Multi-voxel magnetic resonance spectroscopy at 3 T in patients with idiopathic generalised epilepsy

Seizure 19 (2010) 485–492

Contents lists available at ScienceDirect

Seizure journal homepage: www.elsevier.com/locate/yseiz

Multi-voxel magnetic resonance spectroscopy at 3 T in patients with idiopathic generalised epilepsy M.T. Doelken a,*, A. Mennecke a, A. Stadlbauer c, L. Kecskeme´ti a,b, B.S. Kasper b, T. Struffert a, A. Doerfler a, H. Stefan b, Thilo Hammen a a b c

Department of Neuroradiology, University of Erlangen-Nuremberg, Schwabachanlage 6, 91054 Erlangen, Germany Department of Neurology, Epilepsy Center (ZEE), University of Erlangen-Nuremberg, Schwabachanlage 6, Erlangen, Germany Department of Neurosurgery, University of Erlangen-Nuremberg, Schwabachanlage 6, 91054 Erlangen, Germany

A R T I C L E I N F O

A B S T R A C T

Article history: Received 11 February 2010 Received in revised form 23 May 2010 Accepted 9 July 2010

Purpose: The objective of our study was to gain further insight into the extent of local metabolic alterations in patients with idiopathic generalised epilepsy (IGE), respectively, the subgroup with generalised tonic–clonic seizures (GTCS). The extent of regional metabolic involvement perhaps indicates the key structures in generation of seizures and involvement of specific network of dysfunction. Methods: Using the multi-voxel technique at a 3 T MRI Scanner metabolite levels of 25 age-matched healthy controls and 18 patients with GTCS were obtained from the basal ganglia, insular cortex, cingulum, hippocampus and along both hemispheres in the fronto-parietal white and grey matter. Results: Group analysis of GTCS patients versus healthy controls revealed significant (p < 0.05) decrease of tNAA in the cortex of the central region and cingulum, but also in the thalami. Glx was elevated broadly in both hemispheres, in particular in central region, cingulum, insular cortex and left putamen, yet also in the right thalamus. Cho and mI demonstrated a significant coincidental decrease pronounced in the grey and white matter of the central region. Significant metabolic correlation (p  0.05) based on tNAA, respectively, Glx occurred between the thalamus and the central region, cingulum, putamen and medial frontal cortex. In patients with >2 tonic–clonic seizures in the last 12 months a trend towards higher Glx and lower tNAA levels was observed. Discussion: Our results demonstrate the altered metabolic interconnection of cerebral anatomic regions in patients with GTCS, in particular the major role of basal ganglia-central region relay in seizure generation. ß 2010 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.

Keywords: Spectroscopy MRS MRI Epilepsy

1. Introduction MRI techniques as proton magnetic resonance spectroscopy (1H-MRS) give non-invasive insights to diverse functional and structural alterations in patients with neurological disorders as epilepsy.1–9 The multi-voxel technique refers to chemical shift imaging (CSI) which combines features of MR imaging and spectroscopy yielding data from multiple adjacent voxels eventually covering the whole brain.10–12 1H-MRS performed at a higher magnetic field strength as 3 T has the advantages of higher signalto-noise ratio (SNR) and improved spectral resolution in clinically acceptable scan times. Due to the advantages of high field strengths and cumulative availability of 3 T MRI multi-voxel techniques are increasingly used in clinical settings. Commonly

* Corresponding author. Tel.: +49 09131 85 39388; fax: +49 0 9131 85 36179. E-mail address: [email protected] (M.T. Doelken).

distinguishable metabolites include total N-acetyl aspartate (tNAA), choline (tCho), myo-inositol (mI), glutamate (Glu) plus glutamine (Gln) (Glu + Gln = Glx) and creatine (tCr). In patients with idiopathic generalised epilepsy (IGE) most previous studies were implemented using the single-voxel technique (SVS), acquiring a single spectrum from a definite volume of tissue as an integral.5,6,13–15 This acquisition is fairly fast (3 min) and a spectrum is easily obtained, however it provides limited information about regional distribution of brain metabolites. Only few recent studies implemented the multi-voxel technique in clinical settings with IGE.1,3,4 Juvenile myoclonic epilepsy (JME), childhood absence epilepsy (CAE) and epilepsy with generalised tonic–clonic seizures only (GTCS) are the most common syndromes of IGE with distinct electroclinical features and prognosis.16 They are considered to share absence of magnetic resonance imaging (MRI) findings and several clinical features, however differ in their predominant seizure types.17 In previous single- and multi-voxel 1H-MRS

1059-1311/$ – see front matter ß 2010 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.seizure.2010.07.005

486

M.T. Doelken et al. / Seizure 19 (2010) 485–492

studies of patients with IGE, including different subgroups, metabolic alterations were mainly found in the thalamic and prefrontal tNAA levels.1,3–6,13–15 The neurochemical abnormalities are considered to reflect thalamocortical dysfunction, however, metabolic alterations of specific cerebral networks in particular the extend of metabolic alterations in the mentioned subgroups still remain ambiguous.1,3–6 Using the multi-voxel technique at a 3 T scanner our purpose was to gain further insight into the extent of local metabolite alterations in patients with GTCS within grey matter sites of the basal ganglia including thalamus and putamen, insular cortex, hippocampus, cingulum and along both hemispheres in the frontoparietal white and grey matter. Our results might contribute further data regarding the extent of regional metabolic alteration, respectively, involvement of a specific network of dysfunction in patients with GTCS. Additionally, metabolic impairment in the context of seizure frequency was evaluated.

