malignantwww.niuav.com

RESEARCHARTICLEOpenAccessSerumTGF-β1andSMAD3levelsarecloselyassociatedwithcoronaryarterydiseaseCanChen1,2,WeiLei2,WenjiangChen2,JianfengZhong2,XiaoxinGao2,BoLi2,HuailongWang2andCongxinHuang1*AbstractBackground:Coronaryarterydisease(CAD)isoneofthemostcommondiseasesleadingtomortalityandmorbidityworldwide.
Thereisconsiderabledebateonwhetherserumtransforminggrowthfactorβ1(TGF-β1)levelsareassociatedwithlong-termmajoradversecardiovasculareventsinpatientswithCAD,andtodate,nostudyhasspecificallyaddressedlevelsinpatientswithdifferentdegreesofCADseverity.
Methods:SerumTGF-β1andmothersagainstdecapentaplegichomolog3(SMAD3)concentrationswereevaluatedin279patientswithCADand268controlswithoutCAD.
Theclinicalandbiochemicalcharacteristicsofallsubjectswerealsodeterminedandanalyzed.
Results:TGF-β1andSMAD3concentrationsinCADpatientsweresignificantlyhigherthanthoseinthecontrols.
TheserumTGF-β1levelinacutemyocardialinfarction(AMI)wassignificantlyhigherthanthatinbothstableanginapectoris(SAP)andunstableanginapectoris(UAP)(p0.
05).
SMAD3levelsshowednoobviousdifferenceamongAMI,SAP,andUAP.
TGF-β1andSMAD3arepotentialbiomarkersforCAD,andmaybemoreaccuratethanLpa,ApoA1,uricacid,BUN,ortriglycerides(TG).
Conclusions:SerumTGF-β1andSMAD3levelsarecloselyassociatedwithCAD,andmaybecomeusefulbiomarkersfordiagnosisandriskstratification.
Keywords:Coronaryarterydisease,TGF-β1,SMAD3protein,Human,BiomarkersBackgroundCoronaryarterydisease(CAD)isoneoftheleadingcausesofmortalityandmorbidityworldwide,especiallyinmanydevelopedcountries[1,2].
AccordingtotheWorldHealthOrganization,itispredictedthat23.
6milliondeathsperyearwillbebecauseofcardiovascu-lardiseasesby2030[3].
CAD,withacomplexetiology,isconsideredtobetheresultofaninteractionbetweengeneticandenvironmentalfactors[4-6].
Inthepastdecades,manycontributingfactorsincludingsmoking,hypertension,diabetesmellitus(DM),atherosclerosis,obesity,anddiethavebeenestablished,buttheexactetiologyunderlyingCADremainsobscure.
IncreasingevidencesuggeststhatinflammationplaysanimportantroleinthepathogenesisofCAD[7,8].
AnumberofinflammatorycytokinesmediateadversecardiovasculareventsinpatientswithCAD,andtrans-forminggrowthfactor-β1(TGF-β1)hasgarneredpar-ticularattentionbecauseofitsmultiplerolesinimportantpathologicalchangessuchasenhancementofmacrophageandfibroblastchemotaxis,stimulationofextracellularmatrix(ECM)synthesis,andvascularcellproliferationab-normalities[9,10].
AlthoughithasbeenknownthatTGF-β1activatesseveralpathways,includingtheextracellular-regulatedkinasepathway,thenuclearfactor-κBpathway,andthephosphatidylinositol-3-kinasepathway[11,12],itsmainsignalingmechanismislinkedtothemothersagainstdecapentaplegichomolog3(SMAD3)family[13].
TGF-β1bindstoitstypeIreceptorsandformsahetero-mericcomplexwiththetypeIIreceptor,whichsubse-quentlyresultsinphosphorylationandactivationofthe*Correspondence:huangcongxin@vip.
163.
com1DepartmentofCardiovascularMedicine,RenminHospitalofWuhanUniversity,Wuhan430060,ChinaFulllistofauthorinformationisavailableattheendofthearticle2014Chenetal.
;licenseeBioMedCentralLtd.
ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(http://creativecommons.
org/licenses/by/2.
0),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycredited.
TheCreativeCommonsPublicDomainDedicationwaiver(http://creativecommons.
org/publicdomain/zero/1.
