Epilepsy Research (2012) 99, 281—292

jou rn al h om epa ge: www.elsev ier .com/ locate /ep i lepsyres

EEG-LORETA endophenotypes of the commonidiopathic generalized epilepsy syndromes

B. Clemensa,∗, S. Puskásb, M. Besenyeia, M. Emric, G. Opposits c, S.A. Kisc,K. Hollódyd, A. Fogarasie, I. Kondákor f, K. Füle f, K. Benseg, I. Feketeb

a Kenézy Hospital Ltd., Department of Neurology, Debrecen, Hungaryb University of Debrecen, Medical and Health Science Center, Department of Neurology, Debrecen, Hungaryc University of Debrecen, Institute of Nuclear Medicine, Debrecen, Hungaryd University of Pécs, Department of Pediatrics, Pécs, Hungarye Epilepsy Center, Bethesda Children’s Hospital, Budapest, Hungaryf Teaching County Hospital of Bács-Kiskun, Department of Neurology, Kecskemét, Hungaryg Teaching County Hospital of Bács-Kiskun, Department of Child Neurology, Kecskemét, Hungary

Received 5 July 2011; received in revised form 14 November 2011; accepted 11 December 2011Available online 11 January 2012

KEYWORDSEEG;LORETA;Endophenotype;Absence;Juvenile myoclonicepilepsy

SummaryObjective: We tested the hypothesis that the cortical areas with abnormal local EEG synchro-nization are dissimilar in the three common idiopathic generalized epilepsy (IGE) phenotypes:IGE patients with absence seizures (ABS), juvenile myoclonic epilepsy (JME) and epilepsy withgeneralized tonic—clonic seizures exclusively (EGTCS).Patients and methods: Groups of unmedicated ABS, JME and EGTCS patients were investigated.Waking EEG background activity (without any epileptiform potentials) was analyzed by a sourcelocalization method, LORETA (Low Resolution Electromagnetic Tomography). Each patient groupwas compared to a separate, age-matched group of healthy control persons. Voxel-based, nor-malized broad-band (delta, theta, alpha, and beta) and very narrow band (VNB, 1 Hz bandwidth,from 1 to 25 Hz) LORETA activity (=current source density, A/m2) were computed for each per-son. Group comparison included subtraction (average patient data minus average control data)

and group statistics (multiple t-tests, where Bonferroni-corrected p < 0.05 values were acceptedas statistically significant).Results: Statistically not significant main findings were: overall increased delta and theta broad band activity in the ABS and JME groups; decrease of alpha and beta activity in the EGTCSgroup. Statistically significant main findings were as follows. JME group: bilaterally increasedtheta activity in posterior (temporal, parietal, and occipital) cortical areas; bilaterally

∗ Corresponding author at: Kenézy Kórház Ltd., Department of Neurology, Bartók Béla út 3, 4031, Debrecen, Hungary.Tel.: +36 52 511 777; fax: +36 51 511 729.

E-mail address: bela.clemens@yahoo.com (B. Clemens).

0920-1211/$ — see front matter © 2011 Elsevier B.V. All rights reserved.doi:10.1016/j.eplepsyres.2011.12.008

282

increased activity in the medial aactivity in the precuneus, poste21—22 Hz VNBs. ABS group: bilateand medial temporal limbic areacentral gyrus and parts of the rigactivity was found in the frontal cin the right parahippocampal gyruDiscussion: Increased theta activiJME. Increased theta activity in tStatistically not significant findingin JME and ABS.Significance: EEG-LORETA endop

seizs res

I

C(gcetdspcltosttmlemcwai

fotmsfia(GMslttct2

eA2stassAsnAaTaiIai

miaes(ptasdStiitoaSBs

erate absence and myoclonic© 2011 Elsevier B.V. All right

ntroduction

hildhood absence epilepsy (CAE), juvenile absence epilepsyJAE), juvenile myoclonic epilepsy (JME) and epilepsy witheneralized tonic—clonic seizures exclusively (EGTCS) areommon epilepsy syndromes classified as idiopathic gen-ralized epilepsy (IGE). By definition, this term implieshat they are genetically determined nonlesional disor-ers, in which the seizures start in both hemispheresimultaneously (ILAE, 1989). From the point of neuro-hysiology, the ‘‘mild diffuse epileptogenic state’’ of theortex was appointed as the pathological basis of seizureiability, while the generalized seizures were realized asransient, abnormal, bilateral-synchronous thalamo-corticalscillations (Gloor, 1979). Clinical and EEG reports demon-trating focal epileptogenic abnormality or focal onset ofhe so-called generalized seizures were not weighty enougho influence mainstream thinking until 1998 when twoagnetic resonance imaging (MRI) studies disclosed morpho-

ogical cerebral abnormalities in IGE patients (Woermannt al., 1998; Savic et al., 1998). After this breakthrough,odern imaging studies increasingly made the scientific

ommunity to re-conceptualize IGE as a group of disordershere structural abnormalities may contribute to etiology,nd the role of focal abnormalities should be emphasized inctogenesis (Koepp and Duncan, 2004).

Two issues are particularly important within this newramework. First, which part of the cortex is the seizure-nset area and which parts of the brain are involved inhe course of the seizures? Second, what cortical abnor-ality is responsible for ictogenicity, in other words, the

eizure-prone state of the brain in IGE syndromes? Mostunctional imaging studies addressed the first issue lead-ng to increasing knowledge regarding the ictal anatomy ofbsence, myoclonic, and generalized tonic—clonic seizuresBlumenfeld et al., 2003; Holmes et al., 2004, 2010;otman, 2008; Stefan et al., 2009; Carney et al., 2010;oeller et al., 2010; Sakurai et al., 2010). However, most

tudies demonstrated that a few, topographically dissimi-ar cortical sites are involved at seizure onset, in additiono the syndrome-specific, main seizure onset area. Fur-

hermore, the seizures are preceded by electromagnetichanges in topographically non-contiguous areas far beyondhe seizure onset zone (Aarabi et al., 2008; Amor et al.,009). These results indicate that IGE is similar to the focal

iiaa

B. Clemens et al.

nd basal prefrontal area in the 8 Hz VNB; bilaterally decreasedrior cingulate and superior parietal lobule in the 11 Hz andrally increased theta activity emerged in the basal prefrontal

s. Decreased activity was found at 19—21 Hz in the right post-ht superior and medial temporal gyri. EGTCS group: decreasedortex and the postcentral gyrus at 10—11 Hz, increased activitys at 16—18 Hz.ty in the posterior parts of the cortex is the endophenotype forhe fronto-temporal limbic areas is the endophenotype for ABS.s might indicate diffuse biochemical abnormality of the cortex

henotypes may correspond to the selective propensity to gen-ures in the ABS and JME syndromes.erved.

pilepsies insofar as the epileptogenic zone (Lüders andwad, 1991) or epileptic network (Spencer, 2002; Nair et al.,004) that is the anatomical—physiological substrate of theeizure-prone condition is topographically more extendedhan the seizure onset zone. This important matter is alwaysddressed when surgical treatment of focal epilepsy is con-idered (Lüders et al., 2006). However, its neurobiologicalignificance reaches far beyond planning resective surgery.s to the IGE syndromes, the anatomical—physiological sub-trate of ictogenicity remains a neglected issue that hasever been critically revised in the light of recent findings.t best, the results of the imaging studies were interpreteds being potentially related to ictogenicity in some way.his is actually correct because morphological, metabolicnd biochemical abnormalities do not permit any conclusivenference regarding the neuronal substrate of ictogenicity.n other words, morphological, metabolic and biochemicallterations do not automatically indicate that the abnormal-ty contributes to ictogenicity.

