The heart of epilepsy: Current views and future concepts

Review

The heart of epilepsy: Current views and future concepts

S. Shmuely

, M. van der Lende

, R.J. Lamberts

, J.W. Sander

, R.D. Thijs

*

aStichtingEpilepsieInstellingenNederland—SEIN,Heemstede,TheNetherlands

bNIHRUniversityCollegeLondonHospitalsBiomedicalResearchCentre,UCLInstituteofNeurology,QueenSquare,LondonWC1N3BG,UK

cDepartmentofNeurology,LUMCLeidenUniversityMedicalCentre,Leiden,TheNetherlands

ARTICLE INFO

Articlehistory:

Received1September2016 Accepted3October2016

Keywords:

Epilepsy Seizures Cardiac Cardiovascular Comorbidity Arrhythmia

ABSTRACT

Cardiovascular(CV)comorbiditiesarecommoninpeoplewithepilepsy.Severalmechanismsexplainwhy theseconditionstendtoco-existincludingcausalassociations,sharedriskfactorsandthoseresulting fromepilepsyoritstreatment.

Variousarrhythmiasoccurringduringandafterseizureshavebeendescribed.Ictalasystoleisthemost commoncause.Theconversephenomenon,arrhythmiascausingseizures,appearsextremelyrareand hasonlybeenreportedinchildrenfollowingcardioinihibitorysyncope.Arrhythmiasinepilepsymaynot onlyresultfromseizureactivitybut alsofromasharedgenetic susceptibility.Variouscardiacand epilepsygenescouldbeimplicatedbutfirmevidenceisstilllacking.Severalantiepilepticdrugs(AEDs) triggeringconductionabnormalitiescanalsoexplaintheco-existenceofarrhythmiasinepilepsy.

Epidemiologicalstudieshaveconsistentlyshownthatpeoplewithepilepsyhaveahigherprevalenceof structuralcardiacdiseaseandapoorerCVriskprofilethanthosewithoutepilepsy.SharedCVriskfactors, geneticsandetiologicalfactorscanaccountforasignificantpartoftherelationshipbetweenepilepsyand structuralcardiacdisease.SeizureactivitymaycausetransientmyocardialischaemiaandtheTakotsubo syndrome.Additionally,certainAEDsmaythemselvesnegativelyaffectCVriskprofileinepilepsy.

Herewediscussthefascinatingborderlandofepilepsyandcardiovascularconditions.Thereview focusesonepidemiology,clinicalpresentationsandpossiblemechanismsforsharedpathophysiology.It concludeswithadiscussionoffuturedevelopmentsandacallforvalidatedscreeninginstrumentsand guidelinesaidingtheearlyidentificationandtreatmentofCVcomorbidityinepilepsy.

ã2016BritishEpilepsyAssociation.PublishedbyElsevierLtd.Allrightsreserved.

1.Introduction

Wellover100yearsago,theoccurrenceofasystoleduringthe courseofanepilepticseizurewasdescribed:“Heutteredacryand was seen to be rubbing his hands together. His pulse was immediatelyexaminedforbutwasnotpalpable”[1].Sincethen numerousassociationsbetweenepilepsyandCVconditionshave beenidentified,includingthisclassicalexampleofictalasystole.

Co-existingconditions formanimportantpartoftheoverall burden of epilepsy [2–5]. Several mechanisms of association betweenepilepsyandcomorbidconditionshavebeendescribed:

associationscanbeexplainedbycauseoreffect,asharedriskfactor maycausebothconditions,orthemechanismoftheassociationis unknownorspurious(i.e.coincidental)(Fig.1)[3,5].

This review serves to discuss the fascinating borderland betweenepileptology and cardiologyand focuses onthemajor developmentsoverthelast25yearsandonfuturedevelopments.

Weusethecomorbidityframework(Fig.1)[3]toreviewallcardiac conditions known, and alleged, to be linked to epilepsy.

Associationswithcardiacarrhythmiasarediscussedfirst,followed by an overview of all structural cardiac conditions related to epilepsy.

2.Epilepsyandcardiacarrhythmias

Various arrhythmias have been described, occurring during (ictal)orafter(postictal)seizures.Sinustachycardiaisthemost commonictalpattern,seeninupto80%ofallseizures[6]andin 82%ofpeoplewithepilepsy[7],butusuallywithoutsymptoms.

Themostfrequentclinicallyrelevantarrhythmiaisictalasystole, occurring in 0.318% (95% CI 0.316–0.320%) of people with refractoryfocalepilepsyadmittedforvideo-EEG[8].Ictalasystole, bradycardia and AV block predominantly occurin people with

*Correspondingauthorat:StichtingEpilepsieInstellingenNederland—SEIN,P.O.