Table 1 Summary of clinical data of patients with GTCS [group I <2 seizures in the last 12 months; group II >2 seizure in the last 12 months; * = 1/5 patient with previous Levetiracetam medication had no current medication due to intolerance of Levetiracetam; SD = standard deviation].

2. Patients and methods

cortex (Fig. 1e). The last multi-voxel MRS section (Fig. 1c) in the hippocampi comprised the anterior portion of the hippocampus (H1) as well as the medial (H2) and posterior portion (H3) (Fig. 1f). The positioning of the multi-voxel MRS sections was performed on voxels graphically prescribed from T2-weighted images (TR 4000 ms, TE 94 ms, slices 19–21, FOV 200, Matrix 218  384). The voxels were then arranged into symmetrical pairs on each side of the midline by an experienced neuroradiologist. All measurements were performed by the same neuroradiologist to ensure reliable and comparable positioning of the sections. The caudo-mesial portion of frontal lobes was avoided owing to susceptibility artefacts generated from this region. Fig. 2a–c shows examples of acquired spectra. For all three multi-voxel MRS sections a 2D-MRS sequence was used (point-resolved spectroscopy (PRESS), TE = 30 ms, TR = 1700 ms, averages 128). Corresponding unsuppressed water spectra with equal TE and TR were additionally acquired. Before data collection the automatic shim provided by the manufacturer was carried out to optimise field homogeneity. In some cases manual post-shimming was performed.

2.1. Subjects–inclusion criteria 18 patients with diagnosis of GTCS16 (10 men, 8 women; age range 18–39 years, average 28 years) were included. Demographic and clinical data were obtained through interviews with the patients and by reviewing hospital charts. Diagnosis of idiopathic epilepsy, respectively, GTCS was based on seizure history and semiology, EEG and interictal  ictal video-EEG. Estimation of the frequency of seizures was based on questioning of the patients. The status of seizure control was defined as follows: group I <2 tonic– clonic seizures in the last 12 months, group II >2 tonic–clonic seizures in the last 12 months. Only patients with negative findings in highfield MRI were enclosed in the study. To exclude missed subtle pathological findings in conventional MRI as, e.g. cortical dysplasia or heterotopia a voxel-based morphometric analysis with a T1-weighted MRI volume data set was additionally performed using a technique based on free available statistical parametric mapping software (SPM) assembled by Huppertz et al.18 Major clinical data of the eventually enclosed patients are listed in Table 1. 1H-MRS was shifted if the last tonic–clonic seizure was within the last 14 days. The normal reference group comprised 25 age-matched healthy young adults (12 men, 13 women; age range 19–40 years, average 27 years), who were equally examined with multi-voxel 1H-MRS at the same 3 T scanner. All subjects were free from neurological or psychiatric diseases, and their brain MRI scans were normal. All subjects participated voluntarily and informed consent was obtained for each subject. Studies have shown that metabolite values do not significantly vary in adolescents and adults.19–21 Therefore, we recruited adults from a limited age range, thus expecting no significant variation of the metabolites due to age. The study was approved by the local ethics committee. 2.2. Multi-voxel MR spectroscopy Multi-voxel 1H-MRS was performed at a 3 T Siemens Magnetom Trio (Siemens Medical Solutions, Erlangen Germany) with an 8 channel head array coil. Both patients suffering from epilepsy and healthy adults were equally examined using multi-voxel MRS sections acquired according to Fig. 1. The multi-voxel MRS sections in Fig. 1a–f consisted of 16  16 voxels. The volume of interest (VOI) of the multi-voxel MRS sections was 180 mm  180 mm  10 mm, thus the volume of one voxel was (180/16)2  10 = 1.3 cm3. The first multi-voxel section (Fig. 1a and b) encompassed parts of the precuneus, gyrus cinguli (referred to as cingulum), as well as frontal and parietal grey and white matter (Fig. 1d). The second multi-voxel section (Fig. 1a and b) included the thalamus, putamen and insular

Clinical data

GTCS

Gender male/female Age (years) Hand preference Duration of epilepsy (years) Medical treatment in the last 12 months 5*  Levetiracetam 5  Lamotrigine 2  Topiramate Seizure frequency under medication in the last 12 months II. 9/18 >2 tonic–clonic seizures/12 months

10/18 (56%); 8/18 (44%) Mean 28; min 18; max 39 17/18 (94%) right handed 12  SD 4 6  Valproat acid

