Introduction

Juvenile myoclonic epilepsy (JME) is a syndrome ofidiopathic generalized epilepsy (IGE) with an age-related seizure onset and typical clinical featuresconsisting of bilateral, single or repetitive arrhyth-mic, irregular myoclonic jerks, predominantly in thearms. Jerks may cause some patients to fall suddenly.No disturbance of consciousness is noticeable. Oftenthere are generalized tonic–clonic seizures. The sei-zures usually occur after awakening and are often

precipitated by sleep deprivation. Interictal and ictalelectroencephalograms (EEG) have rapid generalized,often irregular spike-waves and polyspike-waves;there is no close phase correlation between EEGspikes and jerks. Frequently, the patients are pho-tosensitive. Response to appropriate drugs is good[18].

Neuroimaging studies in JME have been rarelyperformed. Voxel-based morphometry (VBM) withmagnetic resonance imaging (MRI) in JME patientshowed increased gray matter concentration (GMC)

Woo Suk TaeEun Yeon JooSun Jung HanKyung-Han LeeSeung Bong Hong

CBF changes in drug naive juvenilemyoclonic epilepsy patients

Received: 18 January 2006Received in revised form:20 October 2006Accepted: 25 October 2006Published online: 12 March 2007

j Abstract Purpose The role ofthalamus and brainstem in gener-alized epilepsy has been suggestedin previous studies. The aim of thepresent study was to assess regio-nal cerebral blood flow (rCBF)abnormality in juvenile myoclonicepilepsy (JME) patients. Methods99mTc-ethylcysteinate dimer brainsingle photon emission computedtomography (SPECT) was per-formed in 19 drug naive JMEpatients and 25 normal controlswith the similar age and genderdistribution. Differences of rCBFbetween a JME group and a nor-mal control group were examinedby the statistical parametric map-ping of brain SPECT images usingindependent t test. The regressionanalyses in SPM were also per-formed between rCBF and the ageof seizure onset or the diseaseduration in JME group. ResultsCompared to normal controls, the

JME group showed a significantrCBF reduction in bilateral thal-ami, red nucleus, midbrain, pons,left hippocampus, and in thecerebelli (FDR corrected p < 0.01)whereas rCBF increase in the leftsuperior frontal gyrus (uncor-rected p < 0.001 but FDR cor-rected p > 0.05). Disease durationwas negatively correlated withrCBF in bilateral frontal cortices,caudate nuclei, brainstem andcerebellar tonsils. ConclusionsOur results suggest that abnormalneural networks in the thalamus,hippocampus, brainstem and cer-ebellum are associated with JME.

j Key words juvenile myoclonicepilepsy Æ cerebral blood flow ÆSPECT Æ thalamus Æbrainstem

ORIGINAL COMMUNICATIONJ Neurol (2007) 254:1073–1080DOI 10.1007/s00415-006-0491-6

JON

2491

W.S. Tae, MS Æ S.J. Han, MDS.B. Hong, MD, PhD (&)Dept. of NeurologySamsung Medical Center & Center forClinical Medicine, SBRI, SungkyunkwanUniversity School of Medicine50 Irwon-dong, Gangnam-guSeoul, KoreaTel.: +82-2/34103592Fax: +82-2/34100052E-Mail: sbhong@smc.samsung.co.kr

K.-H. Lee, MDDept. of Nuclear MedicineSamsung Medical Center & Center forClinical Medicine, SBRI, SungkyunkwanUniversity School of MedicineSeoul, Korea

E.Y. Joo, MDDept. of NeurologyCollege of Medicine, Ewha WomansUniversitySeoul, Korea

in the left mesiofrontal cortex, which may be relatedto dystopic neurons and microdysgenesis [47]. On thecontrary, another VBM study in JME patients showeddecreased gray matter concentration in frontal lobe(bilateral gyri rectus, anterior orbital gyri, left anteriormiddle frontal gyrus, right anterior superior frontalgyrus) [41]. Magnetic resonance spectroscopy studyreported low concentrations of thalamic N-acetylas-partate (NAA) in JME [4, 30]. A FDG-PET study inJME patients demonstrated that thalamic glucosemetabolism is increased interictally and is correlatedwith the amount of spike-wave activity [21]. Theysuggest that the thalamus has an important role in thegeneration of spontaneous generalized spike-wavedischarges in JME. However, cerebral blood flowchanges has not been studied in JME patients.

