The effect of vagus nerve stimulation on cardiorespiratory parameters during rest and exercise

The

effect

of

vagus

nerve

stimulation

on

cardiorespiratory

parameters

during

rest

and

exercise

Daphne

M.

Mulders

a

,

Cecile

C.

de

Vos

a,b,c,

*

,

Ilse

Vosman

d

,

Michel

J.A.M.

van

Putten

a,b a_{MIRA,InstituteforBiomedicalTechnologyandTechnicalMedicine,UniversityofTwente,TheNetherlands}

b

DepartmentofClinicalNeurophysiology,MedischSpectrumTwente,Enschede,TheNetherlands

c

DepartmentofNeurosurgery,MedischSpectrumTwente,Enschede,TheNetherlands

d

SEINEpilepsieInstellingen,Zwolle,TheNetherlands

1. Introduction

Vagusnervestimulation(VNS)wasfirstintroducedin1988,and isnowanacceptedtreatmentoptionforpatientswithrefractory epilepsy,whoarenotsuitablecandidatesforresectivebrainsurgery or for whom surgery has failed. VNS consists of intermittent electricalstimulationoftheleftcervicalvagusnerve,providedbyan implantedpulsegeneratorandastimulationelectrode.Themain purposeofVNSistoreduceseizurefrequency,whichwouldideally resultinatotalterminationofseizureoccurrence.Longtermstudies thatassessedtheefficacyofVNS,concludedthatover50%seizure

reduction was accomplished in 20–63% of the patients after treatmentfor6monthsto6years[1–3].

Besideseizurefrequencyreductionandotherpositiveeffectsof VNS[4],severalsideeffectshavebeenreported.Themostcommonly reportedside-effectsarehoarseness,coughinganddyspnea,withan incidence reported ashighas66%of patients [5]. Mostlaryngeal symptomshave beencontributedtovocalcord adduction,caused by stimulation of the laryngeal branchesof the vagus nerve [6]. RespiratorycomplicationsbyVNSduringwakearerarelyreported, whilealterationsofrespiratory patternsduringsleep appeartobe morecommoninbothadults[7,8]andchildren[9,10].Duringsleep, thesepatients showanincreased breathing frequency(BF) anda decreasedtidal volume (TV), asa directeffect of VNS.The exact mechanismofthiseffectisnotknown,butseveralmechanismshave beenproposed,suchasperipheraleffectsonupperairway muscula-ture innervated bythe vagus nerveor more central mechanisms influencingupperairwaypatencyandrespiratoryeffort[7].

ARTICLE INFO Articlehistory: Received17July2015

Receivedinrevisedform28September2015 Accepted8October2015

Keywords:

Vagusnervestimulation(VNS) Sideeffects

Bradycardia,Tidalvolume,Breathing frequency

ABSTRACT

Purpose:Vagusnervestimulation(VNS)hasbeensuccessfullyappliedtoreduceseizurefrequencyin numerouspatientswithepilepsy.However,varioussideeffects,includingdyspneaandbradycardiahave beenreported,thatappearexerciserelatedinsomepatients.Thispilotstudyaimstoobtaininsightinthe cardiorespiratoryeffectsofVNSduringbothrestandexercise.

Methods:PatientswithaVNSdevicewhoexperiencesideeffectsduringexercisearecomparedwith patientswithoutsideeffects.Respiratoryandcardiacparametersmeasuredduringrestandexercise includeheartrate,breathingfrequencyandtidalvolume.

Results:Sixty-twoepisodesofVNSinfivepatientswithandfivepatientswithoutsideeffectswere recorded.Inaddition,fivecontrolsubjectshavebeenmeasured.Duringrest,allsubjectsshowedstable valuesforthecardiorespiratoryparameters.Duringthefirstminutesofexercise,heartrate,breathing frequencyandtidalvolumeincreased.Thereafter,asteadystatewasreachedagainforallsubjects. DuringVNSepisodes,eightoutof10patientsshowedasmallbutconsistentdecreaseinheartrate,along withanincreaseinbreathingfrequencyineightoutofninepatients.Tidalvolumesdecreasedduring VNSepisodes.Theseeffects,inducedbyVNS,occurredduringbothrestandexercise.Magnitudeofthese effectsvariedbetweenpatients,butwasnotnecessarilyrelatedtotheintensityoftheexperiencedside effects.