I. 9/18 2 tonic–clonic seizures/12 months

2.3. Post-processing The following brain metabolites were selected for further analysis: totals NAA (tNAA, the sum of N-acetyl aspartate (NAA) and N-acetylaspartylglutamate (NAAG)), glutamate (Glu), glutamate (Glu) plus glutamine (Gln) (Glu + Gln = Glx), total Cho (tCho, predominantly glycerophosphocholine (GPC) and phosphocholine (PCh)), myo-Inositol (mI) and total Cr (tCr, the sum of Cr and phospocreatine (PCr)). tNAA, Glu, Glx, tCho and tCr were quantified using LCModel,22 which is a user-independent frequency domain spectral fitting program (Fig. 2d). LCModel quantifies metabolite values referencing to the unsuppressed water peak. Water scaling and fitting were carried out automatically. Spectra were corrected for T1- and T2-relaxation according to previous studies.23,24 Estimated uncertainties by Cramer-Rao Lower Bounds (CRLB) served as main guidelines for judging the spectra of absolute metabolite values. Only metabolite spectra with LCModel estimated uncertainty of <15% standard deviation (SD) (SD < 20% for Glu and Glx) and spectra with a signal to noise ratio (SNR) above four were included in this study. Spectra with a Full Width at Half Maximum (FWHM) >0.065 ppm in the central region and a FWHM >0.08 ppm in the thalamus & insular cortex & hippocampus were not included in the study. LCModel provides estimates of the individual values of Glu and Gln, however, owing to substantial overlap even at 3 T the sum of the values of these two components (Glx) is more reliably determined than the individual components. As to Glu more than 60% of the spectra met the study criteria mentioned above (e.g. SD < 20%), therefore

M.T. Doelken et al. / Seizure 19 (2010) 485–492

487

Fig. 1. This figure shows the arrangement of three multi-voxel spectra (a–c) into symmetric pairs on each side of the midline including frontal and parietal grey and white matter, parts of the pre-cuneus, gyrus cinguli (referred to as cingulum), as well as the thalamus, putamen, insular cortex and hippocampus. (a–c) Show reference lines and correct positioning of the multi-voxel spectra. (d–f) Indicate the voxel inclusion in relation to brain anatomy [Abbreviations: cr-C = central region Cortex; cr-W = central region white matter; ci = cingulum; pcu = pre-cuneus; lf-C = lateral frontal Cortex; f-W = frontal lobe white matter; mf-C = medial frontal Cortex; In = insular; Th = thalamus; P = putamen; Hip = Hippocampus].

it was included as an additional parameter in our evaluation. Gln as a single value was not included because less than 15% of the spectra met the study criteria (SD < 20%). For further details see Hammen et al.25 The amount of CSF in a voxel was found by manually segmenting CSF on T2-weighted images and was expressed as a percentage of voxel volume. The correction of each metabolite for CSF dilution was carried out by the formula: Ccorr = (Cmeas  100)/ (100 CSFperc). Cmeas refers to measured value of metabolite of the whole voxel. CSFperc corresponds to the percentage of CSF in the voxel. We did not perform automatic morphometric segmentation on the precise amount of grey and white matter in a voxel. However, the included voxels in the respective anatomic region in both hemispheres were carefully selected by one experienced neuroradiologist preceding evaluation with LCModel. The location of a voxel, e.g. in the peripheral frontal and parietal region, was manually considered within the cortex when comprising >60% grey matter. The number of spectra averaged for each anatomic region in both cerebral hemispheres was (Fig. 1): thalamus = 2; putamen = 1; insular cortex = 3, pre-cuneus = 2, cingulum = 4; central region white matter = 3; frontal lobe white matter = 3; central region cortex = 3; medial frontal cortex = 3; lateral frontal cortex = 3.

2.4. Statistics Statistical analysis was carried out by SPSS 15.0 and R 2.6.2.26 Normal distribution of the metabolite concentrations within the anatomic regions was checked by the Shapiro–Wilk-Test. Differences between patients and controls with respect to age were assessed using Student’s t-test (Table 1). For group analysis between healthy controls and epilepsy patients within the anatomic regions we used one-dimensional analysis of variance (ANOVA) based on p < 0.05 (Figs. 3 and 4). Individual regional metabolic level was considered as abnormal if exceeding a plus or minus 1.5 standard deviation (SD) shift of the mean of the healthy control group (Table 2). The metabolic correlation within the epileptic network (hippocampus, thalamus, putamen and central region) was tested for tNAA and Glx with Pearson’s correlation (p-value 0.05). Additionally, Pearson’s correlation for tNAA and Glx against frequency of seizures was applied. Analysis of variance (ANOVA) was performed for tNAA and Glx for concurrent medical treatment. 3. Results 3.1. Group analysis GTCS versus healthy controls Table 1 summarises the clinical data of the enclosed patients with GTCS. Figs. 3 and 4 delineate the overall extent of metabolic

488

M.T. Doelken et al. / Seizure 19 (2010) 485–492

Fig. 2. Examples of attained multi-voxel spectra of the cingulum (a), thalamus (b) and hippocampus (c). Additionally, LCModel evaluation of metabolite values with baselinefit is delineated (d).

