- Email: [email protected]
Imaging the cortical effect of lamotrigine in patients with idiopathic generalized epilepsy: A low-resolution electromagnetic tomography (LORETA) study
Epilepsy Research (2008) 81, 204—210
journal homepage: www.elsevier.com/locate/epilepsyres
Imaging the cortical effect of lamotrigine in patients with idiopathic generalized epilepsy: A low-resolution electromagnetic tomography (LORETA) study Béla Clemens a,∗, Pálma Piros a, Mónika Bessenyei b, Márton Tóth e, Katalin Hollódy c, István Kondákor d a
Kenézy Gyula Memorial Hospital, Neurological Department, Debrecen, Hungary Kenézy Gyula Memorial Hospital, Department of Child Neurology and Psychiatry, Debrecen, Hungary c University of Pécs, Department of Pediatrics, Pécs, Hungary d County Hospital, Department of Neurology, Kecskemét, Hungary e University of Pécs, Neurology Clinic, Hungary b
Received 25 June 2007; received in revised form 7 June 2008; accepted 12 June 2008 Available online 23 July 2008
KEYWORDS Lamotrigine; Idiopathic generalized epilepsy; LORETA
Summary Purpose: Anatomical localization of the cortical effect of lamotrigine (LTG) in patients with idiopathic generalized epilepsy (IGE). Methods: 19 patients with untreated IGE were investigated. EEG was recorded in the untreated condition and 3 months later when LTG treatment abolished the seizures. 19-channel EEG was recorded, and a total of 2 min artifact-free, waking EEG was processed to low-resolution electromagnetic tomography (LORETA) analysis. Activity (that is, current source density, A/m2 ) was computed in four frequency bands (delta, theta, alpha, and beta), for 2394 voxels that represented the cortical gray matter and the hippocampi. Group differences between the untreated and treated conditions were computed for the four bands and all voxels by multiple t-tests for interdependent datasets. The results were presented in terms of anatomical distribution and statistical significance. Results: p < 0.01 (uncorrected) changes (decrease of activity) emerged in the theta and the alpha bands. Theta activity decreased in a large cluster of voxels including parts of the temporal, parietal, occipital cortex bilaterally, and in the transverse temporal gyri, insula, hippocampus, and uncus on the right side. Alpha activity decreased in a relatively smaller cortical area involving the right temporo-parietal junction and surrounding parts of the cortex, and part of the insula on the right side.
∗ Corresponding author at: Kenézy Gyula Memorial Hospital, Neurological Department, Bartók Béla út 3, 4031 Debrecen, Hungary. Tel.: +36 52 511 777; fax: +36 52 511 729. E-mail address: [email protected] (B. Clemens).
0920-1211/$ — see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2008.06.002
Imaging lamatrigine’s action on the cortex
205
Conclusions: LTG decreased theta activity in several cortical areas where abnormally increased theta activity had been found in a prior study in another cohort of untreated IGE patients [Clemens, B., Bessenyei, M., Piros, P., Tóth, M., Seress, L., Kondákor, I., 2007b. Characteristic distribution of interictal brain electrical activity in idiopathic generalized epilepsy. Epilepsia 48, 941—949]. These LTG-related changes might be related to the decrease of seizure propensity in IGE. © 2008 Elsevier B.V. All rights reserved.
Introduction Lamotrigine (LTG) is an anticonvulsive drug that is effective in most epilepsy syndromes. Concerning focal (localizationrelated) epilepsy, the principal mechanism of action of LTG is decreasing the ion fluxes in voltage-activated sodium and calcium channels (Leach et al., 1995). On the other hand, the mechanism of action of LTG against generalized seizures with generalized spike-wave (GSW) paroxysms has not been clarified yet. GSW-type seizures are bilateral from the onset and involve widespread, cortico-reticular, thalamo-cortical, and cortico-cortical networks (Gloor, 1979; Steriade and Amzica, 1994; Steriade and Contreras, 1995). Despite considerable efforts of pharmacological research, the complexity of these neuronal interactions has not permitted the identification of the molecular mechanisms. Similarly, the cortical site of action of the so-called antiabsence drugs including LTG remains hidden. Thus, it seems that complementary approaches are required to understand the effect of LTG in generalized epilepsy. The ‘‘network approach’’ and the identification of the cortical site of action of these drugs are potentially useful, complementary possibilities. As to the former, we remember the physiologist’s recommendation that ‘‘investigation of intact brains is the preferable approach to global brain functions and . . . paroxysmal activities’’ (Steriade, 2001). There is no doubt that the disturbed network interactions underlying seizure liability and its modification in idiopathic generalized epilepsy (IGE) belong to this category of events. In fact, LTG did not modify a lot of elementary neuronal processes (including thalamic T-Ca2+ currents) in intact thalamo-cortical loops but decreased pathological oscillations and corresponding GSW-type seizures in the same series of experiments (Gibbs et al., 2002). Measuring electromagnetic oscillations at a large scale was successfully applied to human idiopathic generalized epilepsy: LTG decreased pathologically synchronized EEG background activity in a use-dependent manner (Clemens et al., 2007a). However, dampening thalamo-cortical oscillations cannot fully explain the mechanism of a drug against GSW-type seizures. The full antiabsence effect of ethosuximide is only exerted when the drug binds to a specific cortical site, the primary somatosensory cortex (Manning et al., 2004). Given the leading role of the cortex in initiating GSW-type seizures in human IGE, a cortical site for LTG binding might be postulated. A transcranial magnetic stimulation study disclosed the effect of LTG on the primary motor cortex (Manganotti et al., 1999). However, nothing is known about the effect of LTG on the remaining part of the cortical mantle. The goal of the present study was to localize the neuronal sources (generators) the activity of which was altered by LTG treatment. Our prior study suggested that LTG decreases EEG activ-
ity mainly in the posterior parts of the cortex (Clemens et al., 2007a). Albeit the distribution of EEG spectral power on the scalp may roughly refer to the LORETA-localized sources of greatly abnormal activity (Thatcher et al., 2005), exact localization of the sources from the raw EEG is impossible because of the so-called ‘‘inverse problem’’ of EEG (computing the localization of the sources from the potential field is a mathematical problem with an infinite number of equally probable solutions). Thus, no concrete hypothesis concerning the anatomy of the cortical action of LTG was generated beforehand of this study.
Patients and methods Inclusion criteria were newly diagnosed, unmedicated patients with ‘‘IGE with variable phenotypes’’ (Panayiotopoulos, 2005); age of onset after the eighth year of age. Exclusion criteria were permanent medication except oral contraceptives, alcohol or drug abuse, any medical condition that is known to influence EEG activity, generalized tonic-clonic seizure in 5 days before EEG investigation. The study design was approved by the Local Research Ethics Committee of Kenézy Gyula Kórház. 19 patients were recruited (12 females, 7 males; age, 10—29 years; average, 16.2 years). The patients were diagnosed and treated as usual. No diagnostic procedure or treatment was indicated, missed, or postponed for study purposes. Neurological and EEG investigation were done at the first visit, treatment started the next day. The initial daily dose of LTG was 25 mg and was progressively increased until seizure control was reached. At the end of the titration the daily dose of LTG was between 100 and 300 mg. No patient experienced symptoms or complaints indicating adverse effects of the drug. The first EEG was recorded in drug-free state. The second was done 3 months later when the patients took the effective daily dose of LTG. The EEG recordings were carried out in the morning, after a night of sufficient sleep, in the same semi-isolated room, with the same Brain Quick (Micromed, Mestre, Italy) digital equipment, by trained personnel, according to recommended Standards for Quantitative EEG (QEEG) studies (Nuwer et al., 1994). Silver—silver chloride electrodes were placed according to the 10—20 system, fixed by appropriate adhesive and conductive gel. Impedances did not exceed 5 k. 19-channel EEG was recorded against linked ears reference. Additional bipolar derivations were used to differentiate between EEG and eye movement potentials and to detect EMG activity. In the EEG derivations the filters were set at 0.1 and 30 Hz. Sampling frequency was 128 per second. 12-bit on-line digitization was used. 30 min EEG was recorded in waking-relaxed, eyes-closed condition. The state of vigilance was controlled by the EEG technician who gently aroused the patient when the alpha rhythm disappeared. The first 2 min of artifact-free editable EEG activity 60 epochs (each 2 s) were selected for QEEG analysis. Our standard epoch selection protocol includes: (1) the presence of continuous physiological (‘‘waking’’ or ‘‘resting’’) alpha activity with alpha voltage maximum in posterior regions, (2) the absence of artifacts, epileptiform potentials, and other nonstationary elements, (3) the absence of patterns indicating drowsiness or arousal. Epoch
206 selection and analysis was done blindly by one of the authors and controlled by another author (KI). This epoch selection method results in fairly reproducible results (John et al., 1983; Nuwer, 1988). The reported electrographic definition of the relaxed-waking state refers to a narrow range of vigilance (Bente, 1979), as supported by EEG and functional MRI (fMRI) co-registration (Laufs et al., 2006a).
