Article
Reference
Electrode location and clinical outcome in hippocampal electrical stimulation for mesial temporal lobe epilepsy
BONDALLAZ, Percy, et al.
Abstract
To study the clinical outcome in hippocampal deep brain stimulation (DBS) for the treatment of patients with refractory mesial temporal lobe epilepsy (MTLE) according to the electrode location.
BONDALLAZ, Percy, et al . Electrode location and clinical outcome in hippocampal electrical stimulation for mesial temporal lobe epilepsy. Seizure , 2013, vol. 22, no. 5, p. 390-395
PMID : 23540624
DOI : 10.1016/j.seizure.2013.02.007
Available at:
http://archive-ouverte.unige.ch/unige:48386
Disclaimer: layout of this document may differ from the published version.
1 / 1
Electrode location and clinical outcome in hippocampal electrical stimulation for mesial temporal lobe epilepsy
Percy Bondallaz
a, Colette Boe¨x
c, Andrea O. Rossetti
b, Giovanni Foletti
b, Laurent Spinelli
c, Serge Vulliemoz
c, Margitta Seeck
c, Claudio Pollo
a,*
aDepartmentofNeurosurgery,CentreHospitalierUniversitaireVaudois,1011Lausanne,Switzerland
bDepartmentofNeurology,CentreHospitalierUniversitaireVaudois,1011Lausanne,Switzerland
cEpilepsyUnit,DepartmentofNeurology,UniversityHospitalofGeneva,1211Geneva,Switzerland
1. Introduction
Epilepsyisafrequentneurologicaldiseasethataffects0.5–1%of thepopulation.1About30%ofpatientshaveapharmacologically intractable form of epilepsy.2 Mesial temporal lobe epilepsy (MTLE) is a particularly common form of pharmacoresistant epilepsy.3Surgicalresectionoftheamygdalo-hippocampalstruc- turesaloneortogetherwiththeanteriorportionoftemporallobeis aneffectivetreatmentofMTLE.4,5However,ablativesurgeryisnot possible in up to 30% of patients in whom resection of the amygdalo-hippocampalcomplexwillresultinsevereneurological impairments such as memory deficits,2,6 or in cases involving bitemporalepilepticfoci.Inthesepatientselectricalstimulationof the amygdala and hippocampus has been proposed as an alternativetreatment.7–10
Previousstudieshavehighlightedtheefficacyofhighfrequency deepbrainstimulation(DBS)toreduceepilepticactivityeitherby targetingintracerebralstructuresbelievedtohaveatriggeringrole intheepilepticnetwork,suchasthethalamus,thesubthalamic nucleus, the caudate nucleus, and the cerebellum or thevagal nerve.11–13 Alternatively, theictal onset zone maybe targeted, withthehypothesisthatstimulationmayinterferewithseizure initiation.Thelatterstrategyhasbeendescribedtobesuitableto controlseizuresinpatientswithMTLE.Inthesecasesinvestiga- tions using intracranial electrodes14,15 have strongly suggested that seizure onset and propagation involve the amygdala and hippocampus.
Clinically, it has been shown that hippocampal stimulation using depth electrodes significantly reduces interictal EEG spikes16,17 and improves seizure outcome in patients with temporallobeepilepsy.7–10,16,18,19
However,responsesarevariable intermsofseizurefrequencyreductionleadingtotheneedfora betterunderstanding of the mechanism by which DBSreduces seizurefrequency,aswellasidentificationofoptimaltargetsand optimizationof stimulation parameters. Onehypothesis is that DBSmayactthroughlocalinhibitionofneuronsadjacenttothe ARTICLE INFO
Articlehistory:
Received24September2012
Receivedinrevisedform13February2013 Accepted14February2013
Keywords:
Deepbrainstimulation Temporallobeepilepsy Hippocampalsclerosis Pharmacoresistance Outcome
ABSTRACT
Purpose:Tostudytheclinicaloutcomeinhippocampaldeepbrainstimulation(DBS)forthetreatmentof patientswithrefractorymesialtemporallobeepilepsy(MTLE)accordingtotheelectrodelocation.
Methods:EightMTLEpatientsimplantedinthehippocampusandstimulatedwithhigh-frequencyDBS wereincludedinthisstudy.Fiveunderwentinvasiverecordingswithdepthelectrodestolocalizeictal onsetzonepriorto chronicDBS.Positionoftheactivecontactsoftheelectrodewascalculatedon postoperativeimaging.Thedistancesto theictalonsetzoneweremeasuredaswellasatlas-based hippocampusstructuresimpactedbystimulationwereidentified.Bothwerecorrelatedwithseizure frequencyreduction.
