. J Am Coll Cardiol.1997;29:630-634.
6. Bossé Y, Mathieu P, Pibarot P. Genomics: The next step to elucidate the etiology of calcific aortic valve stenosis. J Am Coll Cardiol. 2008;51:1327-1336.
7. Thanassoulis G, Campbell CY, Owens DS, Smith JG, Smith AV, Peloso GM, et al. Genetic associations with valvular calcification and aortic stenosis. N Engl J Med. 2013;368:503-512.
8. Glader CA, Birgander LS, Soderberg S, Ildgruben HP, Saikku P, Waldenstrom A, et al.Lipoprotein(a), chlamydia pneumoniae, leptin and tissue plasminogen activator as risk markers for valvular aortic stenosis. Eur Heart J. 2003;24:198-208.
9. Day N, Oakes S, Luben R, Khaw KT, Bingham S, Welch A, et al. Epic-norfolk: Study design and characteristics of the cohort. European prospective investigation of cancer. Br J Cancer.1999;80 Suppl 1:95-103.
10. Voight BF, Kang HM, Ding J, Palmer CD, Sidore C, Chines PS, et al. The metabochip, a custom genotyping array for genetic studies of metabolic, cardiovascular, and anthropometric traits. PLoS Genet. 2012;8:e1002793.
11. Http://genome.Sph.Umich.Edu/wiki/exome_chip_design. Accessed 02/04/2013.
12. Gurdasani D, Sjouke B, Tsimikas S, Hovingh GK, Luben RN, Wainwright NW, et al.Lipoprotein(a) and risk of coronary, cerebrovascular, and peripheral artery disease: The epic-norfolk prospective population study. Arterioscler Thromb Vasc Biol. 2012;32:3058-3065.
13. Bozbas H, Yildirir A, Atar I, Pirat B, Eroglu S, Aydinalp A, et al. Effects of serum levels of novel atherosclerotic risk factors on aortic valve calcification. J Heart Valve Dis. 2007;16:387-393.
14. Greif M, Arnoldt T, von Ziegler F, Ruemmler J, Becker C, Wakili R, et al. Lipoprotein (a) is independently correlated with coronary artery calcification. Eur J Intern Med. 2013;24:75-79.
15. Guerra R, Yu Z, Marcovina S, Peshock R, Cohen JC, Hobbs HH. Lipoprotein(a) and apolipoprotein(a) isoforms: No association with coronary artery calcification in the dallas heart study. Circulation. 2005;111:1471-1479.
eloso GM, et al. GGGGeneneneneMMMMedededed. 2020202013131313;3;3;3368686868:5:5:5:503030303-5-5-5-51111
Waldensnsnsnstrtrtrtromomomom AAAA eeeetttt alalalan(a), chlamydia pneumoniae, leptin and tissue plasminogen activator as risk marr
O dt ru
,not ping arra for genetic st dies of metabolic cardio asc lar and anthropomet
n(a)a)a)a), chchchhlalalaamymymydiiia a a pneumoniae, leptin and tttisisii ssue plasminogen n aaca tivator as risk marr aaaorororortic stennnnosssisissis. EuEuuEur rrr HeHeHH ararrt t tt JJJJ. . . . 202020200300 ;2;2;2;24:44 19988-20008.8..8. JJJJJJJJ
Oaaakekekes s S, Lubububenn R, KKKhawww KKT,T, BBiniinghgg amaam S, WeWeelclch AA, eeet al. EpEpEpicc-nnnorfofofolklkl : SSStudddyy dteristtticicicsss ofofofo theheh cc hohohororttt. Eurroppopopean prprprosososospeppp ctctttiivivee ininnvevevevestststigigigatatatiooioion n ofofff ccanceccecerrr. Br JJJ CCCanceruppl 1:95-103.
BBF, Kang g HMHHMHM,,, DiDDiDingngngng JJJ, , , PaPPaPalmlmlmlmererer CCCCD,D,D, SSSSididdidoroororeee C,C,C,C CCCChihhihineneenessss PSSPSPS, ,, etetett aaaalll. .. ThThThThe ee memmemetabochip,ot iin ffo iti st ddiie fof et bab loliic drdiio ll dd hth t by guest on June 27, 2018
http://circgenetics.ahajournals.org/D
ownloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
15
16. Sharma A, Kasim M, Joshi PH, Qian Z, Krivitsky E, Akram K, et al. Abnormal lipoprotein(a) levels predict coronary artery calcification in southeast asians but not in caucasians: Use of noninvasive imaging for evaluation of an emerging risk factor. J Cardiovasc Transl Res. 2011;4:470-476.
