BiomarkerPro?lesinHeartFailurePatientsWithPreservedand
ReducedEjectionFraction
JasperTromp,MD;MohsinA.F.Khan,PhD;IJsbrandT.Klip,MD;SvenMeyer,MD;RudolfA.deBoer,MD,PhD;TinyJaarsma,RN,PhD;
HansHillege,PhD;DirkJ.vanVeldhuisen,MD,PhD;PetervanderMeer,MD,PhD;AdriaanA.Voors,MD,PhD
Background-—Biomarkersmayhelpustounraveldifferencesintheunderlyingpathophysiologybetweenheartfailure(HF)patients
withareducedejectionfraction(HFrEF)andapreservedejectionfraction(HFpEF).Therefore,wecomparedbiomarkerpro?lesto
characterizepathophysiologicaldifferencesbetweenpatientswithHFrEFandHFpEF.
MethodsandResults-—Weretrospectivelyanalyzed33biomarkersfromdifferentpathophysiologicaldomains(in?ammation,
oxidativestress,remodeling,cardiacstretch,angiogenesis,arteriosclerosis,andrenalfunction)in460HFpatients(21%HFpEF,left
ventricularejectionfraction≥45%)measuredatdischargeafterhospitalizationforacuteHF.Theassociationbetweenthese
markersandtheoccurrenceofall-causemortalityand/orHF-relatedrehospitalizationsat18monthswascomparedbetween
patientswithHFrEFandHFpEF.Patientswere70.6C611.4yearsoldand37.4%werefemale.PatientswithHFpEFwereolder,more
oftenfemale,andhadahighersystolicbloodpressure.Levelsofhigh-sensitiveC-reactiveproteinweresigni?cantlyhigherin
HFpEF,whilelevelsofpro-atrial-typenatriureticpeptideandN-terminalpro-brainnatriureticpeptidewerehigherinHFrEF.Linear
regressionfollowedbynetworkanalysesrevealedprominentin?ammationandangiogenesis-associatedinteractionsinHFpEFand
mainlycardiacstretch–associatedinteractionsinHFrEF.Theangiogenesis-speci?cmarker,neuropilinandtheremodeling-speci?c
marker,osteopontinwerepredictiveforall-causemortalityand/orHF-relatedrehospitalizationsat18monthsinHFpEF,butnotin
HFrEF(Pforinteraction<0.05).
Conclusions-—InHFpEF,in?ammationandangiogenesis-mediatedinteractionsarepredominantlyobserved,whilestretch-
mediatedinteractionsarefoundinHFrEF.Theremodelingmarkerosteopontinandtheangiogenesismarkerneuropilinpredicted
outcomeinHFpEF,butnotinHFrEF.(JAmHeartAssoc.2017;6:e003989.DOI:10.1161/JAHA.116.003989.)
KeyWords:biomarkerheartfailuremultimarkerpathophysiology
T
hedifferenceinpathophysiologybetweenheartfailure
withareducedejectionfraction(HFrEF)andheartfailure
withapreservedejectionfraction(HFpEF)remainspoorly
understood,andeffectivetreatmentoptionsarecurrentlynot
availableforHFpEF.
1–4
Therefore,abetterunderstandingof
thepathophysiologyofHFpEFisrequired,whicheventually
mayhelptoimproveoutcome.
Patient-speci?cbiomarkerpro?lesareusefulforthe
purposeofmonitoringdiseaseseverityandprogression,to
guidetherapy,butalsoforcharacterizingthepathophysiology
ofHF.
5–9
Wehypothesizethatdifferencesinbiomarkerlevels
andcorrelativeassociationsbetweenHFrEFandHFpEFmay
provideimportantinsightsintospeci?cactivitiesofpatho-
physiologicalprocesses.
5–9
TheaimofthisstudywastocharacterizeHFpEFandHFrEF
usinganetworkanalysisonanextensivesetof33biomarkers
ofvariouspathophysiologicalpathways.Therefore,weinves-
tigateddifferencesinbiomarkerlevels,patternsof
FromtheDepartmentofCardiology,UniversityMedicalCenterGroningen,UniversityofGroningen,TheNetherlands(J.T.,M.A.F.K.,I.T.K.,S.M.,R.A.d.B.,H.H.,D.J.v.V.,
P.v.d.M.,A.A.V.);HeartFailureResearchCenter,AcademicMedicalCenter,Amsterdam,TheNetherlands(M.A.F.K.);DepartmentofCardiology,HeartCenter
Oldenburg,EuropeanMedicalSchoolOldenburg-Groningen,CarlvonOssietzkyUniversityOldenburg,Oldenburg,Germany(S.M.);DepartmentofSocial-andWelfare
Studies,FacultyofMedicalandHealthSciences,Link€opingUniversity,Link€oping,Sweden(T.J.).
AccompanyingTablesS1throughS5andFiguresS1throughS6areavailableathttp://jaha.ahajournals.org/content/6/4/e003989/DC1/embed/inline-suppleme
ntary-material-1.pdf
DrTrompandDrKhancontributedequallytothiswork.
Correspondenceto:AdriaanA.Voors,MD,PhD,DepartmentofCardiology,UniversityMedicalCenterGroningen,Hanzeplein1,Groningen9713GZ,The
Netherlands.E-mail:a.a.voors@umcg.nl
ReceivedJune17,2016;acceptedJanuary4,2017.
a2017TheAuthors.PublishedonbehalfoftheAmericanHeartAssociation,Inc.,byWileyBlackwell.ThisisanopenaccessarticleunderthetermsoftheCreative
CommonsAttribution-NonCommercialLicense,whichpermitsuse,distributionandreproductioninanymedium,providedtheoriginalworkisproperlycitedandis
notusedforcommercialpurposes.
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation1
ORIGINALRESEARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
correlations,andpredictivevalueofbiomarkersinpatients
withHFpEFandHFrEF.
Methods
StudyDesignandPopulation
Measurementsofbiomarkerswereperformedinasubcohort
oftheCoordinatingstudyevaluatingOutcomesofAdvising
andCounselinginHeartfailure(COACH)trialofwhich
rationale,design,andresultshavebeenpreviously
described.
10,11
Inshort,theCOACHtrialstudiedtheeffects
ofadditionalintensivenurse-ledsupportontheprognosisof
1023chronicHFpatients.AhospitaladmissionforHF(NYHA
II-IV)inclusioncriteriafortheCOACHtrialincludedand
patientshadtobeatleast18yearsofage.Patientswere
excludediftheyunderwentanintervention(percutaneous
transluminalcoronaryangioplasty,coronaryarterybypass
graft,hearttransplantation,valvereplacement)intheprevious
6monthsoriftheyhadaplannedinterventioninthefollowing
3months.Additionally,patientswereexcludediftheyhadan
ongoingevaluationforhearttransplantation.
10
Leftventricular
ejectionfraction(LVEF)measurementswereavailablein832
patients.Biomarkersweremeasuredinbloodcollectedfrom
460patientsshortlybeforedischargebetween8:00AMand
4:00PM,afterpatientshadbeenclinicallystabilizedandwere
consideredwellenoughtogohome.Baselinecharacteristics
ofthecurrentsubstudywerecomparabletotheentireCOACH
study(TableS1).ThestudycomplieswiththeDeclarationof
Helsinki,localmedicalethicscommitteesapprovedthestudy,
andallpatientsprovidedwritteninformedconsent.
StudyandLaboratoryMeasurements
HFpEFwasde?nedashavingaLVEF≥45%,measurementsof
high-sensitiveC-reactiveprotein(hs-CRP),pentraxin-3,growth
differentiationfactor,solublereceptorofadvancedglycation
end-products,interleukin-6,tumornecrosisfactora,tumor
necrosisfactor–associatedreceptor1a,myeloperoxidase,
syndecan-1,periostin,ST-2,osteopontin,pro-atrial-typenatri-
ureticpeptide(proANP),vascularendothelialgrowthfactor
receptor(VEGFR),angiogenin,end-terminalproc-typenatri-
ureticpeptide,neuropilin-1,endothelialcell-selectiveadhe-
sionmolecule,neutrophilgelatinase-associatedlipocalin,
d-dimer,WAP4-disul?decoredomainproteinHE4,mesothe-
lin,polymericimmunoglobulinreceptor,prosaposin,andTROY
weremeasuredbyAlereSanDiego,Inc,(SanDiego,CA),
usingenzyme-linkedimmunosorbentassays.Immunoassays
toST2weredevelopedbyAlere.ThisresearchassaybyAlere
hasnotbeenstandardizedtothecommercializedassaysused
inresearchorinclinicaluse.Furthermore,theextenttowhich
thisAlereassaycorrelateswiththecommercialassayisnot
fullycharacterized.Galectin-3wasmeasuredusingELISAby
BGMedicine,Inc.(Waltham,MA).Transforminggrowthfactor-
bandVEGFwereanalyzedusingaquantitativemultiplexed
sandwichELISAsystem,SearchLightwproteomearrays
(AushonBioSystems,Billerica,MA).N-terminalpro-brain
natriureticpeptide(NT-proBNP)wasmeasuredusingthe
ElecsysproBNPELISAbyRocheDiagnostics(Mannheim,
Germany).ErythropoietinawasmeasuredusingtheIMMU-
LITEwerythropoietinELISAbyDiagnosticProductsCorpora-
tion(LosAngeles,CA).Inter-andintra-assaycoef?cientsof
theassaysusedcanbefoundinTableS2.Endothelin-1,
interleukin-6,andcardiac-speci?ctroponinIweremeasured
infrozenplasmasamplescollectedatbaselineusinghigh-
sensitivesinglemoleculecounting(SMC
TM
)technology(RUO,
ErennaC226
ImmunoassaySystem;SingulexInc,Alameda,CA).