Conclusion:ThispilotstudyshowsthatVNScausesanincreaseinbreathingfrequencyandadecreasein tidalvolumeandheartrateinthemajorityofpatients,duringbothrestandexercise.

ß2015BritishEpilepsyAssociation.PublishedbyElsevierLtd.Allrightsreserved.

* Correspondingauthorat: Department ofNeurosurgery,MedischSpectrum Twente,PObox50.000,7500KAEnschede,TheNetherlands.Tel.:+31534872840.

E-mailaddress:c.devos@mst.nl(C.C.deVos).

ContentslistsavailableatScienceDirect

Seizure

j o urn a l hom e pa g e : ww w . e l se v i e r. c om / l oca t e / y se i z

http://dx.doi.org/10.1016/j.seizure.2015.10.004

Cardiac symptoms, such as bradycardia and even asystole havealsobeenreported,duringtheintra-operativedevicetesting

[11],buttherearealsoreportsthatmentionbradyarrhythmias duringregulartreatment,evenyearsafterimplantation[12,13]. Furthermore, changes in heart rhythm variability have been observed, although the direction of the variability changes showed inter-individual differences [13,14]. Cardiac effects of VNShavebeen attributedtoanineffective anodalblockofthe lowerVNSelectrode, causingefferentstimulationofthevagus nerve, which may lead to inhibitionof AVconduction tissue, causingvariousdegreesofAVblock,ormayslowSAnodalfiring, orevenstopit.

Inaddition,severalpatientsexperiencehardlyanysideeffects fromVNSduringrest,butmentionadecreaseinphysicalcondition only during exercise, oftenreported as breathlessnessat those moments the stimulator is active [4,8]. The cause of this phenomenon remains unclear; therefore, the aim of this pilot studyistoevaluatetheeffectofVNSoncardiacandrespiratory functions during rest and exercise, by recording ECG, pulse oximetry and respiratory parameters (e.g. breathing frequency andtidalvolume).

2. Methods 2.1. Subjects

PatientswithaVNSwererecruitedviatheepilepsycenterSEIN andtheneurologydepartmentofMedischSpectrumTwente,The Netherlands.We includedfive epilepsypatients withVNSwho reported shortnessof breath and intensified side effects of the stimulation during exercise to their treating physician, five epilepsy patients with VNS without reported side effects specificallyduringexercise,andfivehealthycontrolsubjects,all of whom had no cardiorespiratory diseases, diabetes or other disorders affecting the autonomic nerve system. The study conformedtotheDeclarationof Helsinkiand wasapprovedby theinstitutionalreviewboardTwente.Allsubjectsgavewritten informedconsent.

2.2. Measurements

MeasurementswereperformedattheExperimentalCenterfor TechnicalMedicineattheUniversityofTwente.Beforethestartof thetest,patientswereaskedtofillina questionnaireregarding theirepilepsy,theVNSinducedsideeffectstheyexperiencedand their exercise habits. The settings of theVNS wereread via a handhelddevice.Apulseoximeterwasplacedontherightmiddle finger(Avant9600,NoninMedicalInc.,MN,USA). Cardiorespira-tory parameters were acquired synchronously. ECG recordings were performed with a Viasys ECG monitor and respiratory parameterswereobtainedusingaJaegerOxyconPro(amaskwith a flow sensor)(both CareFusion,CA,USA). Twobipolarsurface electrodeswereplacedovertheVNSelectrodeontheleftsideof thenecktorecordepisodesof stimulationand on thechest to record ECG to facilitate heart rate analysis (sample frequency: 2048Hz).

