Ventricular fibrillation waveform characteristics differ according to the presence of a previous myocardial infarction: A surface ECG study in ICD-patients

ContentslistsavailableatScienceDirect

Resuscitation

j o ur na l h o me pa g e:ww w . e l s e v i e r . c o m / l o c a t e / r e s u s c i t a t i o n

Clinical

Paper

Ventricular

fibrillation

waveform

characteristics

differ

according

to

the

presence

of

a

previous

myocardial

infarction:

A

surface

ECG

study

in

ICD-patients

夽

Judith

L.

Bonnes

_,

_Jos

_Thannhauser

_,

_Mathilde

_C.

_Hermans

_,

_Sjoerd

_W.

_Westra

_,

Thom

F.

Oostendorp

_,

_Gjerrit

_Meinsma

_,

_Menko

_Jan

_de

_Boer

_,

_Marc

_A.

_Brouwer

_,

Joep

L.R.M.

Smeets

a_{DepartmentofCardiology,RadboudUniversityMedicalCenter,Nijmegen,TheNetherlands}

b_{DepartmentofCognitiveNeuroscience,DondersCenterforNeuroscience,RadboudUniversityMedicalCenter,Nijmegen,TheNetherlands}

c_{FacultyofElectricalEngineering,MathematicsandComputerSciences,DepartmentofAppliedMathematics,UniversityofTwente,Enschede,The}

Netherlands

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received11June2015

Receivedinrevisedform22July2015

Accepted20August2015

Keywords:

Ventricularfibrillation

Fourieranalysis

Myocardialinfarction

Implantablecardioverterdefibrillator

Electrocardiography

a

b

s

t

r

a

c

t

Background:Characteristicsoftheventricularfibrillation(VF)waveformreflectarrestdurationandhave beenincorporatedinstudiesonalgorithmstoguideresuscitativeinterventions.Findingsinanimals indicatethatVFcharacteristicsarealsoaffectedbythepresenceofapreviousmyocardialinfarction (MI).Asstudiesinhumansarescarce,weassessedtheimpactofapreviousMIonVFcharacteristicsin ICD-patients.

Methods:ProspectivecohortofICD-patients(n=190)withdefibrillationtestingattheRadboudumc (2010–2013).VFcharacteristicsofthe12-lead surfaceECGwerecomparedbetweenthreegroups: patientswithoutahistoryofMI(n=88),withapreviousanterior(n=47)andapreviousinferiorMI (n=55).

Results:Ascomparedtoeachoftheothergroups,themeanamplitudeandamplitudespectrumareawere lower,forananteriorMIinleadV3andforaninferiorMIinleadsIIandaVF.Acrossthethreegroups,the bandwidthwasbroaderintheleadscorrespondingwiththeinfarctlocalisation.Incontrast,thedominant andmedianfrequenciesonlydifferedbetweenpreviousanteriorMIandnohistoryofMI,beinglowerin theformer.

Conclusions:TheVFwaveformisaffectedbythepresenceofapreviousMI.Amplitude-relatedmeasures werelowerandVFwaslessorganisedintheECG-lead(s)adjacenttotheareaofinfarction.AlthoughVF characteristicsofthesurfaceECGhavesofarprimarilybeenconsideredaproxyforarrestdurationand metabolicstate,ourfindingsquestionthisparadigmandmayprovideadditionalinsightsintothefuture potentialofVF-guidedresuscitativeinterventions.

©2015ElsevierIrelandLtd.Allrightsreserved.

Introduction

Ventricularfibrillation(VF)isthefirstobservedcardiacrhythm in about 30% of out-of-hospital cardiac arrests (OHCAs).1 _{As a}

夽 ASpanishtranslatedversionoftheabstractofthisarticleappearsasAppendix

inthefinalonlineversionathttp://dx.doi.org/10.1016/j.resuscitation.2015.08.014.

∗ Correspondingauthorat:RadboudUniversityMedicalCenter,Departmentof

Cardiology,616,P.O.Box9101,6500HBNijmegen,TheNetherlands.

E-mailaddresses:judithbonnes@gmail.com,Judith.Bonnes@radboudumc.nl

(J.L.Bonnes).