Box540,2130AMHoofddorp,TheNetherlands.Fax:+31235588159.

E-mailaddress:rthijs@sein.nl(R.D. Thijs).

http://dx.doi.org/10.1016/j.seizure.2016.10.001

1059-1311/ã2016BritishEpilepsyAssociation.PublishedbyElsevierLtd.Allrightsreserved.

ContentslistsavailableatScienceDirect

Seizure

j o u r n a l h o m e p a g e : w w w . e l s ev i er . c o m / l o c a te / ys e i z

temporal lobeepilepsy(Table 1)[8]. Clinically,ictal asystole is characterisedbysuddenlossoftoneduringadyscognitiveseizure [9].Thecirculatorypatternresemblesvasovagalsyncopewitha transient,progressiveandself-limitingslowingoftheheartrate and decrease of blood pressure [9–11]. For many years, ictal asystole was thought to be a possible mechanism underlying suddenunexpecteddeathinepilepsy(SUDEP).Thisappearstobe unlikely:allbutonereportedcasesofarofictalasystolewereself- limiting[8].Inthisonecasesuccessfulresuscitationwasstarted after44sofasystoleandtheeventwasclassifiedasnear-SUDEP [12].Thelongestictalasystolereportedsofar,however,lasted96s andappearedself-limiting[13].Whetheraneventisclassifiedas near-SUDEP or not will depend on interventions of medical personnel:promptresuscitationinresponsetoictalasystolewill likelyleadtomoreclassifiedasnear-SUDEPcases.Whilethereare noreportsoffatalictalasystole,itremainsdebatablewhetherictal asystolecancauseSUDEP.

The precisemechanism ofictal asystole isunknown.It may result from epileptic activity directly stimulating the central autonomicnetworks[6,14].Forexample,focalstimulationofparts ofthelimbicsystem(i.e.amygdala,cingulategyrus)mayprovoke asystole [6,15–17]. Alternatively, seizure-induced fear and cate- cholaminerelease[18] mayevokeavasovagalresponsecausing cardioinhibitionandvasodilation[19].

Ictalasystoleisassumedtobeself-limiting,butmaycausefalls andinjuriesduetoseizure-inducedsyncope[20].Propertrialsare lackingbutretrospectivestudiessuggestthatimprovingseizure controlmaypreventictalasystole[21–23].Italsoseemsadvisable to withdraw negative inotropic drugs and to consider the implantationofalooprecordertomonitorpossiblefutureevents in individuals in whom ictal asystole has been noted. If the asystolicepisodespersist,cardiacpacemakerimplantationshould beconsideredtoreducetheriskoftrauma[20,21,23,24].

Incontrasttoictalasystole,postictalasystoleislesscommon, associated with convulsive rather than focal (temporal lobe) seizuresandhasahigherfatalityrate:7of13reportedpostictal asystolecasesdiedfromSUDEP[8].Allfatalcaseshadaconvulsive seizurewithimmediatepostictalgeneralisedEEGsuppressionand astutteringcourseoftransientapnoeaandasystoleresultingina terminalapnoeafollowedbyaterminalasystole[25].

The mechanism underlying this sequence of postictal EEG suppression, apnoea, and terminal asystole has not yet been elucidated. Excessive inhibition causing brainstem depression mightplayarole[26].Recentworkintwoanimalmodels(mice carrying mutations in the KCNA1 gene or the SCN1A gene) demonstratedthatseizuresinitiatedbydirectcorticalstimulation mayevokeaspreadingdepressioncausingbrainsteminhibition andcardiorespiratorycollapse[27].

Anotherrare(post)ictalarrhythmiaisventriculartachycardia/

ventricularfibrillation(VT/VF).SofarthreecasesofpostictalVT/VF leadingto(near)SUDEPhavebeenreported[8].AllVT/VFoccurred directlyfollowing a convulsiveseizure. Nocardiaclesionswere found in the case reports. There may be a publication bias, however,ascaseswithseizure-triggeredVT/VFandcardiaclesions may not qualify as SUDEP and thus may be less likely to be reported. The mechanism of seizure-induced VT/VF is unclear.