Forthefirststageofthemeasurement,thesubjectswereseated inachairfor20min,readingamagazine.Duringthesecondstage of the measurement, the subjects exercised for 20min on the Ergoselect100 cycleergometer(ergolineGmbH,Germany),ata constant level that could be maintained comfortably for the durationofthetest.ThesubjectswithVNShadtheirstandardVNS settingsprogrammedandhadatleastthreeepochsofstimulation duringboththerestandtheexercisestage,sothatintotalatleast 60VNSepisodescouldberecorded.

2.3. Analysis

ForeverysubjectwithVNS,epochsof50sjustprior,duringand directly after stimulation were selected, during both rest and exercise. Respiratory parameters of interest were tidal volume (TV), breathing frequency(BF)and saturation, andthesevalues wereaveragedforeachepoch.ObtainedECGdatawasfilteredwith abandpassButterworthfilterfrom0.5to45Hz(MATLAB,2011a, theMathWorks,Inc.).PeakdetectionwasdoneusingtheMATLAB functionfindpks.Detectedpeakswerecheckedvisuallyandmissed peaks were added or erroneouspeaks were deleted manually. Heartrate,asmeasuredbybeatsperminute,wascalculatedand valueswereaveragedforeachepoch.FromthesameECGsignal, heartratevariability(HRV)wasdeterminedinthetimedomain,as thestandarddeviationoftheheartrateinoneepoch.

First,thevaluesofpatientswithVNSwerecomparedtohealthy subjects, to analyze whether the parameters were within the normalrangeduringrestandexercise.Second,inordertoavoid differencesinweighting,averageparameterswerecalculatedfor eachindividualpatientbeforeaveraging.Third,theaverageeffect ofVNSon thecardiorespiratoryparameters forallpatientswas analyzed, disregarding subgroup or phase of the test (rest or exercise),tostudytheglobaleffectofVNS.Thenthespecifictest phasewastakenintoaccount,toinvestigatewhethertherewere differencesbetween theeffectof VNSduring restandexercise. Last,thesubgroupswereseparated,toinvestigatewhetherthere weredifferencesbetweenpatientswhodidanddidnotreportthe side effects during exercise.During both restand exercise, the average valueoftheparameters oftheepochsduringVNSwas comparedwiththeaveragevalueoftheparametersduringeither theepochpreorpostVNS.

Eventhoughwehaveadecent numberofVNSepisodes,the numberofsubjectsinthispilotstudywassmall.Thereforewedid not test for statistical significance, but identified whether cardiorespiratoryresponsestoVNSwereconsistentand reproduc-ibleinpatientswithandwithoutsideeffects,andduringrestand exercise.

3. Results

Tenpatients witha vagusnerve stimulatorandfive healthy subjectswereincluded.The patientswitha VNSwereall good responderstoVNStherapy,andweredividedintwogroups:group 1consistedofpatientswhoreportedsideeffectsduringexercise and group2 consistedofpatients whoreported no side effects duringexercise.Averageageofallsubjects(epilepsypatientsand healthy subjects) was 44 (20–78) years and for the epilepsy patientsthemeantimesinceVNSimplantationwas45(8–132) months.

AnoverviewofsubjectcharacteristicscanbefoundinTable1. Subject 1 (group 1) was measured without the mask, so no respiratory parameters are available. Subject 6 (group 2) terminatedtheexerciseprematurely,soonlyoneVNSepisode during exercise isavailable for analysis.Furthermore, epochs that contained large artifacts hindering proper analysis were excluded.

3.1. Globaleffectofexercise

Analysisinbothhealthysubjectsandpatientsshowsconstant valuesfortidalvolume(TV),breathingfrequency(BF),heartrate (HR)andsaturationduringthe20minrestperiod.Whenexercise isinitiated,TV,BFandHRvaluesincrease,untiltheyreachanew stablevalue,whichismaintainedfortheremainderofthe20-min exercise. All values are withinnormal range and as expected duringlowimpactexercise,healthysubjectsandpatientsalike.