1 _{Address:GeertGrootepleinZuid10,6525GANijmegen,TheNetherlands.}

2 _{Address:Kapittelweg29,6525ENNijmegen,TheNetherlands.}

3 _{Address:P.O.Box217,7500AEEnschede,TheNetherlands.}

potentialstrategy toimprovetheratherpooroutcomes,theVF signalitselfhasbecomesubjectofstudy.2–4

It has been demonstrated that the VF waveform can be relatedtosurvivalandfavourableneurologicaloutcome.5–7_Given

that the VF waveform reflects arrest duration and myocardial metabolic state, VF characteristics have also been studied to predict shock success, and may be used to guide the decision whethertooptforastrategyofimmediatedefibrillationor CPR-first.2–5 _{Based on this idea, a randomised trial was designed}

in which first shock delivery was guided by a VF-based algo-rithm incorporated into an automatic external defibrillator.8

This strategy did not result in improved outcomes. The basis for this study was derived from the many observational stud-iesthat suggesteda positive associationbetweenVFwaveform http://dx.doi.org/10.1016/j.resuscitation.2015.08.014

characteristicsandshocksuccess,albeitgenerallywithmodest pre-dictivevalues.2–5,9,10

In view of the above, there is room for improvement. The observedresultsmaybeexplainedbythefactthattheappearance oftheVFwaveformdoesnotmerelyreflecttimedelayormetabolic state,butisinfluencedbyotherfactorsaswell.Observationsin ani-malsindicatethatthepresenceofamyocardialinfarction(MI)is associatedwithlesscoarseVF.11,12_{Accordingly,itmaybedifficult}

todifferentiatewhetherfineVFreflectsalongerarrestduration– withalowerchanceofsuccessfuldefibrillation–orisrelatedtoa shortarrestdurationinananimalwithanMI.Evidenceonthistopic inhumansisscarce,althoughtherearestudiesusingintracardiac recordingsthatdemonstratethattheVFwaveformisaffectedby thepresenceofapreviousMI.13,14

Therefore,westudiedalargeseriesofpatientswhounderwent defibrillationtesting after implantablecardioverter defibrillator (ICD)implantationandassessedtheimpactofapreviousMIand itslocalisationonVFwaveformcharacteristicsofthesurface elec-trocardiogram(ECG).

Methods

Patientpopulation

WeidentifiedallfirstICDimplantationswithdefibrillation test-ingattheRadboudUniversityMedicalCenterfromJune2010till December2013.Forthepresentanalysis,westudiedpatientswith apreviousanteriorMI,apreviousinferiorMIandthosewithouta historyofMI.Exclusioncriteriawerethefollowing:age<18years, congenitalheartdisease,right-sidedICDimplants,noanalysable 12-leadECG-recordingoftheinducedVF.Inaddition,patientswere excludedincaseofahistoryofMIthatdidnotinvolvethe ante-riororinferiorwall.Patientswithbothanteriorandinferiorwall infarctionswereexcludedaswell.Giventheobservationaldesign ofthestudy,writteninformedconsentwasnotnecessarytoobtain accordingtotheDutchActonMedicalResearchinvolvingHuman Subjects.

ICDimplantationandtesting

The devices implanted were Medtronic® (Minneapolis, MN, USA),StJudeMedical®(St.Paul,MN,USA)orBiotronik®(Berlin, Germany)ICDorcardiac resynchronisationtherapy-defibrillator systems with transvenous single coil leads. Routine defibrilla-tion testing was performed after ICD implantation to test the ability of the implanted device to sense, detect and terminate VFappropriately.Aftersedationwithpropofol,VFwasinduced usingT-waveshock,directcurrentpulsesor50Hzburstpacing. ThepresenceofVF,definedasarapid(around300bpm)grossly irregularventricularrhythmwithmarkedvariabilityinQRScycle length,morphology,andamplitude,wasconfirmedonsurface ECG-recordings.15_{Sequential shocksweredelivered(15–25–35joule)}

untilVFwasterminated.IncaseofpersistingVFafterthethird shock,anexternaldefibrillationshockwasdelivered.

Dataacquisition

Demographic,clinicalandechocardiographicparameterswere collectedfrompatientrecords.

During defibrillation testing, a standard 12-leadECG of the inducedVFwasrecorded(samplingfrequency1000Hz;16-bitA/D converter)withBARD®LabSystem(Lowell,MA,USA).LeadI,II,aVF, V1,V3andV6wereselectedforVFanalysis,astheserepresentthe mainelectricalvectorsandincludeuni-andbipolarleads.