Results:Thedistancesbetweenactiveelectrodelocationandestimatedictalonsetzonewere114.3or 9.12.3mmforpatientswitha>50%or<50%reductioninseizurefrequency.Inpatients(N=6)showinga
>50% seizure frequencyreduction, 100%had the active contacts located<3mmfrom thesubiculum
(p<0.05).The2non-responderspatientswerestimulatedoncontactslocated>3mmtothesubiculum.
Conclusion:DecreaseofepileptogenicactivityinducedbyhippocampalDBSinrefractoryMTLE:(1) seemsnotdirectlyassociatedwiththevicinityofactiveelectrodetotheictalfocusdeterminedby invasiverecordings;(2)mightbeobtainedthroughtheneuromodulationofthesubiculum.
ß2013BritishEpilepsyAssociation.PublishedbyElsevierLtd.Allrightsreserved.
*Correspondingauthorat: DepartmentofNeurosurgery,UniversityHospital Bern,CH-3010Bern,Switzerland.Tel.:+410316320810;fax:+410313822414.
E-mailaddress:[email protected](C.Pollo).
ContentslistsavailableatSciVerseScienceDirect
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
1059-1311/$–seefrontmatterß2013BritishEpilepsyAssociation.PublishedbyElsevierLtd.Allrightsreserved.
http://dx.doi.org/10.1016/j.seizure.2013.02.007
areaofelectrodeimplantation,therebymodulatingtheactivityof cerebral structures triggering seizure onset. Alternatively, DBS may have an effect on the network of neuronal projections connecting several cerebral structures.20 Since mesialtemporal lobestructuresarepotentiallyinvolvedinepilepticnetworks,the targeting of ictal foci in this region may also affect adjacent networks.
Wepreviouslypublishedastudythatfocusedontheefficiency ofhippocampalstimulationonreducingseizurefrequencyandon the influence of stimulation parameters. One unresolved issue concernstheimpactofelectrodepositioningonseizuretreatment, which mayin turnproveinformative for targetingpractices in general.
Therefore,inthepresentstudy,weretrospectivelyanalyzed(1) thedistancebetween theimplantedDBSstimulatingcontact(s) relativetotheictalonsetfocusdeterminedinvasively,and(2)the anatomicalstructurespossibly influenced byelectrical stimula- tion. These two parameters were compared with the clinical outcome.
2. Methods
2.1. Patientsandinclusioncriteria
EightpatientswithintractableMTLEepilepsywereselectedfor DBStreatment betweenJune2002andApril2008aspreviously described10(5womenand3men,medianage:31.5years,range:
25–47). The criteria for patient selection to proceed withDBS includedpharmaco-resistance and proven MTLE seizure origin.
Resectivesurgeryisusuallyproposedasthetreatmentofchoicein these patients. DBS was considered in patients with either concerns for possible post-operative significant worsening of memory,particularlyverbalmemory,orwhenbilateralepilepto- geniczonesweresuspected.Detailsofinclusioncriteriaandofthe presurgical protocol were published previously10 and include high-resolutionbrainMRI,video-EEGtelemetry,interictalpositron emission tomography (PET), ictal and interictal single photon emissioncomputerizedtomography(SPECT),aswellasneuropsy- chological and psychiatric examinations. High-resolution MRI showedahippocampalsclerosisin2patients;theremaining6had non-lesionalMTLE(Table1).
ThestudywasapprovedbythelocalEthicsCommitteeofthe University Hospitals ofGeneva and Lausanne,and an informed consentwasobtainedfromeachpatient.
2.2. Identificationofictalfocus
In5of8patients(Pt4,5,7,8,9),theEEGictalonsetfocuswas estimated by invasive recordings using intracerebral depth
electrodesinsertedperpendiculartotheskullsurfaceatamygda- lar, anteriorandposterior hippocampallevelsin both temporal lobes as previously described.10 Epileptogenic ictal focus was assignedtothecontact(numbered1to8)recordingmaximalictal activity(pathologicalwaveform).Ahigh-resolutionCTscanwas then co-registered with a T1-weighted MRI acquired under stereotacticconditions (CRW,Radionics1,Burlington, MA,USA) and processed using the Framelink 5.1 software on a Stealth workstation(MedtronicInc,Minneapolis,MN,USA).Thepostop- erative imaging was realigned to the anterior commissure- posteriorcommissure(AC–PC)coordinatessystembyidentifying theanteriorandposteriorcommissuresand3midlinelandmarks.
Origin wasset at the midcommissural point. Three orthogonal planesofviewwerethenusedtolocalizetheelectrodecontact.Its coordinateswerecalculatedandexpressedas(x)mmlateraltothe midline,(y)mmantero-posteriorand(z)mmsupero-inferiortothe mid-commissuralplane.