ThefirstVNSepisodeduringexercisewasinitiated5minafterthe start of the exercise phase, when the new plateau had been reached.

3.2. Respiratoryparameters

Eight outof 10 patients reported hoarsenessduring VNS,of whichthefivepatientsingroup1reportedaformof breathless-nessduringVNS.Norespiratorydataisavailableforsubject1,so respiratoryparametersareanalyzedfor57VNSepisodesinnine patients.Ingeneral,VNScausesasmallbutreproducibledecrease intidalvolumeandanincreaseinbreathingfrequency(Table2). Regarding the saturation, no clinically significant effects of stimulationorexerciseareobserved.

Duringrest,eightoutofninepatientsshowanincreaseinBF duringVNS.While exercising,six outof ninepatients showan increaseinBFduringVNS: twopatients withand fourpatients withoutreportedsideeffects.Thethreepatientswhodonotshow theincreaseinBFduringVNSwhenexercising,showadecreasein BF,andtwoofthembelongtothegroupwiththereportedside effects(group1).

In both subgroups and both stages, average TV decreases duringVNS.Inrest,eightoutofninepatientsshowadecreasein TV during VNS, whereas seven out of nine patients show thisduringexercise.One patientwithoutreportedsideeffects (group2)showsinbothstagesasmallincreaseinTVduringVNS (1%and3%).

3.3. Cardiacparameters

Withregardtotheeffects ofVNSoncardiacparameters,the heartrate(HR)hasbeenanalyzed.The50sepochsbefore,during andafterVNSthatwereusedforcardiacanalysiswerefirstvisually inspected,andepochscontainingtoomuchnoiseorartifacts(e.g. duetomovement)tobeanalyzedwerediscarded.

For62VNSepisodes,theHRwascalculated.Overall,patients showedaconsistentdecreaseinHRduringVNS,comparedwith epochsprior andafterVNS (Fig.1).During rest,nineout of10 patientsshowasmalldecreaseinHRduringVNS,ascomparedto theepochpriortoVNS.ComparedtotheepochafterVNS,eightout of10patientshaveasubsequentincreaseinHRafterVNS.During exercise,sixoutof10patientshaveadecreasedHRduringVNS, comparedtotheepochbeforeVNS.Thesesixpatientsalsohada decreaseinHRduringrest,andtheyareevenlydistributedoverthe twosubgroups.ComparedtotheepochafterVNS,sevenoutof10 patientshaveasubsequentincreaseinHRafterVNS.

Additionally,tocheckforconsistencyinHRinbetweentwoVNS episodes,HRvaluesofVNSepochshavealsobeencomparedwith epochsnotdirectlyprecedingorsucceedingthestimulationepoch. Comparisonbetweenthestimulationepochsandepochsrandomly in between two subsequent stimulation epochs also shows decreasedHRduringVNSinallpatients.

Theanalyzedepochsof50saretooshorttoreliablyassessthe heart ratevariability withinan epoch, sowe cannot reporton whetherHRVisconsistentlyaffectedbyVNS.

Table1

Subjectcharacteristicsandstimulatorsettings.EpilepsyisyearsofepilepsyandVNSrepresentsmonthssinceimplantation.VNSsettingsduringthemeasurement:output current/frequency/pulseduration/on-time/off-time.

Subject Sex Age [y] Epilepsy [y] VNS[months] SettingsofVNS: [mA/Hz/ms/s/min] Stimulationepisodes Rest Exercise Group1:withside-effects

1 F 34 23 8 2.00/30/500/60/3 3 2

2 F 41 27 120 2.25/30/250/30/3 6 5

3 F 63 61 9 1.00/20/500/60/5 3 3

4 F 21 21 24 1.75/30/500/60/5 3 2

5 M 67 23 24 1.75/20/250/60/5 3 3

Group2:withoutside-effects

6 F 78 72 29 2.00/30/250/60/5 3 1 7 F 52 26 132 1.25/20/130/30/3 4 5 8 M 52 25 10 2.25/30/500/60/5 3 3 9 M 20 8 78 1.75/30/500/60/5 2 3 10 M 44 30 11 2.00/20/500/60/3 3 2 Group3:controls 11 F 27 n/a 12 M 24 13 M 27 14 F 52 15 F 72 Table2

Overviewofaveragerespiratoryvalues(standarddeviation)foreachsubgroup.Pre,duringandpost:before,duringandafterstimulation,respectively.Group1:withside effects,Group2:withoutsideeffects,Group3:healthycontrolsubjects.SpO2:oxygensaturation,BF:breathingfrequency,TV:tidalvolume.