The likely polygenic nature of JME has beenstressed [32, 40], and subcortical nuclei have beenreported to be involved in primary generalized sei-zures. Low concentrations of thalamic N-acetylas-partate (NAA) were reported in JME [4, 30], andthalamocortical network was emphasized for thegenesis of generalized epilepsy [5]. The critical role ofbrainstem in generalized seizures has also been re-ported in the animal studies [11, 28, 34]. Antiepilepticdrug (AED) affects regional cerebral blood flow(rCBF), cerebral glucose metabolism, and neuro-transmitter systems [6, 15, 39, 43, 44]. Valproate(VPA) is the most commonly used antiepileptic drugin the treatment of JME. Regional changes in Gamma-aminobutyric acid (GABA) level, aspartic acid andglutamic acid concentrations have been found in thecortex, striatum, hippocampus, and cerebellum in ratsafter VPA treatment [6]. VPA has been reported todecrease cerebral metabolic rate for glucose(CMRGlc) and CBF in 10 normal volunteers [15].Thus, drug naive patients are better than medicatedpatients for the studies of drug effect on cerebralblood flow or metabolism.

Statistical parametric mapping (SPM) is an auto-mated, voxel-wise software for neuroimage analysis[14], and a useful tool for brain mapping of functionalchanges in brain diseases [13]. To investigate rCBFchanges in JME, we performed 99mTc-ehtylcysteinatedimer (ECD) brain single photon emission computedtomography (SPECT) in drug naı̈ve JME patients andnormal controls. Then, the difference of rCBF betweentwo groups was mapped by a SPM.

Patients and methods

j Patients

We prospectively selected 19 patients [mean age ± SD:19.0 ± 4.1 years (ranged 15–26), 5 males, 14 females]

who were diagnosed as having JME based on theInternational League Against Epilepsy classification ofepilepsy and epileptic syndrome [33]. The age of onsetwas 13.6 ± 1.8 (mean ± SD; ranged 12–18) years, andthe duration of disease was 6.2 ± 6.3 (ranged 1–13)years. All patients had myoclonic jerks characterizedby short, bilateral, and usually symmetric synchro-nous muscle contractions predominantly involvingthe shoulders and arms with preservation of con-sciousness and less frequent or rare generalized tonic–clonic seizure (GTCS). The frequency of myoclonicjerks was 81.4 ± 150.3 per year whereas that of GTCSwas 0.7 ± 0.4 (range: 0–8) per year. Myoclonic jerkswere single or repetitive and they were usuallyarrhythmic. The amplitude and the force of the jerksvary. Some jerks occurred unilaterally. GTCS wereoften preceded by generalized mild to moderatemyoclonus of increasing frequency and intensity.

In all patients, neurological examination was nor-mal, but EEG showed generalized polyspike-waves,spike-waves, polyspikes or spikes with normal alpharhythms. To remove the effects of AED on rCBF, onlyJME patients with no history of AED medication wereincluded in this study.

We recruited similarly age and gender distributed25 normal subjects [mean age ± SD: 20.4 ± 3.3 years(ranged 15–26), 9 males, 16 females] with no familialor personal history of neurological or psychiatricdiseases, or head trauma. There were no structural orsignal abnormalities on brain MRI in either the pa-tients or normal controls. Informed consent, accord-ing to the protocol of the Institutional Review Board,Samsung Medical Center, was obtained from all nor-mal controls and patients for permission to admin-ister a radioactive substance for SPECT scan.

j 99mTc-ECD brain SPECT

99mTc- ECD (925 MBq) was injected intravenously forthe SPECT studies. Brain SPECT images were ob-tained for 15 minutes at 30 minutes after the ECDinjection using a three-headed Triad XLT systemequipped with low-energy and high-resolution colli-mators (Trionix Research Laboratory, Twinsburg,OH, USA). The transaxial system resolution of thiscamera was 6.9 mm full width at half maximum.Images were reconstructed using a filtered back-pro-jection and a Butterworth filter. Attenuation correc-tion was performed using Chang’s method(attenuation coefficient = 0.12 cm)1). The recon-structed spatial resolution of SPECT was3.56 · 3.56 · 3.56 mm (x, y, z, respectively).

99mTc-ECD was injected intravenously when thepatient concerned had been free of seizure at least forthe previous 24 hours, and we also confirmed that no

1074

myoclonic seizures or generalized tonic clonic sei-zures had occurred from ECD injection to the end ofSPECT scanning. The brain SPECT was performed atleast 2 days after last GTCS had occurred. BrainSPECT was performed at 2–3,654 (mean ± SD:346 ± 900) days after last GTCS had occurred.

j SPM analysis

SPM99 (Wellcome Department of Imaging Neurosci-ence, Institute of Neurology, University College Lon-don, UK) and MATLAB 6.5 (The MathWorks, Natick,MA) were used for image processing. All SPECTimages of normal controls and JME patients werespatially normalized to the standard SPECT templatein SPM99 with default options, and smoothed using a14 mm full width at half maximum (FWHM) Gauss-ian kernel. The count of each voxel was normalized tothe total count of the brain (proportional scaling) toremove differences in global CBF between individuals.The unpaired t-test was used for group comparisonbetween the controls and JME patients.