Estimatedglomerular?ltrationratewasbasedonthe
simpli?edModi?cationofDietinRenalDisease.
12
StudyEndPoints
Forstudyingtherelationshipbetweenbiomarkerlevelsand
outcome,theprimaryendpointoftheCOACHtrialwasused.
Thisendpointisacombinedendpointconsistingofall-cause
mortalityand/orHF-relatedrehospitalizationsat18months.
Anindependentendpointcommitteeadjudicatedtheend
point.
StatisticalAnalysis
Continuousvariablesarepresentedasmedianswith
interquartilerangeormeansC6SDwhereappropriate.Cate-
goricalvariablesarepresentedasnumberswithpercentages.
Baselinecharacteristicsandbiomarkerconcentrationsat
baselinewerestrati?edaccordingtoHFrEFandHFpEF.
IntergroupdifferencesweretestedusingStudentttestor
Mann–WhitneyUtestforcontinuousvariablesorv
2
testfor
categoricalvariables.Principalcomponent(PC)analysiswas
performedtocorrectformultiplecomparisonswithHFrEFand
HFpEFascategoricalvariables,usinganestablishedstatis-
ticalmethoddescribedelsewhere.
13
Thismethodisoften
usedin-omicsbasedstudies,wherethereisanatural
correlationbetweenmarkersbecauseofthefactthatthese
oftenbelongtosimilarpathophysiologicalprocesses.
14
Indeed,alsoforthe33biomarkersemployedinthisstudy,
biomarkersareclearlyinterrelated,belongingtoseveral
similarpathophysiologicalprocesses(Figure1).Inthissitu-
ationtheBonferronicorrectioncanbeconsideredtoo
conservative.
15
Here,thePC-basedcorrectionhasbeen
suggestedtobemoreeffective.
14,15
Additionally,thismethod
hasbeenpreviouslysuccessfullyusedincorrectingfor
multiplecomparisonsinpairwisecorrelations.
13
Atotalof
21PCs,ofwhichtheeigenvaluescumulativelyexplained
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation2
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
>95%ofthevariationobservedinthedatasetwhen
comparingHFrEFwithHFpEF,werefound.Thecorrected
signi?cancelevelformultipletestingwasthussetatP<0.05/
21,equatingtoanadjustedP-valuecut-offof0.00238.To
correctformultiplecomparisonforinterbiomarkercorrela-
tions,0.05/[PC9(PCC01)/2]wasusedfortheadjustedPcut-
offvalue,wherePCisthenumberofprincipalcomponents
found.Tostudythein?uenceofclinicalconfounderson
biomarker-leveldifferencesbetweenHFrEFandHFpEF,logis-
ticregressionwasperformed.Here,HFpEFiscodedas1and
HFrEFas0.Anoddsratioabove1signi?esthathigherlevels
areassociatedwithHFpEF.Associationswerecorrectedfor
age,sex,estimatedglomerular?ltrationrate,ahistoryof
diabetesmellitus,andotherclinicalcovariatesthatsigni?-
cantlydifferedbetweenHFrEFandHFpEF.Next,aSpearman’s
rankcorrelationcoef?cientwascalculatedforeachpossible
biomarkerpairintheHFrEFcohortofpatientsandthe
procedurewasrepeatedforHFpEF.Thisresultedin2setsof
R-valueswithassociatedP-valuesforbothHFrEFandHFpEF.
Toadjustformultipletesting,onlythosecorrelationspassing
theadjustedP-valuecut-offcalculatedfromthePCanalysis
weredeemedstatisticallysigni?cantandsubsequently
retained.Thesesigni?cantcorrelationcoef?cientsforHFrEF
andHFpEFwerethengraphicallydisplayedasheatmapswith
associateddiseasedomainsforallbiomarkers.Network
analysiswasperformedtoanalyzeassociationsbetween
biomarkersinHFrEFandHFpEF.First,allsigni?cantassoci-
ationsfoundwithinHFrEFandHFpEFwereseparately
depictedascircularnetworks.Next,signi?cantassociations
betweenbiomarkersexclusivetoHFrEFandHFpEFwere
identi?ed.Toascertainwhethertheseassociationswere
signi?cantlydifferent,theFishersz-transformationtestwas
usedtocompareR-valuesbetweenHFrEFandHFpEF.TheP-
valuesfromtheseassociationswerecorrectedusingthePC
analysismethoddescribedabove.
Foroutcomeanalysis,aunivariableinteractiontestwas
performedbetweenthe(log2-transformed)biomarkerandHF
status(HFrEFversusHFpEF).Theinteractiontestwasthen
bootstrappedwith1000iterationstovalidatetheresults.
Followingthis,amultivariableinteractiontestwasperformed
correctingfortheCOACHriskengine.TheCOACHriskengine
includessex,age,pulsepressure,diastolicbloodpressure,
historyofstroke,historyofdiabetesmellitus,estimated
glomerular?ltrationrate,atrial?brillation,myocardialinfarc-
tion,peripheralarterialdisease,andlevelsofNT-proBNPand
sodiumandispoweredfortheprimaryendpointusedinthis
study,aspublishedelsewhere.
16
Therelationshipoftheprimary
endpointwithbiomarkers,showingasigni?cantinteraction
Figure1.HeatmapsdepictingcorrelationbetweenbiomarkersinHFrEF(A)andHFpEF(B).Biomarkercorrelationsthatdidnotpassthe
correctedP-value(0.05/21)areblack.Redentailsanegativecorrelation,greenentailsapositivecorrelation.BUNindicatesbloodureanitrogen;
CRP,C-reactiveprotein;EPO,erythropoietin;ESAM,endothelialcell-selectiveadhesionmolecule;GDF-15,growthdifferentiationfactor15;
HFpEF,heartfailurewithpreservedejectionfraction;HFrEF,heartfailurewithreducedejectionfraction;IL-6,interleukin6;MPO,
myeloperoxidase;NGAL,neutrophilgelatinase-associatedlipocalin;NT-proBNP,N-terminalpro-brain-typenatriureticpeptide;NT-proCNP,amino
terminalpro-C-typenatriureticpeptide;PIGR,polymericimmunoglobulinreceptor;proANP,pro-atrial-typenatriureticpeptide;PSAP,prostate-
speci?cacidphosphatase;RAGE,receptorofadvancedglycationend-products;ST-2,suppressionoftumorigenicity2;TGF-b,transforming
growthfactorb;TNF-a,tumornecrosisfactora;TNF-a-R1a,tumornecrosisfactorareceptor1a;VEGF,vascularendothelialgrowthfactor;
VEGFR,vascularendothelialgrowthfactorreceptor;WAP4C,WAP4disul?decoredomainprotein.
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation3
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
withHFstatusandoutcome,wasthengraphicallydepicted
usingKaplan–Meiercurves.Tocorrectforpotentialoptimism
andgiventhelimitedsamplesize,webootstrappedthe
estimateswith1000iterations.
17
Thesigni?canceofadiffer-
encebetweentertilesofbiomarkerlevelsandassociationwith
outcomewastestedusingtheLog-ranktest.Univariableand
multivariableassociationsofbiomarkerswithoutcomewere
testedusingtheCoxregression.Testsperformedwere2-tailed
andaP-valueof<0.05wasconsideredstatisticallysigni?cant.
AllstatisticalanalyseswereperformedusingSTATAversion
13.0(StataCorpLP,CollegeStation,TX)andR,version3.2.3.