SpO2[%](SD) BF[/min](SD) TV[l](SD)

Pre During Post Pre During Post Pre During Post Rest Group1 97.3(1.2) 97.1(1.4) 97.4(0.9) 17.7(4.0) 19.2(3.0) 18.7(3.6) 0.52(0.25) 0.49(0.27) 0.48(0.21) Group2 96.8(1.0) 96.9(1.2) 96.8(1.0) 19.7(1.2) 22.4(2.9) 20.7(1.7) 0.40(0.15) 0.37(0.13) 0.40(0.14) Group3 97.5(0.6) 18.8(2.3) 0.51(0.07) Exercise Group1 96.5(1.7) 96.9(2.2) 96.9(1.9) 26.5(4.0) 25.9(4.3) 24.6(4.5) 1.19(0.48) 1.13(0.54) 1.14(0.60) Group2 97.0(0.2) 96.7(0.6) 96.6(0.6) 22.8(3.1) 24.4(3.2) 22.9(2.8) 1.16(0.34) 1.10(0.32) 1.16(0.32) Group3 97.4(1.4) 22.4(1.6) 1.06(0.19)

4. Discussion

Themainpurposeofthispilotstudywastoinvestigatetheeffect ofVNSoncardiacandrespiratoryfunctionsduringrestandexercise, and subsequently, whether there were differences between patients who reported shortness of breath during exercise and thosewho didnot.The main finding ofthis pilot studyis that individualpatientsgenerallyshowreproduciblechangesinseveral cardiorespiratory parametersduring VNSepisodes. Overall, this pilot study shows a decrease in tidal volume and increase in breathingfrequencyrelatedtoVNSinthemajorityofthepatients andaconsistentdecreaseinbreathingfrequencywhileexercisingin twopatientswithreportedbreathlessnessduringexercising.Inall butonepatientheartratedecreasesduringVNS.Althoughpatients showcardiorespiratorychangesduringexercise,thesechangesare comparablewithorevensmallerthanthechangesduringrest.No specificVNSinduceddifferencesarefoundbetweenthetwogroups, therebynotexplainingwhysomepatientswithVNSexperience breathlessnessduringexercise,whileothersdonot.

Forthispilotstudy,patientsweredividedintotwosubgroups, basedonwhetherornottheyreportedVNSinducedshortnessof breath during exercise.However, this separation proved to be challenging,since sideeffects ingeneral wereexperiencedin a gradualmanner,orwerenotperceivedassideeffects,ordidnot counterweightagainstthebenefitsofVNS.Itisthereforepossible thatapatientunderstatedthesymptomsandendedupingroup2, whileheorsheshouldhavebeeningroup1,orviceversa.

Webelievethatthepatientswhoparticipatedinthispilotstudy arerepresentativeforthe(Dutch)healthyVNSpatientpopulation andallpatientsbikeindailylife.Bylettingthembikeattheconstant speedtheywouldbikenormally,wehaveusedanexercisewhichis perceivedequallyintensiveforallsubjects.Aninterestingdifference though,isthatthepatientswhoreportednoshortnessofbreath during exercise (group 1) might have been in better physical conditionthan the patients who didreport shortness ofbreath (group2),reflectedbyonaveragealowerheartrateduringrestand whileexercising.Thereisnocardiorespiratorydataavailablefrom thesepatientspriortoVNSimplantation,soitisunknownwhether thisdifferencebetweenthetwogroupswasalreadyapparentor perhapscausedbyVNSinducedsideeffects.Thiswillbestudiedin ournewprospectivestudy.