0/)appliestothedatamadeavailableinthisarticle,unlessotherwisestated.
Chenetal.
BMCCardiovascularDisorders2014,14:18http://www.
biomedcentral.
com/1471-2261/14/18TGF-β1downstreamsignalingmoleculesSMAD2andSMAD3.
ActivatedSMAD2orSMAD3heterodimerizeswithSMAD4andthenintroducestheabovecomplexintothenucleus,whereitregulatestheexpressionofitstargetgene[14,15].
TGF-β1/SMAD3-dependentpathwaysplayapivotalroleinmediatingdifferentbiologicaleffectsofTGF-β1suchascellproliferation,immunesuppression,andinflammation[12,16-18].
Inthecardiovascularsystem,myocardialTGF-β1ex-pressionismarkedlyactivatedinpatientswithhyper-trophicordilatedcardiomyopathy,andinexperimentalmodelsofmyocardialhypertrophyandmyocardialin-farction[19].
OtherstudiessuggestthatTGF-βsignal-ingmaybecrucialforrepressionofinflammatorygeneexpressioninhealinginfarctsmediatedbyaninflam-matoryinfiltrate[20].
TGF-β1iscrucialinthepatho-genesisofinfarcthealing,cardiacremodeling,andinterstitialfibrosis.
Therehavebeenfewclinicalinves-tigationsofTGF-β1/SMAD3signalsinlargepatientgroups.
Thereisconsiderabledebateoveranycorrel-ationofTGF-β1levelswithmajoradversecardiovas-culareventsinpatientswithCAD.
Inparticular,nostudyhasspecificallyaddressedpatientswithdifferentdegreesofCAD[9].
Wecollectedserumsamplesfrom279patientswithCADandfrom297controls,andassessedforanyasso-ciationbetweenserumTGF-β1/SMAD3levelsandthepresenceandseverityofCAD.
Weevaluatedthepoten-tialclinicalapplicationofmeasurementsofthesecyto-kinesforCADdiagnosis,comparingtheirutilitywiththecurrentlyusedbiochemicalindicators.
MethodsStudypopulationWritteninformedconsentwasobtainedfromthepopu-lationinvolvedinthisstudyandthestudyprotocolwasapprovedbytheEthicsCommitteeofAffiliatedHospitalofGuangdongMedicalCollege.
Atotalof279patientswhowerereferredforcoronaryangiographybytheirattendingphysiciansfromMaytoAugust2013,andwhowerefoundtohaveatleast50%stenosisinatleastonecoronaryartery,wereincluded.
Theirageswereabout69.
30years.
The268controlswereconsecutivesubjectsundergoingroutinemedicalexaminationsatthePhysicalExaminationCenter.
Thosewhohadanyclinicalmanifestationsoramedicalhistoryofheartdisorders,afamilyhistoryofcoronaryarterydisease,orabnormalECGwereexcluded.
Exclusioncri-teriaalsoincludedacuteorchronicinfectionsorinflam-matorydiseases,severehepaticorrenaldysfunction,malignanttumors,orhematologicdisorders.
TheCADgroupwasdividedintothreesubgroups:acutemyocar-dialinfarction(AMI),unstableanginapectoris(UAP),andstableanginapectoris(SAP).
PhysiologicalevaluationandbloodcollectionAllstudyparticipantsunderwentastandardclinicalexamination.
Bodymassindex(BMI)wascalculatedastheindividual'sbodymassdividedbythesquareoftheirheight.
Smokingwasdefinedashavingsmokedforatleast1yearcurrently.
Hypertensionwasdefinedasasys-tolicbloodpressureof≥140mmHgoruseofantihyper-tensivetherapy.
Diabeteswasdiagnosedaccordingtononfastingglucoselevels≥11.
1mmol/L.
Approximately2mLofvenousbloodwascollectedfromeachsubject,andthenkeptintubeswithoutpreservativesat4°Covernight.
Serumaliquotswereobtainedaftercentrifugationat1000gfor40minat4°C.
CoronaryangiographyAllcoronaryangiographieswereperformedusingstand-ardtechnique.
SignificantCADwasconsideredtobepresentiftherewasaninternalluminalstenosis≥50%intheleftmaincoronaryartery,rightcoronaryartery,and/ortheirmajorbranches.