To test the prior hypothesis of thalamus andbrainstem dysfunction in idiopathic generalized epi-lepsy [4, 5, 11, 28, 30, 34], small volume correction(SVC) was adjusted in these areas.

The regression analyses in SPM were performedbetween rCBF and the age of seizure onset or thedisease duration in JME group. To test the age effecton rCBF, Pearson correlation analysis between ageand rCBF was conducted in normal controls.

In SPM analyses, the height threshold was consid-ered to be significant at the level of false discovery rate(FDR) corrected p < 0.05 [16]. The extent thresholdfor whole brain was set to kE > 125 [27]. The FDRcorrection was performed using FDR toolbox (http://www.sph.umich.edu/~nichols/FDR). The heightthreshold for SVC analysis was rigorously set to FWEcorrected p < 0.05 without extent threshold to localizebrain regions with hypoperfusion because FDR cor-rected p < 0.05 showed too large areas with hypop-erfusion to localize and all of inter-connected clustersdisappeared at a time when a significance level wasincreased (p value 0.05 fi 0.01) in FDR analysis.

To explore the changes of frontal lobe gray matterconcentration in previous study [47], thresholdedcontrast image was presented with the significancelevel of uncorrected p < 0.05 [19]. For volume basedanalysis of thalamus, the normalized mean intensitiesof bilateral thalami by whole brain intensity wereextracted with predefined volume of interest (VOI) inall JME patients. Two-tailed Pearson correlationanalysis using SPSS version 11.5 (SPSS Inc, Chicago,III) was performed between disease duration and themean intensities of thalami. Data normality was ver-

ified using the Kolmogorov-Smirnov test. The statis-tical significance level was set to p < 0.05.

The regional definitions for the VOI and SVCanalyses used Maldjian atlas [26]. The regions ofclusters were named by Henri M. Duvernoy’s atlas [9,10].

Results

j SPM analysis of brain SPECT images between JMEpatients and normal controls

The SPM result shows decreased rCBF in bilateralthalami (ventral posterolateral thalamic nucleus,pulvinar), brain stem (colliculi, pons, and medulla),and cerebellum in JME group compared to the normalcontrol group. In the limbic area, JME group showeddecreased rCBF in the left hippocampus, both para-hippocampal gyri, both lingual, fusiform gyri, and leftgyrus rectus of the frontal lobe (Fig. 1, Table 1). Theclusters of brain regions with decreased rCBF arewidely connected (Fig. 1). The overall size of clustersis 25,054 voxels.

In the SVC analysis for thalamus and brainstem,rCBF in bilateral pulvinars, red nucli, colliculi, pons,and medulla also significantly decreased.

rCBF was increased in the left superior frontalgyrus (local maxima: x, y, z = )6, 15, 67, Z = 3.84,kE = 198), right postcentral gyrus (local maxima: x, y,z = 51, )21, 45, Z = 3.37, kE = 150 voxels), rightsuperior frontal gyrus (local maxima: x, y, z = 22, 17,34, Z = 3.33, kE = 40 voxels), but these results are notsignificant after the correction for the multiple com-parison problems (Fig. 2). The frontal lobe showed nosignificant decrease of rCBF at uncorrected p < 0.05.

j Correlation analysis between rCBF and clinicalvariables in JME group

The disease duration was negatively correlated withrCBF in the bilateral frontal cortices, bilateral cau-date nuclei, left putamen, right anterobasal temporalgyrus, bilateral brainstem, and cerebellar tonsils (seeFig. 1C). The age of seizure onset was not corre-lated with rCBF. The correlation analysis betweenage and rCBF in normal controls to explore ageeffect on rCBF showed no significant positive ornegative correlations at the significance of uncor-rected p < 0.05.

Although the correlation between rCBF of thala-mus and disease duration was not found by SPM atFDR corrected p < 0.05, rCBF of thalamus was neg-atively correlated with disease duration at the level ofuncorrected p < 0.005 (T = 2.9) as similar as the re-

1075

sult of VOI analysis (left: r = )0.51, p = 0.025; right:r = )0.64, p = 0.003). Seizure frequencies of myo-clonic jerks or GTCS or the time since last GTCS werenot correlated with rCBF changes in our JME patients.