2.3. Surgicalprocedure
Surgicalplanningandprocedurewereperformedaspreviously described.10 The Pisces-Quad 3487A electrode and the Soletra 7426 stimulator (Medtronic Inc, Minneapolis, MN, USA) were implantedinthefirst5patients.The4cylinder-shapedcontactsof the Pisces-Quadelectrode are 3mm in length and 1.27mm in diameter.Theintercontactdistanceis6mm,andtheelectrodeis 30mmintotallength.The3remainingpatientsreceivedtheSub Compact Octad 3876 electrode and the Restore stimulator (MedtronicInc.,Minneapolis,MN,USA).TheSubCompactOctad electrodeis34.5mmintotallengthwith8contacts(3mmlength, 1.27mm diameter, 1.5mm intercontact distance). The DBS electrodeswereplacedparasagittalyintheamygdalo-hippocam- pal complex so that the distal contact (contact 0) could be implanted in the area of the amygdala. Internalization of the electrodeandconnectiontotheneurostimulatorwasperformed 3–4days after the implantation procedure to provide EEG recordings.
2.4. Stimulationparametersandfollow-up
Thesetting ofpost-implantationstimulationparameters and neurological evaluations were performed as previously de- scribed.10 All patients were stimulated at high-frequency, i.e.
130Hz, and with pulse width of 0.45ms. The amplitude of stimulation(0.5–2V)andthenumberofcontactsstimulated(bi- or quadripolar)were, however,differentacrosspatients. In the quadripolarconfiguration,the4contactsweresetascathodes,and thecaseboxoftheneurostimulatorwassetastheanode.Inthe bipolar configuration, the cathode was set on the contact
Table1
Clinicalcharacteristicsofpatients.
Patient Sex Age/onset Follow-up (months)
HS Side Ictalfocus Interictal focus
Stimulation contact
Amplitude (V)
Outcome(%reduction inseizurefrequency)
Pt1 F 37/24 74 Yes Left – C1 quad 1 67
Pt2 F 32/3 50 Yes Right – C2 quad 1 88
Pt3 F 44/4 46 No Right – C0 quad 0.5 72
Pt4 F 31/25 45 No Left LAH1-2 C1 C0–C1 0.5 84
Pt5 M 47/21 42 No Right RAH3 n.i. C0–C1 1 100
Pt7 M 31/14 34 No Left LAH2 C2 C1–C2 1 0
C2–C3 1 0
Pt8 M 25/13 11 No Left LA1a C2 C1–C2 1.5 22
Pt9 F 26/13 10 No Left LAH2 C0 off 0 100
C4 off 0 100
HS:hippocampalsclerosis,quad:quadripolarstimulation,LAH:leftanteriorhippocampus,RAH:rightanteriorhippocampus,n.i.:notidentified,LA:leftamygdala,off:not stimulated,C:electrodecontact.
aSecondaryfocus.
P.Bondallazetal./Seizure22(2013)390–395 391
correspondingtothemaximalinterictalepileptogenicactivity,as determinedfromintracranialEEGrecordingsbeforetheinternali- zationoftheneurostimulator,andtheanodewassetonthecontact closesttothesecondmajorinterictalepileptogenicsite.Aftera3 monthoff-period,patientswerefollowedforamediandurationof 43.5months(range10–74months).Quantificationoftheclinical outcome and efficacy of stimulation was performed by the evaluation of the ratio between the improvement in seizure frequency after implantation compared to pre-implantation baselines determined prospectively as the mean number of seizurespermonthduringthethreemonthspriortoimplantation (accordingtopatients’self-reports).IntroductionofanynewAED wasnotallowedafterimplantationinordertodeterminetheeffect ofDBS,butminorchangesinmedicationdosageswereacceptedas previously described.10 Characteristics of patients, stimulation parametersandclinicaloutcomearesummarizedinTable1.For the following analysis, patients were then split into 2 groups accordingtotheir rateofseizurefrequency reduction(i.e.>or
=50%and<50%).
2.5. Determinationofthedistancebetweentheactivecontact(s)ofthe DBSelectrodeandtheestimatedictalfocus
PostoperativeimagingwasprocessedusingtheFramelink5.1 softwareon a Stealth workstation(Medtronic Inc, Minneapolis, MN,USA)and realignedtotheAC–PC coordinatessystemwith origin set at the midcommissural point. The electrode contact imageartifactwaslocalizedinthe3orthogonalplanesofview.The centeroftheartifactwasidentifiedasthecenteroftheelectrode (Fig.1A)accordingtoa previousstudy ofDBSin patients with Parkinson’s disease.21 Its coordinates (x), (y) and (z) were calculatedasexplainedabove.In 5patients, AC–PCcoordinates weresubtractedineachplane(dx,dy,dz)todeterminethedistance between theestimatedictal focusand the implantedelectrode contacts. Euclidian distance in 3D space was then calculated (squarerootof(dx2+dy2+dz2)).These4 parameters,aswellas clinical outcome, were used for further analysis. In each case, distancesbetweentheestimatedictalfocusandallcontactsofthe
electrode were calculated in order to estimate the minimal distancetotheelectrode.