impairment in the respective anatomic region of the GTCS patients. The mean levels of the single metabolites in Figs. 3 and 4 are displayed by a colour code and matched with axial T2w brain image (green/red = reduction/increase of the metabolite in comparison to healthy controls – see colour bar p values). The hippocampi are not shown due to inconspicuous results regarding all metabolites. In summary, group analysis of GTCS patients versus healthy controls revealed significant (p < 0.05) decrease of tNAA in the cingulum and cortex of the pre- and postcentral region (in the following referred to as ‘‘central region’’). Additionally, a decrease (p < 0.05) of tNAA was observed in the thalami. Glx was elevated broadly in both hemispheres, most of all in central region, cingulum, insular cortex and left putamen. A significant increase (p < 0.05) was seen also in the right thalamus. As to Glu a significant (p > 0.05) increase was noticed in the white matter of the central region and, corresponding to Glx, pronounced in the left putamen and both insular cortex. Cho and mI demonstrated a significant coincidental decrease (p < 0.05) in particular in the grey and white matter of the central region. Moreover, Cho was significantly decreased (p < 0.05) in the left thalamus.

in these regions (central region cortex up to 27.3%; central region white matter up to 42.9%, thalamus up to 16.7%). Additionally, Glx was elevated in the medial frontal cortex (up to 53.8%). Further significant increase of Glx was noticed also in the putamen (up to 20%) and insular cortex (up to 25%). Glu was increased in the central region (up to 17.6%), respectively in the frontal lobe white matter (up to 14.3%) as well as in the insular cortex (up to 6.2%), putamen (up to 30%) and thalamus (up to 33.3). Cho and mI demonstrated a significant concomitant decrease mainly in the grey and white matter in the central region (up to 27.8% mI; up to 16.7% Cho), in the medial frontal cortex (up to 15.4% Cho; up to 23.1% mI) and in the thalamus (up to 16.7% Cho; up to 8.3% mI).

3.2. Individual results in comparison to healthy controls

3.4. Relation to age, duration of epilepsy and seizure frequency

Table 2 indicates the individual regional results of the patients with GTCS in comparison to healthy controls. In summary, tNAA was eminently reduced in patients in the grey and white matter in the central region (up to 17.6%), in the cingulum (up to 38.9%) and thalamus (up to 23.1%), while Glx showed corresponding increase

There was no significant correlation between tNAA or Glx and age, respectively duration of epilepsy. However, in patients with >2 tonic–clonic seizures in the last 12 months a trend towards higher Glx values and lower tNAA levels was observed in the thalamus (tNAA p = 0.08; Glx p = 0.07), insular cortex (tNAA

3.3. Correlations within the epileptic network Significant correlation within the epileptic network with p  0.05 based on tNAA, respectively Glx occurred between the thalamus and the central region, cingulum and putamen. As to Glx significant correlation was also noticed between the thalamus and the medial frontal cortex. In contrast, no significant correlation was observed in controls between thalami and other structures.

M.T. Doelken et al. / Seizure 19 (2010) 485–492

489

Fig. 3. This figure displays the overall extent of hemispheric metabolic impairment in the respective anatomic regions of the GTCS patients in comparison to healthy controls. The mean levels of single metabolites are displayed by a colour code (colour bar on right bottom side of the figure; EP = epilepsy patients; HC = healthy controls) and matched with axial T2w brain images: green/red = reduction/increase of the metabolite in comparison to healthy controls (see colour bar p values). An decrease of tNAA with corresponding increase of Glx in the cingulum and cortex of the pre- and postcentral region (referred to as ‘‘central region’’) was noticed. Additionally, Cho and mI were in particular reduced in the central region. Further details are presented within the text. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

p = 0.06; Glx p = 0.02), putamen (tNAA p = 0.03; Glx p = 0.08), cingulum (tNAA p = 0.05; Glx p = 0.01), central region grey matter (tNAA p = 0.04; Glx p = 0.002) and medial frontal cortex (tNAA p = 0.03; Glx p = 0.02). 3.5. Relation to concurrent medical therapy We did not find significant difference in tNAA and Glx in any region when comparing patients with different medical treatments (ANOVA, p > 0.05). 4. Discussion In agreement with recent reports applying both single- and multi-voxel techniques in different anatomic brain regions at 1.5 T scanners our results at 3 T document the major role of ‘‘basal ganglia-central region relay dysfunction’’, respectively, the extent of (sub-)network involvement in patients with GTCS.1,4,5,15 However, if the metabolic alterations, e.g. in the central region are primary or secondary to the origin of the

seizures remains uncertain, particularly because of further partial metabolic abnormalities in other regions (potential sub-networks). There is relative inhomogeneity in published IGE groups, sometimes including IGE sub-syndromes that were not classifiable. Patients with GTCS in this study were selected according to the ILAE classification.17 Additionally, we performed a voxel-based morphometric analysis to exclude missed subtle pathologic processes in conventional MRI as, e.g. cortical dysplasia or heterotopia.18 Of course, microdysgenesis cannot be excluded using this method as well. In fact, neuropathologic studies actually demonstrated distributed cortical and subcortical microdysgenesis in patients with IGE,27 which could form the basis of generalised involvement of the brain during seizure. There are several theories proposed to explain the pathophysiology of generalised epilepsy.28–30 Only few 1H-MRS reports addressed the key structures in patients with GTCS, respectively the extent of metabolic alteration within brain networks.1,5,15 We were unable to find a clinical study as to GTCS patients covering such a wide brain range.