Convulsive seizures may exert proarrhythmogenic effects by triggering thesympathetic nervoussystem, asreflectedby the peakincatecholaminesandelectrodermalactivity[18,28].Atthe same time, convulsive seizures may increase cardiac oxygen deprivation by inducing sinus tachycardia [7] and respiratory impairmentcausinghypoxemia[29].Ithasalsobeenfoundthat ECG-markers of suddencardiac death suchas QTc-lengthening and/or shortening [30,31], and T-wave alternans are more prevalent[23] duringandafterconvulsiveseizures.Thevarious factors might interact as seizure-related cardiac repolarization abnormalities appeared more frequent in seizures with ictal hypoxemiacomparedtothosewithout[32].

Thoughseizure-inducedVT/VFappearstoberare,aprospective community-basedstudyofout-of-hospitalcardiacarrestsdueto ECG-documented VT/VF showed that VT/VF risk in those with epilepsywasthreetimesashighasthegeneralpopulation[33].A furtheranalysisofthosecaseswithepilepsyandVT/VFshowed that most were not seizure-related, but rather occurredin the contextofeitherpre-existingheartdiseaseorastheimmediate resultof anacute myocardialinfarction[34]. Pre-existingheart diseasewasastrongerpredictorforVT/VFinpeoplewithepilepsy thanmarkersofepilepsyseverity.Inaminorityofcases,however, VT/VF was unexplained and a diagnosis of (near) SUDEP was established.ItthusappearsthatsuddencardiacarrestandSUDEP arepartiallyoverlappingdiseaseentities.

The increased risk of non-seizure related VF/VT episodes in people epilepsy may be explained by high cardiovascular Fig.1.Mechanismsofassociationbetweenepilepsyandcomorbidconditions.The

dottedlineindicatesthatanassociationdoesnotreallyexist.Figureoriginatesfrom Gaitatzisetal.[3],permissiontoreproducecopyrightedmaterialgrantedbyJohn Wiley&Sons.

comorbidity[3,35].Peoplewithepilepsymayhaveapropensityfor suddencardiacdeathasreducedheartratevariability,ameasureof cardiacsympathovagalbalancethatisalsoariskmarkerofsudden cardiac death, progressively worsens over time in people with refractory,butnotinthosewithwell-controlled,epilepsy[36].In addition,other markers of sudden cardiac deathsuch as early repolarizationpatternandQTc-prolongationaremorefrequently foundintheinterictalECGsofpeoplewithepilepsythaninthose withoutepilepsy[37].

Another mechanism explaining the association between arrhythmiasandepilepsyisasharedgeneticriskfactor.Arapidly increasingnumberofgenespotentiallylinkingepilepsytocardiac arrhythmiashasbeenidentified.Herewediscusssomerelevant examples;startingwiththegenespredominantlyknownfortheir cardiacfunctionsandthenthe‘epilepsygenes’.

Several genetic ion channel mutations are thought to be expressedinthebrainaswellasintheheart,andmightthuscause seizuresandcardiacarrhythmias.Thefirstreportedgeneticlink betweenepilepsyandcardiacarrhythmiaswas thediscoveryof cardiacsodiumchannelgeneSCN5Ainthebrain[38].Subsequent- ly,morepathogenicvariantsinthelongQT(LQT)genefamily(i.e.

KCNQ1,KCNH2 and SCN5A) were associated with a “seizure phenotype” (e.g. self-reported diagnosis of epilepsy and AED use)[39–44].Micemodelsindicatedthatother,non-LQT,cardiac channelopathygenesincluding RYR2 (associatedwithcatechol- aminergicpolymorphicventriculartachycardia)[45],andHCN1-4 [46,47]potentiallypredisposetoepilepsy.

SeveralpostmortemstudiessuggestthattheLQTandnon-LQT cardiacgenemutationsaremorecommoninSUDEPvictims[48– 50]. As ictal recordings are lacking, it remains questionable whetherthefataleventswerecausedbyarrhythmias.Thesame appliestotheidentificationof‘epilepsygenes’inthepost-mortem cohorts [48,49]. These mutations could be markers explaining epilepsy severity or a genetically mediated liability to fatal

seizures. In certain epilepsy syndromes, SUDEP risk seems particularlyhigh.

ThemostrecognizedexampleistheDravetsyndrome(DS),a severeepilepsysyndromewithhighprematuremortality,caused by SCN1A mutation [51]. In mutant SCN1A knock-out mice, postictalbradycardiaandseizure-triggeredventricularfibrillation wererecordedbefore a deathresemblingSUDEP [52,53].In DS subjects,markersassociatedwiththeriskofsuddencardiacdeath (decreasedHRVand increasedQT-dispersion)havebeenfound [54,55].Ictalproofis, however,lackingand isthesubjectof an ongoingstudy(ClinicalTrials.govIdentifier:NCT02415686).