17. Grainger DJ, Kirschenlohr HL, Metcalfe JC, Weissberg PL, Wade DP, Lawn RM. Proliferation of human smooth muscle cells promoted by lipoprotein(a). Science. 1993;260:1655-1658.
18. Xu S, Liu AC, Gotlieb AI. Common pathogenic features of atherosclerosis and calcific aortic stenosis: Role of transforming growth factor-beta. Cardiovasc Pathol. 2010;19:236-247.
19. Klezovitch O, Edelstein C, Scanu AM. Stimulation of interleukin-8 production in human thp-1 macrophages by apolipoprotein(a). Evidence for a critical involvement of elements in its c-terminal domain. J Biol Chem. 2001;276:46864-46869.
20. Côté C, Pibarot P, Després JP, Mohty D, Cartier A, Arsenault BJ, et al. Association between circulating oxidised low-density lipoprotein and fibrocalcific remodelling of the aortic valve in aortic stenosis. Heart. 2008;94:1175-1180.
21. Boerwinkle E, Leffert CC, Lin J, Lackner C, Chiesa G, Hobbs HH. Apolipoprotein(a) gene accounts for greater than 90% of the variation in plasma lipoprotein(a) concentrations. J Clin Invest. 1992;90:52-60.
22. Qi Q, Workalemahu T, Zhang C, Hu FB, Qi L. Genetic variants, plasma lipoprotein(a) levels, and risk of cardiovascular morbidity and mortality among two prospective cohorts of type 2 diabetes. Eur Heart J. 2012;33:325-334.
23. Clarke R, Peden JF, Hopewell JC, Kyriakou T, Goel A, Heath SC, et al. Genetic variants associated with lp(a) lipoprotein level and coronary disease. N Engl J Med. 2009;361:2518-2528.
24. Raal F, Scott R, Somaratne R, Bridges I, Li G, Wasserman SM, Stein EA. Low-density lipoprotein cholesterol-lowering effects of amg 145, a monoclonal antibody to proprotein convertase subtilisin/kexin type 9 serine protease in patients with heterozygous familial hypercholesterolemia: The reduction of ldl-c with pcsk9 inhibition in heterozygous familial hypercholesterolemia disorder (rutherford) randomized trial. Circulation. 2012;126:2408-2417.
25. McGowan MP, Tardif JC, Ceska R, Burgess L, Soran H, Gouni-Berthold I, et al.Randomized, placebo-controlled trial of mipomersen in patients with severe hypercholesterolemia receiving maximally tolerated lipid-lowering therapy. PLoS ONE.2012;7:e49006.
26. Ladenson PW, Kristensen JD, Ridgway EC, Olsson AG, Carlsson B, Klein I, et al. Use of the thyroid hormone analogue eprotirome in statin-treated dyslipidemia. N Engl J Med.2010;362:906-916.
et al. AAAsA so iicii tttatiiioi n bebebebetling ooooffff ththththeeee aoaoaoaortrtrtrticicicic vvvvaaalv
o
nkle E, Leffert CC, Lin J, Lackner C, Chiesa G, Hobbs HH. ol oprotein go l2
Wcardiovascular morbidi and mortali amo two ro ective cohorts of 2
H t J 2012;33:325 334J
osis... HeHeHeH ararartt. 2020200808080 ;94:1175-1180.
nknkkkleee E, Leffertt CCCCCC, LLLiL n J,JJJ Lacaacknknkneer CCC, Chieiesa GGG,,,, HoHoHobbbbss HHH. ApApApoolipopopopo roooteeein(a(a(a( ) gor gggrerereater thahahan 90%% ooof thehehe varriaiatiiioon innn pplaasmamaa lliiipopoprotettein(aa( ) cooncncncentrtrtraata iooonnns. JJJ Cl2;90:000 5252522-66600.0
Workalemahu T,T,T, ZZZhhhah nggg CCC, ,, HuHH FFFB,B,B, QiQiQii LL. GeGGG netic variiiiants,,, plplllasma lippoppprotein(a( ))ccardiovascscululululararar mmmmororrorbibbb didididitytytyty aaaandndndnd mmmmororoortataatalillilitytyyty aaaamomoomongngg ttttwowwowo pppprororospspspspecececctitititivevevve ccccohohohorororortststs of type 2
HH t JJ 20201212 3;333:323255 333344JJ by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
16
27. Boden WE, Probstfield JL, Anderson T, Chaitman BR, Desvignes-Nickens P, Koprowicz K, et al. Niacin in patients with low hdl cholesterol levels receiving intensive statin therapy. N Engl J Med. 2011;365:2255-2267.