Results
PatientCharacteristics
The460patientsinthiscohorthadameanageof
70.6C611.1yearsand37.4%werefemale.Mostpatients
wereinNYHAclassIII(52%)withameanLVEFof32.5C614.0%
(Table1).Ninety-sixpatientshadHFpEF(21%).Patientswith
HFpEFinthiscohortwererelativelyolder(74.5yearsversus
69.6years,P<0.001)andmoreoftenfemale(51.0%versus
33.8%,P=0.002).Additionally,patientswithHFpEFwere
foundtohaveahighersystolicbloodpressure(126.6mmHg
Table1.BaselineCharacteristics
TotalCohort
(n=460)
HFrEF(LVEF<45%)
(n=364)
HFpEF(LVEF≥45%)
(n=96)PValue
LVEF(%)32.5C614.026.7C68.554.4C67.5NA
DemographicsandHFcharacteristics
Age,y70.6C611.169.6(11.2)74.5(10.0)<0.001
Femalesex,n(%)172(37.4%)123(33.8%)49(51.0%)0.002
NYHAclass(atdischarge)II/III/IV,%44/52/442/54/455/41/40.064
PreviousHFhospitalization,n(%)155(33.7%)118(32.4%)37(38.5%)0.260
Clinicalsigns
BMI,kg/m
2
27.0C65.626.8C65.528.0C65.70.08
SystolicBP,mmHg117.9C621.3115.6C620126.6C623.1<0.001
DiastolicBP,mmHg68.9C612.368.9C612.468.9C612.10.980
eGFR,mL/minper1.73m
2
54.9C620.555.1C620.453.8C621.10.580
Heartrate,bpm74.2C613.474.7C613.872.2C611.80.110
Medicalhistory,n(%)
Myocardialinfarction187(40.7%)161(44.2%)26(27.1%)0.002
Hypertension191(41.5%)143(39.3%)48(50.0%)0.058
Diabetesmellitus135(29.3%)104(28.6%)31(32.3%)0.048
COPD130(28.3%)99(27.2%)31(32.3%)0.320
Atrialfibrillation/flutter209(45.4%)159(43.7%)50(52.1%)0.140
Anemia128(27.8%)92(25.3%)36(37.5%)0.017
Medication,n(%)
ACE-inhibitor/ARB378(82.2%)311(85.4%)67(69.8%)<0.001
b-Blocker312(67.8%)255(70.1%)57(59.4%)0.005
Diuretic440(95.7%)350(96.2%)90(93.8%)0.300
Statin183(39.8%)153(42.0%)30(31.2%)0.055
Digoxin155(33.7%)120(33.0%)35(36.5%)0.052
Laboratory
Hemoglobin,g/dL8.5(7.7,9.2)8.6(7.8,9.3)8.1(7.2,8.8)<0.001
Sodium,mEq/L138.6C64.3138.6C64.4138.6C64.20.973
Potassium,mEq/L4.2(3.9,4.6)4.3(3.9,4.6)4.1(3.7,4.6)0.214
ACEindicatesangiotensin-convertingenzyme;ARB,angiotensinIIreceptorblocker;BMI,bodymassindex;BP,bloodpressure;COPD,chronicobstructivepulmonarydisease;eGFR,
estimatedglomerular?ltrationrate;HF,heartfailure;HFpEF,heartfailurewithapreservedejectionfraction;HFrEF,heartfailurewithareducedejectionfraction;LVEF,leftventricular
ejectionfraction;NA,notavailable;NYHA,NewYorkHeartAssociation.
P-valuelowerthanthesigni?cancetreshholdof0.05.
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation4
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
versus115.6mmHg,P<0.001)comparedtopatientswith
HFrEF.Furthermore,patientswithHFpEFusedfewer
angiotensin-convertingenzymeinhibitors(55.2%versus
76.9%,P<0.001)andb-blockers(59.4%versus70.1%,
P<0.001)atdischarge.
BiomarkerLevelsinHFWithReducedand
PreservedEjectionFraction
PCanalysisrevealed21principalcomponentsthataccounted
foracumulativeproportionofvarianceof95%betweenHFrEF
andHFpEF,whichweresubsequentlyusedforadjustingtheP-
valuesigni?cancethreshold(P<0.05/21;FigureS1).Table2
showsthebaselinebiomarkerconcentrationsstrati?ed
accordingtoHFrEFandHFpEFwhereP-valuesshownare
correctedformultipletesting.Levelsofhs-CRPwerehigherin
HFpEF(3.6mg/Lversus2.1mg/L,P=0.001)andlevelsof
pentraxin-3werehigherinHFrEF(3.9ng/mLversus3.2ng/
mL,P=0.009).LevelsofcardiacstretchmarkersNT-proBNP
(2988pg/mLversus1948pg/mL,P<0.001)andproANP
(21.9pg/mLversus17.0pg/mL)werehigherinHFrEF.
Additionally,theangiogenesis-speci?cmarkerVEGFR
(0.8ng/mLversus0.7ng/mL,P=0.009)washigherin
HFrEF.Afteradjustingformultiplecomparisons,levelsof
hs-CRP(P=0.022)remainedsigni?cantlyhigherinHFpEF,
whilethecardiacstretchmarkersNT-proBNP(P<0.001)and
proANP(P=0.042)remainedsigni?cantlyhigherinHFrEF.
BiomarkerassociationswithHFrEFandHFpEFareshown
inTableS3.Whencorrectingforclinicalcovariates(age,sex,
estimatedglomerular?ltrationrate,systolicbloodpressure,a
historyofmyocardialinfarction;diabetesmellitus;atrial
?brillationandanemia),higherlevelsofhs-CRP(oddsratio:
1.29;95%CI1.09–1.52,P=0.003)remainedassociatedwith
HFpEF,whilehigherlevelsofNT-proBNP(oddsratio:0.68;
95%CI0.57–0.82,P<0.001)andproANP(oddsratio:0.69;
95%CI0.53–0.88,P=0.003)remainedassociatedwithHFrEF.
Afteradditionallycorrectingforb-blockerandangiotensin-
convertingenzymeinhibitor/angiotensinIIreceptorblocker
use,thestatisticalassociationsforthese3markersremained
(TableS3).
BiomarkerAssociationsandNetworkAnalysis
HeatmapsfortheassociationbetweenbiomarkersinHFrEF
andHFpEFaredepictedinFigure1.Figure2showsthe
graphicaldepictionofbiomarkernetworksinHFrEFand
HFpEF.Resultsfromthecorrelationanalysisandassociated
heatmapsrevealthatcorrelationsbetweenbiomarkersin
HFpEFaremoreassociatedwithremodelingandin?amma-
tion,whileinHFrEFangiogenesisisamoreprominentfeature
(Figure1).Networkanalysisfurthershowedmyeloperoxidase
tobeinvolvedininteractionsinbothHFrEFandHFpEF.
Additionally,renalmarkerneutrophilgelatinase-associated
lipocalinandbloodureanitrogenaswellasin?ammation
markerreceptorofadvancedglycationend-productswere
involvedinbiomarkerassociationsinHFpEF.
Whenexaminingtheexclusiveinteractionsbetween
biomarkersinHFrEFandHFpEF,HFpEFrevealedinteractions,
whichweremainlyassociatedwithin?ammation(interleukin-
6;pentraxin-3;Table3,correctedP-valuefordifference
<0.05).Incontrast,HFrEFshowedexclusiveinteractionsthat
wereNT-proBNPmediated(Table3),indicatingthatbiomar-
kerinteractionsaremoreassociatedwithcardiacstretchin
HFrEFandin?ammationinHFpEF.Insensitivityanalysiswith
ade?nitionofHFrEFatLVEF≤40%andade?nitionofHFpEF
atLVEF≥50%,exclusiveassociationsinHFpEFremained
in?ammationmediated,whileNT-proBNPmediatedassocia-
tionsinHFrEF(TableS4).
Outcome
Ofthetotalcohort,41%reachedtheclinicalendpointof
deathand/orHFrehospitalization(41%HFrEFversus44.8%
HFpEF,P=0.659,FigureS2).NT-proBNPwasfoundtobe
equallypredictiveinHFrEFandHFpEF(TableS5).Asigni?cant
interactioninbothunivariableandmultivariableanalysiswas
foundforHFstatusandneuropilinaswellasosteopontin
(bothP<0.05).Bothbiomarkerswerefoundonlytobe
predictiveinHFpEF(Figures3and4,TableS5).Interaction
betweenneuropilin(P=0.007)andosteopontin(P=0.018)and
HFstatusfortheprimaryendpointremainedfollowing
sensitivityanalysisforade?nitionofHFpEFofLVEF≥50%.
Afterbootstrappingwith1000iterations,theinteractionwith
HFstatusfortheprimaryendpointstayedsigni?cantforboth
osteopontin(P=0.002)andneuropilin(P=0.011)inunivariable
analyses.Alsoinmultivariableanalyses,theinteraction
remainedsigni?cantforosteopontin(P=0.016)andneuropilin
(P=0.015).
WhenexaminingtherelationshipwithHFrehospitalizations
andall-causemortalityseparatelyinunivariableanalysis,we
seethatosteopontinispredictiveforbothHFrehospitalizations
(P=0.007)andall-causemortality(P=0.031)separately,butnot
inHFrEF(FiguresS3andS4).Neuropilinwaspredictivein
univariableanalysisforall-causemortalityinbothHFrEF
(P=0.003)andHFpEF(P=0.023).However,neuropilinwasonly
predictiveofHFrehospitalizationsinHFpEF(P=0.026)andnot
inHFrEF(P=0.026)(FiguresS5andS6).
Discussion
Inthisstudy,wedemonstrateadistinctbiomarkerpro?lefor
HFpEFandHFrEFpatientsbyusinganovelapproach
employingnetworkanalysistoidentifyexclusiveinteractions
withinthe2diseaseentities.HigherlevelsofHs-CRPand
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation5
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
Table2.BaselineMarkersStrati?edtoHFrEFandHFpEF
TotalCohort(n=460)HFrEF(n=364)HFpEF(n=96)PValuePValue
Inflammation
hs-CRP,mg/L2.3(0.9,5.2)2.1(0.8,4.7)3.6(1.8,7.0)0.001
?
0.022
?
Pentraxin-3,ng/mL3.7(2.5,5.6)3.9(2.7,5.8)3.2(2.4,4.7)0.009
?