Otherless wellstudiedexamplesof‘epilepsygenes’possibly mediating SUDEP risk include KCNA1 and SCN8A. KCNA1 is expressed in the vagal nerve as well as in the brain, and is associated withseizures,cardiacarrhythmias,vagalhyperexcit- abilityandprematuredeathinKCNA1nullmice[56].Mutationsin thisgenewerefoundinaSUDEPcasewithepilepticencephalopa- thyandsuspectedcardiacarrhythmias[57].

A novel pathogenicSCN8A mutationwas identified through whole-genome sequencing in a family affected by epileptic encephalopathyand SUDEP[58].Beforethen,SCN8Amutations hadonlybeenlinkedtoepilepsyinmice[59].TheSCN8Agene encodesasodiumchannelthatisexpressedinheartandbrainof miceandrats,andplaysaroleinexcitation-contractioncoupling, actionpotentialpropagationandpacemaking[60,61].

Wepreviouslydiscussedhowseizuresmaycausearrhythmias.

Whether the converse phenomenon exists is a subject of controversy. The major complication is the fact that syncopal eventsareeasilymistakenforepilepsy.Ratesofmisdiagnosisin epilepsy are high (upto 71%), and syncope is the commonest imitator[62].Thisis understandable, asvarioussymptomsand signs are seen in both conditions [11,62–64]. Notably, jerking movements or signs indicative of cerebral standstill (complete flatteningoftheEEG)suchasrovingeyemovementsorstertorous Table1

Reported(post)ictalcardiacarrhythmias.FDS—focaldyscognitiveseizure;FAS—focalautonomicseizure;fbCS—focalseizureevolvingtobilateralconvulsiveseizure;GTCS—

generalisedtonicclonicseizure;LT—lefttemporal;RT—righttemporal;BT—bitemporal;Gen—generalised;Nonloc—non-localising;PGES—postictal generalizedEEG suppression;*inpeoplewithrefractoryfocalepilepsyadmittedforavEEGrecording.FormoredetailsseevanderLendeetal.[8].

Seizurerelatedarrhythmia Reportedinn cases

Associatedseizure types

Reportedinn cases

EEGseizure onset

Reportedinn cases

SUDEPassociation

Ictalasystole 103 99%FDS 97 46%LT 80 Unlikely

1%FAS 31%RT

13%BT 10%other

Postictalasystole 13 85%fbCS 13 20%LT 10 Verylikely,accompaniedorprecededby

PGES/apnea

15%FDS 60%RT

20%other

Ictalbradycardia 25 100%FDS 8 52%LT 21 Unlikely

38%RT 10%other

IctalAVblock 11 90%FDS 10 73%LT 11 Unlikely

10%FAS 18%BT

10%other

PostictalAVblock 2 100%fbCS 2 100%RT 1 Unlikely

Atrialfibrillation 13 46%GTCS 13 33%LT 3 Unlikely

46%fbCS 33%Gen

8%FDS 33%Nonloc

(Post)ictalventricular fibrillation

3 100%GTCS 3 Insufficient

data

0 Probable,butinaminorityofcases

breathing[11]areofteninterpretedassignsspecifictoepilepsy.

Thetruecauseofthesesymptomscanonlybedeterminedwith helpofadetailedhistory(takingintoaccountthecircumstances and otherdiagnostic clues) or a proper investigation (e.g. ictal recordingofvideo,heartrate,bloodpressureandEEG)[65].Two largescalesurveysofupto2000tilt-tabletestsfailedtoidentify anyadultcasewithsyncopal-inducedseizures[66,67].Inchildren, however,afewcaseshavebeenreportedwithacardioinhibitory reflexsyncopefollowedbyvideo-EEGdocumentedclonicseizures [68–70].Thereasonwhythisphenomenononlyappearstoaffect childrenisunknown.Itmaybethattheseizurethresholdislower in children (paralleling febrile seizures that also peak in childhood). Alternatively, the depth of cerebral anoxia may be more profound in children as reflected by prolonged asystolic spells. For clinical management it is important to stress that syncope-inducedseizuresareextremelyrareand probablyonly affectchildren.Thediagnosisrequiresanictalvideo-EEGrecord- ing.

SeveralAEDs,particularlythosewithsodiumblockingproper- tiesareknowntotriggerconductionabnormalitiesorarrhythmias [71]. Atrioventricular (AV) conduction is the most frequent reportedcomplication.ST changes,Brugada-like patterns,atrial fibrillationandQTcprolongationhavealsobeenreportedbutthe associationwithAEDtreatmentislesswellestablished[72–86]. MostclinicallyrelevantarrhythmiaswererelatedtoAEDoverdose.

