Available online at www.sciencedirect.com

www.elsevier.com/locate/yebeh

Epilepsy & Behavior 11 (2007) 460–465

Case Report

Impact of interictal epileptic activity on normal brain functionin epileptic encephalopathy: An electroencephalography–functional

magnetic resonance imaging study

X. De Tiege a,*, S. Harrison b, H. Laufs c,d, S.G. Boyd f, C.A. Clark e, P. Allen g,B.G. Neville a, F. Vargha-Khadem b, J.H. Cross a

a Neuroscience Unit, UCL Institute of Child Health, London, UKb Developmental Cognitive Neuroscience Unit, UCL Institute of Child Health, London, UK

c National Society for Epilepsy, London, UKd Department of Clinical and Experimental Epilepsy, UCL Institute of Neurology, Queen Square, London, UK

e Radiology and Physics Unit, UCL Institute of Child Health, London, UKf Department of Clinical Neurophysiology, Great Ormond Street Hospital for Children NHS Trust, London, UK

g Department of Clinical Neurophysiology, UCL Institute of Neurology, Queen Square, London, UK

Received 9 March 2007; revised 31 May 2007; accepted 2 June 2007Available online 14 September 2007

Abstract

Using electroencephalography (EEG) in combination with functional magnetic resonance imaging (fMRI), we studied a 9.5-year-oldgirl who developed cognitive and behavioral regression in association with intense interictal bilaterally synchronous epileptic discharges(IBSEDs) both during the awake state and during sleep. During runs of IBSEDs, EEG–fMRI demonstrated deactivations in the lateraland medial frontoparietal cortices, posterior cingulate gyrus, and cerebellum together with focal relative activations in the right frontal,parietal, and temporal cortices. The deactivations probably reflect the repercussion of the interictal epileptic activity on normal brainfunction which might cause the neuropsychological regression by inducing repetitive interruptions of neurophysiological function result-ing in a chronic state of specific psychomotor impairment. The relative activations could possibly indicate the source of epileptic activityrapidly spreading to other brain regions.� 2007 Elsevier Inc. All rights reserved.

Keywords: Epilepsy; Children; Epileptic encephalopathy; Cognition; Behavior; Electroencephalography–functional magnetic resonance imaging

1. Introduction

The combination of electroencephalography (EEG) withfunctional magnetic resonance imaging (fMRI) allows non-invasive mapping of brain regions that show changes infMRI signal in response to specific events recorded byEEG [1]. This technique is therefore particularly useful instudying the repercussion of epileptic abnormalities on nor-mal brain function [2,3].

1525-5050/$ - see front matter � 2007 Elsevier Inc. All rights reserved.

doi:10.1016/j.yebeh.2007.06.001

* Corresponding author.E-mail address: xdetiege@ulb.ac.be (X. De Tiege).

The epileptic encephalopathies are age-related condi-tions observed in childhood, where the adverse conse-quences of epilepsy on cognition and behavior arepresumed to be secondary to epileptic discharges and,therefore, potentially reversible with the control of epilepticactivity [4]. The pathophysiological mechanisms by whichintense epileptic activity causes developmental regressionremain a major issue that is still poorly understood [4].

We used EEG–fMRI to examine a girl who developedan epileptic encephalopathy in association with intenseinterictal epileptic activity to determine the potential path-ophysiological link between epileptic activity and develop-mental regression.

Case Report / Epilepsy & Behavior 11 (2007) 460–465 461

2. Case report

A right-handed girl with previously unremarkable med-ical history developed episodic left-sided seizures at age2.5 years. EEG revealed independent right frontal andparietal spike–wave discharges. Cerebral MRI revealedsmall lesions in the left periventricular white matter andfocal subcortical white matter changes in the right parietalregion suggestive of previous hypoxic–ischemic injury.Valproate (VPA) treatment was initiated. At age 5,because of seizure recurrence, VPA was changed to car-bamazepine (CBZ, 20 mg/kg/day), which induced seizureremission. Subsequently, her family and teachersdescribed her as a pleasant happy girl without any devel-opmental difficulty (no baseline neuropsychological evalu-ation available). At age 8, she exhibited clear regression inbehavior and school performance (unreasonable, easilyprovoked, inappropriate language, impulsive, temper tan-trums) not related to antiepileptic drug change. Episodesof behavioral arrest and atonic attacks were alsoobserved. At that time, awake EEG revealed very fre-quent bilaterally synchronous spike–wave discharges(BSSWDs). At age 9, the clinical seizures ceased withthe addition of topiramate (TPM, 4 mg/kg/day) to CBZ,

