DeepDeep BrainBrain StimulationStimulation ff oror EpilepsyEpilepsy

Kyung Jin Lee

St. Mary’s Hospital The Catholic University of Kore a Introduction EpilepsyEpilepsy

! Common chronic neurological disorder that affects 0.5-1% of the population.

! Medically refractory is associated with injury and death, psychosocial dysfunction, and significant cognitive impairment.

! Surgical resection, which is associated with long-term sei zure freedom in 60-80% of patients.

! After temporal lobectomy, the recurrence rate of seizures after a long period of seizure freedom is as high as 15%

! In spite of improvements in surgical technique, 4% of pat ients will suffer death or permanent neurological disabilit y. ! Up to 40% of patients refractory to antiepilepti c medications are not candidates for surgical re section.

! There are unsuitable cases for surgery because of the involvement of the eloquent areas or bila teral or multifocal nature of the ictal onset.

! So, Alternative therapeutic methods are introd uced, such as modulation of cortical excitability through chronic electrical neurostimulation DeepDeep brainbrain StimulationStimulation

! Deep brain stimulation for movement disorder s has meet with widespread success.

! Deep brain stimulation (DBS) has been succes sfully applied for the therapy of several neurol ogical disorders

! Due to the reversibility, programmability, and low risk of complication.

! Recently, stimulation of the nervous system h as been performed in an attempt to treat epile psy. DeepDeep BrainBrain StimulationStimulation (DBS)(DBS)

Established for: ! Parkinson’s Disease (STN, GPi) ! Dystonia (GPi) ! Tremor (Vim )

Emerging treatment for: ! Chronic Pain (PAG/PVG, Thalamus) ! Obsessive compulsive disorder ! Depression (Area 25) ! Epilepsy (Anterior Thalamus) ! Tourette’s Syndrome (CM Thalamus) DeepDeep brainbrain stimulationstimulation

! The exact mechanism of action of DBS in r educion seizure activity is unknown.

- DBS may interfere with synchronized oscillations by release.

- Stimulation-induced modulation of pathological neural networks. 1) Targeting of the CNS structures believed to have a “gating” role in the epileptogenic network, such as the STN or thalamus .

2) Targeting of the ictal onset zone with the theory that stimulation may lead to interference with seizure initiation. AnimalAnimal StudiesStudies

! Numerous animal models have been used to eluci date mechanism of DBS action and its potential us efulness in the treatment of epilepsy.

! Models of Pentylenetetrazol, kainic acid, bicuculli ne, picrotoxin, and kindling to induce seizures.

! To identify new targets, to achieve a better under standing of the mechanism of action in accepted t argets. StimulationStimulation forfor EpilepsyEpilepsy TargetsTargets

! Cerebellum (Cooper IS, 1973; Velasco F, 2005) ! Locus ceruleus (Farber J, 1983) ! Centromedian (CM) thalamic (Velasco F, 1987, 2006; Fisher RS, 1992) ! Anterior thalamus (Mirski MA, 1994 ; Lim et al, 2007) ! (Veliskova J, 1996; Vercueil L, 1998; Hand forth et al, 2006) ! Vagus nerve (approval by FDA in 1997) ! Amygdala (quenching, Weiss, 1995) ! Hippocampus (Velasco F, 2000 ; Vonek et, 2005) ! Trigeminal nerve, caudate nucleus, SubthalamicSubthalamic nucleusnucleus

! STN DBS for the treatment of Parkinson’s disease makes the STN an appealing target.

! The substantia nigra pars reticulata (SNr) plays a role in seizure propagation through its GABAergic nigrotectal progections to the .

! Inhibition of GABAergic SNr results in su ppression of partial and generalized epileptic seiz ures in a rat model of epilepsy. (Iadarola MJ, 198 2) . STN stimulation for epilepsy treatment : Basis from some experimental evidence of the subcortical network that influences cortical excitability- nigral control of epilepsy system.

AnteriorAnterior nucleusnucleus ofof thethe thalamusthalamus

! In 1937, James W. Papez26 described a circuit linking hippocam pal output via the fornix and mammillary nuclei to the ANT. ! The ANT is an appealing target because of its relatively small si ze and projection to limbic structures, ultimately affection wide regions of neocortex. AnimalAnimal studiesstudies

! Observations of enhanced glucose metabolism in the ANT after administration of both pentylenetetrazole and ethosuximide to guinea pigs supports the role of the ANT in propagating seizure activity. - basis from some experimental evidence by the finding of increased metabolic activity in ATN during seizure and high frequency stimulation of ATN.