VFwaveformanalysis

VFwaveformcharacteristicsweredeterminedusingatime seg-mentof4.1spriortofirstshockdelivery(4096time-points).The signalwaspre-processedwitha2Hzhigh-passfilteranda20Hz low-passfilter. Tostudy differentaspects of the VFwaveform, we analysed severalpreviously studiedVFcharacteristics.4,16,17

FromtheECG-signalinthetimedomain,wedeterminedthemean absoluteamplitude.Subsequently,thesignalwasconvertedtothe frequencydomainbyusingafastFouriertransformtovisualise thefrequenciesandcorrespondingamplitudeswhichthesampled VFsignalcontains.4_{Fromtheamplitudefrequencyspectrum,the}

amplitudespectrumarea(AMSA)wascalculatedasthesummed productofindividualfrequenciesandtheircorresponding ampli-tudesoveranintervalof2–20Hz.16,17_{Fromthepowerspectrum,}

wedeterminedthedominantfrequency,whichisthefrequency wherethepowerspectrumattainsitsmaximum.4_{Inaddition,we}

determinedthemedianfrequency,i.e.thefrequencyforwhichthe integratedsignalpowerwasonehalfofthetotalintegratedpower. Finally,wecalculatedthebandwidth,which isthedifferencein frequencycorrespondingtothefirstandthirdquartileofthetotal power,providingameasureofthespreadinfrequencies.17

Defi-nitionsoftheanalysedVFcharacteristicsaredescribedindetailin theAppendix.CalculationswereperformedusingMatlab(version 2011a,Mathworks,Natick,MA,USA).

Outcomemeasuresandstudygroups

TheoutcomemeasuresaretheECG-characteristicsof theVF waveformasdescribedabove.Thesewerecomparedbetweenthree studygroups:patientswithapreviousanteriorMI,aprevious infe-riorMIandwithoutahistoryofMI.Evidenceforthepresenceor absenceofapreviousMIwasbasedonreportsinthemedicalcharts. MIwasdefinedaccordingtotheESCcriteria.18_{Infarctlocalisation}

wasbasedonacombinationofinformationobtainedfromECGs (e.g.areawithST-elevationorpathologicalQwaves)andcoronary angiographies,andwasconfirmedbyimagingreports.18,19_Imaging

wasperformedandanalysedaspartofdailyclinicalpractice fol-lowingtherecommendationsoftheACC/AHA/ESCguidelinesfor ICDtherapyandwasnotpartofaparticularstudyprotocol.18,20

Inshort,forMRIthepresenceofdelayedenhancement,regional wallmotionabnormalitiesandinformationonmyocardial thick-nessandthickeningwereusedinthediagnosisandlocalisationof theMI.Incaseofechocardiography,informationonregional dif-ferencesinwallmotion,myocardialthicknessandthickeningwere used.Fornuclearimaging,thecombinationofapersistentperfusion defectandwallmotionabnormalitieswereusedasindicators.For allmodalities,wallmotionabnormalitieshadtobepresentinat leasttwocorrespondingsegments.

Statisticalanalysis

StatisticalanalysiswasperformedwithIBMSPSSstatistics soft-wareversion20(IBMCorp.,Armonk,NY,USA).Baselinevariables and the VF waveform characteristics were compared between thethreestudygroups. Categoricaldatawereexpressedas fre-quencies(percentages) and analysed using theChi square test. Continuousbaselinevariableswerereportedasmeans_±standard deviations and compared using the analysisof variance analy-sis(ANOVA).TheVFwaveform characteristicswerereportedas medians(interquartileranges)andcomparedbetweenthethree study groups with the Kruskal–Wallis test. Post hoc pairwise comparisonswereperformedusingMann–WhitneyUwith Bon-ferronicorrection.Ap-valueof<0.05wasconsideredstatistically significant.

Table1

Baselinecharacteristicsofthestudypopulation.

NopreviousMIn=88 PreviousanteriorMIn=47 PreviousinferiorMIn=55 p-Value

Age(years) 58±14 64±12 68±9 <0.001 Malegender 57(65) 35(75) 51(93) 0.001 Hypertension 30(34) 18(38) 23(43) 0.59 Diabetes 17(19) 13(28) 13(24) 0.53 Atrialfibrillation 26(30) 11(23) 20(36) 0.36 Secondaryprevention 24(27) 16(34) 27(49) 0.03 CRT-D 33(38) 14(30) 12(22) 0.14 BMI(kg/m2_{) 26±4 27±4 28±4 0.13} LVEF(%) 37±16 31±11 38±14 0.09 LVIDd-index(cm/m2_{) 3.1±0.6 3.2±0.4 3.0±0.4 0.52} QRSduration(ms) 126±31 116±25 125±28 0.16 Creatinine(␮mol/l) 109±151 99±66 100±35 0.85 Betablocker 76(86) 45(96) 48(89) 0.24 ACEinhibitor/ARB 73(83) 41(87) 49(91) 0.41 Aldosteroneblocker 31(35) 28(60) 15(28) 0.003 Diuretics 41(47) 29(62) 21(39) 0.07 Antiplatelet 33(38) 34(72) 44(82) <0.001 Anticoagulation 35(40) 21(45) 20(37) 0.73 Cholesterolreducer 35(40) 41(87) 48(89) <0.001 Amiodarone 9(10) 5(11) 9(17) 0.49

Valuesaren(%)ormeans±standarddeviations.ACEinhibitor:angiotensinconvertingenzymeinhibitor,ARB:angiotensinreceptorblocker,BMI:bodymassindex,CRT-D: cardiacresynchronisationtherapy-defibrillator,LVEF:leftventricularejectionfraction,LVIDd:leftventricularinternaldiastolicdiameter,MI:myocardialinfarction.