0.198
GDF-15,ng/mL2.8(1.9,4.2)2.8(1.9,4.3)2.6(1.9,4.1)0.6701.000
RAGE,ng/mL2.9(1.9,4.8)3.0(1.9,4.9)2.6(1.7,4.0)0.0531.000
IL-6,pg/mL7.0(3.7,12.2)6.7(3.6,11.3)8.2(4.5,13.6)0.1001.000
TNF-a,pg/mL47.9(6.2,119.4)47.3(8.1,109.5)56.7(4.8,194.4)0.3501.000
TROY,ng/mL0.9(0.7,1.5)0.9(0.7,1.4)0.9(0.6,1.6)0.5401.000
TNF-a-R1a,ng/mL3.0(2.1,4.5)3.0(2.1,4.4)3.1(2.2,4.9)0.4901.000
Oxidativestress
MPO,ng/mL20.4(15.6,28.2)20.6(15.6,28.4)19.9(15.2,27.2)0.5301.000
Remodeling
Syndecan-1,ng/mL20.2(14.1,27.5)20.5(14.1,28.1)19.2(14.0,24.6)0.3601.000
Periostin,ng/mL4.6(3.4,6.6)4.7(3.4,6.6)4.5(3.4,6.6)0.5201.000
Galectin-3,ng/mL19.9(15.2,25.7)20.0(14.8,25.9)19.3(15.8,25.3)0.9601.000
ST-2,ng/mL2.5(1.4,5.6)2.4(1.4,5.5)3.1(1.6,6.2)0.1401.000
Osteopontin,ng/mL160.1(108.8,219.5)161.2(108.4,217.1)153.8(110.7,240.5)0.9801.000
TGF-?,ng/mL50.6(34.4,75.1)51.4(35.3,77.5)44.3(30.9,63.3)0.0691.000
Cardiomyocytestretch
NT-proBNP,pg/mL2601(1398–5989)2988.8(1511.0,6708.9)1948.0(855.3,3827.0)<0.001
?
<0.001
?
proANP,ng/mL20.4(12.1–33.3)21.9(13.2,35.4)17.0(10.0,28.2)0.002
?
0.042
?
cTnI,pg/mL14.1(7.3,29.4)13.1(5.8,34.8)0.5621.000
Angiogenesis
VEGF,pg/mL62.8(31.4,148.7)62.5(28.5,139.9)63.0(35.8,162.9)0.2801.000
VEFGR,ng/mL0.8(0.6,1.0)0.8(0.6,1.1)0.7(0.5,1.0)0.009
?
0.255
Angiogenin,lg/mL5.0(3.5,7.4)5.0(3.5,7.5)5.2(3.5,7.3)0.8401.000
NT-proCNP,ng/mL0.024(0.017,0.035)0.023(0.017,0.034)0.024(0.015–0.037)0.4401.000
Neuropilin-1,ng/mL10.0(7.1,13.7)10.1(7.1,14.0)9.6(7.0,13.5)0.7701.000
Arteriosclerosis
ESAM,ng/mL52.9(44.5,64.4)53.8(45.3,64.8)50.2(41.1,63.2)0.0651.000
Renalfunction
NGAL,ng/mL84.6(60.4,119.9)84.2(59.4,119.2)84.7(63.3,122.3)0.4401.000
BUN,mmol/L11.0(8.2,15.5)10.7(8.3,15.6)11.1(7.7,15.1)0.6501.000
Hematopoiesis
EPOa,IU/L9.6(5.1,15.9)9.5(5.0,15.5)10.3(5.2,16.5)0.5601.000
Other
D-Dimer,lg/mL0.5(0.2,1.1)0.5(0.2,1.1)0.6(0.2,1.0)0.7101.000
WAP4C,ng/mL5.7(3.1,10.1)5.8(3.5,10.0)5.3(3.1,10.3)0.9101.000
Mesothelin,ng/mL29.4(22.8,38.7)29.8(22.9,38.8)28.3(22.5,38.0)0.3801.000
PIGR,ng/mL600.6(337.4,952.0)609.0(388.7,952.0)598.7(331.5,943.0)0.3301.000
PSAP,ng/mL68.6(49.2,98.5)68.8(49.8,101.0)67.3(48.0,93.6)0.7601.000
ET-1,ng/mL4.5(3.6,6.1)4.5(3.6,6.1)4.5(3.4,5.7)0.4301.000
BUNindicatesbloodureanitrogen;cTNI,cardiactroponin-I;EPOa,erythropoietin;ESAM,endothelialcell-selectiveadhesionmolecule;ET-1,endothelin-1;GDF-15,growthdifferentiation
factor15;HFpEF,heartfailurewithapreservedejectionfraction;HFrEF,heartfailurewithareducedejectionfraction;hs-CRP,high-sensitiveC-reactiveprotein;IL-6,interleukin6;MPO,
myeloperoxidase;NGAL,neutrophilgelatinase-associatedlipocalin;NT-proBNP,N-terminalpro-brain-typenatriureticpeptide;NT-proCNP,aminoterminalpro-C-typenatriureticpeptide;
PIGR,polymericimmunoglobulinreceptor;proANP,pro-atrial-typenatriureticpeptide;PSAP,prostate-speci?cacidphosphatase;RAGE,receptorofadvancedglycationend-products;ST-2,
suppressionoftumorigenicity2;TGF-b,transforminggrowthfactorb;TNF-a,tumornecrosisfactora;TNF-a-R1a,tumornecrosisfactorareceptor1a;VEGF,vascularendothelialgrowth
factor;VEGFR,vascularendothelialgrowthfactorreceptor;WAP4C,WAP4disul?decoredomainprotein.
CorrectedP-value.
?
P-valuelowerthanthesigni?cancetreshholdof0.05.
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation6
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
lowerlevelsofcardiacstretchmarkersNT-proBNPandpro-
ANParefoundinHFpEF,whichcon?rmpreviousstudies.
8,18
Furthermore,exclusiveinteractionsbetweenbiomarkersin
HFpEFwerefoundtobeassociatedwithin?ammationand
angiogenesis.Incontrast,HFrEFshowedexclusiveinterac-
tionsassociatedwithNT-proBNP.Thisisthe?rststudy
reportingonexclusiveinteractionsbetweenbiomarkersin
HFrEFandHFpEF.Additionally,thisstudyshowedforthe?rst
timethatangiogenesismarkerneuropilinandremodeling
markerosteopontinhaveexclusivepredictivevalueforclinical
outcomeinHFpEF.
Levelsofhs-CRPwerefoundtobehigherinHFpEFpatients
comparedtoHFrEFpatients.Overall,reportswithregardto
differencesinassociationofCRPbetweenHFrEFandHFpEF
havelackedconsensus.
8,19–21
Yet,patientsincludedinthe
previousstudieswereolderandhadrelativelylowlevelsof
NT-proBNP.
8,19,20,22
Regardlessofthedifferenceinlevels,
predictivevalueforhs-CRPwasfoundtobelimitedinboth
HFrEFandHFpEFaftercorrectionforariskmodelinboththis
andanearlierstudy.
21
ThecardiacstretchmarkersproANP
andNT-proBNPwerefoundtobelowerinHFpEF.Thisisthe
?rststudyreportingdifferentiallevelsofproANPinHFrEFand
HFpEF.ThedifferenceinlevelsofNTproBNPbetweenHFrEF
andHFpEFcon?rmsearlierreports.
8,18,23
Arecentstudyusedasimilarnetworkanalysisapproach.
8
However,thenumberofbiomarkersstudiedwaslimitedand
noexclusivecorrelationswereidenti?ed.Whenexamining
exclusivecorrelationsinHFpEFandHFrEFbetweenbiomark-
ers,weidenti?edcorrelationsthatwerein?ammationand
angiogenesisassociatedinHFpEF,whilecorrelationswere
associatedwithNT-proBNPinHFrEF.Therelativelystrong
correlationsbetweenmarkersinbothHFrEFandHFpEF
provideputativeinsightsintopossibledifferencesatthe
pathophysiologicalpathwaylevel.ForHFpEF,correlations
werefoundtobeassociatedwithinterleukin-6andpentraxin-
3.Thisisinlinewithearliersuggestions,inwhichapro-
in?ammatorystatewasproposedtounderliethepathophys-
iologyofHFpEF.
24–29
Incontrast,exclusiveinteractionsin
HFrEFwereassociatedwithNT-proBNP.Assuch,the
pathophysiologyofHFrEFseemstobemoreassociatedwith
cardiacstretchandoxidativestress.