! Bilateral ANT DBS may also prolong the latency of onset of status epilepticus after administration of pilocarpine (Mirski, MA, 1984). ATNATN stimulationstimulation inin RatRat ((MirskiMirski,, 1997)1997)

! HFS (50-100 Hz) raised clonic sz threshold in rats with PT Z-induced seizures

! LFS (8Hz) - proconvulsant

! HFS believed to be inhibitory (effects mimicked by injectio n of GABA agonist muscimol)

! Mirski and Fisher postulate disruption of corticothalamic tr ansmission with ATN stimulation, preventing normal recrui tment and synchronization into a generalized seizure RationaleRationale ofof DBSDBS ofof ATNATN

! The thalamus was chosen for these studies because stimulation of a relatively small anatomic region can influence physiologic activity in more widespread areas of cortex.

! The anterior nucleus of the thalamus projects largely to the cingulate gyrus, and via the cingulate gyrus, to limbic structure and wide regions of neocortex.

! Total 19 patients

STN 3 patients ATN 16 patients

! Removal of electrode in 2 patients : Due to infection PatientPatient selectionselection

! Partial onset seizures with or without secondary ge neralization associated with frequent falls, injuries and impaired quality of life

! Seizures should be refractory to 12-18 months of at least two therapeutically dosed antiepileptic agents

! Failure to respond to surgical resection or vagal ner ve stimulation may also be indications for DBS

! Scalp video-EEG monitoring is necessary to characte ruze seizures and demonstrate bilateral or nonlocali zing findings consistent with preoperative imaging 1. Surgical precedure - MRI guided ATN or STN target with micro re cording - Intraoperative EEG recording with or witho ut stimulation

2. Postoperative management - Post op CT or MRI check - Follow up EEG DirectDirect VisualVisual TargetingTargeting inin MRIMRI Computer superimposition of the schaltenbrand Atlas on a stereotactic MRI with indication of the position in stereotact ic space of AN (ant thalamic nucleus) PostopPostop CTCT MicroMicro recordingrecording

PreoperativePreoperative EEGEEG IntraoperativeIntraoperative EEGEEG (before(before insertion)insertion)

FP1-F3

F3-C3

C3-P3

FP2-F4

F4-C4

C4-P4

T7-C3

C3-CZ

CZ-C4

C4-T8 IntraoperativeIntraoperative EEGEEG (after(after insertion)insertion) FP1-F3

F3-C3

FP2-F4

F4-C4

C4-P4

T7-C3

C3-CZ

CZ-C4

C4-T8

R STN1

R STN2

R STN3 Postop. 4 weeks

Postop. 26 weeks

Postop. 1 year PostoperativePostoperative MRIMRI ResultResult

Seizure frequency

! Preoperative : 120.3 ± 304.1 / month

! Postoperative : 27.6 ± 61.2 / month

! Mean seizure reduction rate(%) : 70.4 ± 33.03 % Preoperative sei Postoperative seMean seizure Total F/U Seizure i Case Age Sex Seizure type zure frequency / izure frequency reduction (month) ntensity month /month (%)

1 28 F Both FTLE, cryptogenic 55 15 1 93 0

2 14 F Multifocal epilepsy 47 1200 230 80 0 Multifocal epilepsy (P-O dominant), 3 23 F 42 5 1 80 1 cryptogenic B FLE, symptomatic / prev meningi 4 49 M 22 100 2 98 2 oma resection (frontal) L hemispheric epilepsy, 5 54 F 21 35 1 97 0 R HS & hemispheric atrophy 6 37 F B C-PLE, cryptogenic 20 30 15 50 1

7 25 F B FLE, cryptogenic 19 16 2 93 0

8 41 M B C-PLE, cryptogenic 18 2 1 50 1 P-OLE, symptomatic 9 30 M 18 20 4 75 0 (B perisylvian PMG) Multifocal epilepsy, B schizencepha 10 42 M 16 5 2 60 0 ly 11 14 M Both FLE 12 95 95 0 1 B FLE, cryptogenic, previous cranie 12 23 M 51 12 0 100 2 ctomy (medulloblastoma) 13 35 M B TLE, symptomatic (B HS) 18 20 1 95 1