Thedegreeofcoronaryatherosclerosiswasfurthercategorizedaccord-ingtothenumberofcoronaryvesselswithsignificantTable1ClinicalandbiochemicalcharacteristicsofthestudypopulationVariablesCADControlpSex(M/F)168/111180/880.
0913Age(years)69.
30±0.
704963.
73±0.
69520.
05)(Figure2).
However,SMAD3levelswerenotsignificantlydifferentamongAMI,SAP,andUAPsubgroups,suggestingthatTGF-β1mayregulatetheprogressionfromanginapectoristoAMI.
CorrelationofTGF-β1andSMAD3levelswithCADoccurrenceanditsriskfactorsToexplainapotentialmechanismofTGF-β1andSMAD3,weevaluatedtheirrelationshipwithCADanditsriskfactorsincludingLp(a),BUN,creatinine,uricacid,bloodglucose,totalcholesterol,TG,HDL,LDL,ApoA1,andApoB.
Therewasaverysignificantcorrel-ationbetweenCADandconcentrationofTGF-β1andSMAD3(p<0.
0001).
Amongtheotherbiochemicalindices[21],TGF-β1significantlycorrelatedwithLp(a)(p<0.
0203)anduricacid(p<0.
0001),respectively,andSMAD3corre-latedverysignificantlywithbloodglucose(p<0.
0001)(Figure3).
TGF-β1andSMAD3asbiomarkersforCADFarhigherTGF-β1andSMAD3concentrationsoccurredinCADpatients(TGF-β1:432.
2±22.
12ng/L;SMAD3:11.
47±0.
6161ng/L)comparedwithcontrols(TGF-β1:220.
1±8.
831ng/L;SMAD3:5.
157±0.
1965ng/L).
TheareaundertheReceiveroperatingcharacteristiccurves(AUCROC)indicatedthatbothTGF-β1(AUCROC:0.
678,95%confidenceinterval[CI]:0.
633–0.
723)andSMAD3(AUCROC:0.
715,95%CI:0.
672–0.
758)werethemorepotentbiomarkersforCAD,comparedwithLpa(AUCROC:0.
574,95%CI:0.
525–0.
623),ApoA1(AUCROC:0.
561,95%CI:0.
512–0.
609),uricacid(AUCROC:0.
663,95%CI:0.
617–0.
710),BUN(AUCROC:0.
581,95%CI:0.
533–0.
630),orTG(AUCROC:0.
545,95%CI:0.
496–0.
594)(Figure4).
DiscussionTGF-β1isamultifunctionalpeptidethatcontrolsprolif-eration,differentiation,andotherfunctionsinmanycelltypes,andithasdualregulationrolesinimmunere-sponseandcellulardevelopment.
TGF-β1actssynergis-ticallywithTGFαininducingtransformation.
Italsoactsasanegativeautocrinegrowthfactor.
DysregulationofTGF-β1activationandsignalingmayresultinapop-tosis.
Forexample,TGF-β1inhibitscancercellgrowthattheearlystageoftumorformation,andpromotesthecancerdevelopmentinthelatephase[22].
PreviousstudieshaveshownplasmaTGF-β1levelstobereducedinpatientswithadvancedatherosclerosisandangiographicallyprovenCAD.
ThelevelsofTGF-β1werethoughttobeinverselyrelatedtothedevelopmentandseverityofcoronarydisease[23-27].
Bycontrast,BorderandRuoslahtifoundthatTGF-β1couldenhanceatherogenesisbymediatingexcessiveECMdeposition[28].
AnotherrecentstudyrevealedthatahighplasmalevelofTGF-βhadasignificantlystrongprognosisintermsofsurvivalwithoutcardiovasculareventsandsur-vivalwithoutcoronaryinterventionscomparedwiththelowTGF-βgroup(bothp<0.
05),suggestingthatplasmaTGF-βmaypotentiallyhavegreatprognosticvalueinpatientswithCAD[29].
SchaanetalmaintainedthatserumTGF-β1wasnotassociatedwithCADoccurrenceafterclinicalandlaboratoryevaluationofTGF-β1inpa-tientswithCADorDM[9].
However,theysampledFigure3CorrelationofTGF-β1orSMAD3withtheriskfactorsofCAD,(A)CorrelationanalysisofTGB-β1withLPa;(B)CorrelationanalysisofTGB-βwithSUA;(C)SMAD3withGlu.