Fig. 1. Sample of the video/EEG telemetry performed at age 9.5 years showingduring sleep that fulfills the diagnostic criterion of continuous spike-and-wave

although behavioral and cognitive difficulties persisted.Two months later, TPM was replaced by clobazam(CLB, 10 mg/day) because of concern about a potentialrole of TPM in the persistence of cognitive and behavioralproblems. Under CBZ and CLB, the patient remained sei-zure free, but no cognitive or behavioral improvementwas noted. At age 9.5 years, 24-hour video-telemetryshowed very frequent interictal BSSWDs (spike index(SI) = 50–75%) during the awake state and subcontinuousBSSWDs during sleep (SI = 85–90%) (see Fig. 1). Interic-tal epileptic activity is here defined as that not accompa-nied by any behavioral change detected by standardmethods of clinical observation [5,6]. At that time, neuro-psychological evaluation and EEG–fMRI were performedwhile the patient was still under CLB and CBZ treatment.CBZ was then replaced by corticosteroids.

3. Methods

3.1. Neuropsychological assessment

Intelligence, memory, executive function, and behavioral status wereassessed using standardized neuropsychological tests and parental ratings(Table 1).

the presence of intense interictal bilaterally synchronous epileptic activitydischarges during slow sleep (spike index > 85%).

Table 1Results of the neuropsychological tests performed at age 9.5 years when the patient had been clinically seizure free for 6 months but had frequent interictalsecondarily generalized spike–wave discharges

Test/subtest Standard score Percentile rank

Wechsler Intelligence Scale for Children-IV (intellectual function, index score mean = 100, SD = 15)Full Scale IQ 74 4Verbal Comprehension Index 81 10Perceptual Reasoning Index 79 8Working Memory Index 77 6Processing Speed Index 80 9

Children’s Memory Scale (memory, index score mean = 100, SD = 15)Verbal Immediate Index 51 0.1Verbal Delayed Index 54 0.1Visual Immediate Index 100 50Visual Delayed Index 78 7

Behavioral Assessment of the Dysexecutive Syndrome for Children (executive function, subtest score mean = 10, SD = 3)Playing Cards Test (inhibition/set shifting) 1 <0.2Water Test (initiation/novel action planning) 8 20–31Key Search Test (planning/organization/self-monitoring) 6 7–12Zoo Map Test 1 (planning with minimal structure) 16 >96Zoo Map Test 2 (planning within rule governed task) 11 57–69Six Part Test (planning, task scheduling, monitoring) 10 45–56

Parental rating Results

Behavior Rating Inventory of Executive Function (executive function)Inhibition Clinically abnormalShifting Clinically abnormalEmotional control Clinically abnormalInitiation Clinically abnormalWorking Memory Clinically abnormalPlanning/Organization Clinically abnormalOrganization of Materials NormalMonitoring Clinically abnormal

Strengths and Difficulties Questionnaire (behavior)Emotional symptoms Clinically abnormalConduct problems NormalHyperactivity BorderlinePeer problems BorderlineProsocial behavior Clinically abnormal

462 Case Report / Epilepsy & Behavior 11 (2007) 460–465

3.2. EEG–fMRI data acquisition

MRI investigations were performed with a 1.5-T Avanto MRI System(Siemens, Erlangen, Germany). An anatomical scan was acquired for thelocalization of activated voxels (three-dimensional T1-weighted fast low-angle shot, TR: 11 ms, TE: 4.94 ms, flip angle: 15�, field of view (FoV):256 · 256 mm2, 1-mm3 isotropic voxels, 176 slices/volume). Whole-braintwo-dimensional echo planar imaging (EPI) fMRI data were acquired dur-ing eyes-closed rest in two 9-minute sessions (TR: 3080 ms, TE: 50 ms,FoV: 192 · 192 mm2, 3-mm3 isotropic voxels, 36 interleaved axial slices).