Results

Studygroup

Intotal,337eligiblepatientsunderwentafirstICD

implanta-tionwithdefibrillationtesting.Twelve-leadECG-recordingsofVF

inductionswereavailablein214patients;theECGcouldnotbe

analysedin8patientsduetoartefacts.Sixteenpatientswitha

pre-viousMIwereexcluded,becausetheyhadevidenceofinfarctionof

(1)boththeanteriorandtheinferiorwall(n=15)or(2)the

pos-teriorwallonly(n=1).Accordingly,atotalof190patientswere

includedinthepresentanalysis.Baselinecharacteristicsdidnot

differbetweenthein-andexcludedpatientgroups(Supplementary

Table1).

Patientcharacteristics

Themeanagewas63years±13and75%(143/190)weremale. In35%(67/190)ofpatients, ICDswereimplantedfor secondary prevention.Themeanleftventricularejectionfraction(LVEF)was 36%±14.In46%(88/190)ofpatients,therewasnoevidenceofa previousMI.ApreviousanteriorMIwaspresentin25%(47/190) andapreviousinferiorMIin29%(55/190).Ofthesubsetwithouta historyofMI,leftventriculardysfunctionwascausedby cardiomy-opathies,hypertensiveorvalvularheartdiseasesin88%(77/88), while5%(4/88)hadacardiacchannelopathyandin8%(7/88)the causeoftheventriculararrhythmiaswasunknown.Baseline char-acteristicsofthestudygroupsarepresentedinTable1.

VFwaveformcharacteristics

Studygroups: InFigs. 1and 2,wepresent themedianswith interquartilerangesoftheanalysedVFcharacteristicsforthestudy groups.Theonly leadwithoutbetween-groupdifferencesin VF characteristicswasV1.Theexactnumericalvaluescanbefound inSupplementaryTable2.

PreviousanteriorMI:Weobservedalowermeanamplitudeand AMSAandabroaderbandwidthforpatientswithaprevious ante-riorMIinleadV3thanforpatientswithapreviousinferiorMIand thosewithoutahistoryofMI,respectively.Whencomparedtothe latter,themeanamplitudeandAMSAwerelowerinleadV6fora previousanteriorMI.Thedominantandmedianfrequencieswere

lowerforpatientswithananteriorMIinleadsV3,IandV6when comparedtothosewithoutahistoryofMI.Theadjustedp-values ofpairwisecomparisonsarepresentedinSupplementaryTable3.

PreviousinferiorMI:Weobservedalowermeanamplitudeand AMSAandabroaderbandwidthforpatientswithaprevious infe-riorMIinleadsIIandaVFthanforpatientswithapreviousanterior MIorthosewithoutahistoryofMI,respectively.Whencompared tothelatter,themeanamplitudeandAMSAwerelowerandthe bandwidthwasbroaderinleadV6.Thedominantandmedian fre-quenciesdidnotdifferfromthoseoftheotherstudygroupsinall leads.Theadjustedp-valuesofthepairwisecomparisonscanbe foundinSupplementaryTable3.

Discussion

To our knowledge, this is the largest surface ECG study in humansanalysingtheimpactofthepresenceofapreviousMIand itslocalisationoncharacteristicsoftheVFwaveform. Amplitude-relatedcharacteristicsoftheinducedVFwaveformweremarkedly lowerintheleadsadjacenttotheareaofinfarction,i.e.inleadV3 forananteriorMI,andinleadsIIandaVFforaninferiorMI.Inthe leadscorrespondingwiththeinfarctlocalisation, thebandwidth wasbroaderaswell,indicatinglessorganisedVF.Thedominant andmedianfrequenciesofVFwereaffectedincaseofananterior MIonly.Ourobservationschallengethecurrentlyusedconceptthat fineVFisamereproxyforalongerarrestdurationand myocar-dialmetabolicstateandindicatethattheunderlyingaetiologyalso playsanimportantroleintheappearanceoftheVFwaveformon thesurfaceECG.