LORETA analysis LORETA is a recently developed method to localize multiple distributed cortical sources of bioelectric activity in the three-dimensional space (Pascual-Marqui et al., 1994). In other words, LORETA demonstrates the synchronously activated neuronal populations underlying EEG activity by computing their cortical localization from the scalp distribution of the electric field. This is called solving the inverse problem of the EEG. The LORETA inverse solution is based on existing neuroanatomical and physiological knowledge and a mathematical constraint called the smoothness assumption (Pascual-Marqui, 2002a). The principles of LORETA and the mathematical tools have been described in details (http://www.keyinst.unizhch/loreta.html). The ‘‘smoothness assumption’’, which means that neighbouring neuronal generators show highly correlated activity in terms of orientation and strength (Pascual-Marqui, 2002a). The smoothness assumption is based on proven neuroanatomical and electrophysiological constraints as described elsewhere, including the mathematical basis of the method (http://www.keyinst.unizhch/loreta.html). In order to mathematically mitigate the disturbing effects of the electrically conducting layers between the cortical surface and the electrodes, LORETA computes the inverse solution within a three-shell spherical head model including scalp, skull, and brain. The brain compartment of this model was restricted to the cortical grey matter and hippocampus, according to the Talairach Brain Atlas digitized at Montreal Neurological Institute (Talairach and Tournoux, 1988). The grey matter compartment was subdivided in 2394 voxels, which allows a spatial resolution of 7 mm. LORETA computes a physically existing dimension, current density (A/m2 ) for each voxel. For the sake of brevity, this is called ‘‘activity’’ in this paper. The consistency of LORETA with physiology and localization has been validated by several authors. Crossmodal validation studies disclosed that LORETA and other functional neuroimaging methods showed the same cortical localization of dysfunction in several neuro-psychiatric conditions (Pascual-Marqui et al., 2002b). Concerning epilepsy, LORETA-defined localization of circumscribed activity corresponded to the localization of the epileptic discharges given by functional MRI, subdural and intracerebral EEG recordings, and the MRI-defined epileptogenic lesion (Lantz et al., 1997; Seeck et al., 1998; Worrell et al., 2000). The ability of LORETA to localize drug-related cortical changes has already been demonstrated (Saletu et al., 2002). We investigated the LORETA solution in four frequency bands separately (delta, 1.5—3.5 Hz; theta, 3.5—7.5 Hz; alpha, 7.5—12.5 Hz; beta, 12.5—25.0 Hz). Details on the computation of LORETA for several frequency bands can be found elsewhere (Frei et al., 2001). Thus, each patient was characterized by (4 × 2394 = 9576) current density values. Delta, theta, alpha, and beta group differences between the
B. Clemens et al. untreated and treated condition were computed for each voxel by parametric statistics. The anatomical localization of the voxels was carried out by the LORETA-KEY software: placing the cursor at any voxel displayed the localization of that voxel in terms of cortical anatomy. A total of 9588 comparisons were done (t-tests for interdependent datasets) without correction for multiple comparisons. The t-values were color-coded and projected onto axial brain slices. The findings were interpreted in terms of statistical significance and topographic distribution. Only the p < 0.01 results (tvalues outside the −3.037 to +3.037 range) were discussed. The uncorrected p = 0.01 threshold was reported to provide sufficient protection against type I error in a prior LORETA investigation (Pizzagalli et al., 2002). It is not sure that this statement is entirely valid for our study. Therefore, we did interpret this threshold as a sharp boundary that separates the meaningful and the unimportant changes. The postulated biological significance of the results was estimated by the degree of statistical significance (p) and the topographical distribution of the data.
Results LTG decreased EEG activity in all frequency bands to a variable degree, but uncorrected p < 0.01 changes emerged only in the theta and alpha bands. Fig. 1 demonstrates the degree of the changes (expressed in t-values) over the entire cortex. Theta activity was reduced in all voxels showing a peculiar postero-anterior distribution. However, only uncorrected p < 0.01 differences are described here as follows. The overwhelming majority of the voxels with decreased theta activity were crowded in a large, topographically contiguous cluster that included parts of the temporal, parietal, occipital cortex and insula. At the basal cortical surface, bilateral, partly asymmetrical theta decrease was found in the temporal cortex including the inferior temporal, fusiform, lingual, and parahippocampal gyri. The hippocampus and the uncus gyri hippocampi were affected on the right side only. Basal parts of the occipital cortex were affected bilaterally. At the cortical convexity, theta activity decreased in the middle and posterior parts of the lateral temporal cortex, involving a greater area in the left than in the right hemisphere. The transverse temporal gyri and the posterior part of the insula showed decreased theta activity on the left but not the right side. The occipital gyri on the lateral surface were involved bilaterally. The postcentral gyrus was completely involved on the left but only marginally on the right side. The angular and supramarginal gyri and part of the superior parietal lobule were involved bilaterally. A small area of the left precentral gyrus at the cortical convexity was the only part of the frontal lobes that showed decreased theta activity. Fig. 2 demonstrates that LTG decreased activity in several areas where pathologically increased activity had been reported in untreated IGE patients by the same authors (Clemens et al., 2007b). Also alpha activity decreased all over the cortex. However, p < 0.01 (uncorrected) changes only occurred in a limited part of the right hemisphere including posterior parts of the superior and medial temporal gyri, parts of the supramarginal and angular gyri, and part of the precuneus. A
Imaging lamatrigine’s action on the cortex topographically separated small cluster of 25 voxels with decreased alpha activity was found in the right insula.
Discussion EEG background activity of untreated IGE patients is characterized by increased synchronization in the delta, theta, and alpha frequency bands (Clemens et al., 2000). The presence of increased synchronization and coherence in the theta band was emphasized in epilepsy (Sarthein et al., 2003), and was demonstrated in IGE, too (Clemens, 2004). Albeit the exact contribution of the enhanced theta oscillations to seizure liability and ictogenesis has not been clarified yet, it is interesting that LTG caused the most prominent decrease of EEG synchronization exactly in the theta band in successfully treated IGE patients (Clemens et al., 2007a). Thus, it was not surprising that LORETA confirmed the decrease of synchronized theta activity in cortical neuronal pools. On the other hand, the anatomical distribution of the cortical effect of LTG was a novel finding that had not been predictable on the basis of our prior investigations. This was the first study to anatomically localize the cortical effect of LTG. The significance of our finding is that the net effect of LTG within the cortex (the cerebral compartment that is critical in determining seizure liability and seizure generation) was grasped and anatomically localized, without
207 knowing precisely the molecular and cellular actions of the drug at an unknown number of sites within the brain. The shortcoming of the study is suboptimal spatial sampling due to the limited number of electrodes, which may cause incorrect localization of small sources of activity. Nevertheless, we do not think that this resulted in significantly inaccuracy in this study. Resting EEG rhythms are generated by largely distributed cortical sources that can be accurately investigated by the standard 10—20 system (Nunez, 1995; Babiloni et al., 2006). In fact, comparative studies disclosed that LORETA analysis results in very similar localization of the generators with 19 and 46 electrodes provided that they are evenly distributed on the scalp (Michel et al., 2004). Statistics is a difficult task in LORETA studies. Unfortunately, software limitations prevented us from performing straightforward correction for multiple comparisons in order to compute a ‘‘hard’’ threshold of significance. However, the anatomically meaningful distribution of the results strongly suggested that they were due to biological causes rather than to by chance generated type-l errors. Fig. 1 combines the statistical and the topographical aspects of the results while only the uncorrected p < 0.01 findings were discussed. Our results posed another problem of interpretation. Which statistical degree of the LTG effect should be judged as biologically relevant in case of the reported topographical distribution? The ‘‘top of the iceberg’’ (that is, the
Figure 1 LORETA-defined decrease of theta and alpha activity. The Z-coordinates (millimeters) of the Talairach Brain Atlas label the selected axial slices. Blue color indicates the decrease of activity. t-Values are color-coded from t = 0 to −6.074 (t = −6.074 refers to p = 0.001). Laterality (R, L) is given in the Figure.