Electroencephalograms were recorded using 12 gold disk electrodesapplied according to the 10–20 international system [7]. All equipment(in-house) was MRI compatible [7]. Image and pulse artifact subtractionwas used to recover the underlying electroencephalogram [8].

The patient was awake throughout the EEG–fMRI.

3.3. Data analysis

Images were slice-time corrected, realigned, normalized, and spatiallysmoothed to two times the original voxel size using SPM5 (http://www.fil.ion.ucl.ac.uk/spm). Individual (onset) and bursts of (onset and offsetdefining boxcars) BSSWDs were visually identified, convolved with thecanonical hemodynamic response function (HRF), and used as separateregressors for an event-related fMRI analysis (SPM5). Based on previous

EEG–fMRI studies, we expected widespread cortical blood oxygen level-dependent (BOLD) signal decrease associated with BSSWDs [7,9,10]. Wetherefore estimated the same model with global scaling [11]. The assump-tion guiding this analysis was that a regional increase in BOLD signallinked to the epileptic activity could have been masked by a global signaldecrease associated with BSSWDs and that brain regions accurately exhib-iting signal increase during epileptic activity could have not been identifiedif global scaling was omitted [11]. Motion parameters obtained fromrealignment were always included as confounds.

4. Results

4.1. Neuropsychological assessment

Results are summarized in Table 1. Overall intelligencewas about 2 SD below normal without a significant differ-ence between verbal and nonverbal abilities. Immediateand delayed verbal memory (auditory stimuli) was morethan 3 SD below normal, whereas nonverbal memory (visualstimuli) was in the normal range in the immediate conditionand about 1.5 SD below normal in the delayed condition.Performance was restricted in some domains of executive

Case Report / Epilepsy & Behavior 11 (2007) 460–465 463

function (inhibition/set shifting, planning/organization).Behavioral ratings were clinically abnormal for executivefunction, emotion regulation, and prosocial behavior.

4.2. EEG–fMRI results

Sixty-three individual BSSWDs and 52 bursts of interic-tal BSSWDs (duration: 2.5–30 seconds) were recorded dur-

Fig. 2. (A) Sample of the electroencephalogram acquired during fMRI showingof discharges. (B) Widespread decrease in BOLD signal (blue) displayed onprefrontal cortex (anterior frontal, orbitofrontal, dorsolateral, and ventrolatercortex. (C) Decreases in BOLD signal (blue) in the medial frontal cortex, thretrosplenial cortices) displayed on a left sagittal paramedian slice of the patientafter global scaling in the right superior frontal (left), postcentral (left and rig

ing fMRI (spike index = 40%). Individual events were notassociated with significant changes in BOLD signal. With-out global scaling and during bursts of interictal BSSWDs,significant decreases in BOLD signal were found exclu-sively in the prefrontal cortex, posterior cingulate gyrus,precuneus, and cerebellum bilaterally and in the right pari-etal lobe (pcorrected < 0.05) (Figs. 2A–C). According to ourhypothesis, the model incorporating global scaling revealed

an individual bilaterally synchronous epileptic discharge followed by a runthe patient’s brain surface rendering. Deactivations were observed in theal cortices) and the cerebellum bilaterally and in the right lateral parietale precuneus, and the posterior cingulate gyrus (posterior cingulate and

’s individual brain. (D) Relative focal increases in BOLD signal (red) foundht), and superior temporal gyri (right).

464 Case Report / Epilepsy & Behavior 11 (2007) 460–465

additional focal relative signal increases in the rightsuperior frontal, postcentral, and superior temporal gyri(pcorrected < 0.05) (Fig. 2D).

5. Discussion

This typical case of epileptic encephalopathy belongingto the spectrum of epilepsies with continuous spikes andwaves during sleep (CSWS) provides specific localization-related insight into how intense interictal epileptic activitymight cause developmental regression in this type of epilep-tic disorder. Altered brain function in the regions identifiedusing EEG–fMRI could explain the cognitive and behav-ioral deterioration observed in this child.