4.1. Respiratoryparameters

Therespiratorysideeffects ofVNS,duringbothwakefulness andsleep,havebeeninvestigatedforyears.Especiallyinchildren,

therespiratorypatternisoftenaffectedduringsleep.Zaaimietal. showedincreasedBFanddecreasedTVduringVNSin10children, sometimescombinedwithslightlydecreased oxygensaturation

[10].TheseresultsareconcordantwiththoseofNagarajanetal. (2003)[18].Inadults,Banzettetal.foundnosystematicchangein TV norBFduring wake,but foundan increasedend-expiratory volumeinpatientswhoreceivedhighVNSstimulationintensity

[8]. They attributed this effect to A-fiber afferent stimulation. Thesefiberscarryafferentvisceralinformation,aswellaslaryngeal motorstimuli.Binksetal.(2001)[19]showedthatthethreshold for A-fiberactivationis ratherlow (0.25–1.5mA) intheir adult subjects. The specific A

d

-fibers carry information from the pulmonary stretch receptors, which play role in the Hering– Breuer reflex. Increased sensory activity of these pulmonary stretchlung-afferentsresultsininhibitionofthecentral inspira-tory drive and thus inhibition of inspiration and initiation of expiration.Thisreflexisthoughttohaveasmalleffectonbreathing rateanddepthduringrest,whileitmayinfluencerateanddepth whenTVexceeds1l,asduringexercising.

Thelattermightexplainwhyinthispilotstudythemajorityof patients showedan increasein BFduring VNS,whilepatient 3 (group1)andpatient6(group2)showedadecreaseinBFduring exerciseduringVNS,astheirTVwasbelow1l.However,patient2 alsoshowedadecreaseinBF,whileherTVexceeded1l.Sinceeight outof10patientsreportedhoarsenessasasideeffectoftheirVNS, itis likelythattheirstimulationparametersarehighenoughto elicit afferent A-fiber activation. That would also explain the decreaseinTVandincreaseinBF.However,thetwopatientswho didnotexplicitlyreporthoarsenessalsoshowthedecreaseinTV and increase in BF. They did not experience hoarseness as a problem.

Furthermore, althoughthereare consistentand reproducible effectsofVNSonrespiratoryparameters,itisnotcertainwhether theyareclinicallyrelevantand doubtfultheycouldleadtothe feelingofbreathlessnessduringexercise,asthereappearstobeno difference inVNSinducedrespiratory effects betweengroups 1 and2.

Inthisstudy,respiratoryvalueswereobtainedevery5sviaa mask,whichisarelativelylowsamplingratetostudyrespiratory differencesbetweenstimulationONandOFF.Amorecontinuous signalwould bepreferable, toallowbreath-for-breath analysis. Furthermore,themaskwasperceivedasuncomfortablebysome subjects, even though it did not impede respiration. It is also possiblethattheuseofthemaskmadethesubjectsmoreconscious of their breathing, resulting in more controlled breathing and thereby influencing theparameterswe studied.Alternatively,a chestband,capableofmeasuringvariablessuchastidalvolume andbreathingfrequencycouldbeused[8,10].

4.2. Cardiacparameters

Vagalinfluenceoncardiacfunctioncanbeanundesirableside effectin epilepsypatients,butisalsoexploredasa treatment option for patients with heart failure [15]. Even though stimulationisappliedtotheleftvagusnerveinepilepsypatients, in ordertodiminishthecardiac interference,(intra-operative) bradycardiasandventricularasystolehavebeenreported[1,11– 13,16].Themajorityofpatientsinthecurrentstudyshowedthat VNSresultedinasmallbutreproducibledecreaseinHR.However, noteverypatientshowedthissameresponse,asonepatientin group 1 showed a small but reproducible increase in HR. Furthermore,therearetwopatientswhoshowedbothdecreases andsomeincreasesinHRduringVNS.ApreviousstudybyFrei andOsorioalsodemonstratedtheoccurrenceofbradycardiaas well as a combination of bradycardia and tachycardia during stimulation[14].