***p<0.
0001,*p<0.
05.
Chenetal.
BMCCardiovascularDisorders2014,14:18Page5of7http://www.
biomedcentral.
com/1471-2261/14/18fewerthan30cases.
Wedemonstratedsignificantposi-tivecorrelationbetweenserumTGF-β1/SMAD3levelsandCADbasedon547subjects.
TherehasbeennostudystratifyingpatientsaccordingtoCADseverity.
ItisprobablethattherelevanceofTGF-β1inCADcanonlybedetectedinseveredisease,orhighconcentrationsofTGF-β1fromsevereCADmayhaveaparadoxicaleffect[27,30,31].
InthecorrelationofTGF-β1andSMAD3levelswiththeriskfactorsofCAD,TGF-β1iscloselyrelatedtoLp(a)anduricacid,whichbothareconsideredtobemarkersofatherosclerosis,indicatingthatthiscytokinemaycontributetotheestablishmentofCADbyregulat-ingatherogenesis.
SMAD3levelscorrelatedcloselywithbloodglucose,implyingthatSMAD3mayinfluenceCADoccurrenceandthedevelopmentofDM.
Interest-ingly,ithasbeenreportedthatriskfactorsincludingLp(a),uricacid,BUN,triglycerides,andApoA1canbeuse-fulbiomarkersfordifferentclinicaldiagnoses[21,32],butwefoundthatTGF-β1andSMAD3weremoresen-sitivebiomarkersforCADthanthosefactorsmentionedabove.
ConclusionSerumTGF-β1andSMAD3levelsarecloselyassociatedwithCAD.
Theunderlyingmechanismmaybetheirregulatoryeffectsonatherosclerosisandbloodsugar.
Alimitationofthepresentstudyisthatthecontrolpopulationdiffersfromthecases,intermsofageandcardiovascularriskfactors.
Themeanagedifference(~5years)shouldbeacceptableinanepidemiologicstudyofCADaccordingtoapreviousreport[6,9].
CADpredominatesinanolderpopulation.
CompetinginterestsTheauthorshadnoconflictsofinteresttodeclareinrelationtothisarticle.
Authors'contributionsWLandCCperformedthestudyandwrotethemanuscript.
WC,JZ,andWLperformedthestudyand/orcontributedtodataanalysisandinterpretation.
XG,BL,andHWreviewed/approvedtheresearchprotocol.
CHtakesfullresponsibilityfortheworkasawhole,includingthestudydesign,accesstodata,andthedecisiontosubmitandpublishthemanuscript.
Allauthorsreadandapprovedthefinalmanuscript.
AcknowledgmentsThispaperwassupportedbytheDoctorScientificResearchFoundationinGuangdongMedicalCollege(no.
10201B01202andXB1322)andtheNaturalScienceFoundationofGuangdongProvince(S2013040012115andS2013010015074)andtheNationalNaturalScienceFoundationofChina(no.
81270212andno.
81300035).
Authordetails1DepartmentofCardiovascularMedicine,RenminHospitalofWuhanUniversity,Wuhan430060,China.
2DepartmentofCardiovascularMedicine,TheAffiliatedHospitalofGuangdongMedicalCollege,Zhanjiang524000,China.
Received:23October2013Accepted:11February2014Published:17February2014Figure4Receiveroperatingcharacteristic(ROC)curvesforTGF-β1andSMAD3.
TheareaunderthecurveofTGF-β1was0.
678(95%confi-denceinterval[CI],0.
633–0.
723).
TheareaunderthecurveofSMAD3was0.
715(95%CI,0.
672–0.
758).
Chenetal.
BMCCardiovascularDisorders2014,14:18Page6of7http://www.
biomedcentral.
com/1471-2261/14/18References1.
Lloyd-JonesD,AdamsR,CarnethonM,DeSimoneG,FergusonTB,FlegalK,FordE,FurieK,GoA,GreenlundK,HaaseN,HailpernS,HoM,HowardV,KisselaB,KittnerS,LacklandD,LisabethL,MarelliA,McDermottM,MeigsJ,MozaffarianD,NicholG,O'DonnellC,RogerV,RosamondW,SaccoR,SorlieP,StaffordR,SteinbergerJ,etal:AmericanHeartAssociationStatisticsCommitteeandStrokeStatisticsSubcommittee.