490

M.T. Doelken et al. / Seizure 19 (2010) 485–492

Fig. 4. This figure displays the overall extent of metabolic impairment in the basal ganglia of the GTCS patients in comparison to healthy controls. The mean levels of single metabolites are displayed by a colour code (colour bar on right bottom side of the figure; EP = epilepsy patients; HC = healthy controls) and matched with axial T2w brain images: green/red = reduction/increase of the metabolite in comparison to healthy controls (see colour bar p values). tNAA and Cho were partially decreased in thalamus while Glx was partially increased the thalamus, putamen and insular cortex. Further details are presented within the text. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

4.1. Metabolic alterations of tNAA, Cho and mI In agreement with previous studies including patients with GTCS we found a reduction of tNAA and Cho in the thalamus,1,5,15 however also significant reduction in the cingulum (tNAA, Cho, mI) and normal mean values in the medial frontal cortex unlike Glx. In patients suffering from JME, in contrast to GTCS, significantly reduced levels of tNAA in the medial frontal lobe were reported.2,4– 6,32 Thus, our data further support the aspect of different forms of IGE, which to a certain extent demonstrate different pathophysiologies or at least involvement of unequal (sub)-networks.5,13 As to tNAA a decrease is not necessarily due to reduced cell density.33 The dysfunctional neurometabolism might arise from ‘‘minor’’ regional microdysgenesis on the basis of gene mutations.34 4.2. Metabolic alterations of Glx and Glu Significant increase of Glx has proven valuable in evaluation of patients with IGE.4,6,15,35 Elevation of Glx, respectively, Glx/Cr was described as in our study in the thalamus, central region, medial frontal cortex and cingulum.4,15 The origin of epilepsy is believed

to be caused by an inhibitory-excitatory imbalance. Glutamate (Glu) is the major excitatory neurotransmitter in the brain and elevation is known to have an excitatory effect.36 Glx (Glu + Gln) is commonly determined, because the fitting error of the sum is much smaller than those of the single compounds. Owing to substantial overlap even at 3 T the sum of Glx is more reliably determined than the individual components. However, our study criteria admitted the inclusion of Glu although the single compound still has to be addressed with caution for reasons stated above. Thus, a substantial interpretation of different result with respect to elevation of Glx and normal Glu values, e.g. in the thalamus is not actually possible. Nevertheless, we found concomitant elevation of Glx and Glu in the putamen, insular cortex and interestingly in the fronto-parietal white matter. Glu is assumed to play an important role in the modulation of thalamic activity initiated by cortico-thalamic projections.15 Neurons and glia are tightly coupled with respect to Glu and Gln metabolism,37 though Gln is only synthesised in astrocytes. Disturbance of excitatory mediator turnover might contribute to seizure generation. Glial dysfunction could cause low rates of neuron-glia cycling resulting in a down regulation of the Gln synthetase. Fractional

M.T. Doelken et al. / Seizure 19 (2010) 485–492

491

Table 2 Summary of GTCS patients with abnormal higher or lower metabolic values using a 1.5 standard deviation cut-off from the mean of the control group. Each field shows the included number of patients depending on the inclusion criteria (see methodical part) and underneath the corresponding percentage. tCr is not shown owing to unremarkable results [Abbreviations: r-cr-C = right central region Cortex; l-cr-C = left central region Cortex; r-cr-WM = right central region White Matter; l-cr-WM = left central region White Matter; r-Ci = right cingulum; l-Ci = left cingulum; r-pCu = right pre-cuneus; l-pCu = left pre-cuneus; r-lf-C = right lateral frontal Cortex; l-lf-C = left lateral frontal Cortex; r-fWM = right frontal lobe White Matter; l-f-WM = left frontal lobe White Matter; r-mf-C = right medial frontal Cortex; l-mf-C = left medial frontal Cortex; r-In = right insular; lIn = left insular; r-Th = right thalamus; l-Th = left thalamus; r-P = right putamen; l-P = left putamen]. Region

tNAA low

tNAA high

Glx low

Glx high

Glu low

Glu high

tCho low

tCho high

mI low

mI high

r-cr-C l-cr-C r-cr-WM l-cr-WM r-Ci l-Ci r-pCu l-pCu r-lf-C l-lf-C r-fl-WM l-fl-WM r-mf-C l-mf-C r-In l-In r-Th l-Th r-P l-P

3/17 2/18 2/18 2/18 7/18 3/17 0/16 1/16 3/18 1/18 1/18 1/18 0/13 2/12 0/17 1/16 2/12 3/13 0/12 0/14

0/17 0/18 3/18 1/18 0/18 0/17 0/16 1/16 0/18 0/18 1/18 1/18 2/13 0/12 0/17 0/16 1/12 1/13 1/12 1/14

0/11 0/11 0/14 0/14 0/17 0/17 0/14 1/16 0/17 0/14 0/13 0/13 1/13 0/13 0/16 0/15 0/12 0/10 0/12 0/11

3/11 3/11 6/14 3/14 3/17 2/17 0/14 1/16 1/17 2/14 4/13 3/13 3/13 7/13 4/16 3/15 2/12 1/10 2/12 2/11