On the other hand, abnormal patterns of the electro-agnetic oscillations of the brain are direct markers of

ctal and interictal epileptic malfunctioning (McCormicknd Contreras, 2001; Timofeev and Steriade, 2004). Micro-lectrophysiology detects epileptic malfunctioning at smallpatial scales while EEG and magnetoencephalographyMEG) detect it within great neuronal ensembles. Multi-le pieces of experimental and clinical evidence indicatehat EEG- and MEG-detected local cortical dysfunction andltered cortical connectivity characterize the neuronal sub-trate of ictogenicity albeit the concepts are far from beingefinite (Spencer, 2002; Nair et al., 2004; Timofeev andteriade, 2004). As a corollary, both local cortical dysfunc-ion and remote cortical connectivity should be investigatedn order to understand the neuronal substrate of ictogenic-ty (McCormick and Contreras, 2001; Steriade, 2001). Abouten years ago we suggested that the seizure-prone statef the brain can be grasped by means of EEG backgroundctivity analysis in IGE syndromes (Clemens et al., 2000).ince then, further evidence confirmed this statement.riefly, the main electrophysiological characteristics of theeizure-prone state are increased cortical excitability andncreased EEG synchronization. These traits are interrelated

n IGE models (van Gelder et al., 1983; Kostopoulos, 1986)nd in human IGE. Seizures, increased cortical excitabilitynd increased EEG synchronization in the delta and theta

omS1IfrrbgateErtcoeprait(ttrtd

pyiitctloes(on

L

LddLboEiccmtpwp

EEG-LORETA endophenotypes of IGE

frequency bands coexist in untreated IGE patients (Reutensand Berkovic, 1992; Clemens et al., 2000) while success-ful drug treatment is associated with seizure freedom,decreased cortical excitability and decreased EEG synchro-nization in the delta and theta bands as compared to theuntreated state (Reutens et al., 1993; Clemens et al., 2007b;Clemens, 2008). In the present study we investigated thesources of normal and abnormal EEG background activity inorder to identify cortical areas with abnormal EEG activityin IGE syndromes. This is labeled as ‘‘local cortical dysfunc-tion’’ while investigating ‘‘remote connectivity’’ will be theobject of forthcoming studies.

Among the IGE syndromes, the most distinctive part ofthe phenotype is the seizure type. Genetic studies suggestthat the prevailing seizure type is more strongly relatedto the genetic failure than the epilepsy syndrome (Durneret al., 2001; Greenberg et al., 2005). CAE and JAE are genet-ically related to each other but not to JME, while generalizedtonic—clonic seizures may occur in all IGE patients and fam-ilies (Winawer et al., 2003; Marini et al., 2004). Based onclinical and genetic dissimilarity, we tested the hypothesisthat the cortical areas with abnormal local EEG synchroniza-tion are dissimilar for the three distinct phenotypes, IGEpatients with absence seizures: ABS (CAE and NAE), JME andEGTCS.

Patients and methods

Unmedicated CAE, JAE, JME or EGTCS patients were enrolled inthe order they visited one of the Epilepsy Outpatient Services atfour locations in Hungary. Inclusion criteria were: proven diagnosisof one of these syndromes (Panayiotopoulos, 2005); age of onsetafter the sixth year of age. As to ensure the diagnosis and to createhomogenous groups, the presence of the characteristic EEG general-ized spike—wave (GSW) paroxysms was required, and patients withintermediary phenotypes (for example, the coexistence of frequentabsences and myoclonic seizures) were not included. Exclusion cri-teria were: neurological symptoms, mental retardation, remarkablecognitive deficit, medication except oral contraceptives, alcoholor drug abuse, any medical condition that is known to influenceEEG activity; generalized tonic—clonic seizures in 3 days beforeEEG investigation. The presence of absence and myoclonic seizuresdid not exclude the patient. The study design was approved by theLocal Research Ethics Committee of Kenézy Kórház Ltd. The patientswere diagnosed and treated according to generally accepted stan-dards for IGE. No diagnostic procedure or treatment was indicated,missed, or postponed for study purposes. Cranial MRI was carriedout in the patients with generalized tonic—clonic seizures but notis those with typical CAE, JAE and JME. Primary aim of this studywas to compare patient groups with control groups. Each patientgroup (ABS, JME and EGTCS) was compared to a separate group ofage-matched controls. EEG records of healthy control persons wereselected from our own normative EEG database.

One might argue that if background EEG abnormalities reflectseizure propensity they may progressively increase in the timeperiod (perhaps a few seconds) preceding a GSW paroxysm. The sec-ondary aim of the study was to investigate whether the presumedEEG abnormalities do or do not change as a function of distance tothe next GSW paroxysm.

EEG recording and sample selection

EEG recordings were carried out in the morning, after a nightof sufficient sleep, in a semi-isolated room, with the same type

oLlc

283

f digital equipment, by trained personnel, according to recom-ended standards for quantitative EEG studies (Nuwer et al., 1994).

ilver—silver chloride electrodes were placed according to the0—20 system, fixed by appropriate adhesive and conductive gel.mpedances did not exceed 10 kOhm. 21-channel EEG was recordedrom standard scalp sites and the earlobes against Fpz samplingeference. EEG was recomputed against a mathematical linked earseference. Additional bipolar derivations were used to differentiateetween EEG and eye movement potentials and to detect myo-enic activity. In the EEG derivations the filters were set at 0.1nd 33.6 Hz, sampling rate was 128 per second, 12 bit on-line digi-ization was used. 30 min EEG was recorded in the waking-relaxed,yes-closed condition. The state of vigilance was controlled by theEG technician who gently aroused the patient when the poste-ior alpha rhythm disappeared. The ‘‘best’’ 60 epochs (each 2 s, aotal of 2 min EEG activity) were selected for comparing patient andontrol groups. Epoch selection and analysis was done blindly with-ut knowing the name and diagnosis of the persons. Our standardpoch selection protocol includes: (1) the presence of continuoushysiological alpha activity with alpha voltage maximum in poste-ior regions; (2) the absence of artifacts, epileptiform potentials,nd other nonstationary elements; and (3) the absence of patternsndicating drowsiness or arousal. This electrographic definition ofhe relaxed-waking state refers to a narrow range of vigilance levelBente, 1979). In addition, two reproducibility measures were usedo minimize the effect of short- and long-term variability withinhe samples. Only samples with at least 95% of average split-halfeliability and test—retest reliability (calculated as the average ofhe 19 channels) entered further analysis. All steps of sampling andata analysis were the same for the patients and the controls.

As to compare near-GSW and far-GSW epochs in the sameatients, we selected ‘‘preictal’’ epochs (2 s before the GSW parox-sms) and ‘‘interictal’’ epochs (10 s before the GSW paroxysms)n JME and ABS patients who had at least 10 GSW paroxysmsn their record. Only epochs that corresponded to the selec-ion criteria (as described above) were selected as to avoid theonfounding effect of pre-GSW vigilance changes. Furthermore,he distance of the ‘‘preictal’’ and ‘‘interictal’’ epochs was ateast 10 s after a GSW paroxysm, as to avoid the contaminationf post-GSW effects the existence of which cannot be a priorixcluded. Final control of all the selected epochs was done by theenior author, by means of the NeuroGuide software Version 2.5.6http://www.appliedneuroscience.com/) that allowed transmissionf the samples to the joined LORETA (Low Resolution Electromag-etic Tomography) software.

ORETA analysis

ORETA is a recently developed method to localize multipleistributed cortical sources of bioelectric activity in the three-imensional space (Pascual-Marqui et al., 1994). In other words,ORETA demonstrates the synchronously activated EEG generatorsy computing their cortical localization from the scalp distributionf the electric field. This is called solving the inverse problem of theEG. The LORETA inverse solution is based on existing neuroanatom-cal and physiological knowledge and a mathematical constraintalled the smoothness assumption (Pascual-Marqui, 2002). LORETAomputes the inverse solution within a three-shell spherical headodel including scalp, skull, and brain. The brain compartment of

his model was restricted to the cortical gray matter and hippocam-us. The gray matter compartment was subdivided in 2394 voxels,hich allows a spatial resolution of 7 millimeters. LORETA com-utes current source density (A/m2) for each voxel. For the sake

f brevity, this is called ‘‘activity’’ in this paper as usual in theORETA literature. The consistency of LORETA with physiology andocalization has been validated for a lot of normal and pathologicalonditions (Pascual-Marqui et al., 2002). Comprehensive evaluation

2

oPtstiuc8sa

L

PVadSsvGtBiddsFw5rt

CCaiwwm

R

1(afJ2nlrsdmaEGanrno

anaa

L

Ve(abaa3

iid7y

ectadamicna

Sb

Id3g3Vooo(ifm1bagiq

84

f the LORETA method is available in reviews (Grova et al., 2008;lummer et al., 2008) and the LORETA literature is available athe website http://www.uzh.ch/keyinst/loreta.htm. In the presenttudy we explored the frequency spectrum from 1 to 25 Hz by split-ing it into 1 Hz very narrow bands (VNBs). For example, activityn the 1 Hz VNB was the average of the 0.5 and the 1.0 Hz val-es. In addition, the VNB values were compressed into the fourlassic EEG frequency bands (delta: 1—3 Hz, theta: 4—7 Hz; alpha:—12 Hz; beta: 13—25 Hz). Localization is given in anatomical terms,pecifying the lobe, the gyrus and the Brodmann area (BA) of thebnormality.