Fig.1.Averageheartrate(HR)inbeatsperminute(bpm),pre,duringandpostVNS forbothsubgroups,duringbothrestandexercise.Group1:withsideeffects,group 2:withoutsideeffects.Errorbarsrepresentstandarddeviation.

Even though the decrease in HR we found is small, it is reproducibleinmajorityofthepatientsandclearlyrelatedtoVNS. Itishoweverprobablytoosmalltoplayanapparentroleintheside effectsexperiencedduringexercise.Nonetheless,theeffectcould beofimportance,especially asitappearstobepresentin both subgroupsduringbothrestandexercise.Thiseffectonheartrate indicatesthatmostlikelynotallefferentstimulationofthevagus nerve is blocked by the VNS anodal block: one of the VNS electrodesisdesignedtoblockefferentstimulationofthevagus nerve.Failureof thisanodalblockcouldresultina decreasein heartrate[17]andcouldalsoberesponsibleforthehoarseness manypatientsreport.

AlthoughwefoundadecreaseinTVandanincreaseinBFand HRinpatientswhoexperiencedsideeffectsduringexercise,the sameeffectsonTVandBFwerealsopresentinthegroupwhodid notreportsideeffectsduringexercise.Besides,itturnedoutthat theeffectofVNSonTVandBFwaspresentduringbothrestand exercise,anddifferenceswereprimarilygradual.Itisunlikelythat theVNS inducedchangesin TV,BFand HRduringexerciseare exclusivelyresponsibleforthefeelingofbreathlessnessinsome patients.

Noneofthesubjectshadahistoryofcardiorespiratorydiseases, diabetesorotherdisordersthataffecttheautonomicnervesystem thatcouldhaveinfluencedtheresults.However,allpatientsinthis pilotstudyusedmedicationandalthoughunlikely,itcannotbe ruledoutthatsomeoftheanticonvulsantsorotherdrugshave beenofinfluenceonthemeasuredparameters,e.g.heartrate.The cardiorespiratoryparametersobtainedinpatientswere compara-ble to the parameters that were measured in healthy control subjects.

Although weanalyzed over60 VNS episodes,only a limited numberofpatientsparticipatedinthispilot.Severalreproducible cardiorespiratoryeffectsofVNShavebeenfoundinallpatients,but the number of patients is too small to differentiate between patientswhoreportbreathlessnessduringexerciseandthosewho donot,especiallysincethereseemsmainlyagradualdifferencein theexperiencedsideeffects.

5. Conclusion

We haveinvestigatedtheeffectsofVNSoncardiorespiratory parameters,duringbothrestandexercise.Thebreathingfrequency increased,tidalvolumesdecreasedandtheheartratedecreased consistentlyduring VNSepisodesin about80% of ourpatients. Theseeffectswerepresentduringbothrestandexercise,andthere wasnocleardistinctionbetweenthepatientswhoreportedVNS inducedsideeffectsduringexerciseandthosewhodidnot. Disclosure

WeconfirmthatwehavereadtheJournal’spositiononissues involved in ethical publication and affirm that this report is consistentwiththoseguidelines.

Conflictsofinterest

Noneoftheauthorshasanyconflictofinteresttodisclose. Acknowledgments

WewouldliketothankM.J.DriesseMD,whoimplantedthe VNSdevices,forhiscontinuoussupportandtime.

References

[1]ArdeschJJ,BuschmanHPJ,Wagener-SchimmelLJJC,vanderAaHE,HagemanG. Vagusnervestimulationformedicallyrefractoryepilepsy:along-term follow-upstudy.Seizure2007;16:579–85.

[2]deVosCC,MelchingL,vanSchoonhovenJ,ArdeschJJ,deWeerdAW,van LambalgenHCE,etal.Predictingsuccessofvagusnervestimulation(VNS) frominterictalEEG.Seizure2011;20:541–5.