Animalstudies

Animalstudiesprimarilyinvestigatedtheimpactofaprevious anterior wallinfarction,inducedby ligationof theleftanterior descendingarteryandreportedlowerVFfrequency characteris-ticsforthissubsetwhencomparedtocontrols.21,22_Weobserved

lowerdominantand medianfrequencies inpatientswitha pre-viousanteriorMI aswell,but notincase ofa previousinferior MI.Whilewefocusedonearly,short-duration,inducedVF,animal studiesanalysedtheVFsignalatdifferenttimepointsduringthe arrest.21–23_{Inthesestudies,ithasalsobeendemonstratedthat–}

Fig.1. Amplitude-relatedVFwaveformcharacteristicsintheanalysedsurfaceECG-leads.VFwaveformcharacteristicsarepresentedasmedianswithinterquartileranges.

**p-Value<0.001forcomparisonsbetweenthethreestudygroups.AMSA:amplitudespectrumarea,Hz:hertz,MI:myocardialinfarction,mV:millivolt.Pairwisecomparisons

showedsignificantdifferencesbetweenthefollowingstudygroups:Meanamplitude:LeadII:inferiorvs.anteriorMI(p<0.001);inferiorvs.noMI(p<0.001).LeadaVF:inferior

vs.anteriorMI(p<0.001);inferiorvs.noMI(p<0.001).LeadV3:anteriorvs.inferiorMI(p=0.001);anteriorvs.noMI(p<0.001).LeadV6:anteriorvs.noMI(p=0.001);

inferiorvs.noMI(p<0.001).AMSA:LeadII:inferiorvs.anteriorMI(p=0.001);inferiorvs.noMI(p<0.001).LeadaVF:inferiorvs.anteriorMI(p=0.001);inferiorvs.noMI

(p<0.001).LeadV3:anteriorvs.inferiorMI(p<0.001);anteriorvs.noMI(p<0.001).LeadV6:anteriorvs.noMI(p=0.001);inferiorvs.noMI(p=0.001).

settingofanacute,ongoingMIseemstohaveevenlowerfrequency characteristics.21,23 _{Mappingstudiesinanimalsprovided insight}

intothepotentialelectrophysiologicalbackgroundoftheobserved lowerfrequencycharacteristics.IndogheartswithapreviousMI,it wasobservedthatduringVFthemeansizeofactivationwavefronts waslargerthanincontrols.Largerwavefrontshavebeenassociated withlowerdominantfrequenciesontheECG.24,25

Withregardtotheamplitude-relatedVFcharacteristics, obser-vations in animals were less uniform.11,12,21,22,26 _{This may be}

related to thefact that some studies analysed paddle ECGs or recordingsfromneedleelectrodes(y-axis)–moreorless corre-spondingtoleadIIofthesurfaceECG–whilethelocalisationofthe (previous)MIwasnotintheinferiorbutintheanteriorwall.12,21,26

Hypothetically,aneffectonamplitude-relatedcharacteristicsmay havegoneundetectedinthesestudiesduetothelackofchestleads. Humanstudies

OHCA-setting:Studies ontheimpactofan MI onVF charac-teristicsinOHCA-patientsarehamperedbyalackofinformation ontheexactunderlyingaetiology,especiallyinthenon-surviving patients. Nonetheless, for patients with VFin the setting of a

suspectedacuteMI,thewaveformshowedsignificantdifferences when compared to patientswithout acute MI, including lower AMSAvalues.27 _{In a study onshock successprediction, it was}

described that patientswith a prior MI demonstrated a lower medianfrequency,butnotalowerAMSA.5_{However,intheabsence}

ofdataontheactualarrestaetiology,itremainsuncertaintowhat extentthebaselinecharacteristics,ontheonehand,andtheactual underlyingaetiology,ontheother,weretheprimarydriversbehind theobserveddifferencesintheVFwaveform.

Defibrillationtesting:Intheabove-describedcontext,VF induc-tionduringdefibrillationtestingprovidesauniquesettingtostudy theVFwaveform undercontrolledconditions.13,14,17 _Asforthe

impactofapreviousMI,asmallstudyanalysinglimbleadsreported atrendtowardslowerVFfrequenciesinischaemicthanin nonis-chemicheartdisease.17_{Withregardtoinfarctlocalisation,astudy}

ofintracardiacrecordedVFreportedahigherfractionofenergyin thelowfrequencyregionforpatientswithapreviousinferiorMI thanforananteriorMI.13

Infollow-upofourpreviousintracardiacECGstudythat demon-strateddifferencesintheVFwaveformrelatedtothehistoryofa previousMI,wenowconductedacomprehensive12-leadsurface ECGstudytoexpandthecurrentknowledgeontheimpactofanMI