208 p < 0.001 clusters) or the more broadly outlined areas of less statistical significance should be considered as biologically significant? We cannot answer this question at present. The lack of relevant papers prevented hypothesis generation concerning the expected degree and the exact anatomical distribution of the effect of LTG. Hence, limiting the discussion to the uncorrected p < 0.01 changes was arbitrary, decided intuitively rather than based on scientific evidence. The main question of interest is whether topographic overlap exists between the LORETA-defined LTG effect and the cortical areas that are involved in epileptic dysfunction in IGE. The bulk of evidence suggests the leading role of the medial and basal frontal cortex in initiating generalized seizures (Bancaud et al., 1974; Holmes et al., 2004). However, seizure onset is a hardly definable term in case of generalized seizures. For example, the mediobasal frontal distribution of ictal activity of absence seizures is invariably seen only ‘‘by the second to fourth spike after onset of the epileptiform discharge’’, and the very onset of the seizure is characterized by focal irritative activity at variable sites of the frontal, temporal, and parietal cortex (Holmes et al., 2004). Also interictal GSW discharges typically show electric fields with negativity maxima over the frontal areas and a gradual decline of voltage along the temporal and parietal cortex, suggesting a widespread area with ictogenic property. These results suggest that other, yet not specified cortical areas may contribute to the seizure generation in IGE. The indisputable effect of LTG in IGE (including this cohort of patients) and its reported effect in the posterior parts of the cortex raise the possibility that also some posterior cortical areas might have a significant but underes-
B. Clemens et al. timated role in ictogenesis. Some pieces of evidence argue for this possibility as follows. Having postulated that the ictogenic property of the cortex is not evenly distributed across the scalp, we localized five cortical ‘‘clusters’’ (clusters of voxels) where current source density was pathologically increased in untreated IGE patients (Clemens et al., 2007b). Out of them, the two clusters located in the posterior half of the brain anatomically fairly correspond to cortical areas where LTG decreased theta activity (Fig. 2). The ‘‘temporo-basal cluster’’ in the untreated patients is topographically very similar to the distribution of the LTG-related decrease of theta activity in the medial basal temporal cortex. Temporal negativity in the early evolution of GSW discharge was reported in voltage mapping and LORETA studies (Rodin and Ancheta, 1986; Holmes et al., 2004). However, the contribution of the temporal cortex to the neurophysiology of IGE has been a completely neglected matter. Another area of increased activity, the ‘‘medial parietooccipital cluster’’ shows a remarkable overlap with the LTG-related theta decrease in the medial parieto-occipital area (Fig. 2). The overall physiological significance of this area in epilepsy is its capability dissipates that local epileptic activity can easily spread from this site to many other sites of the brain. Electroencephalographic evidence indicates preferential spread of epileptic activity from the medial and lateral parietal cortex to many limbic and neocortical areas (Sveinbjornsdottir and Duncan, 1993). Occipital epileptic foci can give rise to bilateral, frontally preponderant interictal discharges (Williamson and Spencer, 1986). Pieces of evidence suggest that the medial parieto-
Figure 2 Top row (from Clemens et al., 2007b; with permission of Blackwell Publisher): untreated IGE patients. Red color: uncorrected p < 0.01 increase of theta activity, in the medial temporal area (temporo-basal cluster) and at the medial surface of the left hemisphere (medial parieto-occipital cluster). Bottom row, blue color: LTG-related decrease of theta activity in about the same areas. The relevant Talairach coordinates and laterality (R, L) is given in the Figure.
Imaging lamatrigine’s action on the cortex occipital area has a probable but unexplored role in the neurophysiology of IGE, too. Besides the main frontal negativity, a medial parieto-occipital negativity in the early course of the GSW complex was described by the authors who used voltage mapping analysis (Rodin and Ancheta, 1986; Coppola, 1988; Hughes et al., 1990; Ferri et al., 1995). This area was called the ‘‘posterior center of gravity of the ictal discharge’’ (Rodin, 1999). Further arguments supporting the contribution of this area to the ictogenic condition in IGE are the excess of excitatory neurotransmission in this area (Simister et al., 2003), the patients whose absence seizures arise from the posterior parts of the cortex (Gomez and Westmoreland, 1987), and the midparietal, mid-occipital spike component frequently seen at the beginning of the generalized photoparoxysmal response in IGE patients. We suggest that the topographic overlap between increased theta activity in the ‘‘medial parietooccipital cluster’’ and the drug-related theta decrease in the same area can hardly be a per chance finding. Albeit the relationship between cortical EEG activity and seizure control is a complex matter involving subcortical physiology and genetic influences, our results rise the possibility that LTG counteracts seizure generation by dampening theta activity at specific cortical sites. Nowadays, advanced neuroimaging methods (in particular, fMRI) seem to be valuable tools to address neurophysiological problems that previously had been the privilege of EEG. fMRI studies of absence seizures disclosed bilateral activation in the thalamus, mesial midfrontal regions, insula and cerebellum, while bilateral deactivation was found in the anterior frontal and parietal cortex and the posterior cingulate (Archer et al., 2003; Gotman et al., 2005). We demonstrated that LTG greatly decreased theta activity (p < 0.001, uncorrected) in one part of this network, the posterior cingulate and the surrounding medial parietal cortex. In addition to its role in elaborating absence seizures this cortical area is part of an extended neuronal system the activity of which is related to the ‘‘default state of the brain’’, a fascinating concept that has been refined recently (Raichle and Snyder, 2007). Absence seizures cause fMRI deactivation in this system (Gotman et al., 2005; Laufs et al., 2006b). The posterior cingulate area seems to be the ‘‘key coupling site between the brainstem system for arousal and cortical systems for cognitive processing and awareness’’ (Vogt and Laureys, 2005). Increased theta activity in scalp-recorded EEG generally indicates decreased vigilance and impaired cognition. Thus, it is possible that the established beneficial effect of LTG on vigilance and attention (Aldenkamp et al., 2002) is mediated by decreasing theta activity in this part of the cortex. Another point of interest is the emerging role of fMRI in pharmacological studies (Ianetti and Wise, 2007). Scalp EEG only reflects cortical activity, which, nevertheless, is under the control of cortico-cortical and subcortico-cortical influences. The proportion of cortically and subcortically driven activity cannot be separated on EEG grounds. Thus, we do not state that our findings reflect the direct action of LTG on the cortex. It is possible that some of the reported effect is mediated via subcortical sites. fMRI studies may shed light to the subcortical actions of LTG. Integrating EEG and fMRI methods highlighted several aspects of IGE pathophysiology (Archer et al., 2003; Gotman et al., 2005; Laufs
209 et al., 2006b), and also might highlight the neuronal mechanisms by which LTG exerts its clinical effects on seizures and vigilance.
Acknowledgement This work has been supported by Hungarian Science Foundation (OTKA) grant #T 048338.