24
However,using
networkanalysisfordeterminingunderlyingpathophysiolog-
icaldifferencesbetweendiseaseentitiesusingbiomarkersis
arelativelynovelapproach.Futurestudiesshouldcon?rm
AB
Figure2.NetworkanalysisdepictingassociationsbetweenbiomarkersinHFrEF(A)andHFpEF(B).Associationsshownarethosethatpassed
theP-valuecutoff(0.05/21).Nodesizeandcolorarebasedontheclusteringcoef?cient.Theedgebetweennesswasusedasacriterionforthe
edges.BUNindicatesbloodureanitrogen;CRP,C-reactiveprotein;EPO,erythropoietin;ESAM,endothelialcell-selectiveadhesionmolecule;
GDF-15,growthdifferentiationfactor15;HFpEF,heartfailurewithpreservedejectionfraction;HFrEF,heartfailurewithreducedejection
fraction;IL-6,interleukin6;MPO,myeloperoxidase;NGAL,neutrophilgelatinase-associatedlipocalin;NT-proBNP,N-terminalpro-brain-type
natriureticpeptide;NT-proCNP,aminoterminalpro-C-typenatriureticpeptide;PIGR,polymericimmunoglobulinreceptor;proANP,pro-atrial-type
natriureticpeptide;PSAP,prostate-speci?cacidphosphatase;RAGE,receptorofadvancedglycationend-products;ST-2,suppressionof
tumorigenicity2;TGF-b,transforminggrowthfactorb;TNF-a,tumornecrosisfactora;TNF-a-R1a,tumornecrosisfactorareceptor1a;VEGF,
vascularendothelialgrowthfactor;VEGFR,vascularendothelialgrowthfactorreceptor;WAP4C,WAP4disul?decoredomainprotein.
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation7
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
these?ndingsaswellascombinethemwithdatafrom
experimentalstudiestoexaminewhetherthepathophysio-
logicalrelationshipsfoundinclinicaldataalsotranslateto
pathophysiologicaldifferencesinanexperimentalsetting.
Furthermore,mostbiomarkersarenotcardiacexclusive.
5
This
makesitrelativelydif?culttodiscernwhetherbiomarker
differencesfoundinaclinicalstudyarethecauseor
consequenceofHF.Tooptimizeinterpretabilityofbiomarker
studies,futurestudiesshouldbefocusedonbiomarkersthat
arehighlycardiacspeci?c.Secondly,whenbiomarkerdiffer-
encesarefound,experimentalstudiesshouldvalidatethe
?ndingsanddiscernpossibleunderlyingpathophysiological
processes.
Thisstudyalsoshoweddifferentialassociationwith
outcomeofangiogenesismarkersneuropilinandremodeling
markerosteopontin,whichwerebothfoundtobemore
predictiveinHFpEF.Resultswithregardtodifferential
associationwithoutcomeshouldbeinterpretedinan
explanatorycontextofthepathophysiology,inwhichan
increaseinlevelsofacertainbiomarkercanbedetrimentalin
1diseaseentityandnotnecessarilyintheotherthrough
biologicalinvolvementorre?ectinganunderlyingpathway.
Indeed,osteopontinwasreportedearliertobeinvolvedin
prognosisinHF.
30
However,adifferentialinvolvement
betweenHFrEFandHFpEFhasnotbeenpreviouslyreported.
Earlierexperimentalstudiesfoundadirectinvolvementof
osteopontinandcardiacremodeling,whichinturnwasfound
tocausediastolicdysfunction.
31
Neuropilinisidenti?edasacoreceptorofvascular
endothelialgrowthfactorreceptor2(VEGFR-2).
32
Inamurine
modelofcardiacpressureoverload,animalsthatwere
heterozygousforneuropilinshowedhighermortalityrates.
33
Thisisthe?rststudyreportingthepredictivevalueof
neuropilininHFforthecombinedendpoint.Here,wefound
Table3.InteractionWithinHFrEFandHFpEF
Biomarker
HFpEFHFrEF
PValue
(Difference)
PValue
(Difference)RPValueRPValue
HFpEF
IL-6D-Dimer0.3650.030
?
0.1491.0000.001
?
0.021
?
Pentraxin-3VEGFC00.3440.029
?
C00.1541.0000.002
?
0.043
?
PeriostinVEGFC00.4380.001
?
C00.1121.000<0.001
?
0.001
?
NGALPSAP-B10.3960.010
?
0.1381.000<0.001
?
0.007
?
HFrEF
NT-proBNPIL-60.1351.0000.363<0.001
?
0.001
?
0.023
?
NT-proBNPEPO-A0.1471.0000.36<0.001
?
0.001
?
0.025
?
EPO-Aindicateserythropoietin;HFpEF,heartfailurewithapreservedejectionfraction;HFrEF,heartfailurewithareducedejectionfraction;IL-6,interleukin6;NGAL,neutrophilgelatinase-
associatedlipocalin;NT-proBNP,N-terminalpro-brain-typenatriureticpeptide;PSAP,prostate-speci?cacidphosphatase;VEGF,vascularendothelialgrowthfactor.
CorrectedP-value.
?
P-valuelowerthanthesigni?cancetreshholdof0.05.
Figure3.Kaplan–Meiercurvesdepictingtherelationshipwithoutcomeofosteopontinintertiles,strati?edtoHFrEFandHFpEF.HFpEF
indicatesheartfailurewithpreservedejectionfraction;HFrEF,heartfailurewithreducedejectionfraction.
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation8
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
thatneuropilinwaspredictiveofHFrehospitalizationsin
HFpEF.Additionally,inmultivariableanalysis,neuropilinonly
heldpredictivepowerinHFpEF.Thissuggeststhatneuropilin
ismorere?ectiveofHFseverityinHFpEFandnotinHFrEF.
Essentially,neuropilinisassociatedwithangiogenesis.This
againemphasizestheimportanceofangiogenesismarkersin
HFpEFcomparedtoHFrEF.
24
Inearlierstudiesasigni?cantassociationbetweenout-
comeandHFstatuswasfoundforend-terminalproc-type
natriureticpeptideandgalectin-3withade?nitionofHFpEFof
LVEF>40%.
34,35
These?ndingswerecon?rmedinthisstudy.
Additionally,anearlierpublicationfoundsigni?cantpredictive
valueofsyndecan-1inHFpEFbutnotinHFrEF.
6
Thefactthat
nosigni?cantinteractionwasfoundinthisstudyfor
syndecan-1andtheprimaryendpointcanpotentiallybe
explainedbythelimitedpowerofthisstudyforHFpEF
patientsatade?nitionofLVEF>45%,andtheprevious
publicationforsyndecan-1correctedforastepwisebased
modelforsyndecan-1insteadoftheCOACHriskmodel.
Theclinicalimplicationsofthisstudyare2-fold.First,this
studycharacterizestheunderlyingpathophysiologyof
patientswithHFpEFtobeassociatedwithin?ammationand
endothelialfunction.Thiscon?rmsearlierstudieswithregard
toHFpEFandendorsestheearlierproposedtheorybyPaulus
etal.
24
Secondly,thisstudypropagatesanovelmethodfor
utilizingnetworkanalysistoanalyzeawidearrayofbiomark-
ersindiscerningtheunderlyingpathophysiologyofdisease
entitiesinHF.
8
Thismethodologyprovidesapossiblestep
forwardindissectingtheHFsyndrome.
5,36
StrengthsandLimitations
ThestrengthsofthisstudyaretherelativelyhighlevelsofNT-
proBNPofboththeHFrEFandHFpEFpatientsandthelarge
numberofavailablebiomarkers.ByhavingrelativelyhighNT-
proBNPlevels,theHFpEFpatientsinthisstudyrepresenttrue
HFpatientsandhavearelativelylownumberoffalse
positives.Secondly,thelargenumberofbiomarkersfrom
differentdiseasedomainsavailableinthisstudyprovidefora
moreunbiasedapproachtowardsdiscerningunderlying
pathophysiologicalpathways.
However,thecurrentanalysisisapost-hocanalysis,
leadingtoapossibleselectionbias.Secondly,sincepatients
includedareofEuropeandescentandrelativelyold,thislimits
extrapolationtopatientsofdifferentageandorigin.Also,
pharmacologicaltreatmentduringhospitalizationmighthave
in?uencedbiomarkerlevelsandassociationsbetweenHFrEF
andHFpEF.Furthermore,thechoiceforbiomarkerswas
restrictedbylimitedbaselinesampleavailability,withthe
resultthatseveralinterestingmarkerscouldnotbestudied.
Therefore,thisstudyisnotanexhaustivestudyofbiomarker-
leveldifferencesinHFrEFandHFpEFandshouldbeconsid-
eredexploratoryandhypothesisgenerating.Also,someofthe
biomarkersmeasuredhadrelativelyhighcoef?cientsof
variation.Therefore,somepossibleinterestinginteractions
anddifferencesbetweenbiomarkersinHFrEFandHFpEFmay
havebeenmissed.Mostimportantly,resultsfromthisstudy
shouldbevalidatedinaseparatecohort.
ThesamplingofpatientsinCOACHwasperformedat
dischargeafterrecompensation.Sincenodataareavailable
ontreatmentduringadmissionforHFprevioustodischarge,
thismightconfoundsomeofthereported?ndings.Inthis
context,patientsintheCOACHtrialcoveragrayarea
betweenacutedecompensatedandchronicHFpatients.The
?ndingsinthisstudyshouldberegardedasexplanatoryinthe
contextofthepathophysiologyofHFpEFandHFrEF,actingas
astepping-stoneforfurtherresearch.
Conclusions
BiomarkerlevelsdifferinHFpEFandHFrEF,mainlyinthe
domainsofcardiacstretchandin?ammation.Interactionsin
Figure4.Kaplan–Meiercurvesdepictingtherelationshipwithoutcomeofneuropilinintertiles,strati?edtoHFrEFandHFpEF.HFpEF
indicatesheartfailurewithpreservedejectionfraction;HFrEF,heartfailurewithreducedejectionfraction.