Heartdiseaseandstrokestatistics—2009update:areportfromtheAmericanHeartAssociationStatisticsCommitteeandStrokeStatisticsSubcommittee.
Circulation2009,119:480–486.
2.
MininoAM:DeathintheUnitedStates,2009.
NCHSDataBrief2011,64:1–8.
3.
WorldHealthOrganizationCardiovascularDiseases(CVDs):WhoFactSheet.
;2011.
4.
KathiresanS,SrivastavaD:Geneticsofhumancardiovasculardisease.
Cell2012,148:1242–1257.
5.
StylianouIM,BauerRC,ReillyMP,RaderDJ:Geneticbasisofatherosclerosis:insightsfrommiceandhumans.
CircRes2012,110:337–355.
6.
TayebiN,KeT,FooJN,FriedlanderY,LiuJ,HengCK:Associationofsinglenucleotidepolymorphismrs6903956onchromosome6p24.
1withcoronaryarterydiseaseandlipidlevelsindifferentethnicgroupsoftheSingaporeanpopulation.
ClinBiochem2013,46:755–759.
7.
SugamuraK,KeaneyJFJr:Reactiveoxygenspeciesincardiovasculardisease.
FreeRadicBiolMed2011,51:978–992.
8.
TousoulisD,KampoliAM,PapageorgiouN,AndroulakisE,AntoniadesC,ToutouzasK,StefanadisC:Pathophysiologyofatherosclerosis:theroleofinflammation.
CurrPharmDes2011,17:4089–4110.
9.
SchaanBD,QuadrosAS,Sarmento-LeiteR,DeLuccaG,JrBA,BertoluciM:'Correction:'Serumtransforminggrowthfactorbeta-1(TGF-beta-1)levelsindiabeticpatientsarenotassociatedwithpre-existentcoronaryarterydisease.
CardiovascDiabetol2007,6:19.
10.
GraingerDJ:TGF-betaandatherosclerosisinman.
CardiovascRes2007,74:213–222.
11.
BakinAV,TomlinsonAK,BhowmickNA,MosesHL,ArteagaCL:Phosphatidylinositol3-kinasefunctionisrequiredfortransforminggrowthfactorbeta-mediatedepithelialtomesenchymaltransitionandcellmigration.
JBiolChem2000,275:36803–36810.
12.
CaraciF,BattagliaG,BuscetiC,BiagioniF,MastroiacovoF,BoscoP,DragoF,NicolettiF,SortinoMA,CopaniA:TGF-beta1protectsagainstAbeta-neurotoxicityviathephosphatidylinositol-3-kinasepathway.
NeurobiolDis2008,30:234–242.
13.
EdlinRS,TsaiS,YamanouchiD,WangC,LiuB,KentKC:Characterizationofprimaryandrestenoticatheroscleroticplaquefromthesuperficialfemoralartery:potentialroleofSmad3inregulationofSMCproliferation.
JVascSurg2009,49:1289–1295.
14.
MiyazawaK,ShinozakiM,HaraT,FuruyaT,MiyazonoK:TwomajorSmadpathwaysinTGF-superfamilysignalling.
GenesCell2002,7:1191–1204.
15.
XuF,LinSH,YangYZ,GuoR,CaoJ,LiuQ:Theeffectofcurcuminonsepsis-inducedacutelunginjuryinaratmodelthroughtheinhibitionoftheTGF-β1/SMAD3pathway.
IntImmunopharmacol2013,16:1–6.
16.
DerynckR,ZhangYE:Smad-dependentandSmad-independentpathwaysinTGF-betafamilysignalling.
Nature2003,425:577–584.
17.
ZhuY,CulmseeC,KlumppS,KrieglsteinJ:Neuroprotectionbytransforminggrowthfactor-beta1involvesactivationofnuclearfactor-kappaBthroughphosphatidylinositol-3-OHkinase/Aktandmitogen-activatedproteinkinase-extracellular-signalregulatedkinase1,2signalingpathways.
Neuroscience2004,123:806–897.
18.
GoelSA,GuoLW,ShiXD,KundiR,SovinskiG,SeedialS,LiuB,KentKC:Preferentialsecretionofcollagentype3versustype1fromadventitialfibroblastsstimulatedbyTGF-β/Smad3-treatedmedialsmoothmusclecells.