[3]ElliottRE,MorsiA,KalhornSP,MarcusJ,SellinJ,KangM,etal.Vagusnerve stimulationin436consecutivepatientswithtreatment-resistantepilepsy: long-termoutcomesandpredictorsofresponse.EpilepsyBehav2010;20: 57–63.

[4]TecomaES,IraguiVJ.Vagusnervestimulationuseandeffectinepilepsy:what havewelearned?EpilepsyBehav2006;8:127–36.

[5]FahyBG.Intraoperativeandperioperativecomplicationswithavagusnerve stimulationdevice.JClinAnesth2010;22:213–22.

[6]ArdeschJJ,SikkenJR,VeltinkPH,vanderAaHE,HagemanG,BuschmanHPJ. Vagusnervestimulationforepilepsyactivatesthevocalfoldsmaximallyat therapeuticlevels.EpilepsyRes2010;89:227–31.

[7]MarzecM,EdwardsJ,SagherO,FromesG,MalowBA.Effectsofvagusnerve stimulation on sleep-related breathing in epilepsy patients. Epilepsia 2003;44:930–5.

[8]BanzettRB,GuzA,PaydarfarD,SheaSA,SchachterSc,LansingRW. Cardiore-spiratory variablesand sensationduring stimulationoftheleft vagus in patientswithepilepsy.EpilepsyRes1999;35:1–11.

[9]KhuranaDS,ReumannM,HobdellEF,NeffS,ValenciaI,LegidoA,etal.Vagus nervestimulationinchildrenwithrefractoryepilepsy:unusualcomplications andrelationshiptosleep-disorderedbreathing.ChildsNervSyst2007;23: 1309–12.

[10]ZaaimiB,GrebeR,BerquinP,WalloisF.Vagusnervestimulationinduces changesinrespiratorysinusarrhythmiaofepilepticchildrenduringsleep. Epilepsia2009;50:2473–80.

[11]ArdeschJJ,BuschmanHPJ,VanDerBurghPH,Wagener-SchimmelLJJC,vander AaHE,HagemanG.Cardiacresponsesofvagusnervestimulation: intraoper-ativebradycardiaandsubsequentchronicstimulation.ClinicalNeurol Neu-rosurg2007;109:849–52.

[12]IriarteJ,UrrestarazuE,AlegreM,Macı´asA,Go´mezA,AmaroP,etal.Late-onset periodicasystoliaduringvagusnervestimulation.Epilepsia2009;50:928–32.

[13]A˚markP,Sto¨dbergT,WallstedtL.Lateonsetbradyarrhythmiaduringvagus nervestimulation.Epilepsia2007;48:1023–4.

[14]FreiMG,OsorioI.Leftvagus nervestimulationwiththeneurocybernetic prosthesishascomplexeffectsonheartrateandonitsvariabilityinhumans. Epilepsia2001;42:1007–16.

[15]SabbahHN.Electricalvagusnervestimulationforthetreatmentofchronic heartfailure.CleveClinJMed2011;78:S24–9.

[16]AliII,PirzadaNA,KanjwalY,WannamakerB,MedhkourA,KoltzMT,etal. Completeheartblockwithventricularasystoleduringleftvagusnerve stim-ulationforepilepsy.EpilepsyBehav2004;5:768–71.

[17]ZaaimiB,GrebeR,WalloisF.Animalmodeloftheshort-termcardiorespiratory effectsofintermittentvagusnervestimulation.AutonNeurosci2008;143: 20–6.

[18]NagarajanL,WalshP,GregoryP,StickS,MaulJ,GhoshS.Respiratorypattern changesinsleepinchildrenonvagalnervestimulationforrefractoryepilepsy. CanJNeurolSci2003;3:224–7.

[19]BinksAP,PaydarfarD,SchachterSC,GuzA,BanzettRB.Highstrength stimu-lationofthevagusnerveinawakehumans:alackofcardiorespiratoryeffects. RespiPhysiol2001;127:125–33.