28. Nordestgaard BG, Chapman MJ, Ray K, Boren J, Andreotti F, Watts GF, et al.Lipoprotein(a) as a cardiovascular risk factor: Current status. Eur Heart J. 2010;23:2844-2853.
Table 1. Baseline characteristics of EPIC-Norfolk participants
Controls (n=17,435) Cases (n=118) p-value
Age, y 59.1 (9.2) 65.9 (6.8) < 0.001
Male, % 7,658 (43.9) 71 (60.2) < 0.001
Body mass index, kg/m2 26.1 (3.8) 26.9 (3.7) 0.022
Smoking, %
Current 1,967 (11.3) 15 (12.7)
Past 7,239 (41.5) 62 (52.5) 0.024
Never 8,229 (47.2) 41 (34.7)
Diabetes mellitus 345 (2.0) 5 (4.2) 0.087
Systolic blood pressure, mmHg 134.7 (18.2) 141.6 (19.0) < 0.001
Diastolic blood pressure, mmHg 82.0 (11.1) 83.2 (10.8) 0.3
Lipoprotein(a), mg/dL 11.6 (6.2-27.6) 16.2 (6.3-44.5) 0.025
Data are presented as mean (standard deviation), median (interquartile range) or as percentage (number).
181811 )))) p-ppp vavaalulululueeee
)59.1 (9.2) 65.9 (6.8) 0.001
%
59.1 (9.2) 65.9 (6.8))) 0.001
7,7 656588 (4(( 3.3.9) 71111 (((6060.2)))) < 0.000 00000001
innnddded x, kg/m2 26.11 (3.88) 226.9.9 (3.7)7)) 000.002222
%
1,1,1,1,969696967777 (1(1(1(11.1.1.1 3)3)3)3 15151515 (((12121212.7.7.7))))
7,7,7,7,23323239999 (4(4(4(41.1.1.5)5)5)5) 62626262 (52525252.5.5.5)))) 0.024
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
17
Table 2. Lipoprotein(a) levels and risk of aortic valve stenosis (AVS) in EPIC-Norfolk
Median lipoprotein(a) levels are presented as median (interquartile range). Model 1 is adjusted for sex, age and smoking. Model 2 is model 1 additionally adjusted for low-density lipoprotein cholesterol levels.
Tertile 1 Tertile 2 Tertile 3 P-value <50 mg/dL >50 mg/dL P-value
Lipoprotein(a) range,
mg/dL 2.0-7.7 7.8-19.2 19.2-174.9 2.0-50 50-174.9
Lipoprotein(a) levels,
mg/dL
4.8
(3.7-6.3) 11.6
(9.4-14.8)
40.9
(27.7-57.6)
10.1
(5.7-19.0)
66.1
(56.9-84.1)
Event rate33/5.824
(0.57%)
30/5.815
(0.52%)
55/5.796
(0.95%)
92/15.435
(0.6%)
26/2.000
(1.3%)
Hazard ratios:
Unadjusted 1.00 0.91
(0.55-1.48)
1.66
(1.08-2.56) 0.01 1.00
2.15
(1.39-3.32) 0.001
Model 1 1.00 0.85
(0.52-1.40)
1.57
(1.02-2.42) 0.03 1.00
2.17
(1.40-3.35) 0.001
Model 2 1.00 0.82
(0.50-1.35)
1.44
(0.93-2.23) 0.07 1.00
1.98
(1.25-3.09) 0.002
10.1
(5(5(5(5 777.7 111-19999.0)0)0)0)
0 91 1 66
30/5.8151151
(0(0(0( .5.55.52%2%2%2%) )
5555555 //5/5/ .7969699
(0(0(0(0.99995%5%5%5%) ) ) )
92929292/1/1/1/ 5.55.434343435555
(0(0(0(0.6.6.66%)%)%)%)
00 9191 11 6666
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
18
Table 3. Association between rs10455872, lipoprotein(a) levels and aortic valve stenosis (AVS) risk in EPIC-Norfolk.