Discussion

The present study showed that rCBF decreased inlimbic areas, thalamus, brainstem, and the cerebellum

Table 1 Decreased blood flow in AED naive juvenile myoclonic epilepsy

Location Side Coordinates (mm)* Peak Z FDR corrected p SVC

x y z FWE corrected p

ThalamusPulvinar L )10 )25 )2 3.60 0.007 0.002Pulvinar R 18 )31 3 3.76 0.006 0.002

BrainstemRed nucleus L )10 )20 )2 3.42 0.010 0.042Red nucleus R 8 )20 )7 3.51 0.009 0.01Superior/inferior colliculi B 0 )29 )6 4.06 0.004 0.003Pons B 0 )27 )32 3.60 0.007 0.01Medulla B 0 )39 )37 4.34 0.003 0.001

CerebellumPosterior lobe, cerebellar tonsil R 18 )49 )40 4.77 0.003 –Posterior lobe, inferior semi-lunar lobule R 0 )62 )32 4.56 0.003 –Posterior lobe, inferior semi-lunar lobule L )22 )64 )39 4.34 0.003 –Anterior lobe, culmen L )18 )44 )15 3.72 0.006 –

Nucleus accumbens, subcallosal gyrus L )8 9 )17 3.20 0.014 –Hippocampus, parahippocampal gyrus L )22 )20 )14 3.68 0.006 –Parahippocampal gyrus R 18 )37 )2 5.01 0.003 –Posterior Insula L )34 )29 13 3.90 0.005 –Inferior temporal, parahippocampal gyri R 32 4 )37 3.63 0.007 –Superior temporal gyrus L )30 8 )32 3.30 0.012 –Lingual, fusiform gyri L )24 )43 )3 3.65 0.006 –Lingual, fusiform gyri R 28 )46 )16 3.64 0.007 –

L, left; R, right; B, both; FDR, false discovery rate; FWE, familywise error for the multiple comparison problem*Talairach coordinate, SVC, small volume correction for the areas of thalamus, brain stem

Fig. 1 SPM result of brain SPECT images betweenJME patients and normal controls. (A) JME groupshowed decreased perfusion in the thalamus,brainstem (red nucleus, superior colliculus, pons,medulla), cerebellum left insula, left hippocampus,both parahippocampal gyri, both lingual, fusiformgyri, and left gyrus rectus of the frontal lobe (FDRcorrected p < 0.05). The color bar indicates T value.(B) Small volume correction adjusted result with themask of the thalamus and brainstem. Upper row: Thecoronal images show decreased perfusion in ventralposterolateral thalamic nucleus (VPL), pulviar (Pul) ofthalamus, and the brainstem. Middle row: The sagittalimages show decreased perfusion in the rostral/caudal raphe nuclei (RN) of pons. Lower row: The axialimages clearly show decreased perfusion in thesuperior colliculus (SC), inferior colliculus (IC), andpons (FWE corrected p < 0.05). (C) Correlationanalysis between rCBF and the disease duration inJME group. The rCBF was negatively correlated withthe disease duration in bilateral frontal cortices,caudate nuclei, left putamen, right anterobasaltemporal lobe, bilateral brainstem, and cerebellartonsils (FDR corrected p < 0.05). The color barindicates T value

1076

in JME patients, whereas the rCBF non-significantlyincreased in left superior frontal gyrus. These brainregions correspond to the cortico–ponto–cerebello–thalamo–cortical motor pathway (Fig. 1, Table 1).

j Decreased rCBF in thalamus

Abnormally low concentrations of thalamic N-acet-ylaspartate (NAA) were reported in the 20 IGE pa-tients including 12 JME patients. Moreover, decreasedNAA represents neuronal loss or metabolic dysfunc-tion. The NAA/Creatine ratio in the thalamus of JMEpatients was negatively correlated with seizure dura-tion, suggesting progressive thalamic neuronal dys-function [4, 30]. The negative correlation between themean rCBF of thalamus and disease duration in thepresent study also support the previous findings.

The pulvinar is the most enlarged in the primatebrain, especially in the human brain, and its devel-opment seems to parallel the enlargement the asso-ciation regions of the parietal–occipital–temporalcortex. It is extensively interconnected with wide-spread regions of the parietal, temporal, and theoccipital lobes, as well as with the superior colliculusand other nuclei of the brain stem related to vision[2]. Our result showed rCBF decrease in the pulvinar,lingual and fusiform gyri (see Fig. 1). These areas arerelated with visual function [1, 37], and the dysfunc-tion of these areas may be associated with thephotosensitivity of JME [3, 18].

j Decreased rCBF in brainstem and cerebellum

The rCBF was significantly decreased in the midbrainincluding superior/inferior colliculi in our JME pa-tients. The superior colliculus and the nearby inter-collicular nucleus are parts of a midbrainanticonvulsant zone capable of suppressing general-ized tonic–clonic seizures [28], and are also importantcomponents of the brainstem seizure network [34].