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation9
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
HFpEFwerefoundtobeassociatedwithin?ammationand
angiogenesis,whileinteractionsinHFrEFwereassociated
withcardiacstretch.Theangiogenesismarkerneuropilinand
remodelingmarkerosteopontinwerefoundtoonlyhold
predictivevalueinHFpEF,possiblyre?ectingunderlying
pathophysiologicalprocesses.Resultsofthisstudyshould
becon?rmedinprospectivebiomarkerstudies.
SourcesofFunding
COACHwassupportedbygrant2000Z003fromtheNether-
landsHeartFoundationandbyadditionalunrestrictedgrants
fromBiositeFranceSAS,Jouy-en-Josas,France(brainnatri-
ureticpeptide),RocheDiagnosticsNederlandBV,Venlo,the
Netherlands(N-terminalprohormonebrainnatriureticpep-
tide),BGMedicineInc,Waltham,MA(galectin-3assays)and
NovartisPharmaBV,Arnhem,theNetherlands.
Disclosures
Tromp,Khan,Klip,Meyer,deBoer,Jaarsma,Hillege,van
Veldhuisen,andvanderMeerhavenothingtodisclosewith
regardtothismanuscript.Voorsreceivedresearchgrants
fromAlere,Singulex,andSphingotec.
References
1.OwanTE,HodgeDO,HergesRM,JacobsenSJ,RogerVL,Red?eldMM.Trends
inprevalenceandoutcomeofheartfailurewithpreservedejectionfraction.N
EnglJMed.2006;355:251–259.
2.vanVeldhuisenDJ,Cohen-SolalA,B€ohmM,AnkerSD,BabalisD,RoughtonM,
CoatsAJS,Poole-WilsonPA,FlatherMD.Beta-blockadewithnebivololin
elderlyheartfailurepatientswithimpairedandpreservedleftventricular
ejectionfraction:datafromSENIORS(StudyofEffectsofNebivololInterven-
tiononOutcomesandRehospitalizationinSeniorsWithHeartFailure).JAm
CollCardiol.2009;53:2150–2158.
3.YusufS,PfefferMA,SwedbergK,GrangerCB,HeldP,McMurrayJJ,Michelson
EL,OlofssonB,OstergrenJ;CommitteesCI.Effectsofcandesartaninpatients
withchronicheartfailureandpreservedleft-ventricularejectionfraction:the
CHARM-PreservedTrial.Lancet.2003;362:777–781.
4.PittB,PfefferMA,AssmannSF,BoineauR,AnandIS,ClaggettB,Clausell
N,DesaiAS,DiazR,FlegJL,GordeevI,HartyB,HeitnerJF,KenwoodCT,
LewisEF,O’MearaE,Probst?eldJL,ShaburishviliT,ShahSJ,SolomonSD,
SweitzerNK,YangS,McKinlaySM;InvestigatorsT.Spironolactoneforheart
failurewithpreservedejectionfraction.NEnglJMed.2014;370:1383–
1392.
5.BraunwaldE.Biomarkersinheartfailure.NEnglJMed.2008;358:2148–2159.
6.TrompJ,vanderPolA,KlipIT,deBoerRA,JaarsmaT,vanGilstWH,VoorsAA,
vanVeldhuisenDJ,vanderMeerP.Fibrosismarkersyndecan-1andoutcome
inpatientswithheartfailurewithreducedandpreservedejectionfraction.Circ
HeartFail.2014;7:457–462.
7.DemisseiBG,ClelandJG,O’ConnorCM,MetraM,PonikowskiP,TeerlinkJR,
CotterG,DavisonB,GivertzMM,Bloom?eldDM,DittrichH,vanderMeerP,
vanVeldhuisenDJ,HillegeHL,VoorsAA.Optimizingclinicaluseof
biomarkersinhigh-riskacuteheartfailurepatients.EurJHeartFail.
2016;3:269–280.
8.Sanders-vanWijkS,vanEmpelV,DavarzaniN,MaederMT,HandschinR,
P?stererME,Brunner-LaRoccaHP.Circulatingbiomarkersofdistinct
pathophysiologicalpathwaysinheartfailurewithpreservedvs.reducedleft
ventricularejectionfraction.EurJHeartFail.2015;17:1006–1014.
9.SchmitterD,CotterG,VoorsAA.Clinicaluseofnovelbiomarkersinheart
failure:towardspersonalizedmedicine.HeartFailRev.2013;19:369–381.
10.JaarsmaT,VanDerWalMH,HogenhuisJ,LesmanI,LuttikML,VeegerNJ,Van
VeldhuisenDJ.DesignandmethodologyoftheCOACHstudy:amulticenter
randomisedCoordinatingstudyevaluatingOutcomesofAdvisingandCoun-
sellinginHeartfailure.EurJHeartFail.2004;6:227–233.
11.JaarsmaT,vanderWalMHL,Lesman-LeegteI,LuttikM-L,HogenhuisJ,Veeger
NJ,SandermanR,HoesAW,vanGilstWH,LokDJA,DunselmanPHJM,Tijssen
JGP,HillegeHL,vanVeldhuisenDJ.Effectofmoderateorintensivedisease
managementprogramonoutcomeinpatientswithheartfailure:Coordinating
StudyEvaluatingOutcomesofAdvisingandCounselinginHeartFailure
(COACH).ArchInternMed.2008;168:316–324.
12.SmildeTDJ,vanVeldhuisenDJ,NavisG,VoorsAA,HillegeHL.Drawbacks
andprognosticvalueofformulasestimatingrenalfunctioninpatientswith
chronicheartfailureandsystolicdysfunction.Circulation.2006;114:1572–
1580.
13.AuroK,JoensuuA,FischerK,KettunenJ,SaloP,MattssonH,NiironenM,KaprioJ,
ErikssonJG,Lehtim€akiT,RaitakariO,JulaA,TiitinenA,JauhiainenM,SoininenP,
KangasAJ,K€ah€onenM,HavulinnaAS,Ala-KorpelaM,SalomaaV,MetspaluA,
PerolaM.Ametabolicviewonmenopauseandageing.NatCommun.2014;5:4708.
14.GaoX,StarmerJ,MartinER.Amultipletestingcorrectionmethodforgenetic
associationstudiesusingcorrelatedsinglenucleotidepolymorphisms.Genet
Epidemiol.2008;32:361–369.
15.JohnsonRC,NelsonGW,TroyerJL,LautenbergerJA,KessingBD,WinklerCA,
O’BrienSJ.Accountingformultiplecomparisonsinagenome-wideassociation
study(GWAS).BMCGenom.2010;11:724.
16.PostmusD,vanVeldhuisenDJ,JaarsmaT,LuttikML,LassusJ,MebazaaA,
NieminenMS,HarjolaVP,LewseyJ,BuskensE,HillegeHL.TheCOACHrisk
engine:amultistatemodelforpredictingsurvivalandhospitalizationin
patientswithheartfailure.EurJHeartFail.2012;14:168–175.
17.SmithGCS,SeamanSR,WoodAM,RoystonP,WhiteIR.Correctingfor
optimisticpredictioninsmalldatasets.AmJEpidemiol.2014;180:318–324.
18.vanVeldhuisenDJ,LinssenGCM,JaarsmaT,vanGilstWH,HoesAW,Tijssen
JGP,PaulusWJ,VoorsAA,HillegeHL.B-typenatriureticpeptideandprognosis
inheartfailurepatientswithpreservedandreducedejectionfraction.JAmColl
Cardiol.2013;61:1498–1506.
19.MatsubaraJ,SugiyamaS,NozakiT,SugamuraK,KonishiM,OhbaK,
MatsuzawaY,AkiyamaE,YamamotoE,SakamotoK,NagayoshiY,KaikitaK,
SumidaH,Kim-MitsuyamaS,OgawaH.Pentraxin3isanewin?ammatory
markercorrelatedwithleftventriculardiastolicdysfunctionandheartfailure
withnormalejectionfraction.JAmCollCardiol.2011;57:861–869.
20.WisniackiN,TaylorW,LyeM,WildingJPH.Insulinresistanceandin?ammatory
activationinolderpatientswithsystolicanddiastolicheartfailure.Heart.
2005;91:32–37.
21.MichowitzY,ArbelY,WexlerD,ShepsD,RogowskiO,ShapiraI,BerlinerS,
KerenG,GeorgeJ,RothA.PredictivevalueofhighsensitivityCRPinpatients
withdiastolicheartfailure.IntJCardiol.2008;125:347–351.
22.NiethammerM,SieberM,vonHaehlingS,AnkerSD,MunzelT,HorstickG,
Genth-ZotzS.In?ammatorypathwaysinpatientswithheartfailureand
preservedejectionfraction.IntJCardiol.2008;129:111–117.
23.MaederMT,RickenbacherP,RickliH,Abb€uhlH,GutmannM,ErneP,
VuilliomenetA,PeterM,P?stererM,Brunner-LaRoccaH-P.N-terminalpro
brainnatriureticpeptide-guidedmanagementinpatientswithheartfailureand
preservedejectionfraction:?ndingsfromtheTrialofIntensi?edversus
standardmedicaltherapyinelderlypatientswithcongestiveheartfailure
(TIME-CHF).EurJHeartFail.2013;15:1148–1156.