References Aldenkamp, A.P., Arends, J., Bootsma, H.P.R., Diepman, L., Hulsman, J., Lambrechts, D., Leenen, L., Majoie, M., Schellekens, A., de Vocht, J., 2002. Randomized double-blind parallel-group study comparing cognitive effects of a low dose lamotrigine with valproate. Epilepsia 43, 19—26. Archer, J.S., Abbott, D.F., Waites, A.B., Jackson, G.D., 2003. fMRI deactivation of the posterior cingulate during generalized spike and wave. Neuroimage 20, 1915—1922. Babiloni, C., Frisoni, G., Steriade, M., Bresciani, L., Binetti, G., Del Percio, C., Geroldi, C., Miniussi, C., Nobili, F., Rodriguez, G., Zappasodi, F., Carfagna, T., Rossini, P.M., 2006. Frontal white matter volume and delta sources negatively correlate in awake subjects with mild cognitive impairment and Alzheimer’s disease. Clin. Neurophysiol. 117, 1113—1129. Bancaud, J., Talairach, J., Morel, P., Bresson, M., Buser, P., 1974. Generalized epileptic seizures elicited by electrical stimulation of the frontal lobe is man. Electroenceph. Clin. Neurophysiol. 37, 275—282. Bente, D., 1979. Vigilance and evaluation of psychotropic drug effect on EEG. Pharmacopsychiatry 12, 137—147. Clemens, B., Szigeti, G., Barta, Z., 2000. EEG frequency profiles of idiopathic generalized epilepsy syndromes. Epilepsy Res. 42, 105—115. Clemens, B., 2004. Pathological theta oscillations in idiopathic generalised epilepsy. Clin. Neurophysiol. 115, 1436—1441. Clemens, B., Piros, P., Bessenyei, M., Hollódy, K., 2007a. Lamotrigine decreases EEG synchronization in a use-dependent manner in patients with idiopathic generalized epilepsy. Clin. Neurophysiol. 118, 910—917. Clemens, B., Bessenyei, M., Piros, P., Tóth, M., Seress, L., Kondákor, I., 2007b. Characteristic distribution of interictal brain electrical activity in idiopathic generalized epilepsy. Epilepsia 48, 941—949. Coppola, R., 1988. Topographic display of spike-wave discharges. In: Myslobodsky, M.S., Mirsky, A.F. (Eds.), Elements of Petit Mal Epilepsy. Peter Lang Publ. Inc., New York, pp. 105—129. Ferri, R., Iliceto, G., Caelucci, V., 1995. Topographic EEG mapping of 3/s spike-and-wave complexes during absence seizures. Ital. J. Neurol. Sci. 16, 541—547. Frei, E., Gamma, A., Pascual-Marqui, R.D., Lehmann, D., Hell, D., Vollenweider, F.X., 2001. Localization of NMDA-induced brain activity in healthy volunteers using low resolution brain electromagnetic tomography (LORETA). Hum. Brain Map. 14, 152—165. Gomez, M.R., Westmoreland, B.F., 1987. Absence seizures. In: Lüders, H.O., Lesser, P. (Eds.), Epilepsy Electroclin. Synd. Springer, London, pp. 105—130. Gibbs, J.W., Yun-Fu Zhang, Ahmed, H.S., Coulter, D.A., 2002. Anticonvulsant actions of lamotrigine on spontaneous thalamocortical rhythms. Epilepsia 43, 342—349. Gloor, P., 1979. Generalized epilepsy with spike and wave discharge: a reinterpretation of its electrographic and clinical manifestations. Epilepsia 20, 571—588. Gotman, J., Grova, C., Bagshaw, E., Kobayashi, E., Agakhani, Y., Dubeau, F., 2005. Generalized epileptic discharges show thala-
210 mocortical activation and suspension of the default state of the brain. PNAS 102, 15236—15240. Holmes, M.D., Brown, M., Tucker, D.M., 2004. Are ‘‘generalized’’ seizures truly generalized? Evidence of localized mesial frontal and frontopolar discharges in absence. Epilepsia 45, 1568—1579. Hughes, J.R., Miller, J.K., Hughes, C.A., 1990. Topographic mapping of different types of bilateral spike-and-wave complexes. J. Epilepsy 3, 67—74. Ianetti, G.D., Wise, R.G., 2007. BOLD functional MRI in disease and pharmacological studies: room for improvement. Magn. Reson. Imag. 25, 978—988. John, E.R., Prichep, L., Ahn, H., Easton, P., Fridman, J., Kaye, H., 1983. Neurometric evaluation of cognitive dysfunctions and neurological disorders in children. Progress Neurobiol. 21, 290—329. Lantz, G., Michel, C.M., Pascual-Marqui, R.D., Spinelly, L., Seeck, M., Seri, S., Landis, T., Rosen, I., 1997. Extracranial localization of intracranial interictal epileptiform activity using LORETA (low resolution electromagnetic tomography). Electroenceph. Clin. Neurophysiol. 102, 414—422. Laufs, H., Holt, J.L., Elfont, R., Krams, M., Paul, J.P., Krakow, K., Kleinschmidt, A., 2006a. Where the BOLD signal goes when alpha EEG leaves. Neuroimage 31, 1408—1418. Laufs, H., Lengler, U., Hamandi, K., Kleinschmidt, A., Krakow, K., 2006b. Linking generalized spike-and-wave discharges and resting state brain activity by using EEG/fMRI in a patient with absence seizures. Epilepsia 47, 444—448. Leach, M.J., Lees, G., Ridall, D.R., 1995. Lamotrigine Mechanisms of Action. In: Levy, R.H., Mattson, R.H., Meldrum, B.S. (Eds.), Antiepileptic drugs, fourth edition. Raven Press, New York, pp. 861—902. Manganotti, P., Bongiovanni, L.G., Zanette, G., Turazzini, M., Fiaschi, A., 1999. Cortical excitability in patients after loading dose of lamotrigine: a study with magnetic brain stimulation. Epilepsia 40, 316—321. Manning, J.P., Richards, D.A., Leresche, N., Crunelli, V., Bowery, N.G., 2004. Cortical-area specific block of genetically determined absence seizures by ethosuximide. Neuroscience 123, 5—9. Michel, C.M., Murray, M.M., Lantz, G., Gonzales, S., Spinelli, L., Grave de Peralta, R., 2004. EEG source imaging. Clin. Neurophysiol. 115, 2195—2222. Nunez, P.L., 1995. Quantitative states of neocortex. In: Nunez, P.L. (Ed.), Neocortical Dynamics and Human EEG Rhythms. University Press, Oxford, pp. 1—18. Nuwer, M., 1988. Quantitative EEG. 1. Techniques and problems of frequency analysis and topographic mapping. J. Clin. Neurophysiol. 5, 1—43. Nuwer, M., Lehmann, D., Lopes da Silva, F., Matsuoka, S., Sutherling, W., Vibert, J.F., 1994. IFCN guidelines for topographic and frequency analysis of EEGs and EPs. Report of an IFCN Committee. Electroenceph. Clin. Neurophysiol. 91, 1—5. Panayiotopoulos, C., 2005. Idiopathic generalized epilepsies. In: Panayiotopoulos, C. (Ed.), The Epilepsies. Seizures, Syndromes, and Management. Springer, pp. 271—348. Pascual-Marqui, R.D., Michel, C.M., Lehmann, D., 1994. Low resolution electromagnetic tomography. A new method for localizing electrical activity in the brain. Int. J. Psychophysiol. 18, 49—65. Pascual-Marqui, R.D., 2002a. Functional imaging with lowresolution brain electromagnetic tomography (LORETA): a review. Methods Find Exp. Clin. Pharmacol. 24 (Suppl. C), 91—96.