Fig.2.Frequency-relatedcharacteristicsoftheVFwaveformintheanalysedsurfaceECG-leads.VFwaveformcharacteristicsarepresentedasmedianswithinterquartile

ranges.*p-Value<0.05forcomparisonsbetweenthethreestudygroups.Hz:hertz,MI:myocardialinfarction.Pairwisecomparisonsshowedsignificantdifferencesbetween

thefollowingstudygroups:Dominantfrequency:LeadI:anteriorvs.noMI(p=0.026).LeadV3:anteriorvs.noMI(p=0.015).LeadV6:anteriorvs.noMI(p=0.009).Median

frequency:LeadI:anteriorvs.noMI(p=0.019).LeadV3:anteriorvs.noMI(p=0.013).LeadV6:anteriorvs.noMI(p=0.021).Bandwidth:LeadII:inferiorvs.anteriorMI

(p=0.054);inferiorvs.noMI(p=0.012).LeadaVF:inferiorvs.anteriorMI(p=0.063);inferiorvs.noMI(p=0.017).LeadV3:anteriorvs.inferiorMI(p=0.017);anteriorvs.no

MI(p=0.001).LeadV6:inferiorvs.noMI(p=0.031).

anditslocalisationontheVFwaveform.14_{Wefoundthatamplitude}

characteristicswerelowerinV3forananteriorMIandinIIandaVF foraninferiorMI,i.e.intheleadsadjacenttotheareaofinfarction. Incontrast,thedominantandmedianfrequencieswereloweronly

inthepresenceofapreviousanteriorMI(I,V3,V6).Anexplanation forthisfindingcouldbethatthepatientswithanteriorMIshad largerinfarctions,whichmayhaveaffectedtheVFwaveformtoa greaterextentthanthesmallerinferiorinfarctions.Thishypothesis

issupportedbyfactthattheLVEFwasslightlylowerinthesubset withapreviousanteriorMIthanintheotherstudygroups.Inlead V6,weobservedloweramplitudesandlessorganisedVFbothfor patientswithapreviousanteriorandinferiorMIwhencompared tothesubsetwithoutanMI.Thismayindicatethatinvolvementof thelateralwalloccurredinbothgroupswithapreviousMI. Implications

Ourfindingsthatbothamplitudeandfrequencycharacteristics ofearlyVFareaffectedbyapriorMIanditslocalisationwarranta morecomprehensiveconceptoftheVFwaveformthatgoesbeyond anindicatorofarrestduration.Inaddition,previousstudieshave shownthattheVFwaveformhaspromisingpredictivevaluefor longer-termclinicaloutcomes.5–7

ThefirstimplicationofourfindingsisthatfineVFduringOHCA doesnot necessarily reflect longerarrest duration, with a low chanceofsuccessfuldefibrillation,butmaybeinfluencedbythe underlyingaetiologyaswell.Forexample,inpatientswitha pre-viousinferiorMI,alowAMSAatthepaddleECG(corresponding tolimbleadII)maybetheresultofthepreviousMIratherthan alongarrestduration.Theissuemaybeovercomewiththeuse ofmultipleleads,reflectingdifferentrecordingdirections,butthe potentialbenefitofsuchanapproachneedsfurtherstudy.Aslead V1doesnotseemtobeinfluencedbythepresenceofapriorMI, onemighthypothesisethatVFcharacteristicsinthatleadaremore reliableindicatorsofarrestdurationandlessaffectedbyleft ven-triculardisease.Inthecontextthatthenumberandchoiceofleads affectsstudyfindings,theuseofonlyoneleadduringOHCAseems acomplicatingfactor.

Further study is warranted to see whether our findings on short-duration,electricallyinducedVFapplytotheOHCA-setting. SignaldifferencesbetweenspontaneousandinducedVFhavebeen described,butanimalstudiesandanOHCAcohorthavereported thatapreviousMIaffectstheVFwaveformintheacutesettingas well.11,12,21,22,27–29

AppreciatingthatotherfactorsthantimeaffecttheVF wave-form,the modest prediction in earlier OHCA studies on shock successcouldpossiblybeoptimised.Morecomprehensivestudies onVFcharacteristicsarewarrantedtounravelthefuturepotential ofVF-guidedresuscitativeinterventions.

Limitations

This study describes the impact of a previous infarction in patientswith short-duration, electrically induced VF.This may limitinferencestotheOHCA-settingwithspontaneousVFoflonger duration.Second,aboutathirdofpatientswasexcluded,as digi-tal12-leadECG-recordingswerenotavailable.Giventhesimilar baselinecharacteristics(SupplementaryTable1),itisunlikelythat thisselectionhasaffectedourfindings.Third,for aneven more academicapproach,theuseofasingleimagingtechnique(MRI), includingsystematicquantificationofinfarctextent,wouldhave beenpreferable.