24.PaulusWJ,Tsch€opeC.Anovelparadigmforheartfailurewithpreserved
ejectionfraction:comorbiditiesdrivemyocardialdysfunctionandremodeling
throughcoronarymicrovascularendothelialin?ammation.JAmCollCardiol.
2013;62:263–271.
25.ShahKB,KopWJ,ChristensonRH,DiercksDB,HendersonS,HansonK,LiS-Y,
deFilippiCR.PrognosticutilityofST2inpatientswithacutedyspneaand
preservedleftventricularejectionfraction.ClinChem.2011;57:874–882.
26.KalogeropoulosA,GeorgiopoulouV,PsatyBM,RodondiN,SmithAL,Harrison
DG,LiuY,HoffmannU,BauerDC,NewmanAB,KritchevskySB,HarrisTB,
ButlerJ.In?ammatorymarkersandincidentheartfailureriskinolderadults:
theHealthABC(Health,Aging,andBodyComposition)study.JAmCollCardiol.
2010;55:2129–2137.
27.CollierP,WatsonCJ,VoonV,PhelanD,JanA,MakG,MartosR,BaughJA,
LedwidgeMT,McDonaldKM.Canemergingbiomarkersofmyocardial
remodellingidentifyasymptomatichypertensivepatientsatriskfordiastolic
dysfunctionanddiastolicheartfailure?EurJHeartFail.2011;13:1087–
1095.
28.MentzRJ,KellyJP,vonLuederTG,VoorsAA,LamCSP,CowieMR,KjeldsenK,
JankowskaEA,AtarD,ButlerJ,FiuzatM,ZannadF,PittB,O’ConnorCM.
Noncardiaccomorbiditiesinheartfailurewithreducedversuspreserved
ejectionfraction.JAmCollCardiol.2014;64:2281–2293.
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation10
BiomarkersPro?lesinHeartFailureTrompetal
ORIGI
N
AL
RESE
ARCH
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
29.TerMaatenJM,DammanK,VerhaarMC,PaulusWJ,DunckerDJ,ChengC,
vanHeerebeekL,HillegeHL,LamCSP,NavisG,VoorsAA.Connecting
heartfailurewithpreservedejectionfractionandrenaldysfunction:therole
ofendothelialdysfunctionandin?ammation.EurJHeartFail.2016;6:588–
598.
30.RosenbergM,ZugckC,NellesM,JuengerC,FrankD,RemppisA,GiannitsisE,
KatusHA,FreyN.Osteopontin,anewprognosticbiomarkerinpatientswith
chronicheartfailure.CircHeartFail.2008;1:43–49.
31.LC19opezB,GonzC19alezA,LindnerD,WestermannD,RavassaS,BeaumontJ,
GallegoI,ZudaireA,BrugnolaroC,QuerejetaR,LarmanM,Tsch€opeC,DC19?ezJ.
Osteopontin-mediatedmyocardial?brosisinheartfailure:aroleforlysyl
oxidase?CardiovascRes.2013;99:111–120.
32.ChaudharyB,KhaledYS,AmmoriBJ,ElkordE.Neuropilin1:functionand
therapeuticpotentialincancer.CancerImmunolImmunother.2014;63:81–99.
33.LiF,ZhaoH,LiaoY,TakashimaS,AsanoY,ShintaniY,HoriM,Kitakaze
M.Highermortalityinheterozygousneuropilin-1miceafter
cardiacpressureoverload.BiochemBiophysResCommun.2008;370:
317–321.
34.LokDJ,KlipIT,VoorsAA,LokSI,Bruggink-AndrC19edelaPortePW,HillegeHL,
JaarsmaT,vanVeldhuisenDJ,vanderMeerP.PrognosticvalueofN-terminal
proC-typenatriureticpeptideinheartfailurepatientswithpreservedand
reducedejectionfraction.EurJHeartFail.2014;16:958–966.
35.deBoerRA,LokDJ,JaarsmaT,vanderMeerP,VoorsAA,HillegeHL,van
VeldhuisenDJ.Predictivevalueofplasmagalectin-3levelsinheart
failurewithreducedandpreservedejectionfraction.AnnMed.2011;43:
60–68.
36.BraunwaldE.Anothersteptowardpersonalizedcareofpatientswithheart
failure.EurJHeartFail.2015;17:988–990.
ORIGI
N
AL
RESE
ARCH
DOI:10.1161/JAHA.116.003989JournaloftheAmericanHeartAssociation11
BiomarkersPro?lesinHeartFailureTrompetal
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
SupplementalMaterial
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
TableS1.Differencesbetweenentirecohortandsubcohort
Totalcohort(n=1023)Subcohort(n=460)
Treatmentallocation
Controlgroup3332,2
Basicsupport33,332,4
Intensivesupport33,635,4
Demographicsandclinicalsigns
Age(years)70.8±1170.6±11.1
Femalesex(%)37,537,4
BMI(kg/m226.9±5.327.0±5.6
SystolicBP(mmHg)118.3±21.0117.9±21.3
Heartrate(bpm)74.6±13.474.2±13.4
LVEF(%)33.7±14.432.5±14.0
PreviousHFhospitalization32,733,7
NYHAclass,II/III/IV(%)50.9/45.7/3.444/52/4
Medicalhistory(%)
Myocardialinfarction42,640,7
Stroke1614,8
Hypertension42,941,5
Atrialfibrillationofflutter4445,4
Diabetes29,329,3
COPD26,228,3
Laboratory
Hemoglobin(g/dL)13.1±2.013.2±2.1
Sodium(mmol/L)139±4138.6±4.3
Creatinine(μmol/L)125.0±53125.7±52.8
eGFR(mL/min/1.73m2)55.2±21.154.9±20.5
BUN(mmol/L)10.7(8.1‐15.2)11.0(8.2-15.5)
Treatmentatdischarge(%)
ACEinhibitororARB82,882,2
Betablocker66,267,8
Diuretic95,895,7
MRA54,156,3
Statin37,939,8
Digoxin30,233,7
Abbreviations:ACE,angiotensinconvertingenzyme;ARB,angiotensinIIreceptorblocker;BMI,bodymassindex;BP,bloodpressure;COPD,
chronicobstructivepulmonarydisease;eGFR,estimatedglomerularfiltrationrate;HF,heartfailure;HFpEF,heartfailurewithapreserved
ejectionfraction;HFrEF,heartfailurewithareducedejectionfraction;NYHA,NewYorkheartassociation.
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
TableS2.Biomarkerassaydata.
BiomarkerIntraAssay%CV
InterAssay
%CVLowCutoffHighCutoffUnits
LTBR13%13%0,02845ng/mL
Mesothelin12%12%6,1120ng/mL
MPO15%14%2800ng/mL
Neuropilin114%15%1900ng/mL
Osteopontin21%22%2,52500ng/mL
Pentraxin310%11%0,07150ng/mL
Periostin12%12%2,31921ng/mL
PIGR16%16%122341ng/mL
PSAP-B14%16%2530ng/mL
ST-29%10%0,28380ng/mL
Syndecan-125%24%2,4393ng/mL
TNFR1A11%13%0,02568ng/mL
Troy15%14%0,04487ng/mL
RAGE9%10%0,01985ng/mL
VEGFR113%12%0,38195ng/mL
NTProCNP11%12%0,0039ng/mL
WAP4C14%14%0,16130ng/mL
ANPpropeptide29%28%1600110000pg/mL
D-Dimer9%10%0,02826ug/mL
ESAM9%9%0,5110ng/mL
GDF-159%10%0,0146,4ng/mL
Angiogenin18%18%17040000ng/mL
CRP17%16%0,06533ug/mL
NGAL19%21%7,51500ng/mL
Biomarker
Lowcut
off
highcut
offInterassaycoefficientofvariation(%)
IL-60.100.8813
cTNI0.20100010
ET10.52507
BG
medicineTotalImprecisionDetectionlimit
Galectin-3
Intra-assayvariability
(%)
Intra-assayvariability
(%)LoBLoDLoQ
Measuring
range
3,25,60,861,131,321,4-94,8
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
TableS3.Logisticregressioncorrectingforclinicalconfounders
Marker
Odds
ratio(95%CI)P-value
Hs-CRP(doubling)1.25(1.08-1.46)0.003
Model11.29(1.09-1.52)0.003
Model21.28(1.08-1.51)0.004
Pentraxin-3(doubling)0.74(0.56-0.98)0.037
Model10.81(0.60-1.09)0.165
Model20.83(0.62-1.12)0.212
NT-proBNP(doubling)0.75(0.65-0.87)<0.001
Model10.68(0.57-0.82)<0.001
Model20.74(0.62-0.88)0.001
proANP(doubling)0.72(0.58-0.89)0.002
Model10.66(0.51-0.85)0.001
Model20.69(0.54-.0.89)0.004
VEGF(doubling)1.09(0.97-1.23)0.159
Model11.03(0.90-1.18)0.639
Model21.05(0.92-1.20)0.442
Model1:age,sex,eGFR,systolicbloodpressure,ahistoryofmyocardialinfarction;diabetes;atrialfibrillationandanemia
Model2:Model1+ACE-inhibitors/ARB&Beta-blockerusage
Abbreviations:Hs-CRP,high-senstiveC-reactiveprotein;NT-proBNP,N-terminalpro-brain-typenatriureticpeptide;Pro-ANP,pro-atrial-type
natriureticpeptide;VEGF,vascularendothelialgrowthfactor
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
TableS4.Sensitivityanalysisexclusiveinteractions
HFpEFHFrEF
BiomarkerRp-valueRp-value
HFpEF
IL6D-Dimer0.3610.630.1580.840
Pentraxin-3VEGF-0.3880.21-0.1571.000
PeriostinVEGF-0.476<0.001-0.1021.000
NGALPSAP-B10.3810.210.1471.000
HFrEF
NT-proBNPIL60.2041.0000.378<0.001
NT-proBNPEPO-A0.3151.0000.360<0.001
correctedp-value
Abbreviations:EPO-A,erythropoietin;HFpEF,heartfailurewithapreservedejectionfraction;HFrEF,heartfailurewithareducedejection
fraction;IL-6,Interleukin6;NGAL,neutrophilgelatinase-associatedlipocalin;NT-proBNP,N-terminalpro-brain-typenatriureticpeptide;
PSAP,prostate-specificacidphosphatase;VEGF,vascularendothelialgrowthfactor.