Carbamazepineis,however,knowntocauseAVconductionblocks atlowlevels;thisisalmostexclusivelyreportedinelderlywomen [77,79,87].Rapidadministrationofphenytoinmayalsocausesinus arrestandhypotension;elderlypeopleandthosewithpre-existing heartdiseaseseemmostvulnerabletotheseadverse effects.IV administration should, therefore, be undertaken slowly, with

continuous cardiac monitoring [76,83,86,88]. The above-men- tionedAEDeffectsdonotseemtoplayaroleinictalarrhythmias.

Nevertheless,itisimportanttotaketheseeffectsintoconsider- ation intheselectionofanAED andtomonitoradverseeffects closelyespeciallyinelderlypeopleandthosewithcardiovascular comorbidities.

3.Epilepsyandstructuralcardiacconditions

Epidemiologicalstudieshaveconsistentlyshownthatpeople with epilepsy have a higher prevalence of structural cardiac diseasethan thosewithoutepilepsy[4,5,89–92]. Cardiovascular disease seems tobe a significant contributor to the increased mortality in people with epilepsy, compared with the general population[93–95].

Sharedcardiovascularriskfactorscanaccountfortherelation- ship betweenepilepsyand heart disease,in additiontoshared geneticsandetiologicalfactors.Peoplewithahistoryofepilepsy are more likely to be obese, physically inactive, and current smokers [90] and have a worse cardiovascular risk profile (i.e.

hypertension,hypercholesterolemia,diabetesmellitus,stroke/TIA) thanthegeneralpopulation[35,90,96,97].Unsurprisingly,people withepilepsyhavehigherratesoffatalandnonfatalcardio-and cerebrovasculardiseasethancontrols (mortalityratiosupto5.3 and morbidity ratio up to 7) [35,98,99]. The presence of cardiovasculardisease(e.g. congestiveheart failureand cardiac arrhythmias) was also associated with higher mortalityrisk in peoplewithepilepsy[100].

Epilepsytreatmentcanalsocontributetoapoorercardiovas- cularrisk profilein epilepsy.Useof theenzyme-inducingAEDs phenytoin or carbamazepine may lead to elevated serological Table2

Putativemechanismsofassociationsbetweenepilepsyandcardiacarrhythmias.HRV—heartratevariability;VT—ventriculartachycardia;VF—ventricularfibrillation;AED—

antiepilepticdrugs.

Putativemechanismsofassociationsbetweenepilepsyandcardiacarrhythmias

Mechanismsofassociation Conditions

Directcausal Arrhythmias!seizures

Sharedriskfactor Genetics!epilepsyandarrhythmias

-Important‘heartgenes’:KCNQ1,KCNH2,SCN5A,RYR2

-Important‘epilepsygenes’:SCN1A,KCNA1,SCN8A

Resultant AED!arrhythmias

-Particularlycarbamazepine,phenytoinandlacosamide

Seizures!arrhythmias

-Ictal:tachycardia,asystole,bradycardiaandAVblock

-Postictal:asystole,AVblock,atrialflutterorfibrillationandventricularfibrillation

vascularriskmarkers(e.g.totalcholesterol,LDL,homocysteine), and,thus,resultinacceleratedatherosclerosis[101–104].Certain AEDs(e.g.valproicacid,carbamazepine)arealsoknowntocause weightgainandincreasetheriskofdevelopingnon-alcoholicfatty liverdiseaseandmetabolicsyndrome,leadingtofurtherdeterio- rationofthecardiovascularriskprofile[102].

Theco-occurrenceofepilepsyand(congenital)heartdisease, oftenaccompanied byintellectualdisability, may resultfrom a multiplemalformationsyndrome:geneticdefectsmayaffectthe developmentofbothheartandbrain,orabnormalcardiovascular functionmayleadtopoor(intrauterine)braingrowth[105].

CVdiseasecansometimes(indirectly)causeepilepsythrougha predispositiontostroke[106,107].Strokeisacommonriskfactor forepilepsyandaccountsfor abouta thirdof newlydiagnosed seizuresinpeopleovertheageof60years[107–110].Inparticular, those withischemiceventswith corticalinvolvement, cerebral hemorrhage(i.e. primary hemorrhageor hemorrhagic transfor- mationofischemicstroke)andearlypost-strokeseizures,havean increasedriskofpost-strokeepilepsy[107].