B. Clemens et al. Pascual-Marqui, R.D., Esslen, M., Kochi, K., Lehmann, D., 2002b. Functional imaging with low-resolution brain electromagnetic tomography (LORETA): review, new comparisons, and new validation. Jpn. J. Clin. Neurophysiol. 30, 81—94. Pizzagalli, D.A., Nitschke, J.B., Oakes, T.R., Hendrick, A.M., Horras, K.A., Larson, C.L., Abercrombie, H.C., Schaefer, S.M., Koger, J.V., Benca, R.M., Pascual-Marqui, R.D., Davidson, R.J., 2002. Brain electrical tomography in depression: the importance of symptom severity, anxiety, and melancholic features. Biol. Psychiatry 52, 73—85. Raichle, M.E., Snyder, A.Z., 2007. A default mode of brain function: a brief history of an evolving idea. NeuroImage 37, 1083—1090. Rodin, E., Ancheta, O., 1986. Cerebral electric fields during petit mal absences. Electroenceph. Clin. Neurophysiol. 66, 457—466. Rodin, E., 1999. Decomposition and mapping of generalized spikewave complexes. Clin. Neurophysiol. 110, 1868—1875. Saletu, B., Anderer, P., Saletu-Zyhlarz, G.M., Arnold, O., Pascual-Marqui, R.D., 2002. Classification and evaluation of the pharmacodynamics of psychotropic drugs by single-lead pharmaco-EEG, EEG mapping and tomography (LORETA). Methods Find Exp. Clin. Pharmacol. 24 (Suppl. C), 97—120. Sarthein, J., Morel, A., von Stein, A., Jeanmonod, D., 2003. Thalamic field potentials and EEG: high thalamocortical coherence in patients with neurogenic pain, epilepsy, and movement disorders. Thalamus Relat. Syst. 2, 231—238. Seeck, M., Lazeyras, F., Michel, C.M., Blanke, O., Gericke, C.A., Ives, J., Delavelle, J., Golay, X., Haenggeli, C.A., de Tribolet, N., Landis, T., 1998. Non-invasive epileptic focus localization using EEG-triggered functional MRI and electromagnetic tomography. Electroenceph. Clin. Neurophysiol. 106, 508—512. Simister, R.J., McLean, M.A., Barker, G.J., Duncan, J.S., 2003. A proton magnetic resonance spectroscopy study of metabolites in the occipital lobes in epilepsy. Epilepsia 44, 550—558. Steriade, M., Amzica, F., 1994. Dynamic coupling among neocortical neurons during evoked and spontaneous spike-wave seizure activity. J. Neurophysiol. 72, 2051—2069. Steriade, M., Contreras, D., 1995. Relation between cortical and thalamic cellular events during transition from sleep patterns to paroxysmal activity. J. Neurosci. 15, 623—642. Steriade, M., 2001. Impact of network activities on neuronal properties in corticothalamic systems. Am. Physiol. Soc. 86, 1—39. Sveinbjornsdottir, S., Duncan, J.S., 1993. Parietal and occipital lobe epilepsy: a review. Epilepsia 34, 493—521. Talairach, J., Tournoux, P., 1988. Co-planar Stereotaxic Atlas of the Human Brain: Three-Dimensional Proportional System. G. Thieme, Stuttgart. Thatcher, R.W., North, D., Biver, C., 2005. Evaluation and validity of a LORETA normative EEG database. Clin. EEG Neurosci. 36, 116—122. Vogt, B.A., Laureys, S., 2005. Posterior cingulate, precuneal and retrosplenial cortices: cytology and components of the neural network correlates of consciousness. Prog. Brain Res. 150, 205—217. Williamson, P.D., Spencer, S.S., 1986. Clinical and EEG features of complex partial seizures of extratemporal origin. Epilepsia 27 (Suppl. 2), S46—S63. Worrell, G.A., Lagerlund, T.D., Sharbrough, F.W., Brinkmann, B.H., Busacker, N.E., Cicora, K.M., O’Brien, T.J., 2000. Localization of the epileptic focus by low resolution electromagnetic tomography in patients with a lesion demonstrated by MRI. Brain Topogr. 12, 273—282.
journal homepage: www.elsevier.com/locate/epilepsyres
Imaging the cortical effect of lamotrigine in patients with idiopathic generalized epilepsy: A low-resolution electromagnetic tomography (LORETA) study Béla Clemens a,∗, Pálma Piros a, Mónika Bessenyei b, Márton Tóth e, Katalin Hollódy c, István Kondákor d a
Kenézy Gyula Memorial Hospital, Neurological Department, Debrecen, Hungary Kenézy Gyula Memorial Hospital, Department of Child Neurology and Psychiatry, Debrecen, Hungary c University of Pécs, Department of Pediatrics, Pécs, Hungary d County Hospital, Department of Neurology, Kecskemét, Hungary e University of Pécs, Neurology Clinic, Hungary b
Received 25 June 2007; received in revised form 7 June 2008; accepted 12 June 2008 Available online 23 July 2008
KEYWORDS Lamotrigine; Idiopathic generalized epilepsy; LORETA
Summary Purpose: Anatomical localization of the cortical effect of lamotrigine (LTG) in patients with idiopathic generalized epilepsy (IGE). Methods: 19 patients with untreated IGE were investigated. EEG was recorded in the untreated condition and 3 months later when LTG treatment abolished the seizures. 19-channel EEG was recorded, and a total of 2 min artifact-free, waking EEG was processed to low-resolution electromagnetic tomography (LORETA) analysis. Activity (that is, current source density, A/m2 ) was computed in four frequency bands (delta, theta, alpha, and beta), for 2394 voxels that represented the cortical gray matter and the hippocampi. Group differences between the untreated and treated conditions were computed for the four bands and all voxels by multiple t-tests for interdependent datasets. The results were presented in terms of anatomical distribution and statistical significance. Results: p < 0.01 (uncorrected) changes (decrease of activity) emerged in the theta and the alpha bands. Theta activity decreased in a large cluster of voxels including parts of the temporal, parietal, occipital cortex bilaterally, and in the transverse temporal gyri, insula, hippocampus, and uncus on the right side. Alpha activity decreased in a relatively smaller cortical area involving the right temporo-parietal junction and surrounding parts of the cortex, and part of the insula on the right side.
∗ Corresponding author at: Kenézy Gyula Memorial Hospital, Neurological Department, Bartók Béla út 3, 4031 Debrecen, Hungary. Tel.: +36 52 511 777; fax: +36 52 511 729. E-mail address: [email protected] (B. Clemens).
0920-1211/$ — see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2008.06.002
Imaging lamatrigine’s action on the cortex
205
Conclusions: LTG decreased theta activity in several cortical areas where abnormally increased theta activity had been found in a prior study in another cohort of untreated IGE patients [Clemens, B., Bessenyei, M., Piros, P., Tóth, M., Seress, L., Kondákor, I., 2007b. Characteristic distribution of interictal brain electrical activity in idiopathic generalized epilepsy. Epilepsia 48, 941—949]. These LTG-related changes might be related to the decrease of seizure propensity in IGE. © 2008 Elsevier B.V. All rights reserved.
Introduction Lamotrigine (LTG) is an anticonvulsive drug that is effective in most epilepsy syndromes. Concerning focal (localizationrelated) epilepsy, the principal mechanism of action of LTG is decreasing the ion fluxes in voltage-activated sodium and calcium channels (Leach et al., 1995). On the other hand, the mechanism of action of LTG against generalized seizures with generalized spike-wave (GSW) paroxysms has not been clarified yet. GSW-type seizures are bilateral from the onset and involve widespread, cortico-reticular, thalamo-cortical, and cortico-cortical networks (Gloor, 1979; Steriade and Amzica, 1994; Steriade and Contreras, 1995). Despite considerable efforts of pharmacological research, the complexity of these neuronal interactions has not permitted the identification of the molecular mechanisms. Similarly, the cortical site of action of the so-called antiabsence drugs including LTG remains hidden. Thus, it seems that complementary approaches are required to understand the effect of LTG in generalized epilepsy. The ‘‘network approach’’ and the identification of the cortical site of action of these drugs are potentially useful, complementary possibilities. As to the former, we remember the physiologist’s recommendation that ‘‘investigation of intact brains is the preferable approach to global brain functions and . . . paroxysmal activities’’ (Steriade, 2001). There is no doubt that the disturbed network interactions underlying seizure liability and its modification in idiopathic generalized epilepsy (IGE) belong to this category of events. In fact, LTG did not modify a lot of elementary neuronal processes (including thalamic T-Ca2+ currents) in intact thalamo-cortical loops but decreased pathological oscillations and corresponding GSW-type seizures in the same series of experiments (Gibbs et al., 2002). Measuring electromagnetic oscillations at a large scale was successfully applied to human idiopathic generalized epilepsy: LTG decreased pathologically synchronized EEG background activity in a use-dependent manner (Clemens et al., 2007a). However, dampening thalamo-cortical oscillations cannot fully explain the mechanism of a drug against GSW-type seizures. The full antiabsence effect of ethosuximide is only exerted when the drug binds to a specific cortical site, the primary somatosensory cortex (Manning et al., 2004). Given the leading role of the cortex in initiating GSW-type seizures in human IGE, a cortical site for LTG binding might be postulated. A transcranial magnetic stimulation study disclosed the effect of LTG on the primary motor cortex (Manganotti et al., 1999). However, nothing is known about the effect of LTG on the remaining part of the cortical mantle. The goal of the present study was to localize the neuronal sources (generators) the activity of which was altered by LTG treatment. Our prior study suggested that LTG decreases EEG activ-
ity mainly in the posterior parts of the cortex (Clemens et al., 2007a). Albeit the distribution of EEG spectral power on the scalp may roughly refer to the LORETA-localized sources of greatly abnormal activity (Thatcher et al., 2005), exact localization of the sources from the raw EEG is impossible because of the so-called ‘‘inverse problem’’ of EEG (computing the localization of the sources from the potential field is a mathematical problem with an infinite number of equally probable solutions). Thus, no concrete hypothesis concerning the anatomy of the cortical action of LTG was generated beforehand of this study.