ORETA statistics

atient—control group comparisonoxel-wise LORETA group comparison included calculating simplelgebraic differences between the corresponding groups in order toemonstrate all differences independent of statistical significance.tatistically significant differences were generated in the secondtep. Student-t statistics was carried out by comparing patientersus control datasets composed of normalized individual results.iven that not the voxel-based data but rather the sensors (elec-rodes) are independent (Grave de Peralta Menendez et al., 2004),onferroni correction for multiple comparisons was applied accord-

ng to emphasize and visualize the statistically significant groupifferences at the corrected p < 0.05 level. Broad-band results wereescribed and graphically demonstrated when the VNBs showed veryimilar topographic figures within the corresponding broad-band.or example, ‘‘theta band activity’’ means that the LORETA figuresere very similar in the theta broad-band and the corresponding 4,, 6, 7 Hz VNBs. VNB results were given when the abnormality wasestricted to one or two VNBs but not to the remaining VNBs withinhe broad band.

omparison between ‘‘preictal’’ and ‘‘interictal’’ epochsomparison between ‘‘preictal’’ and ‘‘interictal’’ epochs was donet the group level (ABS preictal versus interictal, JME preictal versusnterictal). Equal numbers of ‘‘preictal’’ and ‘‘interictal’’ epochsere averaged for each eligible patient. Thereafter, group statisticsas carried out by means of paired t-tests including correction forultiple comparisons as described above.

esults

4 ABS patients (average age: 14.4 years), 17 JME patientsaverage age: 19.3 years) and 7 EGTCS patients (averagege: 24.2 years) were analyzed. Seizure frequency was aew to about 50 per day in CAE, one or a few in JAE andME. The EGTCS patients had rare seizures (maximum 1 per

months). Most of the enrolled patients were newly diag-osed except 3 JAE, 2 JME and 2 EGTCS patients who hadong-lasting epilepsy but were off medication for severaleasons. No patient displayed convulsive or nonconvulsivetatus epilepticus, additional focal seizures, neurologicaleficit symptoms or clinically evident cognitive impair-ent. All the patients displayed alpha type EEG background

ctivity being within the normal limits by conventionalEG analysis. All the ABS patients displayed 3—3.5 Hz ictalSW paroxysms, the JME patients showed irregular GSWnd multispike—wave paroxysms with or without myoclo-

ia. 3—4 Hz interictal GSW paroxysms but no seizures wereecorded in the EGTCS patients. Focal spike potentials wereever found in this group. EEG background activity was neverbscured by frequent paroxysms. Cranial MRI was normal in

deas

B. Clemens et al.

ll the 7 EGTCS patients. Control groups were as follows: = 17, average age = 19.5 years for the JME group; n = 14,verage age = 14.0 years for the ABS group; n = 7, averagege = 24.0 years for the EGTCS group.

ORETA results: subtraction of the group averages

oxel-based subtraction is an overview of group differ-nces disregarding the statistical significance of the findingsFig. 1). The ABS group showed greater broad-band LORETActivity than the controls in all cortical areas and frequencyands. The difference was greatest in the theta band. VNBnalysis refined the results disclosing bilaterally decreasedctivity at 10—11 Hz and 19—21 Hz in the central area (BA 2,, 4, 6).

In the JME group, broad-band analysis showed overallncreased delta and theta activity, with greatest differencen the theta band. Alpha and beta activity were slightlyecreased in the precuneus and superior parietal lobule (BA) but slightly increased in the rest of the cortex. VNB anal-sis disclosed overall increased activity at 8 and 9 Hz.

In the EGTCS group, broad-band delta activity was bilat-rally increased in the occipital and posterior basal temporalortex, and bilaterally decreased in the entire frontal cor-ex, insula and anterior temporal area. Increased thetactivity was found in the occipital and temporal cortex,ecreased theta activity in the lateral part of the left centralrea. Broad-band alpha and beta activity were decreased inost parts of the cortical convexity but slightly increased

n the basal, frontal and temporal cortex. VNB analysis dis-losed marked diffuse decrease of activity at 11 Hz but didot add further striking results as compared to broad-bandnalysis.

tatistically significant LORETA group differencesetween patients and controls

n the JME group, statistically significant differences areepicted in Fig. 2. Increased activity was found in the 2 and

Hz VNBs and in the delta broad band in the right fusiformyrus (BA 20, 37), hippocampus, parahippocampal gyrus (BA0, 35, 36). Increased theta activity was found in the 4—7 HzNBs and the theta broad band in an anatomically contigu-us area. This area included the superior, middle and inferiorccipital gyri, lingual gyrus, cuneus (BA 17, 18, 19) in theccipital lobe; the angular gyrus (BA 39), posterior cingulateBA 23, 29, 30) and the posterior part of the precuneus (BA 7)n the parietal lobe; the hippocampus, parahippocampal andusiform gyri (BA 36, 20), the posterior parts of the superior,iddle and inferior temporal gyri in the temporal lobe (BA

9, 22, 37, 39). Increased activity in the 8 Hz VNB emergedilaterally in the prefrontal area including the orbital gyrind part of the medial frontal gyrus (BA 11) and anterior cin-ulate (BA 32). In addition, some of the temporo-occipital,ncreased activity (as described above) persisted at this fre-uency but disappeared at 9 Hz. No statistically significant

ifference emerged in the alpha and beta broad bands. How-ver, symmetrically decreased activity was found at 11 Hznd 21—22 Hz in the same parietal area: the precuneus, rightuperior parietal lobule and posterior cingulate.

EEG-LORETA endophenotypes of IGE 285

Figure 1 Voxel-wise LORETA-activity subtraction figures (patients group averages minus control group averages). Upper row:absence epilepsies (ABS group). Middle row: juvenile myoclonic epilepsy (JME) group. Bottom row: epilepsy with generalizedtonic—clonic seizures exclusively (EGTCS) group. Broad bands (white capitals) are the same for the three groups. Very narrowband (VNB, yellow labeling) results are only given for the selected VNBs for the ABS and EGTCS groups. As to facilitate visualinter-group and inter-band comparison, a common color scale refers to all plots. Red color: more LORETA activity (current sourcedensity) in the patient group than in the control group. Blue color: less LORETA activity in the patient group than in the controlgroup. R and L indicate the right and left hemispheres, respectively. (For interpretation of the references to color in this figure

.)

V3

C

9fsbtfaf

D

T

legend, the reader is referred to the web version of the article

In the ABS group, statistically significant group differ-ences are demonstrated in Fig. 3(top). Topographically verysimilar, statistically significant result appeared in the thetabroad band and in the 5—7 Hz VNBs. Increased activitywas found bilaterally in the temporal lobe: uncus (BA 36),hippocampus, parahippocampal gyrus (BA 34, 35, 36, 37),and the basal prefrontal area including the straight gyrus,basal parts of the inferior and medial frontal gyri (BA 11,47), subcallosal area (BA 25, 34), anterior cingulate (BA25). Decreased activity was found at 19—21 Hz in the rightpostcentral gyrus (BA 43) and parts of the right superiorand medial temporal gyri (BA 21). No statistically signif-icant results emerged in the remaining broad bands andVNBs.