2.6. Determinationofstructuresinthevicinityoftheelectrodes influencedbyDBS
Toidentifystructuresinthevicinityofactiveelectrodecontacts, postoperativeimaging wasco-registeredand adjustedwiththe corresponding template of a neuro-anatomical atlas22 prepared according to the Talairach standard transformation23 (Fig. 1B).
Structures overlapping a 3mm-radius circle centered on the artifact of the electrode contact were considered as possibly influenced by electrical stimulation (Fig. 1C), according to the estimationofthevolumeoftissueactivatedtakenfromdifferent existingfiniteelementmodelsofelectricalpropagationaroundthe electrode.24,25ThenonparametricSpearmancorrelationtestwas usedforthestatisticalanalysisonsmallsamples.
3. Results
3.1. Clinicaloutcomeandstimulationparameters
Postoperativeseizurefrequencieswerecomparedwithapre- implantation baseline period. Six of the8 patients exhibited a reductionofseizurefrequencyof>50%,including2seizure-free patients(i.e.100%reductionofseizurefrequency).The2remaining patientswerenon-responders(i.e.nosignificantchangeinseizure frequency).Reasonsthatcouldexplainsuchgoodresultscompared to other studies have been previously discussed.10 In the first group,Pt1andPt2didnotshowanyreductionwhenstimulatedin abipolarconfigurationwithcontactsC0andC1.Whenstimulated in a quadripolar configuration, they experienced a significant reduction in seizure frequency (67% and 88%, respectively) as publishedpreviously.10Pt3andPt4alsoshowedamajorseizure reduction of 72% and 84% with the quadripolar and bipolar configuration,respectively.Pt5andPt9becameseizurefreewitha bipolarconfiguration;thelatter remainedseizure freeafterthe electrode was implanted, and during the off-period without
Fig.1.MRIanalysisusingthesurgicalnavigationsystemStealthworkstation(MedtronicInc.,Minneapolis,MN,USA).(A)Measurementofelectrodecontact(C1)positionon post-operativebrainT1-weightedMRIofPt3treatedwithDBS,enablingdeterminationofstereotacticcoordinatesbaseduponthestereotacticsurgicalframe.Realignmentof thescantotheAC–PClineisshowninthecoronalplane.ThewhiteasteriskindicatesthecenteroftheDBSelectrodecontact.(B)Overviewofthehippocampalandsubicular regionswithsuperimposedstandardneuro-anatomicalatlas(adaptedfromMaiJK,AssheuerJ,PaxinosG.Atlasofthehumanbrain.SanDiego:AcademicPress;1998).(C)The white3mm-radiuscirclereflectsthebrainareasupposedtobeinfluencedbytheelectrodecontact.CA,cornuammonis;DG,dentategyrus;Ent,entorhinalcortex;FD,fascia dentata;opt,optictract;TLV,temporalhornoflateralventricle;PaS,parasubiculum;PrS,presubiculum;S,subiculum.
stimulation.Pt7 andPt8didnot showsignificantreductionsin seizurefrequenciesduringbipolarstimulation.Findingsregarding Pt7werepreviouslyreported andindicated a seizurereduction duringthefirst6months,butunfortunatelytheelectrodehadtobe reimplantedduetoafractureofthefirstPiscesQuadelectrode.
WiththenewSubCompactOctadelectrode,noseizurereduction wasachievedduringthemonthsofbipolarstimulationat1V.The follow-upforeachpatientisindicatedinTable1.Theoutcomewas notcorrelatedtothefollow-up(Spearmantest,
r
= 0.0599,n.s.).3.2. Distanceofactivecontacttoestimatedictalfocus
Theictalfocuswasestimatedinthe5investigatedpatientswith invasive recordings. Since Pt9 was seizure-free even without stimulation,distancesoftheestimatedictalfocustothecontacts werenotconsideredfortheanalysis.Forthe4remainingpatients, theEuclidiandistancesbetweenthelocationofelectrodecontacts usedfor stimulationand thecontactsregisteringmaximal ictal activityduringpre-surgicalinvasiveinvestigationswereallgreater than6mm(seeTable2).Thesevaluesrangedfrom1.6to7.8mmin thelatero-medial (x)axis, from 0.7 to 14.8mm in theantero- posterior(y)axis,andfrom0.8to7.0mminthesupero-inferior(z) axis.ThemeanEuclidiandistancesare11.04.3or9.12.3mm forpatientswitha>50% or<50%reductioninseizure frequency, respectively.Norelationcouldbeobservedbetweenthedistanceof activeelectrode contactsto the estimatedictalfocus andclinical outcome.Interestingly,theentireelectrodeofPt9islocalizedfarfrom theestimatedictalfocus(>10mm).