Lipoprotein(a) levels are presented as median (interquartile range) and were log-transformed in the Cox proportional hazard model.
AA AG GG P-value AG+GG (vs. AA) P-value
Lipoprotein(a) levels, mg/dL9.7
(5.6-17.5)
45.1
(34.9-57.7)
69.8
(49.9-88.1) <0.001
45.9
(35.4-60.0) <0.001
Event rate75/12.434
(0.6%)
22/2.066
(1.1%)
4/134
(2.9%) <0.001
26/2.200
(1.2%) 0.003
Hazard ratio for AVS,
unadjusted1.00
1.78
(1.11-2.87)
4.83
(1.77-13.20)
1.98
(1.27-3.09)
Hazard ratio for AVS, adjusted
for Lp(a) levels1.00
1.56
(0.90-2.70)
3.83
(1.26-11.65)
1.66
(1.27-3.09)
Hazard ratio for AVS, adjusted
for logLp(a) 1.00
1.63
(0.92-2.70)
4.28
(1.44-12.79)
1.77
(1.01-3.07)
Hazard ratio for AVS, adjusted
for Lp(a) quintiles1.00
1.57
(0.84-2.93)
4.13
(1.35-12.57)
1.71
(0.94-3.11)
000001010101 26222
((((
0
0
0 1.78788
(111.111-11 2.222 8788 ) ))
44.83
(1111 77.77777-7 131313 2.2220)0))0) (11(1.
0001.1.1.1.56565656 3.3.3.3.83838383
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
19
Table 4. Association in cases and controls from the Montreal Heart Institute Biobank for aortic valve stenosis (AVS) with the rs10455872 SNP in the LPA gene region along with other SNPs with P value below 1x10-3 obtained with logistic regression after adjusting for age and gender.
rs number Position on Chr 6 Gene n MAF Odds ratio (95% CI) P value
rs3106164 160850273 SLC22A3 783 0.30 0.68 (0.54-0.85) 0.0008
rs9346817 160902269 LPAL2 783 0.16 0.61 (0.46-0.80) 0.0005
rs9355803 160903178 LPAL2 781 0.16 0.61 (0.46-0.81) 0.0006
rs9355804 160903622 LPAL2 783 0.16 0.61 (0.46-0.80) 0.0005
rs10945673 160910899 LPAL2 783 0.16 0.60 (0.45-0.80) 0.0004
rs7749836 160914343 LPAL2 783 0.16 0.61 (0.46-0.80) 0.0005
rs12212724 160917135 LPAL2 783 0.16 0.61 (0.46-0.80) 0.0005
rs10455872 161010118 LPA 763 0.11 1.57 (1.10-2.26) 0.01
*Chr = Chromosome, MAF = Minor Allele Frequency and CI = Confidence Interval.
0.0.0.0 61616161 ((((0.0.0.0.46464646
LPAL2 781 0 16 0.61 (0.46LPLPLPLPALAAA 2 781 0.16 0.61 (0.46
LPPALA 22 78783 0.0.0.0.16161 00.00 616166 (0..446
LPALALALAL2222 787878783333 0.16161616 0.60 (0.45
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
DOI: 10.1161/CIRCGENETICS.113.000400
20
Figures Legend:
Figure 1. Risk of aortic valve stenosis (AVS) associated with the presence of the rs10455872
allele conferring elevated plasma Lp(a) levels and the equivalent rise (31.1 mg/dL) in plasma
Lp(a) level.
Figure 2. Regional plot of single nucleotide polymorphisms in the IGF2R-SLC22A1-A2-A3-
LPAL2-LPA-PLG region with minor allele frequencies equal or above 5% were genotyped in the
Montreal Heart Institute Biobank and their association with aortic valve stenosis. The x-axis
represents the chromosomal position and the y-axis represents the negative of the logarithm of p-
value for the different genetic association test.