No broad-band differences emerged in the EGTCS group.Statistically significant, VNB-related group differences aredepicted in Fig. 3(bottom). Decreased activity was foundin the frontal cortex at 10 and 11 Hz. The abnormality was

more marked in the left hemisphere including the precentralgyrus (BA 4), superior, middle, inferior and medial frontalgyri (BA 6, 8, 9, 10, 44, 45, 46), and the postcentral gyrus(BA 2, 3). Increased activity was found in the 16—18 Hz

iIct

NBs in a small part of the right parahippocampal gyrus (BA6).

omparison of ‘‘preictal’’ and ‘‘interictal’’ epochs

JME and 7 ABS patients were eligible for computing dif-erence between ‘‘preictal’’ and ‘‘interictal’’ epochs. Notatistically significant differences were found in the broadands and in the 1—9 Hz VNBs. Statistically significant butopographically dissimilar, minor, patchy differences wereound at some frequencies at 10 Hz and in several VNBsbove 15 Hz. None of them was related to the group dif-erences described above.

iscussion

his is the first study to demonstrate group-specific anatom-

cal patterns of abnormal EEG background activity in threeGE phenotypes. We confirmed the hypothesis that theortical areas showing abnormal local EEG synchroniza-ion are topographically dissimilar in the investigated IGE

286 B. Clemens et al.

Figure 2 Voxel-wise comparison of LORETA activity in the JME group and the corresponding control group. Only statisticallysignificant results are depicted. Red color indicates increased LORETA activity (current source density) in the patient group ascompared to the control group. Blue color indicates decreased LORETA activity in the patient group as compared to the controlg tivel .)

pTtibittOs

B

VwbeWdew

nm

Nd

Ioitedmatab2c

roup. R and L indicate the right and left hemispheres, respecegend, the reader is referred to the web version of the article

henotypes: IGE with absence seizures, JME and EGTCS.he study design favored the detection of the abnormalitieshat are common to the patients within each group whilentra-group variability, individual features were suppressedy averaging. In addition, correction for multiple compar-sons abolished some of the group differences. Thus we meanhat the remaining, statistically significant findings reflecthe core patterns of the syndrome-specific abnormalities.n the other hand, also statistically not significant findingseem to carry relevant information and will be discussed.

road-band versus narrow-band analysis

NB-related and broad-band abnormalities were foundithin the 1—25 Hz frequency range. Concerning the broad-and versus VNB issue we confirmed that they should bevaluated together if possible (Shackman et al., 2010).

e confirmed that a single, VNB-related abnormalityoes not necessarily influence broad-band activity (Szavat al., 1994). Furthermore, topographically similar or some-hat asymmetrical current source density distributions in

aidp

ly. (For interpretation of the references to color in this figure

eighboring VNBs were smoothed by averaging, resulting inore interpretable patterns.

europhysiological interpretation of increased andecreased LORETA activity at different frequencies

ncreased EEG synchronization is the common electrophysi-logical denominator across all epilepsy syndromes, at leastn the ictal condition. However, it is present in the interic-al EEG background activity of IGE patients, too (Clemenst al., 2000, 2007a; Willoughby et al., 2003). Traditionally,elta increase in waking records indicates hemispheric whiteatter impairment (Gloor et al., 1977) while the sources

nd significance of theta synchronization remains a mat-er of debate. IGE patients show diffusely distributed deltand theta excess hidden to the naked eye but detectedy EEG spectral analysis and LORETA (Clemens et al.,000, 2007a,b). However, they do not present with clini-al symptoms and MRI signs indicating diffuse white matter

bnormality (Anderson and Hamandi, 2011). In prior stud-es we found that unlike in lesional conditions, increasedelta and theta absolute spectral power are not accom-anied by slowing of the band mean frequencies in IGE

EEG-LORETA endophenotypes of IGE 287

Figure 3 Voxel-wise comparison of LORETA activity in the ABS and EGTCS groups and the corresponding control groups. Onlystatistically significant results are depicted. Red color indicates increased LORETA activity (current source density) in the patientgroup as compared to the control group. Blue color indicates decreased LORETA activity in the patient group as compared to thecontrol group. R and L indicate the right and left hemispheres, respectively. (For interpretation of the references to color in thisfigure legend, the reader is referred to the web version of the article.)

ncmttape

tt1eaHttp

patients. Furthermore, valproate and lamotrigine decreaseddelta and theta power in activity-dependent manner insuccessfully treated IGE patients (Clemens et al., 2007b;Clemens, 2008). Thus we interpreted delta and/or thetaincrease as interictal marker of the seizure-prone state inIGE. This interpretation of increased delta and theta activ-ity is extended to the LORETA results in this paper. However,the present results suggest that theta, rather than delta isthe frequency of main interest in this respect. Alpha andbeta rhythms are physiological components of EEG back-ground activity, the decrease of which is usually interpretedas indicating decreased local or global neurological functions(Markand, 1990; Kozelka and Pedley, 1990).

Core abnormalities in the JME group

Increased LORETA theta activity was confined to the pos-

terior (temporal, parietal, and occipital) parts of thecortex. No remarkable structural or metabolic findings werereported in these areas (Anderson and Hamandi, 2011).Therefore we suggest that a yet not specified, presumably

fnon

onlesional abnormality exists in the posterior part of theortex that is responsible for the propensity to generateyoclonic seizures. The contribution of these cortical areas

o ictogenicity is supported by the occurrence of consis-ently recorded, interictal temporal spikes (Lombroso, 1997)nd the frequent early ictal involvement of the medial tem-oral, sometimes parietal cortex in JME patients (Holmest al., 2010).

On the contrary, increased LORETA activity at 8 Hz inhe prefrontal cortex is topographically overlapping withhe reported gray matter abnormalities (Woermann et al.,999; Kim et al., 2007) white matter derangement (Deppet al., 2008; Vulliemoz et al., 2010) and decreased n-acetyl-spartate concentration (Savic et al., 2000) in JME patients.owever, some of these abnormalities is proportional tohe duration of the disease (Tae et al., 2008) suggestinghat they do not necessarily reflect pre-existent, causativeathology. Unfortunately, the same uncertainty holds true

or the LORETA findings. The majority of our patients wasewly diagnosed, for which reason correlating the degreef the LORETA abnormality with the duration of the ill-ess was not possible. Concerning ictal studies, organized

2

GJpptatpisaatacpUs

aiTm((dbi(bKtImtoia

C

AtfigoatiesrtcTsfsae

zf

tiwrplipJaEavfl(ldts(atatfi(e

felito

edieto

A

Idiltsorl

88

SW discharges first appear in the orbito-frontal cortex inME patients (Holmes et al., 2010). Summing up, a lot ofathological findings anatomically correspond to increasedrefontal EEG (LORETA) synchronization. Thus we suggesthat the prefrontal LORETA abnormality and the enumer-ted imaging findings are interrelated, and LORETA reflectshe baseline (interictal) bioelectric abnormality in this area,erhaps contributing to seizure propensity in JME. However,t is possible that neither posterior, nor prefrontal increasedynchronization per se, but altered long-range interactionsmong these areas (and presumably also subcortical sites)re responsible for this capacity. In fact, an interictal mul-ivoxel magnetic resonance spectroscopic study disclosed

network, in which abnormal biochemical changes syn-hronously occurred at diverse cortical areas (prefrontal,rimary motor, occipital, and insular) and subcortical sites.nfortunately, the temporal lobes were not explored in thistudy (Lin et al., 2009).

Finally, symmetrically decreased LORETA activity at 11nd 21—22 Hz in the precuneus and posterior cingulatendicates decreased physiological function is this area.his abnormality may be related to the reduction in grayatter volume in the posterior cingulate in JME patients

O’Muircheartaigh et al., 2011). The posterior cingulatealone or in conjunction with other parts of the so-calledefault mode system) shows functional MRI inactivationefore (Moeller et al., 2008) and during generalized inter-ctal GSW discharges (Gotman, 2008), myoclonic seizuresLiu et al., 2008) and also during unilateral temporal spikesut not during non-temporal spikes (Laufs et al., 2007;obayashi et al., 2009). Rather than being specific to JME,he dysfunction of the precuneus and posterior cingulate inGE and temporal epilepsy seems to reflect a shared neuronalechanism of these epilepsy syndromes. This argumenta-

ion is in accord with the medial temporal, parietal andccipital LORETA abnormalities in JME that had been real-zed previously as an ‘‘extended’’ frontal syndrome (Koeppnd Hamandi, 2006).

ore abnormalities in the ABS group

bsence seizures can be precipitated by electrical stimula-ion of the medial frontal cortex (Bancaud et al., 1974). Thisnding is frequently quoted as a proof for the frontal ori-in of absences. A less realized but equally important resultf the same investigation is that stimulation precipitatesbsences only if the patients display spontaneous absences,oo. We interpret this finding that, in addition to focallyncreased excitation in the seizure onset zone, somethinglse, a habitual, genetically determined liability is neces-ary to the manifestation of absence seizures. Our resultsise the possibility that increased theta synchronization inhe basal frontal and medial temporal cortex is the EEGorrelate of the propensity to generate absence seizures.he same basal frontal area where bilaterally synchronouspike—wave paroxysms first occur is the seizure onset zone

or absence seizures (Holmes et al., 2004). However, a MEGtudy disclosed that also temporal, insular sites and thenterior cingulate are activated at seizure onset (Sakurait al., 2010). Unlike in the JME group, the seizure onset

pa

t

B. Clemens et al.

one and the main LORETA abnormality are topographicallyairly overlapping in the ABS group.