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
TableS5.Relationshipwithoutcomeofbiomarkers.
HFrEF(n=364)HFpEF(n=96)p-value1p-value2
InflammationHR(95%CI)p-valueHR(95%CI)p-value
hs-CRP,(doubling)0.98(0.88-1.08)0.6841.00(0.77-1.30)0.9820.6150.638
Pentraxin-3,(doubling)0.89(0.71-1.12)0.3361.14(0.73-1.77)0.5690.0740.277
GDF-15,(doubling)1.09(0.81-1.48)0.5632.06(1.16-3.65)0.0140.1800.064
RAGE,(doubling)1.05(0.86-1.28)0.6291.34(0.90-1.98)0.1470.1740.451
Interleukin6,(doubling)1.00(0.87-1.16)0.9431.68(1.14-2.48)0.0080.1360.014
TNF-α,(doubling)1.01(0.96-1.06)0.7070.97(0.88-1.07)0.5800.2820.610
TNF-α-R1a,(doubling)1.31(0.99-1.73)0.0571.47(0.90-2.39)0.1200.6660.653
Oxidativestress
MPO,(doubling)0.89(0.74-1.08)0.2431.12(0.70-1.79)0.6440.5050.276
Remodelling
Syndecan-1,(doubling)1.01(0.82-1.24)0.9551.38(0.99-1.93)0.0590.2440.163
Periostin,(doubling)1.03(0.80-1.33)0.8241.17(0.76-1.79)0.4850.7980.849
Galectin-3,(doubling)0.86(0.59-1.25)0.4252.57(1.19-5.53)0.0160.0700.026
ST-2,(doubling)0.98(0.86-1.11)0.6941.27(0.96-1.67)0.0920.2680.219
Osteopontin,(doubling)0.90(0.72-1.14)0.3981.60(0.98-2.62)0.0620.0040.009
TGF-?,(doubling)1.01(0.91-1.13)0.8341.07(0.89-1.28)0.4650.7020.466
Cariomyocytestretch
NT-proBNP,(doubling)1.28(1.14-1.43)<0.0011.42(1.10-1.84)0.0070.4170.605
proANP,(doubling)1.02(0.82-1.27)0.8401.23(0.85-1.76)0.2680.3640.437
TnI,(doubling)1.16(1.06-1.28)0.0011.07(0.86-1.33)0.5320.3470.269
Angiogenesis
VEGF,(doubling)0.88(0.81-0.96)0.0041.13(0.88-1.47)0.3260.2730.080
VEFGR(doubling)1.19(0.94-1.51)0.1561.37(0.93-2.01)0.1060.9180.603
Angiogenin,(doubling)0.89(0.74-1.06)0.1950.67(0.47-0.95)0.0260.1390.156
NT-proCNP,(doubling)0.96(0.73-1.25)0.7491.69(1.15-2.49)0.0070.2320.042
Neuropilin-1(doubling)1.12(0.85-1.48)0.4252.34(1.40-3.90)0.0010.0170.024
Arteriosclerosis
ESAM,(doubling)1.35(0.79-2.29)0.2681.77(0.85-3.71)0.1270.5710.528
Renalfunction
NGAL,(doubling)0.95(0.69-1.29)0.7290.84(0.46-1.53)0.5690.7020.884
BUN,(doubling)0.91(0.63-1.33)0.6321.03(0.49-2.19)0.9290.6730.816
Haematopoiesis
EPOa,(doubling)1.09(0.96-1.24)0.1971.27(1.00-1.62)0.0490.7450.129
Other
D-Dimer,(doubling)1.09(0.97-1.22)0.1321.31(0.99-1.75)0.0560.4520.331
WAP4C(doubling)1.18(0.93-1.50)0.1641.61(1.09-2.37)0.0160.4130.181
Mesothelin,(doubling)1.19(0.88-1.61)0.2580.93(0.46-1.89)0.8410.5690.803
PIGR(doubling)1.03(0.79-1.34)0.8251.79(1.13-2.83)0.0130.2070.101
PSAP(doubling)1.24(0.96-1.59)0.0991.26(0.79-2.01)0.3240.7860.844
ET-1,(doubling)1.30(0.93-1.80)0.1200.99(0.46-2.13)0.9720.7460.859
TROY(doubling)0.98(0.73-1.31)0.8681.63(1.05-2.54)0.0300.3510.101
1.Univariableinteractionp-value
2.Multivariableinteractionp-value
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
Abbreviations:BUN,bloodureanitrogen;cTNI,cardiactroponin-I;EPOa,erythropoietin;ESAM,endothelialcell-selectiveadhesionmolecule;
ET-1,endothelin-1;GDF-15,growthdifferentiationfactor15;HFrEF,heartfailurewithareducedejectionfraction;HFpEF,heartfailurewith
apreservedejectionfraction;hs-CRP,high-sensitiveC-reactiveprotein;IL-6,Interleukin6;MPO,myeloperoxidase;NGAL,neutrophil
gelatinase-associatedlipocalin;NT-proBNP,N-terminalpro-brain-typenatriureticpeptide;NT-proCNP,aminoterminalpro-C-typenatriuretic
peptide;PIGR,polymericimmunoglobulinreceptor;Pro-ANP,pro-atrial-typenatriureticpeptide;PSAP,prostate-specificacidphosphatase;
RAGE,receptorofadvancedglycationend-products;TGF-b,transforminggrowthfactorbeta;TNF-a,tumornecrosisfactoralpha;TNF-aR1a,
tumornecrosisfactoralphareceptor1a;VEGF,vascularendothelialgrowthfactor;VEGFR,vascularendothelialgrowthfactorreceptor;
WAP4C,WAP4disulfidecoredomainprotein;
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
FigureS1.PCAanalysis
PrincipalComponentAnalysis–PCAplotillustratingthefirsttwoprincipalcomponents,collectivelyaccountingfor43.4%(PC1accountingfor
30.8%,andPC2for12.6%)oftheoverallvarianceinthecombinedHFpEFandHFrEFbiomarkermeasurements.ThePCAwasperformed
usingHFpEFandHFrEFascategoricalvariables,wherebiomarkerlevelsaredisplayedasredandblueforpatientswithHFpEFandHFrEF
respectively.
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
FigureS2.SurivalstratifiedaccordingtoHFrEFandHFpEF
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
FigureS3.OsteopontininHFrEFandHFpEFforHFrelatedhospitalizationsat18months.
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
FigureS4.OsteopontininHFrEFandHFpEFforall-causemortalityat18months.
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
FigureS5.NeuropilininHFrEFandHFpEFforHFrelatedhospitalizationsat18months.
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
FigureS6.NeuropilininHFrEFandHFpEFforall-causemortalityat18months.
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
HansHillege,DirkJ.vanVeldhuisen,PetervanderMeerandAdriaanA.Voors
JasperTromp,MohsinA.F.Khan,IJsbrandT.Klip,SvenMeyer,RudolfA.deBoer,TinyJaarsma,
BiomarkerProfilesinHeartFailurePatientsWithPreservedandReducedEjectionFraction
OnlineISSN:2047-9980
Dallas,TX75231
ispublishedbytheAmericanHeartAssociation,7272GreenvilleAvenue,JournaloftheAmericanHeartAssociationThe
doi:10.1161/JAHA.116.003989
2017;6:e003989;originallypublishedMarch30,2017;JAmHeartAssoc.
http://jaha.ahajournals.org/content/6/4/e003989
WorldWideWebat:
Theonlineversionofthisarticle,alongwithupdatedinformationandservices,islocatedonthe
formoreinformation.http://jaha.ahajournals.orgAccesspublication.VisittheJournalat
isanonlineonlyOpenJournaloftheAmericanHeartAssociationSubscriptions,Permissions,andReprints:The
byguestonMay2,2017
http://jaha.ahajournals.org/
Downloadedfrom
|
|