Patients and methods Inclusion criteria were newly diagnosed, unmedicated patients with ‘‘IGE with variable phenotypes’’ (Panayiotopoulos, 2005); age of onset after the eighth year of age. Exclusion criteria were permanent medication except oral contraceptives, alcohol or drug abuse, any medical condition that is known to influence EEG activity, generalized tonic-clonic seizure in 5 days before EEG investigation. The study design was approved by the Local Research Ethics Committee of Kenézy Gyula Kórház. 19 patients were recruited (12 females, 7 males; age, 10—29 years; average, 16.2 years). The patients were diagnosed and treated as usual. No diagnostic procedure or treatment was indicated, missed, or postponed for study purposes. Neurological and EEG investigation were done at the first visit, treatment started the next day. The initial daily dose of LTG was 25 mg and was progressively increased until seizure control was reached. At the end of the titration the daily dose of LTG was between 100 and 300 mg. No patient experienced symptoms or complaints indicating adverse effects of the drug. The first EEG was recorded in drug-free state. The second was done 3 months later when the patients took the effective daily dose of LTG. The EEG recordings were carried out in the morning, after a night of sufficient sleep, in the same semi-isolated room, with the same Brain Quick (Micromed, Mestre, Italy) digital equipment, by trained personnel, according to recommended Standards for Quantitative EEG (QEEG) studies (Nuwer et al., 1994). Silver—silver chloride electrodes were placed according to the 10—20 system, fixed by appropriate adhesive and conductive gel. Impedances did not exceed 5 k. 19-channel EEG was recorded against linked ears reference. Additional bipolar derivations were used to differentiate between EEG and eye movement potentials and to detect EMG activity. In the EEG derivations the filters were set at 0.1 and 30 Hz. Sampling frequency was 128 per second. 12-bit on-line digitization was used. 30 min EEG was recorded in waking-relaxed, eyes-closed condition. The state of vigilance was controlled by the EEG technician who gently aroused the patient when the alpha rhythm disappeared. The first 2 min of artifact-free editable EEG activity 60 epochs (each 2 s) were selected for QEEG analysis. Our standard epoch selection protocol includes: (1) the presence of continuous physiological (‘‘waking’’ or ‘‘resting’’) alpha activity with alpha voltage maximum in posterior regions, (2) the absence of artifacts, epileptiform potentials, and other nonstationary elements, (3) the absence of patterns indicating drowsiness or arousal. Epoch
206 selection and analysis was done blindly by one of the authors and controlled by another author (KI). This epoch selection method results in fairly reproducible results (John et al., 1983; Nuwer, 1988). The reported electrographic definition of the relaxed-waking state refers to a narrow range of vigilance (Bente, 1979), as supported by EEG and functional MRI (fMRI) co-registration (Laufs et al., 2006a).
LORETA analysis LORETA is a recently developed method to localize multiple distributed cortical sources of bioelectric activity in the three-dimensional space (Pascual-Marqui et al., 1994). In other words, LORETA demonstrates the synchronously activated neuronal populations underlying EEG activity by computing their cortical localization from the scalp distribution of the electric field. This is called solving the inverse problem of the EEG. The LORETA inverse solution is based on existing neuroanatomical and physiological knowledge and a mathematical constraint called the smoothness assumption (Pascual-Marqui, 2002a). The principles of LORETA and the mathematical tools have been described in details (http://www.keyinst.unizhch/loreta.html). The ‘‘smoothness assumption’’, which means that neighbouring neuronal generators show highly correlated activity in terms of orientation and strength (Pascual-Marqui, 2002a). The smoothness assumption is based on proven neuroanatomical and electrophysiological constraints as described elsewhere, including the mathematical basis of the method (http://www.keyinst.unizhch/loreta.html). In order to mathematically mitigate the disturbing effects of the electrically conducting layers between the cortical surface and the electrodes, LORETA computes the inverse solution within a three-shell spherical head model including scalp, skull, and brain. The brain compartment of this model was restricted to the cortical grey matter and hippocampus, according to the Talairach Brain Atlas digitized at Montreal Neurological Institute (Talairach and Tournoux, 1988). The grey matter compartment was subdivided in 2394 voxels, which allows a spatial resolution of 7 mm. LORETA computes a physically existing dimension, current density (A/m2 ) for each voxel. For the sake of brevity, this is called ‘‘activity’’ in this paper. The consistency of LORETA with physiology and localization has been validated by several authors. Crossmodal validation studies disclosed that LORETA and other functional neuroimaging methods showed the same cortical localization of dysfunction in several neuro-psychiatric conditions (Pascual-Marqui et al., 2002b). Concerning epilepsy, LORETA-defined localization of circumscribed activity corresponded to the localization of the epileptic discharges given by functional MRI, subdural and intracerebral EEG recordings, and the MRI-defined epileptogenic lesion (Lantz et al., 1997; Seeck et al., 1998; Worrell et al., 2000). The ability of LORETA to localize drug-related cortical changes has already been demonstrated (Saletu et al., 2002). We investigated the LORETA solution in four frequency bands separately (delta, 1.5—3.5 Hz; theta, 3.5—7.5 Hz; alpha, 7.5—12.5 Hz; beta, 12.5—25.0 Hz). Details on the computation of LORETA for several frequency bands can be found elsewhere (Frei et al., 2001). Thus, each patient was characterized by (4 × 2394 = 9576) current density values. Delta, theta, alpha, and beta group differences between the
B. Clemens et al. untreated and treated condition were computed for each voxel by parametric statistics. The anatomical localization of the voxels was carried out by the LORETA-KEY software: placing the cursor at any voxel displayed the localization of that voxel in terms of cortical anatomy. A total of 9588 comparisons were done (t-tests for interdependent datasets) without correction for multiple comparisons. The t-values were color-coded and projected onto axial brain slices. The findings were interpreted in terms of statistical significance and topographic distribution. Only the p < 0.01 results (tvalues outside the −3.037 to +3.037 range) were discussed. The uncorrected p = 0.01 threshold was reported to provide sufficient protection against type I error in a prior LORETA investigation (Pizzagalli et al., 2002). It is not sure that this statement is entirely valid for our study. Therefore, we did interpret this threshold as a sharp boundary that separates the meaningful and the unimportant changes. The postulated biological significance of the results was estimated by the degree of statistical significance (p) and the topographical distribution of the data.
Results LTG decreased EEG activity in all frequency bands to a variable degree, but uncorrected p < 0.01 changes emerged only in the theta and alpha bands. Fig. 1 demonstrates the degree of the changes (expressed in t-values) over the entire cortex. Theta activity was reduced in all voxels showing a peculiar postero-anterior distribution. However, only uncorrected p < 0.01 differences are described here as follows. The overwhelming majority of the voxels with decreased theta activity were crowded in a large, topographically contiguous cluster that included parts of the temporal, parietal, occipital cortex and insula. At the basal cortical surface, bilateral, partly asymmetrical theta decrease was found in the temporal cortex including the inferior temporal, fusiform, lingual, and parahippocampal gyri. The hippocampus and the uncus gyri hippocampi were affected on the right side only. Basal parts of the occipital cortex were affected bilaterally. At the cortical convexity, theta activity decreased in the middle and posterior parts of the lateral temporal cortex, involving a greater area in the left than in the right hemisphere. The transverse temporal gyri and the posterior part of the insula showed decreased theta activity on the left but not the right side. The occipital gyri on the lateral surface were involved bilaterally. The postcentral gyrus was completely involved on the left but only marginally on the right side. The angular and supramarginal gyri and part of the superior parietal lobule were involved bilaterally. A small area of the left precentral gyrus at the cortical convexity was the only part of the frontal lobes that showed decreased theta activity. Fig. 2 demonstrates that LTG decreased activity in several areas where pathologically increased activity had been reported in untreated IGE patients by the same authors (Clemens et al., 2007b). Also alpha activity decreased all over the cortex. However, p < 0.01 (uncorrected) changes only occurred in a limited part of the right hemisphere including posterior parts of the superior and medial temporal gyri, parts of the supramarginal and angular gyri, and part of the precuneus. A
Imaging lamatrigine’s action on the cortex topographically separated small cluster of 25 voxels with decreased alpha activity was found in the right insula.