Our results indicate that the prefrontal and medialemporal limbic areas equally contribute to the neurophys-ological abnormality. This is not in agreement with theidely held but unbalanced notion that emphasizes the

ole of the frontal but not the temporal structures in theathogenesis of absence epilepsy. The critical review of theiterature shows that most non-ictal findings indicating thectogenic property of the cortex actually point to the tem-oral, not the frontal lobe. In one study, 36—70% of CAE andAE patients display focal EEG abnormalities. Out of the 29bsence patients with focal EEG abnormalities, 21 had focalEG abnormalities in the temporal region, 8 in the centralrea but none in the frontal lobe (Lombroso, 1997). In aery great cohort of CAE and JAE patients the prevalence ofocal interictal spikes was 36—46% in the temporal lobe, fol-owed by far lesser values in the occipital and frontal lobesHughes and Kaydanova, 1997). Postmortem neuropatho-ogical evaluation of 8 idiopathic absence epilepsy personsemonstrated potentially epileptogenic abnormality in 5 ofhem. Microdysgenesis was found at topographically dis-imilar neocortical sites but invariably in the hippocampusMeencke and Janz, 1984). Another, proven epileptogenicbnormality, mossy fiber axonal sprouting was described inhe hippocampus in a mouse model of absence epilepsy (Qiaond Noebels, 1993). In our study, LORETA showed increasedheta activity within both hippocampi. Hippocampal localeld potentials are usually thought not to arrive the scalpLantz et al., 2001), but other authors say it is possible (Attalt al., 2007; Brandeis et al., 2009).

Decreased beta activity, a marker of impaired corticalunction was found in the ABS group at 19—21 Hz in the low-st part of the right postcentral gyrus and in the vis-a-visocated part of the superior temporal gyrus. The potentialmportance of this finding is that a cortical focus exactly inhis area is the established seizure-onset zone in a rat modelf absence epilepsy (Meeren et al., 2002).

Structural quantitative MRI studies carried out in absencepilepsies resulted in conflicting results. Both increased andecreased amount of cortical gray matter was reportedn the orbito-frontal area and other cortical sites (Bettingt al., 2006; Caplan et al., 2009), for which reason discussinghe possible relationship between LORETA and MRI data ismitted in this paper.

bnormalities in the EGTCS group

n contrast to the JME and ABS groups, no broad-bandifferences emerged in the EGTCS group. No VNB-relatedncreased activity emerged in the delta—theta range. Theack of delta—theta increase suggests that the pronenesso GTCSs is not related to genetically determined hyper-ynchronization in any cortical area. The frontal decreasef 10—11 Hz alfa activity and increased activity in theight parahippocampal gyrus are enigmatic findings. Theack of quantitative MRI abnormalities in a group of EGTCS

atients (Betting et al., 2006) further hinders the searchfter structure-function relationship in this group.

In contrast to the absence and myoclonic seizures,he GTCS is the non-specific end stage of non-arrested

ooioofAcdwesbr

bctstbgisad

S

Qdcritbcomccaa(fieanecsbetcsL

EEG-LORETA endophenotypes of IGE

generalized (absence, myoclonic) and focal seizure activ-ity. Furthermore, GTCSs can be precipitated by near-physiological events like sleep deprivation and mentalstress, a lot of neuroactive drugs and toxic agents in non-epileptic patients and in healthy young persons who neverdisplay spontaneous seizures (Rodin, 1984). Summing up,the GTCS has been realized as the non-specific responseof the brain to supracritical, seizure-promoting influences(Aird et al., 1984). Interestingly, the EGTCS group seemsto be clinically and neurophysiologically similar to personswho display sleep deprivation-provoked GTCSs exclusively,no delta—theta LORETA abnormalities but decreased LORETAalpha activity (Puskás et al., 2010).

LORETA abnormalities as endophenotypes

Resting EEG background activity has strong genetic deter-mination (van Beijsterveldt and Boomsma, 1994; Zietschet al., 2007). This is not surprising because the geneticscenario determines the structure, metabolism and homeo-static activity of the brain, which, in turn, determine EEGactivity (John, 2005; Nunez, 1995). No rapid changes charac-terize the morphological basis of EEG at spatial scales mostlyresponsible for fine-tuning the signal (Robinson et al., 2004).For this reason, quantitative EEG variables are stronglydetermined and reproducible in the long run in healthy peo-ple, as demonstrated by several authors, e.g. (Gasser et al.,1985; Kondacs and Szabo, 1999). To what degree this con-stancy refers to patient populations is actually not proven.However, we suppose that in non-progressive, non-lesional,genetic disorders (as IGE syndromes) pathological EEG traitsare as invariant as in healthy persons. This is supported byour unpublished findings. Some IGE patients who had beenevaluated in the first setting but were noncompliant withdrug treatment were evaluated some month or even yearslater. Ictal and interictal raw EEG traces, power spectra andLORETA findings were surprisingly similar in these patients,independent of the time period between the two settings.

The lack of statistically significant and biologically mean-ingful differences between ‘‘preictal’’ and ‘‘interictal’’epochs suggests that the LORETA abnormalities reportedin this study are stabile traits, not state-dependent fea-tures reflecting the momentary degree of seizure propensity.Rather, we think that they reflect specific failures ofthe brain that are responsible for the recurrence of spe-cific seizure types, at least in the ABS and JME groups.In other words, the findings may be realized as fin-gerprints of genetically determined conditions, in otherwords, endophenotypes (Zietsch et al., 2007). For exam-ple, the characteristic interictal epileptiform potential isan endophenotype referring to a genetic failure and a cor-responding epilepsy syndrome, benign rolandic epilepsy ofchildhood (Strug et al., 2009). Abnormalities of EEG back-ground activity might carry even more reliable informationbecause unlike epileptiform transients, they are invariablypresent in the waking record. Given that increased EEGdelta and/or theta synchronization is a peculiar charac-

teristics reflecting the seizure-prone state of the brain inIGE syndromes (for argumentation and references, see theIntroduction of this paper), we concluded that temporo-parieto-occipital theta increase is the EEG-endophenotype

ifica

289

f JME. Whether prefrontal increase of activity at 8 Hz is partf this endophenotype, remains uncertain. Alternatively,t may be related to the postulated structural pathology,r may reflects increased baseline activity in the seizurenset area (Holmes et al., 2010). Theta increase in theronto-temporal limbic areas is the endophenotype of theBS group. On the other hand, no such endophenotypeharacterizes the EGTCS group. The electrophysiologicalichotomy of the investigated IGE phenotypes (ABS and JMEith LORETA endophenotype versus EGTCS without LORETAndophenotype) is supported by genetic studies. EGTCSeems genetically not to be coupled to CAE, JAE or JMEut may occur in any CAE, JAE and JME patient and theirelatives (Marini et al., 2004).

Endophenotypes may help to bridge the gap of knowledgeetween genome and clinical phenotype in diseases withomplex inheritance (de Geus, 2010). IGE syndromes belongo this group of disorders. We postulate that the reported,yndrome-specific EEG endophenotypes may have impor-ance in future research. However, this possibility is limitedy the fact that IGE patients also have individual EEG back-round activity characteristics or abnormalities that maynterfere with the syndrome-specific current source den-ity distribution. The contribution of the syndrome-specificnd the individual EEG features to the final current densityistribution is currently investigated in our laboratory.

tatistically not significant results

uantitative EEG studies result in a plenty of numericalata for which reason most authors interpret the statisti-ally significant results only. We mean that this approachesults in loss of useful information. In our study, subtractionndicated topographically overall, increased delta and par-icularly theta broad-band activity in the ABS and JME groupsut not in the EGTCS group. This finding recalls Gloor’s con-ept of the ‘‘mild diffuse epileptogenic state’’ that wasriginally based on the feline penicillin generalized epilepsyodel (Gloor, 1979). In this model, abnormally increased

ortical excitability is caused by abnormal amino acid con-entrations within the entire cortex, and the biochemicalnd electrophysiological abnormalities changed in close par-llel during the transition of the interictal to the ictal statevan Gelder et al., 1983). The biochemical findings were con-rmed recently by magnetic resonance spectroscopy (Meløt al., 2006). It is possible that diffusely increased deltand/or theta LORETA activity corresponds to this compo-ent of Gloor’s concept, at least for the idiopathic absencepilepsies and JME. Whether a general seizure susceptibilityoncept (Gloor et al., 1982) or the idea of syndrome-specificeizure susceptibility (Berkovic et al., 1998) is closer toiological reality is beyond the scope of this study. How-ver, the oligogenic concept of the IGE syndromes suggestshat they may coexist as the consequences of dissimilar sus-eptibility genes. Our findings raise the possibility that thetatistically not significant but neurobiologically meaningfulORETA findings are related to a diffuse cortical biochem-

cal abnormality while the statistically significant LORETAndings (called endophenotypes) reflect syndrome-specificomponents of the seizure-prone state of the brain in thebsence and JME syndromes.