Using EEG–fMRI, we demonstrated a decrease inBOLD signal in the prefrontal (anterior frontal, dorsolat-eral, ventrolateral, and medial frontal) and parietal (parie-tal lobules and precuneus) cortices, in the posteriorcingulate gyrus (posterior cingulate and retrosplenial corti-ces), and in the cerebellum during bursts of BSSWDs com-pared with vigilant rest. This decrease in BOLD signalcould reflect the occurrence of an epilepsy-induced changein the physiological function of these brain areas either bydecreased synaptic activity or low energy-requiring GABA-mediated inhibition [2], which might render them unavail-able for normal cognitive processes or goal-directed behav-ior [12].

These brain areas are parts of the neuronal networksinvolved in working and delayed memory [13–15] andare parts of thalamocortical and hippocampocortical net-works involved in plasticity and consolidation of memorytraces during sleep [16]. The prefrontal cortex also plays akey role in the neuronal networks sustaining executivefunction, including inhibition control and set shifting[17]. On the basis of these data, the pattern of behavioraland cognitive impairment observed in this patient mightbe related to a mechanism of epilepsy-induced repetitiveimpairment of these neuronal networks that may resultin a chronic state of specific psychomotor impairment inwhich sleep disturbance may play a major role [4,5].Moreover, these bilateral changes could prevent plasticityor compensatory mechanisms, facilitating the appearanceof neuropsychological deficits in the developing brain[18]. Finally, the contribution of cerebellar dysfunctionto the observed neuropsychological deficits is less clear,as its involvement in higher cognitive functions is still amatter of debate.

Previous EEG–fMRI studies have already demonstratednegative frontoparietal changes in BOLD signal duringgeneralized or secondarily generalized spike–wave dis-charges [7,9,10]. These changes have been associated withabsence seizures, but have also been reported in patientswithout obvious clinical manifestations [7,9,10]. Neverthe-less, the potential association between these changes in nor-mal brain function and the occurrence of specificneuropsychological impairments that can be associatedwith this type of interictal epileptic activity, particularly

in the developing brain [5,6], has not been assessed in thesestudies. Through the case described here, we believe we arethe first to tackle this issue.

With the use of global scaling, we found significantfocal relative activations in the right superior frontal,postcentral, and superior temporal gyri. These relativeactivations could possibly indicate the focal onset of epi-leptic activity rapidly spreading to the rest of the brain viacorticocortical and/or thalamocortical connections. Thishypothesis is supported by previous positron emissiontomography (PET) studies performed using [18F]fluorode-oxyglucose (FDG) in children with CSWS, which havedemonstrated a relative increase in metabolism at the siteof the epileptic focus [19]. Interestingly, Hamandi et al.found a left frontal activation associated with widespreadfrontoparietal deactivation in one patient with secondarilygeneralized spike–wave discharges starting from the leftfrontal lobe [7]. These data suggest that EEG–fMRI couldpossibly play a role in the noninvasive presurgical evalu-ation of children with epileptic encephalopathy by identi-fying epileptogenic foci, but further studies are clearlyrequired.

Interestingly, the association between parietal hyperme-tabolism and prefrontal hypometabolism has previouslybeen described, using FDG–PET, in a group of childrenwith CSWS [19]. In that study, the decrease in glucosemetabolism in the frontal cortex was interpreted as a phe-nomenon of remote inhibition induced by altered connec-tivity with highly epileptogenic and hypermetabolicparietal cortex [19].

Taken together, these data suggest that the neuropsy-chological impairment observed in epileptic encephalopa-thy could be related to specific cortical dysfunctionsecondary to the spread of the epileptic activity from focalhypermetabolic foci [19]. Further studies are needed to testwhether this impaired cortical function results from activeinhibition or functional disconnection during epilepticdischarges.