A recent study reported reductions in serotonin 1Areceptor binding in the raphe nucleus, hippocampus,and anterior dorsolateral frontal cortex of JME patients.These observed reductions in serotonin 1A receptorbinding potential (BP) suggest that the serotonin sys-tem is affected in JME [29]. The present study showeddecreased rCBF in areas of the pons corresponding tothe raphe nucleus (Fig. 1B), where the major seroto-nergic pathways arise. Caudal raphe nuclei project tothe cerebellar cortex and spinal cord whereas rostalraphe nuclei project to the thalamus, nucleus accum-bens, hypothalamus, amygdala, hippocampus, and thecortex [20]. This serotonergic pathway of raphe nucleiwas found to be concordant with brain regions of de-creased rCBF in our JME patients (Fig. 1A).

Our study shows a negative correlation betweenmidbrain rCBF and disease duration (Fig. 1C). Thus,decreased rCBF in midbrain structures may not onlybe an underlying dysfunction of these structures butalso the consequences of repetitive seizures in JME.

Lowenthal and Horsley [25] and Sherrington [36]independently found that electrical stimulation of thecerebellum could temporarily abolish decerebraterigidity. In some animal models of epilepsy [22, 42],electrical stimulation of the cerebellum was found tosuppress seizures. Cerebellar stimulating implantswere tried in refractory epilepsy patients with varyingdegrees of success [7, 12, 24, 46]. These observationssuggest that cerebellum has some inhibitory effects onseizure generation. The reduced inhibitory input fromPurkinje cells to cerebellar nuclei could increasethalamocortical activity [17, 35].

The significant decrease of cerebellar rCBF in ourJME patients suggests that the reduction of inhibitoryactivity of the cerebellum may contribute to generateseizures in JME.

j rCBF increase in frontal and parietal lobes

[11C]flumazenil (FMZ) positron emission tomogra-phy (PET) study in 10 IGE patients, which includedfive JME patients, demonstrated a significantly highermean [11C]FMZ volume of distribution (FMZVD) inthe cerebral cortex, thalamus, and cerebellum than incontrols [23]. In subgroup analysis, FMZVD differ-ences were observed in the cerebral cortex and cere-bellum in the JME group, which suggests that afinding of increased FMZVD could reflect microdys-genesis or a state of cortical hyperexcitability.

Another study [47] reported increased gray matterconcentration in the superior frontal gyrus of JME,and they suggested that the finding of increased graymatter concentration reflects microdysgenesis. Thisstudy result is supported by our findings showingincreased rCBF in the bilateral superior frontal andpericentral gyri (Fig. 2). This result is not confirma-tory because no voxel showed statistical significanceafter the correction for multiple comparison prob-lems. However, unthresholded contrast image mayprovide more information for the exploratory analy-ses [19]. Thus, the issue of cortical hyperperfusionduring the interictal period in JME patients warrantsfurther investigation.

j Negative correlation between JME disease durationand rCBF

rCBFs of the frontal cortex, basal ganglia, and brain-stem were found to be negatively correlated with JMEdisease duration in the present study (Fig. 1C), which

1077

suggests a progressive nature of rCBF decrease inJME. Frontal lobe and visuospatial dysfunctions havebeen reported in JME patients [8], but intelligencequotient remains relatively unimpaired [38]. Age doesnot appear to affect the result because there was nosignificant correlation between rCBF and age in nor-mal controls. When we performed SPM analysis afterone possible outlier was excluded, the result was notdifferent from that of all patients.

The correlation between patients’ ages and diseasedurations is very high (r = 0.959, p < 0.001, Pearsoncorrelation coefficient) because of the narrow range ofonset ages (ranged 12–18 years; seven 12 years, five13, two 14, two 15, one for 16, 17 and 18) in ourpatients. SPM result of correlation of rCBF with age inpatients is very similar to the result of rCBF withdisease duration whereas there was no significantcorrelation (at uncorrected p < 0.05) of rCBF with agein normal controls. So the result of negative correla-tion of rCBF with age in patients may be affected bythe disease duration effect. When we re-analyzedstatistically the correlation of rCBF with age in pa-tients with controlling disease duration, no brain re-gions showed significant correlation between rCBFand age.