2

Ii

IiiaaCtsiAlhtrtdtgy

R

A

A

A

A

A

B

B

B

B

B

B

B

C

C

C

C

C

C

D

d

D

G

G

G

G

G

G

G

G

90

ncreased theta synchronization: developmentalssues

nterestingly, most EEG (LORETA) abnormalities were foundn the theta band in the investigated IGE groups. Concern-ng humans, the role of theta activity has been mainlyddressed in the context of cognitive functions like learningnd memory (Basar and Schürmann, 1999; Klimesch, 1999).oncerning resting-state EEG activity, increased theta spec-ral power is a non-specific feature of many diseasetates presenting with impaired cognitive functions. Thetancrease was reported in acquired dementias, typicallylzheimer’s disease, and also to developmental conditions

ike non-specific mental retardation and attention deficityperactivity disorder. Why IGE patients with normal cogni-ive functions display increased resting-state theta activityemains an enigma. Given the above-discussed strong rela-ionship between genes and EEG activity, the recentlyiscovered genetic proximity of some neurodevelopmen-al disorders including non-specific mental retardation andenetic epilepsy (Goh et al., 2007) may be a reasonable butet not proven solution.

eferences

arabi, A., Wallois, F., Grebe, R., 2008. Does spatiotemporal syn-chronization of EEG change prior to absence seizures? Brain Res.1188, 207—221.

ird, R.B., Masland, R., Woodbury, D.M., 1984. The Epilepsies. ACritical Review. Raven Press, New York, pp. 65—70.

mor, F., Baillet, S., Navarro, V., Adam, C., Martinerie, J., Quyen,M.V., 2009. Cortical local and long-range synchronization inter-play in human absence seizure initiation. Neuroimage 45,950—962.

nderson, J., Hamandi, K., 2011. Understanding juvenile myoclonicepilepsy: contributions from neuroimaging. Epilepsy Res.,doi:10.1016/j.eplepsyres.2011.03.008.

ttal, Y., Bhattacharjee, M., Yelnik, J., Cottereau, B., Lefèvre, J.,Okada, Y., Bardinet, E., Chupin, M., Baillet, S., 2007. Modelingand detecting deep brain activity with MEG & EEG. Conf. Proc.IEEE Eng. Med. Biol. Soc. 2007, 4937—4940.

ancaud, J., Talairach, J., Morel, P., Bresson, M., Buser, P., 1974.‘‘Generalized’’ epileptic seizures elicited by electrical stim-ulation of the frontal lobe is man. Electroencephalogr. Clin.Neurophysiol. 37, 275—282.

asar, E., Schürmann, M., 1999. Brain functioning: integrativemodels. In: Basar, E. (Ed.), Brain Function and Oscillations.Integrative Brain Function. Neurophysiology and Cognitive Pro-cesses, vol. II. Springer, pp. 393—406.

ente, D., 1979. Vigilance and evaluation of psychotropic drugeffect on EEG. Pharmacopsychiatry 12, 137—147.

erkovic, S.F., Howell, R.A., Hay, D.A., Hopper, J.L., 1998. Epilep-sies in twins: genetics of the major epilepsy syndromes. Ann.Neurol. 43, 435—445.

etting, L.E., Mory, S.B., Li, L.M., Lopes-Cendes, I., Guerreiro,M.M., Guerreiro, C.A., Cendes, F., 2006. Voxel-based mor-phometry in patients with idiopathic generalized epilepsies.Neuroimage 32, 498—502.

lumenfeld, H., Westerveld, M., Ostroff, R.B., Vanderhill, S.D.,

Freeman, J., Necochea, A., Uranga, P., Tanhehco, T., Smith, A.,Seibyl, J.P., Stokking, R., Studholme, C., Spencer, S.S., Zubal,I.G., 2003. Selective frontal, parietal, and temporal networks ingeneralized seizures. Neuroimage 19, 1556—1566.

B. Clemens et al.

randeis, D., Michel, C.M., Amzica, F., 2009. From neuronal activ-ity to scalp potential fields. In: Michel, C.M., Koenig, T.,Brandeis, D., Gianotti, L.R.R., Wackermann, J. (Eds.), Elec-trical Neuroimaging. Cambridge University Press, New York,pp. 1—24.

aplan, R., Levitt, J., Siddarth, P., Wu, K.N., Gurbani, S., Sankar, R.,Shields, W.D., 2009. Frontal and temporal volumes in ChildhoodAbsence Epilepsy. Epilepsia 50, 2466—2472.

arney, P.W., Masterton, R.A., Harvey, A.S., Scheffer, I.E., Berkovic,S.F., Jackson, G.D., 2010. The core network in absence epilepsy.Differences in cortical and thalamic BOLD response. Neurology75, 904—911.

lemens, B., Szigeti, G., Barta, Z., 2000. EEG frequency profilesof idiopathic generalized epilepsy syndromes. Epilepsy Res. 42,105—115.

lemens, B., Bessenyei, M., Piros, P., Tóth, M., Seress, L., Kondákor,I., 2007a. Characteristic distribution of interictal brain elec-trical activity in idiopathic generalized epilepsy. Epilepsia 48,941—949.

lemens, B., Piros, P., Bessenyei, M., Hollódy, K., 2007b. Lamotrig-ine decreases EEG synchronization in a use-dependent manner inpatients with idiopathic generalized epilepsy. Clin. Neurophys-iol. 118, 910—917.

lemens, B., 2008. Valproate decreases EEG synchronization ina use-dependent manner in idiopathic generalized epilepsy.Seizure 17, 224—233.

eppe, M., Kellinghaus, C., Duning, T., Möddel, G., Mohammadi, S.,Deppe, K., Schiffbauer, H., Kugel, H., Keller, S.S., Ringelstein,E.B., Knecht, S., 2008. Nerve fiber impairment of anterior tha-lamocortical circuitry in juvenile myoclonic epilepsy. Neurology71, 1981—1985.

e Geus, E.J., 2010. From genotype to EEG endophenotype: a routefor post-genomic understanding of complex psychiatric disease?Genome Med. 2, 63.

urner, M., Mehdi, A., Keddache, M.S., Tomasini, L., Shinnar, S.,Resor, S.R., Cohen, J., Harden, C., Moshe, S.L., Rosenbaum, D.,Kang, H., Ballaben-Gil, K., Hertz, S., Labar, D.R., Luciano, D.,Wallace, S., Yohai, D., Kloth, I., Dicker, E., Greenberg, D.A.,2001. Genome scan of idiopathic generalized epilepsy: evidencefor major susceptibility gene and modification genes influencingthe seizure type. Ann. Neurol. 49, 328—335.

asser, T., Bacher, P., Steinberg, H., 1985. Test—retest reliabilityof spectral parameters of the EEG. Electroencephalogr. Clin.Neurophysiol. 60, 312—319.

loor, P., Ball, G., Schaul, N., 1977. Brain lesions that produce deltawaves in the EEG. Neurology 4, 326—333.

loor, P., 1979. Generalized epilepsy with spike and wave discharge:a reinterpretation of its electrographic and clinical manifesta-tions. Epilepsia 20, 571—588.

loor, P., Metrakos, J., Metrakos, K., Andermann, E., van Gelder,N., 1982. Neurophysiological genetic and biochemical nature ofthe epileptic diathesis. In: Broughton, R.J. (Ed.), Henry Gastautand the Marseilles School’s Contribution to the Neurosciences.Elsevier Biomedical Press, Amsterdam, pp. 45—56.

oh, K.I., Cusick, M.E., Valle, D., Childs, B., Vidal, M., Barabási,A.L., 2007. The human disease network. Proc. Natl. Acad. Sci.USA 104, 8685—8690.

otman, J., 2008. Epileptic networks studied with EEG-fMRI. Epilep-sia 49 (Suppl. 3), 42—51.

rave de Peralta Menendez, R., Murray, M.M., Michel, C.M., Mar-tuzzi, R., Gonzalez Andino, S.L., 2004. Electrical neuroimagingbased on biophysical constraints. Neuroimage 21, 527—539.

reenberg, D.A., Cayanis, E., Strung, L., Marathe, S., Durner, M.,Pal, D.K., Alvin, G.B., Klotz, I., Dicker, E., Shinnar, S., Bromfield,

E.B., Resor, S., Cohen, J., Moshe, S.L., Harden, C., Kang, H.,2005. Malic enzyme 2 may underlie susceptibility to adolescent-onset idiopathic generalized epilepsy. Am. J. Hum. Genet. 76,139—146.