Seizureactivitymaynotonlyinducearrhythmiasbutmayalso leadtostructuralcardiacchanges[71,111–113].Epilepticseizures havebeenreportedtoprovokecardiacischaemiaviabothacute andchroniceffectsontheheart(e.g.impairedheartratevariability, cardiacfibrosis,ST-segmentdepressionandincreasedheartrate) [71,114]. Transient myocardial ischaemia as indicated by ST- segmentdepression,was reportedinasmall-scalestudyin40%

of all 15 seizures [114]. Another study, however, failed to demonstratetroponinincreases,suggestingthatthereportedST changesdonotusuallycausemyocardialdamage[115].

SeizuresarethesecondmostfrequentCNSconditionknownto inducethecardiomyopathyknownasTakotsubosyndrome(TTS) [116].TTS mimicsmyocardialinfarctionclinically,electrocardio- graphicallyandchemically[117].Itischaracterizedbyacuteonset of chest pain and dyspnoea, sometimes concomitant with

palpitations, tiredness, oedema, fever, syncope, anxiety, nausea orvomiting[116].TheseizuretypethatmostfrequentlycausesTTS is the generalized tonic-clonic seizure [118,119]. Seizures most likelytriggerTTSbythestress-inducedreleaseofcatecholamines [120].Thisabundantcatecholaminereleasemaybeacontributing factorinfatalstatusepilepticus[121].ArelationshipbetweenTTS andSUDEP,however,doesnotappearlikely[116].

4.Futureconcepts

Significant progress hasbeen madesince the publicationof Russel’s case history: the complex interrelationship between epilepsyandcardiacconditionshasbeenexploredwidelyandthis reviewaimedtocaptureallmajordiscoveriesmadeinthisfield (Table2and3).Manydiscoveriesofcoexistingconditionswere madebyserendipity, andunderlying mechanismsareyettobe uncovered.Treatmentregimensareconsequentlyoftenspecula- tive and lack a personalized approach involving all comorbid conditions. As comorbidity gains recognition we now need to become better at noticing these symptom patterns. Today a substantialgapstillremainsbetweenthespecialties,butasweare nowbecomingawareofalloverlappingsyndromesepileptologists will increasingly need to improve their cardiac skills. Pattern recognitioncanbefosteredbyincorporatingvalidatedscreening instruments and guidelines, aiding the early identification and treatment of cardiovascular comorbidity in epilepsy. Concomi- tantly,afundamentalchangeinthewaycliniciansthinkofepilepsy iscrucial.

Epilepsy will soon be viewed as a collection of individual disorders that share a phenotypeof an abnormal tendencyfor unprovoked epileptic seizures. The number of rare epilepsy syndromes withcardiac phenotypeswill increase substantially.

Epilepsywillbeseenasasymptom-complex,andallcomorbid- ities,eventhemostinconspicuous,shouldbeconsideredaspartof

Table3

Putativemechanismsofassociationsbetweenepilepsyandstructuralcardiacdisease.AED—antiepilepticdrugs;TTS—Takotsubosyndrome.

Putativemechanismsofassociationsbetweenepilepsyandstructuralcardiacdisease

Mechanismsofassociation Conditions

Indirectcausal Cardiaccondition!stroke!epilepsy

Sharedriskfactor Genetic!malformationofcorticalandcardiacdevelopment!epilepsyandcardiovascular comorbidity

Increasedprevalenceofcardiovascularriskfactorsinepilepsy!stroke/cardiacdisease

Resultant AED!poorercardiovascularriskprofile(e.g.arteriosclerosis,weightgain,nonalcoholicfatty liverdiseaseandmetabolicsyndrome)

Seizures!transientmyocardialischaemiaandseizure-triggeredTakotsubosyndrome(TTS)

the stratification and phenotyping in people with epilepsy.

Cardiovascular comorbidities willprovide insight into common mechanismsforepilepsyandgiveawindowintocommongenetic predispositions.Theymayalsoprovideimportantdiagnosticclues.

Channelopathies,forexample,areincreasinglyidentifiedinpeople withepilepsy.Geneticfactorsmayexplainboththeepilepsyand thecomorbiddisorder(s),eveninpeoplewithsporadicepilepsies [122].Genomewidescanningwillbewidelyavailableanddrive theparadigmshiftinepilepsy.Certaingenesmightbeidentifiedas contributingtoSUDEP[48,49],potentiallyallowingthedevelop- mentofindividualisedriskpreventionstrategies.Anothermajor contributortoearlyidentificationofoverlappingsyndromeswill bethe development of new non-invasivetools torecord heart functionathome.Theminiaturisationofsensorswillfavourlong- termhome-basedrecordingsthusaidingtheearlyidentificationof cardiacarrhythmias.