14 18 M B C-PLE, cryptogenic 16 30 30 0 0

15 25 M Rt. T-P LE, Cryptogenic 30 220 30 86 0 • Seizure intensity 0 : no interval change 1 : mild improved ( < 30% intensity compared with preoperative state) 2 : much improved ( > 30% intensity compared with preoperative state)

[ Neurotherapeutics : The Journal of the American Society for Experimental NeuroTherapeutics ]

Changes in regional cerebral blood flo w by deep brain stimulation on bilateral anterior thalamic nuclei in patients with intractable epilepsy ObjectivesObjectives

! There has been no information about the function ally and anatomically correlated metabolic or he modynamic changes after DBS on the ATN from r efractory epilepsy patients.

! To address this issue, functional neuro-imaging with SPECT and group ananalysis by statistical pa rametric mapping (SPM2) were used. ! 13 patients were serially performed 99m to measure rCBF changes with Tc- ECD SPECT scannings Dx of patients # of pts Improvement

Without major anaotomic lesion or prev 8 64.3% ious surgery (NMA group)

With major anaotomic lesion or previou 5 88.8% s surgery (WMA group)

Average 73.7%

• SPM analysis was performed in 8 patients (NMA) - Good Px group (sz reduction > 64.3%): 5 pts - Poor Px group (sz reduction < 64.3%): 3 pts ImagingImaging AcquisitionAcquisition

! SPECT image acquisition - IV injection of 99mTc-ECD (740-925 MBq ) was followed by s canning after 20-30min with multi-detector scanner (ECAM plu s; Siemens, Erlangen, Germany) - 128x128x64 voxel " iso-voxel transformation - At least two images were taken (pre-insertion state & F/U sca nning after at least 3 months (inter-scanning interval: 4.7 ±5.8 M) - At the follow-up (postDBS scanning), all patients were under conditions of continous mode of stimulation - At least 12 hour of sz free period prior to SPECT acquisition

! MRI acquisition - 1.5T Signa excite (GE) - SPGR image with 1.4mm thickness without gap, contiuous 12 8 images, TR 9.7msec, TE, matrix 256°ø256, NEX 1 SPMSPM analysisanalysis SPMSPM analysisanalysis parametersparameters

! Spatial normalization procedure - bilinear interpolation by a 12-parameter linear affine transfor mation - nonlinear 3-dimensional deformation to match scan to a gene ric MNI SPECT template. - A resulting voxel size of 2x2x2 mm was used. ! Smoothing - with an isotropic 14-mm kernel. ! Compare-populations - Whole group (NMA): 2 conditions, 1 scan/condition (paired t- test) - Good vs bad group in NMA group: 2 groups: condition & cova riates - Proportional scaling - Threshold : p<0.001 (uncorrected, cluster and voxel levels) - Voxel threshold: > 10 voxels Results TotalTotal 88 ptspts ((postDBSpostDBS –– preDBSpreDBS))

Regions Hyper o MNI coor Z scor r hypo- dinates e perfusio (x,y,z) n

R IFG (BA9) hyper 36,6,26 4.34 L caudate hyper -18,12,12 4.10 L lentiform Nuc hyper -22, 15,2 3.50

R pons hyper 12, -30, - 4.09 34 L ant. cbll, culm hyper -8,-36,-1 3.81 en 4 L IPL (BA40) hyper -36,-34,2 3.55 8 R SFG (preSMA, hypo 22,26,64 4.36 BA6) L SFG (BA8) hypo -26,44,46 3.94 L precentral G hypo -60,-10,4 3.80 (BA6) 4 R thalamus (bot hypo 10,-8,14 3.59 h ANS) GoodGood 55 ptspts ((postDBSpostDBS –– preDBSpreDBS))

Regions Hyper or MNI coor Z score hypo-perf dinates usion (x,y,z) R MTG & ITG (BA2 hyper 54,-40,-6 4.39 0) 54,-24,-2 0 L fusiform G (BA3 hyper -36,-50,- 4.01 7) 10 L temporal WM hyper -36, -54, 3.64 2 R temporal WM hyper 70,-50,0 3.85

L pons hyper -6, -24, - 3.95 22 L ant. cbll, culmen hyper -2,-30,-1 3.88 8 R precuneus (BA1 hyper 32,-62,38 3.73 9) L IFG (BA11) hypo -36,54,-1 3.80 2 PoorPoor 33 ptspts ((postDBSpostDBS –– preDBSpreDBS))