However, a volumetry study reported that nor-malized volume of right frontal lobe is negativelycorrelated with the disease duration [41]. If graymatter volume goes down with disease duration,partial volume effect on its own could explain rCBFapparently decreasing with disease duration, withoutany functional background.

j Methodological considerations

In this study, one methodological limitation causedby the relatively low spatial resolution of brain SPECTshould be mentioned. The question concerns whetherthe spatial resolution of SPECT is sufficient to reflectrCBF changes in the small sub-structures of the hu-man brain. The intrinsic spatial resolution of theSPECT gamma camera used in this study was 6.9 mm.We measured the widths, heights, thicknesses ofbrainstem substructures in the coronal MRIs of 10JME patients using a brain atlas guide [9]. The meansizes measured were 10.2 · 11.1 · 9.0 mm(width · height · thickness) for right red nucleus,and 16.7 · 10.8 · 9.8 mm for superior/inferior colli-culi. The sizes of red nucleus and superior/inferiorcolliculi were 1.5 or 2 times larger than the SPECTspatial resolution. The finding of blood flow differ-ences in small structures is most likely due to groupaveraging which can bring out anatomical or func-tional features not visible on individual scans.Therefore, the decreased rCBF observed for thesestructures in our study is meaningful. Furthermore,this result concurs with those of previous studies [11,28, 29, 34, 45]. However, the present study, a cross-sectional study of blood flow changes, cannot add tothe elucidation of why seizures in JME are eithergenerated or appear generalized, either.

EEG was not recorded concurrently with SPECTscanning in our patients. A previous study showedthat the summated durations of generalized spike-waves recorded during 40 min (brain uptake time)

Fig. 2 Brain regions showing increased rCBF in JMEpatients. To explore brain regions withhyperperfusion, SPM results at different significancelevels were presented. (A) 3D rendering view of SPMresult at uncorrected p < 0.05, (B) a glass view ofSPM result at uncorrected p < 0.05. (C) A glass viewof SPM result at uncorrected p < 0.001. The extentthresholding was not performed in all images

1078

after 18F-fluoro-2-deoxy-D-glucose (FDG) injection forFDG–PET study in JME patients were correlated withincrease in cerebral metabolism of bilateral thalami,cerebelli and midbrain [21]. But our result did notshow increased perfusion in these brain regions. Be-cause the brain uptake time of radioisotope after theinjection for brain SPECT is very short (about30 seconds to 1 minute) [31], the possibility of gen-eralized spikes occurring during this short period islow. So interictal generalized discharges are less likelyto affect our result.

Conclusions

In summary, we describe a novel finding of decreasedrCBF in the thalamus, limbic areas, brainstem, andcerebellum of drug naive JME patients, and the diseaseduration was negatively correlated with rCBF in

bilateral frontal lobes, basal ganglia, brainstem andcerebellar tonsils. These results suggest that dysfunc-tions in the thalamus, limbic areas, brainstem andcerebellum are related with JME. Thus further studiesin these brain regions are necessary in JME patients.

j Acknowledgments This work was supported by a grant (no.A050462) of the Good Health R&D Project, Ministry of Health &Welfare, Republic of Korea, and by a grant (M103KV010016-06K2201-01610) from Brain Research Center of the 21st CenturyFrontier Research Program funded by the Ministry of Science andTechnology of the Republic of Korea, and by the Samsung Bio-medical Research Institute grant, #SBRI C-A6-435-1.Author contributions Tae WS performed SPM analysis of SPECTimages, and wrote the manuscript. Joo EY and Han SJ gatheredand summarized clinical data and did taking care of patientsduring study period. Lee KH helped designing a study protocoland performed SPECT scan. Hong SB (corresponding author)established a study protocol, recruited JME patients and normalcontrols, and supervised all procedures of study and manuscriptwriting.

References

1. Allison T, McCarthy G, Nobre A, PuceA, Belger A (1994) Human extrastriatevisual cortex and the perception of fa-ces, words, numbers, and colors. CerebCortex 4:544–554

2. Amaral DG (2000) The functionalorganization of perception and move-ment. In: Kandel ER, Schwartz JH,Jessell TM (eds) Principles of neuralscience. McGraw-Hill, New York, pp337–348

3. Appleton R, Beirne M, Acomb B (2000)Photosensitivity in juvenile myoclonicepilepsy. Seizure 9:108–111

4. Bernasconi A, Bernasconi N, NatsumeJ, Antel SB, Andermann F, Arnold DL(2003) Magnetic resonance spectros-copy and imaging of the thalamus inidiopathic generalized epilepsy. Brain126:2447–2454

5. Blumenfeld H (2003) From moleculesto networks: cortical/subcortical inter-actions in the pathophysiology of idi-opathic generalized epilepsy. Epilepsia44(Suppl 2):7–15

6. Chapman AG, Riley K, Evans MC,Meldrum BS (1982) Acute effects ofsodium valproate and gamma-vinylGABA on regional amino acid metab-olism in the rat brain: incorporation of2-[14C]glucose into amino acids. Neu-rochem Res 7:1089–1105