Acknowledgments

The authors thank Danny Flanagan for his role in thedevelopment of the EEG–fMRI technique at GreatOrmond Street Hospital for Children. Xavier De Tiege isfunded by the European Federation of Neurological Soci-ety Fellowship 2005, the Fonds National de la RechercheScientifique (FNRS, Belgium), and the M.E. Horlait-Dap-sens Foundation (Belgium). Helmut Laufs is funded by theDeutsche Forschungsgemeinschaft (LA 1452/3-1).

References

[1] Debener S, Ullsperger M, Siegel M, Engel AK. Single-trial EEG–fMRI reveals the dynamics of cognitive function. Trends Cogn Sci2006;10:558–63.

[2] Gotman J, Kobayashi E, Bagshaw AP, Benar CG, Dubeau F.Combining EEG and fMRI: a multimodal tool for epilepsy research.J Magn Reson Imaging 2006;23:906–20.

Case Report / Epilepsy & Behavior 11 (2007) 460–465 465

[3] Hamandi K, Salek-Haddadi A, Fish DR, Lemieux L. EEG/functional MRI in epilepsy: the Queen Square Experience. J ClinNeurophysiol 2004;21:241–8.

[4] Holmes GL, Lenck-Santini PP. Role of interictal epileptiform abnor-malities in cognitive impairment. Epilepsy Behav 2006;8:504–15.

[5] Aldenkamp AP, Arends J. Effects of epileptiform EEG discharges oncognitive function: is the concept of ‘‘transient cognitive impairment’’still valid? Epilepsy Behav 2004;5(Suppl. 1):S25–34.

[6] Binnie CD. Cognitive impairment during epileptiform discharges: is itever justifiable to treat the EEG? Lancet Neurol 2003;2:725–30.

[7] Hamandi K, Salek-Haddadi A, Laufs H, et al. EEG–fMRI ofidiopathic and secondarily generalized epilepsies. NeuroImage2006;31:1700–10.

[8] Allen PJ, Polizzi G, Krakow K, Fish DR, Lemieux L. Identificationof EEG events in the MR scanner: the problem of pulse artifact and amethod for its subtraction. NeuroImage 1998;8:229–39.

[9] Gotman J, Grova C, Bagshaw A, Kobayashi E, Aghakhani Y,Dubeau F. Generalized epileptic discharges show thalamocorticalactivation and suspension of the default state of the brain. Proc NatlAcad Sci USA 2005;102:15236–40.

[10] Laufs H, Lengler U, Hamandi K, Kleinschmidt A, Krakow K.Linking generalized spike-and-wave discharges and resting state brainactivity by using EEG/fMRI in a patient with absence seizures.Epilepsia 2006;47:444–8.

[11] Laurienti PJ. Deactivations, global signal, and the default mode ofbrain function. J Cogn Neurosci 2004;16:1481–3.

[12] Tassinari CA, Rubboli G. Cognition and paroxysmal EEG activities:from a single spike to electrical status epilepticus during sleep.Epilepsia 2006;47(Suppl. 2):40–3.

[13] Naghavi HR, Nyberg L. Common fronto-parietal activity in atten-tion, memory, and consciousness: shared demands on integration?Conscious Cogn 2005;14:390–425.

[14] Cavanna AE, Trimble MR. The precuneus: a review of its functionalanatomy and behavioural correlates. Brain 2006;129:564–83.

[15] Vogt BA, Vogt L, Laureys S. Cytology and functionally correlatedcircuits of human posterior cingulate areas. NeuroImage2006;29:452–66.

[16] Steriade M, Timofeev I. Neuronal plasticity in thalamocorticalnetworks during sleep and waking oscillations. Neuron2003;37:563–76.

[17] Stuss D, Knight R. Prefrontal cortex: the present and the future. In:Stuss D, Knight R, editors. Principles of frontal lobe func-tion. Oxford: Oxford Univ. Press; 2002. p. 573–97.

[18] Noppeney U, Friston KJ, Price CJ. Degenerate neuronal systemssustaining cognitive functions. J Anat 2004;205:433–42.

[19] De Tiege X, Goldman S, Laureys S, et al. Regional cerebral glucosemetabolism in epilepsies with continuous spikes and waves duringsleep. Neurology 2004;63:853–7.