Conclusions

InthepresentstudyonVFcharacteristicsinalargeseriesof patientsundergoingICDtesting,weobservedthatapreviousMI anditslocalisationaffecttheVFwaveform onthesurface ECG. Amplitude-relatedcharacteristicsweresignificantlylowerandVF waslessorganisedintheleadsadjacenttotheareaofinfarction,i.e. inleadV3forananteriorMIandinleadsIIandaVFforaninferior MI.Frequencycharacteristicswerelower incase of ananterior MIonly.Although VFcharacteristicshave sofarprimarilybeen

consideredaproxyforarrestdurationandmyocardialmetabolic state,our findings warrant more comprehensivestudies on VF tore-evaluatethisparadigmandtostudythefuturepotentialof VF-guidedresuscitativeinterventions.

Funding

None.

Conflictofintereststatement

Prof.DeBoerisamemberoftheEuropeanadvisoryboardon interventionalcardiologyofMedtronic.J.L.Bonnes,J.Thannhauser, M.C. Hermans,S.W. Westra, T.F. Oostendorp, G.Meinsma, M.A. BrouwerandJ.L.R.M.Smeetshavenoconflictsofinteresttodeclare.

AppendixA. Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound, intheonlineversion,athttp://dx.doi.org/10.1016/j.resuscitation. 2015.08.014.

References

1.BerdowskiJ,BergRA,TijssenJG,KosterRW.Globalincidencesofout-of-hospital cardiacarrestandsurvivalrates:systematicreviewof67prospectivestudies. Resuscitation2010;81:1479–87.

2.Strohmenger HU. Predicting defibrillation success. Curr Opin Crit Care 2008;14:311–6.

3.CallawayCW,MenegazziJJ.Waveformanalysisofventricularfibrillationto pre-dictdefibrillation.CurrOpinCritCare2005;11:192–9.

4.ReedMJ,CleggGR,RobertsonCE.Analysingtheventricularfibrillation wave-form.Resuscitation2003;57:11–20.

5.RistagnoG,MauriT,CesanaG,etal.,AziendaRegionaleEmergenzaUrgenza ResearchGroup.Amplitudespectrumareatoguidedefibrillation:avalidation on1617patientswithventricularfibrillation.Circulation2015;131:478–87.

6.IndikJH,ConoverZ,McGovernM,etal.Associationofamplitudespectralareaof theventricularfibrillationwaveformwithsurvivalofout-of-hospitalventricular fibrillationcardiacarrest.JAmCollCardiol2014;64:1362–9.

7.SchoeneP,CoultJ,MurphyL,etal.Courseofquantitativeventricularfibrillation waveformmeasureandoutcomefollowingout-of-hospitalcardiacarrest.Heart Rhythm2014;11:230–6.

8.FreeseJP,JorgensonDB,LiuPY,etal.Waveformanalysis-guidedtreatment versusastandardshock-firstprotocolforthetreatmentofout-of-hospital car-diacarrestpresentinginventricular fibrillation:resultsofaninternational randomised,controlledtrial.Circulation2013;128:995–1002.

9.StrohmengerHU,LindnerKH,BrownCG.Analysisoftheventricularfibrillation ECGsignalamplitudeandfrequencyparametersaspredictorsofcountershock successinhumans.Chest1997;111:584–9.

10.LinLY,LoMT,KoPC,etal.Detrendedfluctuationanalysispredictssuccessful defibrillationforout-of-hospitalventricularfibrillationcardiacarrest. Resusci-tation2010;81:297–301.

11.IndikJH,ShanmugasundaramM,AllenD,etal. Predictorsofresuscitation outcomeinaswinemodelofVFcardiacarrest:acomparisonofVF dura-tion,presenceofacutemyocardialinfarctionandVFwaveform.Resuscitation 2009;80:1420–3.

12.IndikJH,AllenD,ShanmugasundaramM,etal.Predictorsofresuscitationina swinemodelofischemicandnonischemicventricularfibrillationcardiacarrest: superiorityofamplitudespectralareaandslopetopredictareturnof spon-taneouscirculationwhenresuscitationeffortsareprolonged.CritCareMed 2010;38:2352–7.

13.Sánchez-Mu ˜nozJJ,Rojo-AlvarezJL,García-AlberolaA,etal.Effectsofthelocation ofmyocardialinfarctiononthespectralcharacteristicsofventricularfibrillation. PacingClinElectrophysiol2008;31:660–5.

14.BonnesJL,KeuperW,WestraSW,etal.Characteristicsofventricularfibrillation inrelationtocardiacaetiologyandshocksuccess:awaveformanalysisstudyin ICD-patients.Resuscitation2015;86:95–9.