Discussion EEG background activity of untreated IGE patients is characterized by increased synchronization in the delta, theta, and alpha frequency bands (Clemens et al., 2000). The presence of increased synchronization and coherence in the theta band was emphasized in epilepsy (Sarthein et al., 2003), and was demonstrated in IGE, too (Clemens, 2004). Albeit the exact contribution of the enhanced theta oscillations to seizure liability and ictogenesis has not been clarified yet, it is interesting that LTG caused the most prominent decrease of EEG synchronization exactly in the theta band in successfully treated IGE patients (Clemens et al., 2007a). Thus, it was not surprising that LORETA confirmed the decrease of synchronized theta activity in cortical neuronal pools. On the other hand, the anatomical distribution of the cortical effect of LTG was a novel finding that had not been predictable on the basis of our prior investigations. This was the first study to anatomically localize the cortical effect of LTG. The significance of our finding is that the net effect of LTG within the cortex (the cerebral compartment that is critical in determining seizure liability and seizure generation) was grasped and anatomically localized, without
207 knowing precisely the molecular and cellular actions of the drug at an unknown number of sites within the brain. The shortcoming of the study is suboptimal spatial sampling due to the limited number of electrodes, which may cause incorrect localization of small sources of activity. Nevertheless, we do not think that this resulted in significantly inaccuracy in this study. Resting EEG rhythms are generated by largely distributed cortical sources that can be accurately investigated by the standard 10—20 system (Nunez, 1995; Babiloni et al., 2006). In fact, comparative studies disclosed that LORETA analysis results in very similar localization of the generators with 19 and 46 electrodes provided that they are evenly distributed on the scalp (Michel et al., 2004). Statistics is a difficult task in LORETA studies. Unfortunately, software limitations prevented us from performing straightforward correction for multiple comparisons in order to compute a ‘‘hard’’ threshold of significance. However, the anatomically meaningful distribution of the results strongly suggested that they were due to biological causes rather than to by chance generated type-l errors. Fig. 1 combines the statistical and the topographical aspects of the results while only the uncorrected p < 0.01 findings were discussed. Our results posed another problem of interpretation. Which statistical degree of the LTG effect should be judged as biologically relevant in case of the reported topographical distribution? The ‘‘top of the iceberg’’ (that is, the
Figure 1 LORETA-defined decrease of theta and alpha activity. The Z-coordinates (millimeters) of the Talairach Brain Atlas label the selected axial slices. Blue color indicates the decrease of activity. t-Values are color-coded from t = 0 to −6.074 (t = −6.074 refers to p = 0.001). Laterality (R, L) is given in the Figure.
208 p < 0.001 clusters) or the more broadly outlined areas of less statistical significance should be considered as biologically significant? We cannot answer this question at present. The lack of relevant papers prevented hypothesis generation concerning the expected degree and the exact anatomical distribution of the effect of LTG. Hence, limiting the discussion to the uncorrected p < 0.01 changes was arbitrary, decided intuitively rather than based on scientific evidence. The main question of interest is whether topographic overlap exists between the LORETA-defined LTG effect and the cortical areas that are involved in epileptic dysfunction in IGE. The bulk of evidence suggests the leading role of the medial and basal frontal cortex in initiating generalized seizures (Bancaud et al., 1974; Holmes et al., 2004). However, seizure onset is a hardly definable term in case of generalized seizures. For example, the mediobasal frontal distribution of ictal activity of absence seizures is invariably seen only ‘‘by the second to fourth spike after onset of the epileptiform discharge’’, and the very onset of the seizure is characterized by focal irritative activity at variable sites of the frontal, temporal, and parietal cortex (Holmes et al., 2004). Also interictal GSW discharges typically show electric fields with negativity maxima over the frontal areas and a gradual decline of voltage along the temporal and parietal cortex, suggesting a widespread area with ictogenic property. These results suggest that other, yet not specified cortical areas may contribute to the seizure generation in IGE. The indisputable effect of LTG in IGE (including this cohort of patients) and its reported effect in the posterior parts of the cortex raise the possibility that also some posterior cortical areas might have a significant but underes-
B. Clemens et al. timated role in ictogenesis. Some pieces of evidence argue for this possibility as follows. Having postulated that the ictogenic property of the cortex is not evenly distributed across the scalp, we localized five cortical ‘‘clusters’’ (clusters of voxels) where current source density was pathologically increased in untreated IGE patients (Clemens et al., 2007b). Out of them, the two clusters located in the posterior half of the brain anatomically fairly correspond to cortical areas where LTG decreased theta activity (Fig. 2). The ‘‘temporo-basal cluster’’ in the untreated patients is topographically very similar to the distribution of the LTG-related decrease of theta activity in the medial basal temporal cortex. Temporal negativity in the early evolution of GSW discharge was reported in voltage mapping and LORETA studies (Rodin and Ancheta, 1986; Holmes et al., 2004). However, the contribution of the temporal cortex to the neurophysiology of IGE has been a completely neglected matter. Another area of increased activity, the ‘‘medial parietooccipital cluster’’ shows a remarkable overlap with the LTG-related theta decrease in the medial parieto-occipital area (Fig. 2). The overall physiological significance of this area in epilepsy is its capability dissipates that local epileptic activity can easily spread from this site to many other sites of the brain. Electroencephalographic evidence indicates preferential spread of epileptic activity from the medial and lateral parietal cortex to many limbic and neocortical areas (Sveinbjornsdottir and Duncan, 1993). Occipital epileptic foci can give rise to bilateral, frontally preponderant interictal discharges (Williamson and Spencer, 1986). Pieces of evidence suggest that the medial parieto-
Figure 2 Top row (from Clemens et al., 2007b; with permission of Blackwell Publisher): untreated IGE patients. Red color: uncorrected p < 0.01 increase of theta activity, in the medial temporal area (temporo-basal cluster) and at the medial surface of the left hemisphere (medial parieto-occipital cluster). Bottom row, blue color: LTG-related decrease of theta activity in about the same areas. The relevant Talairach coordinates and laterality (R, L) is given in the Figure.
Imaging lamatrigine’s action on the cortex occipital area has a probable but unexplored role in the neurophysiology of IGE, too. Besides the main frontal negativity, a medial parieto-occipital negativity in the early course of the GSW complex was described by the authors who used voltage mapping analysis (Rodin and Ancheta, 1986; Coppola, 1988; Hughes et al., 1990; Ferri et al., 1995). This area was called the ‘‘posterior center of gravity of the ictal discharge’’ (Rodin, 1999). Further arguments supporting the contribution of this area to the ictogenic condition in IGE are the excess of excitatory neurotransmission in this area (Simister et al., 2003), the patients whose absence seizures arise from the posterior parts of the cortex (Gomez and Westmoreland, 1987), and the midparietal, mid-occipital spike component frequently seen at the beginning of the generalized photoparoxysmal response in IGE patients. We suggest that the topographic overlap between increased theta activity in the ‘‘medial parietooccipital cluster’’ and the drug-related theta decrease in the same area can hardly be a per chance finding. Albeit the relationship between cortical EEG activity and seizure control is a complex matter involving subcortical physiology and genetic influences, our results rise the possibility that LTG counteracts seizure generation by dampening theta activity at specific cortical sites. Nowadays, advanced neuroimaging methods (in particular, fMRI) seem to be valuable tools to address neurophysiological problems that previously had been the privilege of EEG. fMRI studies of absence seizures disclosed bilateral activation in the thalamus, mesial midfrontal regions, insula and cerebellum, while bilateral deactivation was found in the anterior frontal and parietal cortex and the posterior cingulate (Archer et al., 2003; Gotman et al., 2005). We demonstrated that LTG greatly decreased theta activity (p < 0.001, uncorrected) in one part of this network, the posterior cingulate and the surrounding medial parietal cortex. In addition to its role in elaborating absence seizures this cortical area is part of an extended neuronal system the activity of which is related to the ‘‘default state of the brain’’, a fascinating concept that has been refined recently (Raichle and Snyder, 2007). Absence seizures cause fMRI deactivation in this system (Gotman et al., 2005; Laufs et al., 2006b). The posterior cingulate area seems to be the ‘‘key coupling site between the brainstem system for arousal and cortical systems for cognitive processing and awareness’’ (Vogt and Laureys, 2005). Increased theta activity in scalp-recorded EEG generally indicates decreased vigilance and impaired cognition. Thus, it is possible that the established beneficial effect of LTG on vigilance and attention (Aldenkamp et al., 2002) is mediated by decreasing theta activity in this part of the cortex. Another point of interest is the emerging role of fMRI in pharmacological studies (Ianetti and Wise, 2007). Scalp EEG only reflects cortical activity, which, nevertheless, is under the control of cortico-cortical and subcortico-cortical influences. The proportion of cortically and subcortically driven activity cannot be separated on EEG grounds. Thus, we do not state that our findings reflect the direct action of LTG on the cortex. It is possible that some of the reported effect is mediated via subcortical sites. fMRI studies may shed light to the subcortical actions of LTG. Integrating EEG and fMRI methods highlighted several aspects of IGE pathophysiology (Archer et al., 2003; Gotman et al., 2005; Laufs
209 et al., 2006b), and also might highlight the neuronal mechanisms by which LTG exerts its clinical effects on seizures and vigilance.