3.3. Identificationofcerebralstructuresimpactedbyelectrode stimulation
Inorder tofurthercharacterize mechanismsunderlyingDBS effectonseizurefrequencyreduction,theamygdalo-hippocampal areasstimulatedbytheelectrodeswereidentifiedoncoronalMRI- slices.Mostoftheactiveelectrodecontactsarelocalizedcloseto theCA1fieldofthehippocampusandthesubiculum(Table3).In patientsshowinga>50%seizurefrequencyreduction,100%had their subiculum localized at less than 3mm from the active contacts.Pt7andPt8presentednosignificantreductionsinseizure frequency and were not stimulated in an area including the subiculum.Theclinicaloutcomewassignificantlycorrelatedwith the proximity of the subiculum to the closest active contact
(Spearmantest,
r
=0.677,p<0.05).Ontheotherhand,allpatients hadtheiractivecontactsclosetotheCA1fieldofthehippocampus, including the two non-responders, and no correlation was observed (r
= 0.5668, n.s.). Furthermore, as described above, Pt1 and Pt2 showed better outcomes when stimulated in a quadripolar configurationincluding thecontact C2,the nearest contacttothesubiculum.4. Discussion
DBS has been shown to be successful in the treatment of refractoryepilepsy,despitethewidespectrumofresultsproduced in clinical experiences in the literature. Pioneering studies concerninghippocampalstimulationforMTLEarebasedonsmall patient populations.8,9,18,19 Its mechanisms of action remain largely unknown. In the present study we first examined the relationshipbetweenelectrodecontactsandestimatedictalonset zonelocationstofurtherinvestigateitsimpactonclinicaloutcome inMTLE.Wedidnotobserveanyclearrelationshipbetweenthe locationofactivecontactsandthepresumedictalonsetfocus.DBS activeelectrodecontactswereallfoundtobepositionedmorethan 6mmfromtheestimatedictalonsetfocus.Theaccuracyofictal onsetfocuslocalizationbyinvasiverecordingsmaybequestioned, especially in theantero-posteriordirectionwherethesampling withdepthelectrodeswasperformedintherangeof1cm,andas therecordedEEG(localfieldpotentials)issupposedtoreflectthe synchronous activity of numerous neurons.26 However, it is reasonabletothinkthattheerrorintheantero-posteriordirection shouldnotexceedtherangeof5mm(i.e.halfthedistancebetween two electrodes). Moreover, as the mean Euclidian distance betweentheictalfocusandthestimulatedcontactsforpatients
with a >50% or <50% reduction in seizure frequency are
comparable, and calculated with thesame probability oferror, wesuggestthattheseizureoutcomeisnotdirectlyrelatedtothe vicinityoftheictalfocusdeterminedwithinvasiveelectrodes.As an illustrative example, Pt9 showed a good outcome without stimulation (probablydue toa micro-lesional effect). Sincethe entireelectrodeofPt9islocalized>10.6mmfromtheestimated ictalfocus,itseemsdifficulttoassociatethisoutcomethrougha directeffectontheictalfocus.
Duetothesmallnumberofpatientsitwasdifficulttoperform statisticalanalyses,butnotrendseemstoseparateonegroupfrom the other in any axis, or according to patients’ characteristics (presurgical seizure frequency, type of seizure, hippocampal sclerosis). However, considering that the current spread from the electrode is presumed to be smaller than 4mm in radius accordingtoDBSmodels,weobservedthatsufficientreductionin seizure frequency was obtained even when the contacts were localizedat higher distances.This suggeststhat indirecteffects couldbeproducedbystimulationofaparticularstructure,orpart of it, potentially involved in the onset or propagation of the epileptic current of mesio-temporal seizure. In our experience, patients with hippocampal sclerosis generally needed a more extended area of stimulation and had more electrode contacts stimulated,comparedtothenon-lesionalcases.Althoughwedo not have a definitive explanation, one speculation is that morphological changes induced by sclerosis may result in less functionaltissuethatcanbestimulatedand/orinanincreasein tissueimpedance.
Interestingly, the contacts presenting maximal ictal activity duringpresurgicalinvasiverecordingswerethecontactsclosestto thesubiculumin3of5patients.Duetothespatialresolution,26itis notpossibletoexcludethatictalactivityofotherstructurescould benonethelessrecordedatthesesites.Therefore,cautionshould betakenbeforedrawingconclusionsregardingthepossibleroleof thesubiculumingenerationorinpropagationofepilepticcurrents.