15.AmericanCollege ofCardiology/AmericanHeartAssociation TaskForceon ClinicalDataStandards(ACC/AHA/HRSWritingCommitteetoDevelopData Standards on Electrophysiology)Buxton AE, Calkins H, Callans DJ, et al. ACC/AHA/HRS 2006 key data elements and definitions for electrophys-iological studies and procedures: a report of the American College of Cardiology/AmericanHeartAssociationTaskForceonClinicalDataStandards (ACC/AHA/HRSWritingCommitteetoDevelopDataStandardson Electrophys-iology).Circulation2006;114:2534–70.

16.PovoasHP,WeilMH,TangW,BiseraJ,KloucheK,BarbatsisA.Predicting thesuccessofdefibrillationbyelectrocardiographicanalysis.Resuscitation 2002;53:77–82.

17.IndikJH,PetersCM,DonnersteinRL,OttP,KernKB,BergRA.Directionof signalrecordingaffectswaveformcharacteristicsofventricularfibrillationin humansundergoingdefibrillationtestingduringICDimplantation. Resuscita-tion2008;78:38–45.

18.Thygesen K,AlpertJS,Jaffe AS,et al.,ESCCommitteeforPractice Guide-lines(CPG).Thirduniversaldefinitionofmyocardialinfarction.EurHeartJ 2012;33:2551–67.

19.ZimetbaumPJ,JosephsonME.Useoftheelectrocardiograminacutemyocardial infarction.NEnglJMed2003;348:933–40.

20.ZipesDP, CammAJ,Borggrefe M,etal. ACC/AHA/ESC2006guidelinesfor managementofpatientswithventriculararrhythmiasandthepreventionof suddencardiacdeath—executivesummary:areportoftheAmericanCollege ofCardiology/AmericanHeartAssociationTaskForceandtheEuropean Soci-etyofCardiologyCommitteeforPracticeGuidelines(WritingCommitteeto DevelopGuidelinesforManagementofPatientswithVentricularArrhythmias andthePreventionofSuddenCardiacDeath)Developedincollaborationwith theEuropeanHeartRhythmAssociationandtheHeartRhythmSociety.Eur HeartJ2006;27:2099–140.AmericanCollegeofCardiology/AmericanHeart AssociationTaskForce;EuropeanSocietyofCardiologyCommitteeforPractice Guidelines;EuropeanHeartRhythmAssociationandtheHeartRhythmSociety.

21.IndikJH,DonnersteinRL,HilwigRW,etal.Theinfluenceofmyocardialsubstrate onventricularfibrillationwaveform:aswinemodelofacuteandpostmyocardial infarction.CritCareMed2008;36:2136–42.

22.IndikJH,DonnersteinRL,KernKB,GoldmanS,GaballaMA,BergRA. Ven-tricularfibrillationwaveformcharacteristicsaredifferentinischemicheart

failure compared with structurally normal hearts. Resuscitation 2006;69: 471–7.

23.JacobsonJT,JohnsonD,HorvathG,GoldbergerJ,KadishA.Effectofunderlying heartdiseaseonthefrequencycontentofventricularfibrillationinthedogheart. PacingClinElectrophysiol2000;23:243–52.

24.DamleRS,RobinsonNS,YeDZ,etal.Electricalactivationduringventricular fibrillationinthesubacuteandchronicphasesofhealingcaninemyocardial infarction.Circulation1995;92:535–45.

25.MandapatiR,AsanoY,BaxterWT,GrayR,DavidenkoJ,JalifeJ.Quantification ofeffectsofglobalischemiaondynamicsofventricularfibrillationinisolated rabbitheart.Circulation1998;98:1688–96.

26.IndikJH,AllenD,GuraM,DameffC,HilwigRW,KernKB.Utilityofthe ven-tricularfibrillationwaveformtopredictareturnofspontaneouscirculation anddistinguishacutefrompostmyocardialinfarctionornormalSwinein ventricularfibrillation cardiacarrest.CircArrhythmElectrophysiol 2011;4: 337–43.

27.OlasveengenTM,EftestølT,GundersenK,WikL,SundeK.Acuteischemicheart diseasealtersventricularfibrillationwaveformcharacteristicsinout-ofhospital cardiacarrest.Resuscitation2009;80:412–7.

28.Sánchez-Mu ˜nozJJ,Rojo-AlvarezJL,García-AlberolaA,etal.Spectralanalysisof intracardiacelectrogramsduringinducedandspontaneousventricular fibrilla-tioninhumans.Europace2009;11:328–31.

29.Lever NA, Newall EG, Larsen PD. Differences in the characteristics of inducedandspontaneousepisodesofventricularfibrillation.Europace2007;9: 1054–8.