CellSignal2013,25:9559–9560.
19.
DobaczewskiM,ChenW,FrangogiannisNG:Transforminggrowthfactor(TGF)-βsignalingincardiacremodeling.
JMolCellCardiol2011,51:600–606.
20.
BujakM,FrangogiannisNG:TheroleofTGF-betasignalinginmyocardialinfarctionandcardiacremodeling.
CardiovascRes2007,74:184–195.
21.
LiuB,ZhangYM:Studyonriskfactorsofprematurecoronaryarterydisease.
ChinJCardiovascRes2009,7:629–632.
22.
LinWT,WuXN,ZhuPP,ZhangZF,ChenJ:Studiesonanti-tumoractivitiesofenteromorphachlorellagrowthfactorinexperimentallivercanceranditsmechanism.
JFujianMedUniv2009,43:227–230.
23.
GraingerDJ,KempPR,MetcalfeJC,LiuAC,LawnRM,WilliamsNR,GraceAA,SchofieldPM,ChauhanA:Theserumconcentrationofactivetransforminggrowthfactor-betaisseverelydepressedinadvancedatherosclerosis.
NatMed1995,1:74–79.
24.
CrookR,LeathamEW,SalomoneOA,Hossein-NiaM,GraingerDJ,KaskiJC:Plasmalevelsofactivetransforminggrowthfactor-betaarereducedinpatientswiththreevesselcoronaryarterydisease.
Heart1996,75:79.
25.
TashiroH,ShimokawaH,YamamotoK,MomoharaM,TadaH,TakeshitaA:Alteredplasmalevelsofcytokinesinpatientswithischemicheartdisease.
CoronArteryDis1997,8:143–147.
26.
LiCG,BethellH,WilsonPB,BhatnagarD,WalkerMG,KumarS:ThesignificanceofCD105,TGFbetaandCD105/TGFbetacomplexesincoronaryarterydisease.
Atherosclerosis2000,152:249–256.
27.
BatumanO,GoD,ClarkLT,SmithEL,ClementsP,FeitA,LedererD:Relationshipbetweencytokinelevelsandcoronaryarterydiseaseinwomen.
HeartDis2001,3:80–84.
28.
BorderWA,RuoslahtiE:Transforminggrowthfactor-betaindisease:thedarksideoftissuerepair.
JClinInvest1992,90:1–7.
29.
TashiroH,ShimokawaH,SadamatuK,YamamotoK:Prognosticsignificanceofplasmaconcentrationsoftransforminggrowthfactor-betainpatientswithcoronaryarterydisease.
CoronArteryDis2002,13:139–143.
30.
StefoniS,CiancioloG,DonatiG,DormiA,SilvestriMG,ColiL,DePascalisA,IannelliS:LowTGF-beta1serumlevelsareariskfactorforatherosclerosisdiseaseinESRDpatients.
KidneyInt2002,61:324–335.
31.
OsI,DjurovicS,SeljeflotI,BergK:Transforminggrowthfactor(TGF)-beta1inverselyrelatedtovascularcelladhesionmolecule-1inpostmenopausalwomenwithcoronaryarterydisease.
ApossiblemechanismfortheputativecardioprotectiveroleofTGF-beta1JInternMed2002,251:223–227.
32.
WenHH,BaiP,NiuLJ,LiYD:Theinfluenceofserumuricacidonacutemyocardialinfarctionanditspredicativevalue.
ActaAcadMedNeiMongol2006,28:27–31.
doi:10.
1186/1471-2261-14-18Citethisarticleas:Chenetal.
:SerumTGF-β1andSMAD3levelsarecloselyassociatedwithcoronaryarterydisease.
BMCCardiovascularDisorders201414:18.
SubmityournextmanuscripttoBioMedCentralandtakefulladvantageof:ConvenientonlinesubmissionThoroughpeerreviewNospaceconstraintsorcolorgurechargesImmediatepublicationonacceptanceInclusioninPubMed,CAS,ScopusandGoogleScholarResearchwhichisfreelyavailableforredistributionSubmityourmanuscriptatwww.
biomedcentral.
com/submitChenetal.
BMCCardiovascularDisorders2014,14:18Page7of7http://www.
biomedcentral.
com/1471-2261/14/18