Table2
Distancesfromelectrodecontactstoestimatedictalfocus.
Patient Stimulated contact
Distance(mm)
dx dy dz 3D
Pt4 C0+ 7.7 6.9 3.9 11.1
C1 7.2 1.8 0.8 7.5
Pt5 C0 4.6 14.8 7.0 17.0
C1+ 1.6 6.7 5.0 8.5
Pt7 C1+ 4.2 8.2 6.2 11.1
C2 4.8 2.7 4.2 7.0
C3+ 5.0 0.8 3.5 6.1
Pt8 C1+ 7.6 0.7 6.7 10.1
C2 7.8 3.6 6.9 11.0
Pt9 C0 6.7 13.0 0.5 14.6
C1 8.2 7.9 1.3 11.4
C2 9.4 4.5 2.4 10.7
C3 11.2 0.1 3.8 11.8
C4 12.6 4.7 5.4 14.5
C5 13.3 8.2 6.0 16.7
C6 14.8 12.0 7.0 20.3
C7 15.9 16.7 8.5 24.6
+,:Polarityofthestimulatedcontacts.
3D:Euclidiandistance.
P.Bondallazetal./Seizure22(2013)390–395 393
Second,atlas-basedanalysisofamygdalo-hippocampalstruc- tures located within a 3mm-radius sphere around the active contacts of stimulation showed that all patients were well- stimulatedintheCAregionofthehippocampus.Interestingly,the electrodecontactsthatwereclosertothesubiculum,ormayhave had a lesional effect on the subiculum during the electrode insertion,wereassociatedwithimportant reductionsinseizure frequency,whereasnosignificanteffectwasobservedwhenthe electrodewaslocatedfartherthan3mmfromthesubiculum.This observationsuggeststhattheefficacyofDBSmightbeassociated withtheinvolvementofthesubiculum,whichalsocarriesaxonsof the perforant pathway, and that the beneficial effects may be obtainedthroughneuromodulationofthisstructure.
Severalstudieshavehighlightedtheroleofthedentategyrus and CA1 region in hippocampal sclerosis models.27,28 More recently,several studies29–33 have demonstrated that the sub- iculumandparahippocampalstructures,butnotthehippocampus itself, play anactive role in the generationand propagation of temporal lobe seizures, even in non-sclerotic hippocampal tissues.34Ourstudyisthefirsttoprovideclinicaldatainhumans supporting a potential involvement of the subiculum in the generationand/orpropagationofseizuresinMTLE.
Therearenodataunderlyingthedirectneuromodulatoryeffect of electrical stimulation on the subiculum in refractory MTLE.
Studies have suggested that changes in GABAergic signaling causing (1) hyperexcitability in thesubiculum, that recalls the GABAergicexcitation29,35ofearlydevelopment,aswellas(2)the
vulnerabilityofGABAergicinterneurons,thatmaygiverisetoan input-specific impairment of inhibition,32 are the mechanisms underlyingdevelopmentofMTLEatacellularlevel.Accordingto these observations, neuromodulatory effects of high-frequency DBS may decrease the excitability of the subiculum and then improvetheinhibitoryeffectofGABAergicpathwaysongenera- tionand/orpropagationofMTLE.
Thereliabilityofourresultscouldbeimprovedbyincreasing theseriessize,especiallywhensubgroupsareconsidered.Further prospective multicentric studies involving a greater number of patientsarenecessarytoprovidemoreconsistentdataconfirming therole ofthesubiculum inelectrical stimulationin refractory MTLE.
Inconclusion,ourresultssuggestthatdecreasesofepilepto- genic activityinduced by hippocampal highfrequency DBS in refractoryMTLEseemnot tobeassociatedwiththevicinityof the activeelectrode to the ictalfocus determined by invasive recordings.Instead,theymightbeassociatedwiththevicinityof theactiveelectrodetothesubiculumandobtainedthroughthe neuromodulationof thisstructure.Further prospectivestudies conductedonalargergroupofpatientsarenecessarytoconfirm the neuromodulatory effect of hippocampal DBS on the subiculum.
Conflictofintereststatement
Noneoftheauthorshasanyconflictofinteresttodisclose.
Table3
Structureslocalizednearthestimulatedcontacts.