0/14 (0) 0/16 (0) 0/17 (0) 0/16 (0) 0/17 (0) 1/16 (6.2) 0/15 (0) 0/15 (0) 0/17 (0) 0/16 (0) 0/14 (0) 0/15 (0) 2/13 (15.4) 0/13 (0) 0/16 (0) 1/16 (6.2) 1/9 (11.1) 0/9 (0) 0/10 (0) 0/10 (0)

1/14 (7.1) 2/16 (12.5) 3/17 (17.6) 2/16 (12.5) 0/17 (0) 0/16 (0) 0/15 (0) 0/15 (0) 1/17 (5.9) 1/16 (6.2) 2/14 (14.3) 1/15 (6.7) 0/13 (0) 1/13 (7.7) 1/16 (6.2) 1/16 (6.2) 3/9 (33.3) 0/9 (0) 1/10 (10) 3/10 (30)

2/17 3/18 3/18 3/18 0/18 1/17 0/16 0/15 3/18 0/18 3/18 1/18 2/13 1/13 0/17 0/16 2/12 1/12 0/12 0/14

0/17 0/18 1/18 1/18 0/18 0/17 0/16 0/15 3/18 0/18 1/18 0/18 0/13 0/13 1/17 2/16 1/12 0/12 1/12 1/14

4/16 5/18 3/18 4/18 3/18 3/17 1/16 1/16 2/18 4/18 3/18 3/18 3/13 2/13 0/17 0/16 1/12 1/12 0/12 0/12

0/16 0/18 0/18 0/18 0/18 0/17 0/16 0/16 1/18 0/18 1/18 0/18 1/13 0/13 0/17 0/16 1/12 0/12 1/12 1/12

(17.6) (11.1) (11.1) (11.1) (38.9) (17.6) (0) (6.2) (16,7) (5.6) (5.6) (5.6) (0) (16.7) (0) (6.2) (16.7) (23.1) (0) (0)

(0) (0) (16.7) (5.6) (0) (0) (0) (6.2) (0) (0) (5.6) (5.6) (15.4) (0) (0) (0) (8.3) (7.7) (8.3) (7.1)

(0) (0) (0) (0) (0) (0) (0) (6.2) (0) (0) (0) (0) (7.7) (0) (0) (0) (0) (0) (0) (0)

(27.3) (27.3) (42.9) (21.4) (17.6) (11.8) (0) (6.2) (5.9) (14.3) (30.8) (23.1) (23.1) (53.8) (25) (20) (16.7) (10) (16.7) (18.2)

failure of the Gln synthesis results in an increased glial Glu level, which could contribute to slower Glu uptake and enhanced Glu transporter reversal. 4.3. The epileptic network The brain is sensitive to changes within certain networks, with a consecutive engagement of the rest of the brain due to interconnected networks.38 In agreement with a recent study using the tNAA/Cr ratio in a subgroup of patients with JME4 we found a correlation of tNAA, respectively Glx in the thalamus and the central region, cingulum and putamen. As to Glx significant correlation was also noticed between the thalamus and the medial frontal cortex. However, we found no significant correlations in the healthy control group, therefore functional interconnection can be assumed. Only one other study including 12 JME and 8 GTCS patients reported a single correlation between thalamic and insular NAA/Cr.1 However, to our best knowledge, no previous study reported corresponding tNAA and Glx correlations in the above mentioned anatomic brain regions. Therefore, in previous studies it was unclear whether the metabolic alterations reflected involvement of a metabolic network or was uncorrelated. The basal ganglia are involved in controlling the cortical excitability,39 also in spread,40,41 modulation42 and persistence of seizures.43 The feedback from the basal ganglia is mainly inhibitory, while the motor cortex sends excitatory output. A basal ganglia to cortex vicious loop with excitatory firing of cortical cells (e.g. resulting from cortical microdysgenesis) and impaired inhibitory feedback of the basal ganglia might be the underlying cause of disease. Excitotoxic lesions, an expected result of hyperexcitability, would be located in the specific networks supporting seizures, rather than distributed to the areas particularly susceptible to diffuse hypoxia or diffuse excitotoxicity, for example in the hippocampus. Therefore, it is of special interest, that we found no hippocampal metabolite alterations in patients with GTCS. 4.4. Clinical data and metabolic alterations We were not able to demonstrate a metabolic relation to age and duration of epilepsy, however this might be due to limited

(11.8) (16.7) (16.7) (16.7) (0) (5.9) (0) (0) (16.7) (0) (16.7) (5.6) (15.4) (7.7) (0) (0) (16.7) (8.3) (0) (0)

(0) (0) (5.6) (5.6) (0) (0) (0) (0) (16.7) (0) (5.6) (0) (0) (0) (5.9) (12.5) (8.3) (0) (8.3) (7.1)

(25) (27.8) (16.7) (22.2) (16.7) (17.6) (6.2) (6.2) (11.1) (22.2) (16.7) (16.7) (23.1) (15.4) (0) (0) (8.3) (8.3) (0) (0)