Advancesinseizuredetectionwilllikelytakeoff.ECGalonewill help to detect a wide variety of seizures but lacks specificity.

CombiningECGwithothermodalitiesincludinganaccelerometry and electrodermal activity will likely improve accuracy and facilitatethewidespreaduseofseizuredetectiondevicesinthose withrefractoryepilepsy[123,124].

Anotherunmetneedrelatestothetreatmentofepilepsy:many AEDs have proarrhythmogenic and arteriosclerogenic effects.

Thoughnon-pharmacological optionsexist,drugtherapyis still themainstayofepilepsytreatmentandotheroptionsareusually only explored after AEDs have failed to successfully control seizures[125].ManynewAEDshavebeenlaunchedinthelasttwo decades,buthavefailedtoimprovetheburdenofsideeffectsor substantiallychangeprognosisforseizurecontrol[126,127].With improvedunderstandingofepileptogenesis,epigeneticdetermi- nantsandpharmacogenomicscomesthehopeforbetter,disease- modifyingorevencurativepharmacologicalandnon-pharmaco- logical treatment strategies. Until then, comorbidity should be consideredwhenprescribingAEDs.

The incorporation of neurocardiology into the paroxysmal spectrumwillrequireacriticalreviewoftheepilepsyservices.We need to validate new instruments to screen for cardiovascular conditions.Modernnon-invasivelong-termECGdevicesmayhelp screenforcardiacconditionsandacardiologistshouldreviewany relevantabnormalities.Incaseswherethereisarelevantfamily historyorabnormalECGfindings,aspecialistcardiacassessment should be done. Identification and adequate treatment of cardiovascular disorders in epilepsy should therefore be an importantpartofepilepsymanagement.

Particularattentionshouldbegiventomodifiableriskfactors suchassmoking,obesity,sedentarylifestyle,highcholesteroland hypertension.Physicians shouldscreen for theserisk factorsin peoplewithepilepsy,providegeneral healthinformation and if necessaryadjust AED treatment. Further studiesare needed to improverisk profiling, thus allowingfor screening in highrisk individuals(with, forexample,implantable looprecorders)and targetedinterventions(e.g.defibrillators).

Conflictofintereststatement

SS,MvdL and RJLreport noconflict of interest.JWS reports personalfeesfromLundbeckandTeva,grantsand personalfees from UCB, Eisai, grants from GSK, WHO and Dutch National EpilepsyFund,outsidethesubmittedwork;hiscurrentpositionis endowedbytheEpilepsySociety,heisamemberoftheEditorial BoardoftheLancetNeurology,andreceivesresearchsupportfrom theMarvinWeil EpilepsyResearchFund.RDTreceives research supportfromtheDutchNationalEpilepsyFund,NUTSOhraFund, Medtronic,andACThomsonFoundation,andhasreceivedfeesfor lecturesfromMedtronic,UCBandGSK.

Funding

ThisworkwassupportedbytheDutchNationalEpilepsyFund [project number 15-10]; and Christelijke Vereniging voor de VerplegingvanLijdersaanEpilepsie(theNetherlands).

Acknowledgements

Thiswork was partlydoneat theUCLH/UCLComprehensive Bio-Medical Research Centre which received a proportion of funding from the Department of Health’s NIHR Biomedical ResearchCentresfundingscheme.WearegratefultoDr.GSBell, forcriticallyreviewingthemanuscript.

References

[1]RussellAE.Cessationofthepulseduringtheonsetofepilepticfits,with remarksonthemechanismoffits.Lancet1906;168:152–4.

[2]ForsgrenL.PrevalenceofepilepsyinadultsinnorthernSweden.Epilepsia 1992;33(3):450–8.

[3]GaitatzisA,SisodiyaSM,SanderJW.Thesomaticcomorbidityofepilepsy:a weightybutoftenunrecognizedburden.Epilepsia2012;53(8):1282–93.

[4]KadimaNKR,ZackM,HelmersS.Comorbidity inadultswith epilepsy—

UnitedStates,2010.MorbMortalWklyRep2013;62(43):849–53.

[5]KeezerMR,SisodiyaSM, SanderJW. Comorbiditiesofepilepsy:current conceptsandfutureperspectives.LancetNeurol2016;15(1):106–15.

[6]SevcencuC,StruijkJJ.Autonomicalterationsandcardiacchangesinepilepsy.