Regions Hyper or h MNI coordin Z score ypo-perfus ates (x,y,z) ion

R MFG (BA6) hypo 14,12,60 4.44 R SFG (preSMA, hypo 18,16,50 4.00 BA6) R SFG (preSMA, hypo 20,18,66 3.97 BA6) R cingulate G hypo 2,14,36 4.28 (BA32) L IFG (BA47) hypo -44,28,-22 3.66 ResultsResults (1)(1)

! Regions of Increased CBF after chronic ATN DBS (All patients)

- Hyperperfusion R inf frontal G, pons L BG, ant cerebellum, inf parietal lobule

- Hypoperfusion Both superior frontal G L precentral G R ant thalamus ResultsResults (2)(2)

! rCBF after chronic ATN DBS: Comparison b/w two groups

- Good Px Prominent hyperperfusion Both temporal, & R parietal minimal hypoperfusion

- Poor Px Prominent hypoperfusion R SFG, MFG & cingulate G, L IFG No significant hyperperfusion DiscussionDiscussion

! The changed rCBF areas after ATN DBS may - Reflect the location of functional improvement by long term ATN DBS

- Be anatomically connected regions to ATN (ACC, insula, temporal or parietal operculum) or indirectly, but functionally reciprocated: cerebellum, pons, etc DiscussionDiscussion # Limitations: Very heterogeneous population ! Despite of a paired-comparison of image data from the same patients, •ƒrCBF may vary with the group’s characteristics (clinical Dx, ther apeutic responses, etc.) " need a well-controlled study ! Too small population group prevents the hasty generalization of t he results ! Weak causal-relationship of rCBF changes and seizure reduction - Effect of DBS itself? - Improved neural function by reduced sz number? - Other -neuronal modulation by adjacent, connected neural networks?

# Prospect for further studies - Setting the prognostic factor of DBS therapy for intractable epilepsy pati ents - In our cases, Frontal & temporal lobe epilepsy were more good results than multifocal epilepsy but questionable. °Ó B TLE might be more plausible candidates for ATN DBS? SANTE (Stimulation of the Anterior Nucleus of the Thalamus in Epilepsy) trial

- The study collected data from 110 patients fro m 17 U.S. centers who were implanted with a DBS system and were monitored for a minimu m of 13 months following implant. SANTESANTE SelectedSelected InclusionInclusion CriteraCritera

! Age 18-65

! Partial or secondarily generalized seizures > 6/month

! Refractory to > 3 AEDs, currently taking 1-4 drugs

! Not a candidate for potentially curative resective surg ery

! IQ > 70

! If VNS in place, willing to remove it at time of implant ation SANTESANTE

! 110 patients implanted ! The primary objective was met : stimulation reduced seizures

Improvement increased in open phase: ! No stimulation related deaths ! No symptomatic hemorrhages (some seen on im aging) ! Results submitted for publication and for FDA StimulationStimulation parameterparameter

CUMC SANTE

! Stimulation settings ! Stimulation settings Control Stimulation Voltage 2-3.0V Voltage 0V 5V Frequency 130-180Hz Frequency 145pps 145pps Pulse width 120-180 Pulse width 90ms 90ms Polarity AN - , Case + Polarity AN – case+ AN-case+ Continuous mode On 1minute 1minute Off 5minute 5minute ConclusionConclusion

! Our study investigating ATN DBS-induced rCBF change at chronic p eriods showed a possible role of various cortical or subcortical stru ctures in the modulating action of ATN DBS.

! Clinically, it seems to be important that short term outcome of ATN DBS reflects long term outcome directly. The correlation between seizure type, characteristics and anticonvulsant effect of ATN DBS did not showed significantly because of small case numbers.

Therefore Longer-term follow-up with larger group of patients is re quired to fully evaluate the safety and effectiveness of this treatment modality.

! Further studies investigating the influence of different stimulation parameters on rCBF in these structures in humans and animals ma y eventually provide practical guidelines and lead to improved clini cal outcome. ConclusionConclusion

1) Resective epilepsy surgery is not an ideal option f or all patients with medically refractory epilepsy.

2) We need a alternative treatment options such as for medically refractory epilepsy.

3) Deep brain stimulation for medically refractory e pilepsy patient is good option.

4) But need to identify the appropriate patient popu lation for DBS, the optimal target, and the best sti mulation parameters, mode . Thanks for your attention !!