(0) (0) (0) (0) (0) (0) (0) (0) (5.6) (0) (5.6) (0) (7.7) (0) (0) (0) (8.3) (0) (8.3) (8.3)

patient number. As to previous studies1,4,5 our results demonstrate that poorly controlled seizures are associated with more altered brain metabolites in the basal ganglia to cortex network. Whether this is related to the severity or threshold of the underlying epileptogenic pathology or to the effect of the seizures themselves remains speculative.1 5. Conclusion Our results demonstrate certain altered metabolic interconnection of cerebral anatomic regions in patients with GTCS, in particular the major role of the basal ganglia-central region relay in seizure generation. However, the primary locations of seizure generation and therewith the underlying substrates (e.g. microdysgenesis; cause–effect issue) remain ambiguous. To what extend different clinical IGE syndromes may depend on specific modifying genes still has to be addressed.44,45 Conflict of interest statement None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. Acknowledgement This research was supported by ‘‘Erlanger Leistungsbezogene Anschubfinanzierung und Nachwuchsfo¨rderung’’(ELAN); Grant No. 09.01.13.1. References 1. Bernasconi A, Bernasconi N, Natsume J, Antel SB, Andermann F, Arnold DL. Magnetic resonance spectroscopy and imaging of the thalamus in idiopathic generalized epilepsy. Brain 2003;126:2447–54. 2. Duncan JS. Brain imaging in idiopathic generalized epilepsies. Epilepsia 2005;46(Suppl. 9):108–11. 3. Fojtikova D, Brazdil M, Horky J, Mikl M, Kuba R, Krupa P, et al. Magnetic resonance spectroscopy of the thalamus in patients with typical absence epilepsy. Seizure 2006;15:533–40.

492

M.T. Doelken et al. / Seizure 19 (2010) 485–492

4. Lin K, Carrete JH, Lin J, Peruchi MM, de Araujo Filho GM, Guaranha MS, et al. Magnetic resonance spectroscopy reveals an epileptic network in juvenile myoclonic epilepsy. Epilepsia 2009;50:1191–200. 5. Savic I, Osterman Y, Helms G. MRS shows syndrome differentiated metabolite changes in human-generalized epilepsies. Neuroimage 2004;21:163–72. 6. Simister RJ, McLean MA, Barker GJ, Duncan JS. Proton MRS reveals frontal lobe metabolite abnormalities in idiopathic generalized epilepsy. Neurology 2003;61:897–902. 7. Wormann FG. The value of neuroimaging in diagnosis of epilepsy. Ther Umsch 2001;58:645–9. 8. Doelken MT, Richter G, Stefan H, Doerfler A, Noemayr A, Kuwert T, et al. Multimodal coregistration in patients with temporal lobe epilepsy – results of different imaging modalities in lateralization of the affected hemisphere in MR imaging positive and negative subgroups. AJNR Am J Neuroradiol 2007;28:449–54. 9. Doelken MT, Stefan H, Pauli E, Stadlbauer A, Struffert T, Engelhorn T, et al. (1)HMRS profile in MRI positive- versus MRI negative patients with temporal lobe epilepsy. Seizure 2008;17:490–7. 10. Hammen T, Stefan H, Eberhardt KE, Huk BH, Tomandl BF. Clinical applications of 1H-MR spectroscopy in the evaluation of epilepsies – what do pathological spectra stand for with regard to current results and what answers do they give to common clinical questions concerning the treatment of epilepsies? Acta Neurol Scand 2003;108:223–38. 11. Schuff N, Ezekiel F, Gamst AC, Amend DL, Capizzano AA, Maudsley AA, et al. Region and tissue differences of metabolites in normally aged brain using multislice 1H magnetic resonance spectroscopic imaging. Magn Reson Med 2001;45:899–907. 12. Ebel A, Maudsley AA. Detection and correction of frequency instabilities for volumetric 1H echo-planar spectroscopic imaging. Magn Reson Med 2005;53:465–9. 13. Savic I, Lekvall A, Greitz D, Helms G. MR spectroscopy shows reduced frontal lobe concentrations of N-acetyl aspartate in patients with juvenile myoclonic epilepsy. Epilepsia 2000;41:290–6. 14. Mory SB, Li LM, Guerreiro CA, Cendes F. Thalamic dysfunction in juvenile myoclonic epilepsy: a proton MRS study. Epilepsia 2003;44:1402–5. 15. Helms G, Ciumas C, Kyaga S, Savic I. Increased thalamus levels of glutamate and glutamine (Glx) in patients with idiopathic generalised epilepsy. J Neurol Neurosurg Psychiatr 2006;77:489–94. 16. ILAE. Commission on classification and terminology of the international league against epilepsy. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989;30:389–99. 17. Engel Jr J. Report of the ILAE classification core group. Epilepsia 2006;47:1558–68. 18. Huppertz HJ, Grimm C, Fauser S, Kassubek J, Mader I, Hochmuth A, et al. Enhanced visualization of blurred grey-white matter junctions in focal cortical dysplasia by voxel-based 3D MRI analysis. Epilepsy Res 2005;67:35–50. 19. Kreis R, Ernst T, Ross BD. Development of the human brain: in vivo quantification of metabolite and water content with proton magnetic resonance spectroscopy. Magn Reson Med 1993;30:424–37. 20. Pouwels PJ, Brockmann K, Kruse B, Wilken B, Wick M, Hanefeld F, et al. Regional age dependence of human brain metabolites from infancy to adulthood as detected by quantitative localized proton MRS. Pediatr Res 1999;46:474–85. 21. Horska A, Kaufmann WE, Brant LJ, Naidu S, Harris JC, Barker PB. In vivo quantitative proton MRSI study of brain development from childhood to adolescence. J Magn Reson Imaging 2002;15:137–43. 22. Provencher SW. Automatic quantitation of localized in vivo 1H spectra with LCModel. NMR Biomed 2001;14:260–4. 23. Baker EH, Basso G, Barker PB, Smith MA, Bonekamp D, Horska A. Regional apparent metabolite concentrations in young adult brain measured by (1)H MR spectroscopy at 3 Tesla. J Magn Reson Imaging 2008;27:489–99.