Epilepsia2010;51(5):725–37.

[7]Eggleston KS, Olin BD, Fisher RS. Ictal tachycardia: the head-heart connection.Seizure2014;23(7):496–505.

[8]vanderLendeM,SurgesR,SanderJW,ThijsRD.Cardiacarrhythmiasduring orafterepilepticseizures.JNeurolNeurosurgPsychiatry2016;87(1):69–74.

[9]SchueleSU,BermeoAC,AlexopoulosAV,LocatelliER,BurgessRC,DinnerDS, etal.Video-electrographicandclinicalfeaturesinpatientswithictalasystole.

Neurology2007;69(5):434–41.

[10]TinuperP,BisulliF,CerulloA,CarcangiuR,MariniC,PierangeliG,etal.Ictal bradycardiainpartialepilepticseizures:autonomicinvestigationinthree casesandliteraturereview.Brain2001;124:2361–71.

[11]vanDijkJG,ThijsRD,vanZwetE,TannemaatMR,vanNiekerkJ,BendittDG, et al. The semiology of tilt-induced reflex syncope in relation to electroencephalographicchanges.Brain2014;137:576–85.

[12]Lanz M,OehlB, BrandtA, Schulze-BonhageA.Seizureinducedcardiac asystoleinepilepsypatientsundergoinglongtermvideo-EEGmonitoring.

Seizure2011;20(2):167–72.

[13]ChailaEBJ,TirupathiS,DelantyN.Ictalbradycardiaandasystoleassociated withintractableepilepsy:acaseseries.BrJCardiol2010;17:245–8.

[14]Leung H, Kwan P, Elger CE. Finding the missing link between ictal bradyarrhythmia,ictalasystole,andsuddenunexpecteddeathinepilepsy.

EpilepsyBehav2006;9(1):19–30.

[15]AltenmullerDM,ZehenderM,Schulze-BonhageA.High-gradeatrioventric- ular block triggered by spontaneous and stimulation-induced epileptic activityinthelefttemporallobe.Epilepsia2004;45(12):1640–4.

[16]OppenheimerSM,GelbA,GirvinJP,HachinskiVC.Cardiovasculareffectsof humaninsularcortexstimulation.Neurology1992;42(9):1727–32.

[17]PoolJL,RansohoffJ.Autonomiceffectsonstimulatingrostralportionof cingulategyriinman.JNeurophysiol1949;12(6):385–92.

[18]SimonRP,AminoffMJ,BenowitzNL.Changesinplasmacatecholaminesafter tonic-clonicseizures.Neurology1984;34(2):255–7.

[19]NilssonD,SuttonR,MelanderO,FedorowskiA.Spontaneousvsnitroglycer- in-induced vasovagal reflexon head-up tilt: are there neuroendocrine differences?HeartRhythm2016;13(8):1674–8.

[20]MoseleyBD,GhearingGR,MungerTM,BrittonJW.Thetreatmentofictal asystolewithcardiacpacing.Epilepsia2011;52(4):e16–9.

[21]BestawrosM,DarbarD,ArainA,Abou-KhalilB,PlummerD,DupontWD,etal.

Ictalasystoleandictalsyncope: insightsintoclinicalmanagement.Circ ArrhythmElectrophysiol2015;8(1):159–64.

[22]KohnoR,AbeH,AkamatsuN,BendittDG.Long-term follow-upofictal asystole in temporal lobe epilepsy: is permanent pacemaker therapy needed?JCardiovascElectrophysiol2016;27(8):930–6.

[23]StrzelczykA,CenusaM,BauerS,HamerHM,MothersillIW,GrunwaldT,etal.

Management and long-term outcome inpatients presenting with ictal asystoleorbradycardia.Epilepsia2011;52(6):1160–7.

[24]Duplyakov D, GolovinaG, Lyukshina N,Surkova E,Elger CE,Surges R.

Syncope,seizure-inducedbradycardiaandasystole:twocasesandreviewof clinicalandpathophysiologicalfeatures.Seizure2014;23(7):506–11.

[25]RyvlinP,NashefL,LhatooSD,BatemanLM,BirdJ,BleaselA,etal.Incidence andmechanismsofcardiorespiratoryarrestsinepilepsymonitoringunits (MORTEMUS):aretrospectivestudy.LancetNeurol2013;12(10):966–77.

[26]MasseyCA,SowersLP,DlouhyBJ,RichersonGB.Mechanismsofsudden unexpecteddeathinepilepsy:thepathwaytoprevention.NatRevNeurol 2014;10(5):271–82.