7. Davis R, Eagle H, Kudsmal J, Gray E,Ryan T, Dusnak A (1982) Update ofchronic cerebellar stimulation forspasticity and epilepsy. Appl Neuro-physiol 45:44–50

8. Devinsky O, Gershengorn J, Brown E,Perrine K, Vazquez B, Luciano D(1997) Frontal functions in juvenilemyoclonic epilepsy. NeuropsychiatryNeuropsychol Behav Neurol 10:243–246

9. Duvernoy HM (1995) The human brainstem and cerebellum. Springer-Verlag,New York

10. Duvernoy HM (1999) The humanbrain. Surface, three-dimensional sec-tional anatomy with MRI, and bloodsupply. Springer-Verlag, New York

11. Faingold CL, Randall ME (1999) Neu-rons in the deep layers of superiorcolliculus play a critical role in theneuronal network for audiogenic sei-zures: mechanisms for production ofwild running behavior. Brain Res815:250–258

12. Fenton GW, Fenwick PBC, Brindley GS,Falconer MA, Polkey CE, Rushton DN(1977) Chronic cerebellar stimulationin the treatment of epilepsy: a pre-liminary report. In: Perry JK (ed) Epi-lepsy, the eighth internationalsymposium. Raven, New York, pp 333–340

13. Frackowiak RSJ, Friston KJ, Frith CD,Dolan RJ, Mazziotta JC (1997) Humanbrain function. Academic Press, USA

14. Friston KJ, Holmes AP, Worsley KJ,Poline JP, Frith CD, Frackowiak RSJ(1995) Statistical parametric maps infunctional imaging: a general linearapproach. Hum Brain Mapp 2:189–210

15. Gaillard WD, Zeffiro T, Fazilat S,DeCarli C, Theodore WH (1996) Effectof valproate on cerebral metabolismand blood flow: an 18F-2-deoxyglucoseand 15O water positron emissiontomography study. Epilepsia 37:515–521

16. Genovese CR, Lazar NA, Nichols T(2002) Thresholding of statistical mapsin functional neuroimaging using thefalse discovery rate. Neuroimage15:870–878

17. Ghez C, Thach WT (2000) The cere-bellum. In: Kandel ER, Schwartz JH,Jessell TM (eds) Principles of neuralscience (ed 4) McGraw-Hill, New York,pp 832–852

18. Janz D, Durner M (1997) Juvenilemyoclonic epilepsy. In: Engel J Jr,Pedley TA (eds) Epilepsy: a compre-hensive textbook. Lippincott-Raven,Philadelphia, pp 2389–2400

19. Jernigan TL, Gamst AC, Fennema-No-testine C, Ostergaard AL (2003) More‘‘mapping’’ in brain mapping: statisti-cal comparison of effects. Hum BrainMapp 19:90–95

20. Kandel ER (2000) Disorder of mood:depression, mania, anxiety disorder. In:Kandel ER, Schwartz JH, Jessell TM(eds) Principles of neural science.McGraw-Hill, New York, pp 1213–1216

21. Kim JH, Im KC, Kim JS, Lee SA, KangJK (2005) Correlation of interictalspike-wave with thalamic glucosemetabolism in juvenile myoclonic epi-lepsy. Neuroreport 16:1151–1155

1079

22. Kitamura M, Yamamoto T, KatayamaY, Nishimoto H, Tsubokawa T, Mori-yasu N (1978) Effects of cerebellarstimulation on cortical and limbic sei-zure discharges [proceedings]. ApplNeurophysiol 42:306

23. Koepp MJ, Richardson MP, Brooks DJ,Cunningham VJ, Duncan JS (1997)Central benzodiazepine/gamma-amin-obutyric acid A receptors in idiopathicgeneralized epilepsy: an [11C]flumaze-nil positron emission tomographystudy. Epilepsia 38:1089–1097

24. Levy LF, Auchterlonie WC (1979)Chronic cerebellar stimulation in thetreatment of epilepsy. Epilepsia 20:235–245

25. Lowenthal M, Horsley V (1897) On therelation between the cerebellar andother centers (namely cerebral andspinal) with special reference to theaction of antagonistic muscles. Proc RSoc Lond 61:20–25

26. Maldjian JA, Laurienti PJ, Kraft RA,Burdette JH (2003) An automatedmethod for neuroanatomic and cyt-oarchitectonic atlas-based interroga-tion of fMRI data sets. Neuroimage19:1233–1239

27. McNally KA, Paige AL, Varghese G,Zhang H, Novotny EJ Jr, Spencer SS,Zubal IG, Blumenfeld H (2005) Local-izing value of ictal-interictal SPECTanalyzed by SPM (ISAS). Epilepsia46:1450–1464