24. Zaaraoui W, Fleysher L, Fleysher R, Liu S, Soher BJ, Gonen O. Human brainstructure resolved T(2) relaxation times of proton metabolites at 3 Tesla. Magn Reson Med 2007;57:983–9. 25. Hammen T, Stadlbauer A, Tomandl B, Ganslandt O, Pauli E, Huk W, et al. Short TE single-voxel 1H-MR spectroscopy of hippocampal structures in healthy adults at 1. 5 Tesla – how reproducible are the results? NMR Biomed 2005;18:195–201. 26. R Development Core Team. A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. R version 2 8 0 2008. 27. Meencke HJ, Janz D. Neuropathological findings in primary generalized epilepsy: a study of eight cases. Epilepsia 1984;25:8–21. 28. Gloor P. Generalized cortico-reticular epilepsies. Some considerations on the pathophysiology of generalized bilaterally synchronous spike and wave discharge. Epilepsia 1968;9:249–63. 29. Meeren HK, Pijn JP, Van Luijtelaar EL, Coenen AM, Lopes da Silva FH. Cortical focus drives widespread corticothalamic networks during spontaneous absence seizures in rats. J Neurosci 2002;22:1480–95. 30. Niedermeyer E. EEG and the differential diagnosis of common generalized (centrencephalic) epilepsy. Electroencephalogr Clin Neurophysiol 1969;26:120. 32. Koepp MJ. Juvenile myoclonic epilepsy – a generalized epilepsy syndrome? Acta Neurol Scand Suppl 2005;181:57–62. 33. Kuzniecky R, Hetherington H, Pan J, Hugg J, Palmer C, Gilliam F, et al. Proton spectroscopic imaging at 4.1 Tesla in patients with malformations of cortical development and epilepsy. Neurology 1997;48:1018–24. 34. Suzuki T, gado-Escueta AV, Aguan K, Alonso ME, Shi J, Hara Y, et al. Mutations in EFHC1 cause juvenile myoclonic epilepsy. Nat Genet 2004;36:842–9. 35. Simister RJ, McLean MA, Barker GJ, Duncan JS. A proton magnetic resonance spectroscopy study of metabolites in the occipital lobes in epilepsy. Epilepsia 2003;44:550–8. 36. Petroff OA, Errante LD, Rothman DL, Kim JH, Spencer DD. Glutamate-glutamine cycling in the epileptic human hippocampus. Epilepsia 2002;43:703–10. 37. Petroff OA, Errante LD, Rothman DL, Kim JH, Spencer DD. Neuronal and glial metabolite content of the epileptogenic human hippocampus. Ann Neurol 2002;52:635–42. 38. Blumenfeld H. From molecules to networks: cortical/subcortical interactions in the pathophysiology of idiopathic generalized epilepsy. Epilepsia 2003;44(Suppl. 2):7–15. 39. Slaght SJ, Paz T, Mahon S, Maurice N, Charpier S, Deniau JM. Functional organization of the circuits connecting the cerebral cortex and the basal ganglia: implications for the role of the basal ganglia in epilepsy. Epileptic Disord 2002;4(Suppl. 3):S9–22. 40. Dupont S, Semah F, Baulac M, Samson Y. The underlying pathophysiology of ictal dystonia in temporal lobe epilepsy: an FDG-PET study. Neurology 1998;51:1289–92. 41. Henry TR, Mazziotta JC, Engel Jr J, Christenson PD, Zhang JX, Phelps ME, et al. Quantifying interictal metabolic activity in human temporal lobe epilepsy. J Cereb Blood Flow Metab 1990;10:748–57. 42. Dematteis M, Kahane P, Vercueil L, Depaulis A. MRI evidence for the involvement of basal ganglia in epileptic seizures: an hypothesis. Epileptic Disord 2003;5:161–4. 43. Deransart C, Depaulis A. The control of seizures by the basal ganglia? A review of experimental data. Epileptic Disord 2002;4(Suppl 3):S61–72. 44. Lu Y, Wang X. Genes associated with idiopathic epilepsies: a current overview. Neurol Res 2009;31:135–43. 45. Durner M, Keddache MA, Tomasini L, Shinnar S, Resor SR, Cohen J, et al. Genome scan of idiopathic generalized epilepsy: evidence for major susceptibility gene and modifying genes influencing the seizure type. Ann Neurol 2001;49:328–35.