Patient Stimulated contact
LaV LaI LaDA BM BL BLVM BLI HiH DG CA3 CA2 CA1 S PrS Ent PHG TLV
Pt1 C0 +++ ++ ++ +
C1 +++ ++ +
C2 +++ + ++
C3 +++ +++ +++
Pt2 C0 ++ ++ + +
C1 +++ + +
C2 ++ + +++ +
C3 ++ + ++
Pt3 C0 ++ ++ ++
C1 +++ +++ ++
C2 + + +++
C3 ++ + + +++ + +
Pt4 C0 ++ ++ ++
C1 +++ ++ +
Pt5 C0 +++ + ++
C1 + +++ ++
Pt9 C0 +++ ++
C1 +++ +
C2 ++ +++ ++ ++
C3 + ++
C4 +++ +++ + ++
C5 +++ + ++
C6 +++ +++ ++ ++
C7 +++ ++ +++
Pt7 C1 ++ +++
C2 ++ +++ + +
C3 ++ ++ +
Pt8 C1 + + + +++
C2 +++
Patientswitha>50%reductioninseizurefrequencyaregroupedintheupperpartofthetableandindicatedinboldtype.Theproximityofthesubiculumtothestimulated contactsisalsoindicatedinboldtype.+to+++:proportionofthestructurelocalizedina3mm-radiusareaaroundtheelectrodecontact.BL:basolateralamygdaloidnucleus, BLI:basolateralamygdaloidnucleusintermediatepart,BLVM:basolateralamygdaloidnucleusventromedialpart,BM:basomedialamygdaloidnucleus,CA1:CA1fieldofthe hippocampus,CA2:CA2fieldofthehippocampus,CA3:CA3fieldofthehippocampus,DG:dentategyrus,Ent:entorhinalcortex,HiH:hippocampalhead,LaDA:lateralamygd.
nucl.dorsalanteriorpart,LaI:lat.amygdaloidnucl.intermed.part,LaV:lateralamygdaloidnucl.ventralpart,PHG:parahippocampalgyrus,PrS:presubiculum,S:subiculum, TLV:temporalhornoflateralventricle.
Acknowledgements
WethankProf.JGVillemureforhispioneeringworkinDBSfor epilepsyandforhavingoperatedthefirstpatients.Thisresearch wassupported by theSwiss National Science Foundation (FNS GrantNos.33CM30-124089,320030-122073).
References
1.HauserWA,AnnegersJF,KurlandLT.Theincidenceofepilepsyandunprovoked seizuresinRochester,Minnesota,1935–1984.Epilepsia1993;34:453–68.
2.KwanP,BrodieMJ.Earlyidentificationofrefractoryepilepsy.TheNewEngland JournalofMedicine2000;342:314–9.
3.EngelJrJ.Mesialtemporallobeepilepsy:whathavewelearned?Neuroscientist 2001;7:340–52.
4.WiebeS,BlumeWT,GirvinJP,EliasziwM.Arandomized,controlledtrialof surgery for temporal-lobe epilepsy. TheNew England Journalof Medicine 2001;345:311–8.
5.Cohen-GadolAA,WilhelmiBG,CollignonF,WhiteJB,BrittonJW,CambierDM, etal.Long-termoutcomeofepilepsysurgeryamong399patientswithnon- lesionalseizurefociincludingmesialtemporallobesclerosis.JournalofNeuro- surgery2006;104:513–24.
6.HelmstaedterC,KurthenM,LuxS,ReuberM,ElgerCE.Chronicepilepsyand cognition:alongitudinalstudyintemporallobeepilepsy.AnnalsofNeurology 2003;54:425–32.
7.VonckK,BoonP,ClaeysP,DedeurwaerdereS,AchtenR,VanRoostD.Long-term deep brain stimulation for refractory temporal lobe epilepsy. Epilepsia 2005;46:98–9.
8.Tellez-ZentenoJF,McLachlanRS,ParrentA,KubuCS,WiebeS.Hippocampal electricalstimulationinmesialtemporal lobeepilepsy.Neurology2006;66:
1490–4.
9.VelascoAL,VelascoF,VelascoM,TrejoD,CastroG,Carrillo-RuizJD.Electrical Stimulationofthehippocampalepilepticfociforseizurecontrol:adouble- blind,long-termfollow-upstudy.Epilepsia2007;48:1895–903.
10.Boe¨xC,SeeckM,Vullie´mozS,RossettiAO,StaedlerC,SpinelliL,etal.Chronicdeep brainstimulationinmesialtemporallobeepilepsy.Seizure2011;20:485–90.
11.IadarolaMJ,GaleK.Substantianigra:siteofanticonvulsantactivitymediatedby gamma-aminobutyricacid.Science1982;218:1237–40.
12.Ben-MenachemE. Vagus-nerve stimulationfor thetreatmentof epilepsy.
LancetNeurology2002;1:477–82.
13.VelascoF,Carrillo-RuizJD,BritoF,VelascoM,VelascoAL,MarquezI,etal.
Double-blind,randomizedcontrolledpilotstudyofbilateralcerebellarstimu- lationfortreatmentofintractablemotorseizures.Epilepsia2005;46:1071–81.
14.SwansonTH.Thepathophysiologyofhumanmesialtemporallobeepilepsy.