5% were ggenotypypppedededed
ve steno iiisis. TTTThehhh xx---aaaaxi
t m
h
the ccchrhrhrhromomomosososommmalaaa position and the y-axissss rrreepresents the negagagative of the logarithm
hee e dddifferent geneneeticc aaassoooocicicic atioooonn teeesst.
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
Figure 1.
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
Figure 2.
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
Kay-Tee Khaw, Manjinder S. Sandhu and Jean-Claude TardifBenoit J. Arsenault, S. Matthijs Boekholdt, Marie-Pierre Dubé, Éric Rhéaume, Nicholas J. Wareham,
Randomization Study and Replication in a Case-Control CohortLipoprotein(a) Levels, Genotype and Incident Aortic Valve Stenosis: A Prospective Mendelian
Print ISSN: 1942-325X. Online ISSN: 1942-3268 Copyright © 2014 American Heart Association, Inc. All rights reserved.
TX 75231is published by the American Heart Association, 7272 Greenville Avenue, Dallas,Circulation: Cardiovascular Genetics
published online April 5, 2014;Circ Cardiovasc Genet.
http://circgenetics.ahajournals.org/content/early/2014/04/05/CIRCGENETICS.113.000400World Wide Web at:
The online version of this article, along with updated information and services, is located on the
http://circgenetics.ahajournals.org/content/suppl/2014/04/05/CIRCGENETICS.113.000400.DC1Data Supplement (unedited) at:
http://circgenetics.ahajournals.org//subscriptions/
is online at: Circulation: Cardiovascular Genetics Information about subscribing to Subscriptions:
http://www.lww.com/reprints Information about reprints can be found online at: Reprints:
document. Permissions and Rights Question and Answer this process is available in the
located, click Request Permissions in the middle column of the Web page under Services. Further information aboutnot the Editorial Office. Once the online version of the published article for which permission is being requested is
can be obtained via RightsLink, a service of the Copyright Clearance Center,Circulation: Cardiovascular Genetics Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:
by guest on June 27, 2018http://circgenetics.ahajournals.org/
Dow
nloaded from
SUPPLEMENTAL MATERIAL
Supplementary Table 1. Clinical characteristics of the nested case-control population from the
Montreal Heart Institute Biobank.
Data are shown as mean (standard deviation) for continuous variables and as number
(percentage) for categorical variables. AVS = Aortic valve stenosis, MI = myocardial infarction,
PCI = percutaneous coronary intervention, CABG = coronary artery bypass grafting and
CAD = coronary artery disease. *Significantly different (p<0.05) from cases shown by Student
unpaired t-tests for continuous variables and chi-square tests for categorical variables.
With AVS
N=405
Without AVS
N=415
Age, years 72.4 (7.6) 69.2 (7.1)*
Male gender 266 (65.7) 273 (65.8)
French Canadian 347 (85.7) 353 (85.1)
Body mass index, kg/m2 29.0 (5.2) 28.6 (5.7)
Systolic blood pressure, mmHg 130 (17) 124 (15)*
Diastolic blood pressure, mmHg 71 (10) 71 (9)
Hypertension 319 (78.8) 269 (64.8)
Current/previous smoking 272 (67.2) 281 (66.9)
CAD
- Previous MI 94 (23.2) 169 (40.8)*
- Previous PCI 83 (20.5) 121 (29.2)*
- Previous CABG 141 (34.8) 111 (26.8)*
Previous stroke 37 (9.1) 58 (14.0)*
Diabetes mellitus 133 (32.8) 96 (23.1)*
Statin use 284 (70.1) 283 (68.2)
Supplementary Table 2. Echocardiographic characteristics of patients with aortic valve stenosis
in the nested case-control population from the Montreal Heart Institute Biobank.
Data are shown as mean (standard deviation) for continuous variables and as number
(percentage) for categorical variables.
Echocardiographic characteristic N Genetic study
MHI Biobank
(N= 405)
Aortic valve stenosis severity
- Unknown 8 (2.0)
- Mild 54 (13.3)
- Moderate 131 (32.3)
- Severe 212 (52.3)
Max transvalvular aortic jet velocity, m/s 329 3.75 (0.91)
Mean transvalvular gradient, mmHg 395 36.8 (17.7)
Aortic valve area, cm2 391 1.03 (0.36)
Left ventricular ejection fraction, % 392 59.4 (9.7)
Indexed left ventricular mass, g/m2 328 112 (33.8)