L

M

M

M

M

M

M

M

M

N

N

N

O

P

P

P

P

P

P

EEG-LORETA endophenotypes of IGE

Grova, C., Daunizeau, J., Kobayashi, E., Bagshaw, A.P., Lina, J.-M.,Dubeau, F., Gotman, J., 2008. Concordance between distributedEEG source localization and simultaneous EEG-fMRI studies ofepileptic spikes. Neuroimage 39, 755—774.

Holmes, M.D., Brown, M., Tucker, D.M., 2004. Are ‘‘generalized’’seizures truly generalized? Evidence of localized mesial frontaland frontopolar discharges in absence. Epilepsia 45, 1568—1579.

Holmes, M.D., Quiring, J., Tucker, D.M., 2010. Evidence thatjuvenile myoclonic epilepsy is a disorder of frontotemporal cor-ticothalamic networks. Neuroimage 49, 80—93.

Hughes, J.R., Kaydanova, Y., 1997. Long-term studies on patientswith absence and bilateral spike—wave complexes: 430 patients,up to 52 years follow-up. Clin. Electroencephalogr. 28, 193—206.

ILAE (Commission on Classification and Terminology), 1989. Proposalfor revised classification of epilepsies and epileptic syndromes.Epilepsia 30, 389—399.

John, E.R., 2005. From synchronous neuronal discharges to subjec-tive awareness? Prog. Brain Res. 150, 143—171.

Kim, J.H., Lee, J.K., Koh, S.B., Lee, S.A., Lee, J.M., Kim, S.I.,Kang, J.K., 2007. Regional grey matter abnormalities in juvenilemyoclonic epilepsy: a voxel-based morphometry study. Neuroim-age 37, 1132—1137.

Klimesch, W., 1999. EEG alpha and theta oscillations reflect cogni-tive and memory performance: a review and analysis. Brain Res.Brain Res. Rev. 29, 169—195.

Kobayashi, E., Grova, C., Tyvaert, L., Dubeau, F., Gotman, J., 2009.Structures involved at the time of temporal lobe spikes revealedby inter-individual group analysis of EEG/fMRI data. Epilepsia 50,2549—2556.

Koepp, M.J., Duncan, J.S., 2004. Epilepsy. Curr. Opin. Neurol. 17,467—474.

Koepp, M.J., Hamandi, K., 2006. Systems and networks in myoclonicseizures and epilepsies. In: Hirsch, E., Andermann, F., Chauvel,P., Engel, J., Lopes da Silva, F., Luders, H. (Eds.), Generalizedseizures: from clinical phenomenology to underlying systems andnetworks. Progress in Epileptic Disorders. John Libbey Eurotext,Montrouge, France, pp. 163—182.

Kondacs, A., Szabo, M., 1999. Long-term intra-individual variabil-ity of the background EEG in normals. Clin. Neurophysiol. 110,1708—1716.

Kostopoulos, G., 1986. Neuronal sensitivity to GABA and glutamatein generalised epilepsy with spike and wave discharges. Exp.Neurol. 92, 20—36.

Kozelka, J.W., Pedley, T.A., 1990. Beta and mu rhythms. J. Clin.Neurophysiol. 7, 191—207.

Lantz, G., Grave de Peralta Menendez, R., Gonzalez Andino, S.,Michel, C.M., 2001. Noninvasive localization of electromagneticepileptic activity. II. Demonstration of sublobar accuracy inpatients with simultaneous surface and depth recordings. BrainTopogr. 14, 139—147.

Laufs, H., Hamandi, K., Salek-Haddadi, A., Kleinschmidt, A.K., Dun-can, J.S., Lemieux, L., 2007. Temporal lobe interictal epilepticdischarges affect cerebral activity in ‘‘default mode’’ brainregions. Hum. Brain Mapp. 28, 1023—1032.

Lin, K., Carrete Jr., H., Lin, J., Peruchi, M.M., de Araújo Filho,G.M., Guaranha, M.S., Guilhoto, L.M., Sakamoto, A.C., Yacu-bian, E.M., 2009. Magnetic resonance spectroscopy reveals anepileptic network in juvenile myoclonic epilepsy. Epilepsia 50,1191—1200.

Liu, Y., Yang, T., Liao, W., Yang, X., Liu, I., Yan, B., Chen, H., Gong,Q., Stefan, H., Zhou, D., 2008. EEG-fMRI study of the ictal andinterictal epileptic activity in patients with eyelid myocloniawith absences. Epilepsia 49, 2078—2086.

Lombroso, C.T., 1997. Consistent EEG focalities detected in subjects

with primary generalized epilepsies monitored for two decades.Epilepsia 38, 797—812.

Lüders, H.O., Awad, I., 1991. Conceptual considerations. In: Lüders,H.O. (Ed.), Epilepsy Surgery. Raven Press, New York, pp. 51—62.

Q

291

üders, H.O., Najm, I., Nair, D., Widdess-Walsh, P., Bingman, W.,2006. The epileptogenic zone: general principles. Epileptic. Dis-ord. 8, S1—S9.

arini, C., Scheffer, I.E., Crossland, K.M., Grinton, B.E., Phillips,F.L., McMahon, J.M., Turner, S.J., Dean, J.T., Kivity, S., Mazarib,A., Neufeld, M.Y., Korczyn, A.D., Harkin, L.A., Dibbens, L.M.,Wallace, R.H., Mulley, J.C., Berkovic, S.F., 2004. Genetic archi-tecture of idiopathic generalized epilepsy: clinical geneticanalysis of 55 multiplex families. Epilepsia 45, 467—478.

arkand, O.N., 1990. Alpha rhythms. J. Clin. Neurophysiol. 7,163—189.

cCormick, D.A., Contreras, D., 2001. On the cellular and networkbases of epileptic seizures. Annu. Rev. Physiol. 63, 815—846.

eencke, H.J., Janz, D., 1984. Neuropathological findings in pri-mary generalized epilepsies: a study of eight cases. Epilepsia25, 8—21.

eeren, H.K., Pijn, J.P., Van Luijtelaar, E.L., Coenen, A.M., Lopesda Silva, F.H., 2002. Cortical focus drives widespread corticotha-lamic networks during spontaneous absence seizures in rats. J.Neurosci. 22, 1480—1495.

elø, T.M., Sonnewald, U., Touret, M., Nehlig, A., 2006. Corticalglutamate metabolism is enhanced in a genetic model of absenceepilepsy. J. Cereb. Blood Flow Metab. 26, 1496—1506.

oeller, F., Siebner, H.R., Wolff, S., Muhle, H., Boor, R., Granert,O., Jansen, O., Stephani, U., Siniatchkin, M., 2008. Changes inactivity of striato-thalamo-cortical network precede generalizedspike wave discharges. Neuroimage 39, 1839—1849.

oeller, F., Levan, P., Muhle, H., Stephani, U., Dubeau, F., Sini-atchkin, M., Gotman, J., 2010. Absence seizures: individualpatterns revealed by EEG-fMRI. Epilepsia 51, 2000—2010.

air, D.R., Mohamed, A., Burgess, R., Lüders, H., 2004. A criticalreview of the different conceptual hypotheses framing humanfocal epilepsy. Epileptic. Disord. 6, 77—83.

unez, P.L., 1995. Quantitative states of neocortex. In: Nunez, P.L.(Ed.), Neocortical Dynamics and Human EEG Rhythms. OxfordUniversity Press, pp. 1—18.

uwer, M., Lehmann, D., Lopes da Silva, F., Matsuoka, S., Sutherling,W., Vibert, J.F., 1994. IFCN guidelines for topographic and fre-quency analysis of EEGs and EPs. Report of an IFCN committee.Electroencephalogr. Clin. Neurophysiol. 91, 1—5.