JournalofClinicalNeurophysiology1995;12:2–22.
15.SpencerSS.Neuralnetworksinhumanepilepsy:evidenceofandimplications fortreatment.Epilepsia2002;43:219–27.
16.VelascoM,VelascoF,VelascoA,BoleagaB,JimenezF,BritoF,etal.Subacute electricalstimulationofthehippocampusblocksintractabletemporallobe seizuresandparoxysmalEEGactivities.Epilepsia2000;41:158–69.
17.Boe¨xC,Vullie´mozS,SpinelliL,PolloC,SeeckM.Highandlowfrequency electricalstimulationinnon-lesionaltemporallobeepilepsy.Seizure2007;
16:664–9.
18.VonckK,BoonP,AchtenE,DeReuckJ,CaemaertJ.Long-termamygdalohippo- campalstimulationforrefractorytemporallobeepilepsy.AnnalsofNeurology 2002;52:556–65.
19.BoonP,VonckK,DeHerdtV,VanDyckeA,GoethalsM,GoossensL,etal.Deep brainstimulationinpatientswithrefractorytemporallobeepilepsy.Epilepsia 2007;48:1551–60.
20.McIntyreCC,SavastaM,Kerkerian-LeGoffL,VitekJL.Uncoveringthemecha- nism(s)ofactionofdeepbrainstimulation:activation,inhibition,orboth.
ClinicalNeurophysiology2004;115:1239–48.
21.PolloC,VingerhoetsF,PralongE,GhikaJ,MaederP,MeuliR,etal.Localization ofelectrodesin thesubthalamicnucleusonmagnetic resonanceimaging.
JournalofNeurosurgery2007;106:36–44.
22.MaiJK,AssheuerJ,PaxinosG.Atlasofthehumanbrain.SanDiego:Academic Press;1998.
23.TalairachJ,TournouxP.Co-planarstereotaxicatlasofthehumanbrain:3- dimensionalproportionalsystem:anapproachtocerebralimaging.NewYork:
ThiemeMedical;1988.
24.MiocinovicS,LempkaSF,RussoGS,MaksCB,ButsonCR,SakaieKE,etal.
Experimentalandtheoreticalcharacterizationofthevoltagedistributiongen- eratedbydeepbrainstimulation.ExperimentalNeurology2009;216:166–76.
25.WalckiersG,FuchsB,ThiranJ-P,MosigJR,PolloC.Influenceoftheimplanted pulsegeneratorasreferenceelectrodeinfiniteelementmodelofmonopolar deepbrainstimulation.JournalofNeuroscienceMethods2010;186:90–6.
26.LachauxJP,RudraufD,KahaneP.IntracranialEEGandhumanbrainmapping.
JournalofPhysiology(Paris)2003;97:613–28.
27.LehmannTN,GabrielS,KovacsR,EilersA,KiviA,SchulzeK,etal.Alterationsof neuronalconnectivityinareaCA1ofhippocampalslicesfromtemporallobe epilepsy patients and from pilocarpine-treated epileptic rats. Epilepsia 2000;41(Suppl.6):S190–4.
28.RaedtR,VanDyckeA,VonckK,BoonP.Celltherapyinmodelsfortemporallobe epilepsy.Seizure2007;16:565–78.
29.CohenI,NavarroV,ClemenceauS,BaulacM,MilesR.Temporallobeepilepsyin vitroontheoriginofinterictalactivityinhuman.Science2002;298:1418–21.
30.Stafstrom CE.The roleof thesubiculum inepilepsy andepileptogenesis.
EpilepsyCurrents2005;5:121–9.
31.FaboD,MagloczkyZ,WittnerL,PekA,ErossL,CzirjakS,etal.Propertiesofin vivo interictalspikegeneration inthehumansubiculum.Brain2008;131:
485–99.
32.KnoppA,FrahmC,FidzinskiP,WitteOW,BehrJ.LossofGABAergicneuronsin thesubiculumanditsfunctionalimplicationsintemporallobeepilepsy.Brain 2008;131:1516–27.
33.HuberfeldG,MenendezdelaPridaL,PalludJ,CohenI,LeVanQuyenM,AdamC, etal.Glutamatergicpre-ictaldischargesemergeatthetransitiontoseizurein humanepilepsy.NatureNeuroscience2011;14:627–34.
34.WoznyC,KiviA,LehmannTN,DehnickeC,HeinemannU,BehrJ.Commenton
‘Ontheoriginofinterictalactivityinhumantemporallobeepilepsyinvitro’.
Science2003;301:463c.
35.FritschyJ-M.LeroˆledelaneurotransmissionGABAergiquedansl’e´pilepsie me´sialedulobetemporal.Epileptologie2003;20:2–7.
P.Bondallazetal./Seizure22(2013)390–395 395