28. Merrill MA, Clough RW, Jobe PC,Browning RA (2003) Role of the supe-rior colliculus and the intercollicularnucleus in the brainstem seizure cir-cuitry of the genetically epilepsy-pronerat. Epilepsia 44:305–314

29. Meschaks A, Lindstrom P, Halldin C,Farde L, Savic I (2005) Regionalreductions in serotonin 1A receptorbinding in juvenile myoclonic epilepsy.Arch Neurol 62:946–950

30. Mory SB, Li LM, Guerreiro CA, CendesF (2003) Thalamic dysfunction injuvenile myoclonic epilepsy: a protonMRS study. Epilepsia 44:1402–1405

31. O’Brien TJ, O’Connor MK, Mullan BP,Brinkmann BH, Hanson D, Jack CR, SoEL (1998) Subtraction ictal SPET co-registered to MRI in partial epilepsy:description and technical validation ofthe method with phantom and patientstudies. Nucl Med Commun 19:31–45

32. Panayiotopoulos CP, Obeid T (1989)Juvenile myoclonic epilepsy: an auto-somal recessive disease. Ann Neurol25:440–443

33. Proposal for revised classification ofepilepsies and epileptic syndromes.Commission on Classification andTerminology of the International Lea-gue Against Epilepsy (1989) Epilepsia30:389–399

34. Raisinghani M, Faingold CL (2003)Identification of the requisite brainsites in the neuronal network subserv-ing generalized clonic audiogenic sei-zures. Brain Res 967:113–122

35. Savic I (2000) PET: central benzodiaz-epine neuroreceptor mapping in gen-eralized epilepsies. Adv Neurol 83:137–144

36. Sherrington CS (1897) Double (anti-drome) activation in the central ner-vous system. Proc R Soc Lond 61:243–246

37. Shibata T, Nishijo H, Tamura R, Mi-yamoto K, Eifuku S, Endo S, Ono T.Generators of visual evoked potentialsfor faces and eyes in the human brainas determined by dipole localization.Brain Topogr 15:51–63

38. Sonmez F, Atakli D, Sari H, Atay T,Arpaci B (2004) Cognitive function injuvenile myoclonic epilepsy. EpilepsyBehav 5:329–336

39. Spanaki MV, Siegel H, Kopylev L,Fazilat S, Dean A, Liow K, Ben-Mena-chem E, Gaillard WD, Theodore WH(1999) The effect of vigabatrin (gam-ma-vinyl GABA) on cerebral blood flowand metabolism. Neurology 53:1518–1522

40. Suzuki T, Delgado-Escueta AV, AguanK, Alonso ME, Shi J, Hara Y, NishidaM, Numata T, Medina MT, Takeuchi T,Morita R, Bai D, Ganesh S, Sugimoto Y,Inazawa J, Bailey JN, Ochoa A, Jara-Prado A, Rasmussen A, Ramos-Peek J,Cordova S, Rubio-Donnadieu F, InoueY, Osawa M, Kaneko S, Oguni H, MoriY, Yamakawa K (2004) Mutations inEFHC1 cause juvenile myoclonic epi-lepsy. Nat Genet 36:842–849

41. Tae WS, Hong SB, Joo EY, Lee JM, KimIY, Kim SI (2006) Structural brainabnormality in juvenile myoclonicepilepsy: volumetry and voxel basedmorphometry. Korean J Radiol 7:162–172

42. Testa G, Pellegrini A, Giaretta D (1979)Effects of electrical stimulation andremoval of cerebellar structures in anexperimental model of generalizedepilepsy. Epilepsia 20:447–454

43. Theodore WH, Bairamian D, NewmarkME, DiChiro G, Porter RJ, Larson S,Fishbein D (1986) Effect of phenytoinon human cerebral glucose metabo-lism. J Cereb Blood Flow Metab 6:315–320

44. Theodore WH, Bromfield E, Onorati L(1989) The effect of carbamazepine oncerebral glucose metabolism. AnnNeurol 25:516–520

45. Toczek MT, Carson RE, Lang L, Ma Y,Spanaki MV, Der MG, Fazilat S, Ko-pylev L, Herscovitch P, Eckelman WC,Theodore WH (2003) PET imaging of5-HT1A receptor binding in patientswith temporal lobe epilepsy. Neurology60:749–756

46. Van Buren J, Wood JH, Oakley J,Hambrecht F (1978) Preliminary eval-uation of cerebellar stimulation bydouble blind stimulation and biologicalcriteria in the treatment of epilepsy. JNeurosurg 48:407–416

47. Woermann FG, Free SL, Koepp MJ,Sisodiya SM, Duncan JS (1999)Abnormal cerebral structure in juvenilemyoclonic epilepsy demonstrated withvoxel-based analysis of MRI. Brain122:2101–210

1080