’Muircheartaigh, J., Vollmar, C., Barker, G.J., Kumari, V., Symms,M.R., Thompson, P., Duncan, J.S., Koepp, M.J., Richardson, M.P.,2011. Focal structural changes and cognitive dysfunction in juve-nile myoclonic epilepsy. Neurology 76, 34—40.

anayiotopoulos, C., 2005. Idiopathic generalized epilepsies. In:Panayiotopoulos, C. (Ed.), The Epilepsies. Seizures, Syndromes,and Management. Springer, pp. 271—348.

ascual-Marqui, R.D., Michel, C.M., Lehmann, D., 1994. Low resolu-tion electromagnetic tomography. A new method for localizingelectrical activity in the brain. Int. J. Psychophysiol. 18,49—65.

ascual-Marqui, R.D., 2002. Functional imaging with low-resolutionbrain electromagnetic tomography (LORETA): a review. MethodsFind. Exp. Clin. Pharmacol. 24, 91—96.

ascual-Marqui, R.D., Esslen, M., Kochi, K., Lehmann, D., 2002.Functional imaging with low-resolution brain electromagnetictomography (LORETA): review, new comparisons, and new vali-dation. Jpn. J. Clin. Neurophysiol. 30, 81—94.

lummer, C., Harvey, A.S., Cook, M., 2008. EEG source localizationin focal epilepsy: where are we now? Epilepsia 49, 201—218.

uskás, S., Bessenyei, M., Fekete, I., Hollódy, K., Clemens, B.,2010. Quantitative EEG abnormalities in persons with ‘‘pure’’epileptic predisposition without epilepsy: a low resolution elec-tromagnetic tomography (LORETA) study. Epilepsy Res. 91,

94—100.

iao, X., Noebels, J.L., 1993. Developmental analysis of hippocam-pal mossy fiber outgrowth in a mutant mouse with inheritedspike—wave seizures. J. Neurosci. 13, 4622—4635.

2

R

R

R

R

S

S

S

S

S

S

S

S

S

T

T

v

v

V

W

W

W

W

92

eutens, D.C., Berkovic, S.F., 1992. Increased cortical excitabilityin generalised epilepsy demonstrated with transcranial magneticstimulation. Lancet 339, 362—363.

eutens, D.C., Berkovic, S.F., Macdonell, R.A.L., Bladin, P.F., 1993.Magnetic stimulation of the brain in generalised epilepsy: rever-sal of cortical hyperexcitability by anticonvulsants. Ann. Neurol.34, 351—355.

obinson, P.A., Rennie, C.J., Rowe, D.L., O’ Connor, S.C., 2004.Estimation of multiscale neurophysiologic parameters by elec-troencephalographic means. Hum. Brain Mapp. 23, 53—72.

odin, E., 1984. Sleep deprivation and epileptological implications.In: Degen, R., Niedermeyer, E. (Eds.), Epilepsy, Sleep and SleepDeprivation. Elsevier, Amsterdam, pp. 293—301.

akurai, K., Takeda, Y., Tanaka, N., Kurita, T., Shiraishi, H.,Takeuchi, F., Nakane, S., Sueda, K., Koyama, T., 2010. General-ized spike—wave discharges involve a default mode network inpatients with juvenile absence epilepsy: a MEG study. EpilepsyRes. 89, 176—184.

avic, I., Seitz, R.J., Pauli, S., 1998. Brain distortions in patientswith primarily generalized tonic—clonic seizures. Epilepsia 39,364—370.

avic, I., Lekwall, A., Greitz, D., Helms, G., 2000. MR spectroscopyshows reduced frontal lobe concentrations of N-acetylaspartatein patients with juvenile myoclonic epilepsy. Epilepsia 41,290—296.

hackman, A.J., McMenamin, B.W., Maxwell, J.S., Greischar, L.L.,Davidson, R.J., 2010. Identifying robust and sensitive frequencybands for interrogating neural oscillations. Neuroimage 51,1319—1333.

pencer, S.S., 2002. Neural networks in human epilepsy: evidenceof and implications for treatment. Epilepsia 43, 219—227.

tefan, H., Paulini-Ruf, A., Hopfengärtner, R., Rampp, S., 2009.Network characteristics of idiopathic generalized epilepsies incombined MEG/EEG. Epilepsy Res. 85, 187—198.

teriade, M., 2001. Impact of network activities on neuronal prop-erties in corticothalamic systems. Am. Physiol. Soc. 86, 1—39.

trug, L.J., Clarke, T., Chiang, T., Chien, M., Baskurt, Z., Li, W.,Dorfman, R., Bali, B., Wirrell, E., Kugler, S.L., Mandelbaum,

D.E., Wolf, S.M., McGoldrick, P., Hardison, H., Novotny, E.J., Ju,J., Greenberg, D.A., Russo, J.J., Pal, D.K., 2009. Centrotempo-ral sharp wave EEG trait in rolandic epilepsy maps to ElongatorProtein Complex 4 (ELP4). Eur. J. Hum. Genet. 17, 1171—1181.

Z

B. Clemens et al.

zava, S., Valdes, P., Biscay, R., Galan, L., Bosch, J., Clark, I.,Jimenez, J.C., 1994. High resolution quantitative EEG analysis.Brain Topogr. 6, 211—219.

ae, W.S., Kim, S.H., Joo, E.Y., Han, S.J., Kim, I.Y., Kim, S.I., Lee,J.M., Hong, S.B., 2008. Cortical thickness abnormality in juve-nile myoclonic epilepsy. J. Neurol. 255, 561—566.

imofeev, I., Steriade, M., 2004. Neocortical seizures: ini-tiation, development and cessation. Neuroscience 123,299—336.

an Beijsterveldt, C.E.M., Boomsma, D.I., 1994. Genetics of thehuman electroencephalogram (EEG) and event-related brainpotentials (ERPs): a review. Hum. Genet. 94, 319—330.

an Gelder, N.M., Siatitsas, I., Menini, C., Gloor, P., 1983. Felinegeneralised penicillin epilepsy: changes of glutamine acid andtaurin parallel to the progressive increase in excitability of thecortex. Epilepsia 24, 200—213.

ulliemoz, S., Vollmar, C., Koepp, M.J., Yogarajah, M.,O’Muircheartaigh, J., Carmichael, D.W., Stretton, J.,Richardson, M.P., Symms, M.R., Duncan, J.S., 2010. Con-nectivity of the supplementary motor area in juvenilemyoclonic epilepsy and frontal lobe epilepsy. Epilepsia,doi:10.1111/j.1528-1167.2010.02770.

illoughby, J.O., Fitzgibbon, S.P., Pope, K.J., Mackenzie, L.,Medvedev, A.V., Clark, C.R., Davey, M.P., Wilcox, R.A., 2003.Persistent abnormality detected in the non-ictal electroen-cephalogram in primary generalized epilepsy. J. Neurol.Neurosurg. Psychiatry 74, 51—55.

inawer, M.R., Rabinowitz, D., Pedley, T.A., Hauser, W.A., Ottman,R., 2003. Genetic influences on myoclonic and absence seizures.Neurology 61, 1576—1581.

oermann, F.G., Sisodiya, S.M., Free, S.L., Duncan, J.S., 1998.Quantitative MRI in patients with idiopathic generalizedepilepsy. Evidence of widespread cerebral structural changes.Brain 121, 1661—1667.

oermann, F.G., Free, S.L., Koepp, M.J., Sisodiya, S.M., Duncan,J.S., 1999. Abnormal cerebral structure in juvenile myoclonicepilepsy demonstrated with voxel-based analysis of MRI. Brain122, 2101—2108.

ietsch, B.P., Hansen, J.L., Hansell, N.K., Geffen, G.M., Martin,N.G., Wright, M.J., 2007. Common and specific genetic influ-ences on EEG power bands delta, theta, alpha, and beta. Biol.Psychol. 75, 154—164.