Acknowledgement This work has been supported by Hungarian Science Foundation (OTKA) grant #T 048338.
References Aldenkamp, A.P., Arends, J., Bootsma, H.P.R., Diepman, L., Hulsman, J., Lambrechts, D., Leenen, L., Majoie, M., Schellekens, A., de Vocht, J., 2002. Randomized double-blind parallel-group study comparing cognitive effects of a low dose lamotrigine with valproate. Epilepsia 43, 19—26. Archer, J.S., Abbott, D.F., Waites, A.B., Jackson, G.D., 2003. fMRI deactivation of the posterior cingulate during generalized spike and wave. Neuroimage 20, 1915—1922. Babiloni, C., Frisoni, G., Steriade, M., Bresciani, L., Binetti, G., Del Percio, C., Geroldi, C., Miniussi, C., Nobili, F., Rodriguez, G., Zappasodi, F., Carfagna, T., Rossini, P.M., 2006. Frontal white matter volume and delta sources negatively correlate in awake subjects with mild cognitive impairment and Alzheimer’s disease. Clin. Neurophysiol. 117, 1113—1129. Bancaud, J., Talairach, J., Morel, P., Bresson, M., Buser, P., 1974. Generalized epileptic seizures elicited by electrical stimulation of the frontal lobe is man. Electroenceph. Clin. Neurophysiol. 37, 275—282. Bente, D., 1979. Vigilance and evaluation of psychotropic drug effect on EEG. Pharmacopsychiatry 12, 137—147. Clemens, B., Szigeti, G., Barta, Z., 2000. EEG frequency profiles of idiopathic generalized epilepsy syndromes. Epilepsy Res. 42, 105—115. Clemens, B., 2004. Pathological theta oscillations in idiopathic generalised epilepsy. Clin. Neurophysiol. 115, 1436—1441. Clemens, B., Piros, P., Bessenyei, M., Hollódy, K., 2007a. Lamotrigine decreases EEG synchronization in a use-dependent manner in patients with idiopathic generalized epilepsy. Clin. Neurophysiol. 118, 910—917. Clemens, B., Bessenyei, M., Piros, P., Tóth, M., Seress, L., Kondákor, I., 2007b. Characteristic distribution of interictal brain electrical activity in idiopathic generalized epilepsy. Epilepsia 48, 941—949. Coppola, R., 1988. Topographic display of spike-wave discharges. In: Myslobodsky, M.S., Mirsky, A.F. (Eds.), Elements of Petit Mal Epilepsy. Peter Lang Publ. Inc., New York, pp. 105—129. Ferri, R., Iliceto, G., Caelucci, V., 1995. Topographic EEG mapping of 3/s spike-and-wave complexes during absence seizures. Ital. J. Neurol. Sci. 16, 541—547. Frei, E., Gamma, A., Pascual-Marqui, R.D., Lehmann, D., Hell, D., Vollenweider, F.X., 2001. Localization of NMDA-induced brain activity in healthy volunteers using low resolution brain electromagnetic tomography (LORETA). Hum. Brain Map. 14, 152—165. Gomez, M.R., Westmoreland, B.F., 1987. Absence seizures. In: Lüders, H.O., Lesser, P. (Eds.), Epilepsy Electroclin. Synd. Springer, London, pp. 105—130. Gibbs, J.W., Yun-Fu Zhang, Ahmed, H.S., Coulter, D.A., 2002. Anticonvulsant actions of lamotrigine on spontaneous thalamocortical rhythms. Epilepsia 43, 342—349. Gloor, P., 1979. Generalized epilepsy with spike and wave discharge: a reinterpretation of its electrographic and clinical manifestations. Epilepsia 20, 571—588. Gotman, J., Grova, C., Bagshaw, E., Kobayashi, E., Agakhani, Y., Dubeau, F., 2005. Generalized epileptic discharges show thala-
210 mocortical activation and suspension of the default state of the brain. PNAS 102, 15236—15240. Holmes, M.D., Brown, M., Tucker, D.M., 2004. Are ‘‘generalized’’ seizures truly generalized? Evidence of localized mesial frontal and frontopolar discharges in absence. Epilepsia 45, 1568—1579. Hughes, J.R., Miller, J.K., Hughes, C.A., 1990. Topographic mapping of different types of bilateral spike-and-wave complexes. J. Epilepsy 3, 67—74. Ianetti, G.D., Wise, R.G., 2007. BOLD functional MRI in disease and pharmacological studies: room for improvement. Magn. Reson. Imag. 25, 978—988. John, E.R., Prichep, L., Ahn, H., Easton, P., Fridman, J., Kaye, H., 1983. Neurometric evaluation of cognitive dysfunctions and neurological disorders in children. Progress Neurobiol. 21, 290—329. Lantz, G., Michel, C.M., Pascual-Marqui, R.D., Spinelly, L., Seeck, M., Seri, S., Landis, T., Rosen, I., 1997. Extracranial localization of intracranial interictal epileptiform activity using LORETA (low resolution electromagnetic tomography). Electroenceph. Clin. Neurophysiol. 102, 414—422. Laufs, H., Holt, J.L., Elfont, R., Krams, M., Paul, J.P., Krakow, K., Kleinschmidt, A., 2006a. Where the BOLD signal goes when alpha EEG leaves. Neuroimage 31, 1408—1418. Laufs, H., Lengler, U., Hamandi, K., Kleinschmidt, A., Krakow, K., 2006b. Linking generalized spike-and-wave discharges and resting state brain activity by using EEG/fMRI in a patient with absence seizures. Epilepsia 47, 444—448. Leach, M.J., Lees, G., Ridall, D.R., 1995. Lamotrigine Mechanisms of Action. In: Levy, R.H., Mattson, R.H., Meldrum, B.S. (Eds.), Antiepileptic drugs, fourth edition. Raven Press, New York, pp. 861—902. Manganotti, P., Bongiovanni, L.G., Zanette, G., Turazzini, M., Fiaschi, A., 1999. Cortical excitability in patients after loading dose of lamotrigine: a study with magnetic brain stimulation. Epilepsia 40, 316—321. Manning, J.P., Richards, D.A., Leresche, N., Crunelli, V., Bowery, N.G., 2004. Cortical-area specific block of genetically determined absence seizures by ethosuximide. Neuroscience 123, 5—9. Michel, C.M., Murray, M.M., Lantz, G., Gonzales, S., Spinelli, L., Grave de Peralta, R., 2004. EEG source imaging. Clin. Neurophysiol. 115, 2195—2222. Nunez, P.L., 1995. Quantitative states of neocortex. In: Nunez, P.L. (Ed.), Neocortical Dynamics and Human EEG Rhythms. University Press, Oxford, pp. 1—18. Nuwer, M., 1988. Quantitative EEG. 1. Techniques and problems of frequency analysis and topographic mapping. J. Clin. Neurophysiol. 5, 1—43. Nuwer, M., Lehmann, D., Lopes da Silva, F., Matsuoka, S., Sutherling, W., Vibert, J.F., 1994. IFCN guidelines for topographic and frequency analysis of EEGs and EPs. Report of an IFCN Committee. Electroenceph. Clin. Neurophysiol. 91, 1—5. Panayiotopoulos, C., 2005. Idiopathic generalized epilepsies. In: Panayiotopoulos, C. (Ed.), The Epilepsies. Seizures, Syndromes, and Management. Springer, pp. 271—348. Pascual-Marqui, R.D., Michel, C.M., Lehmann, D., 1994. Low resolution electromagnetic tomography. A new method for localizing electrical activity in the brain. Int. J. Psychophysiol. 18, 49—65. Pascual-Marqui, R.D., 2002a. Functional imaging with lowresolution brain electromagnetic tomography (LORETA): a review. Methods Find Exp. Clin. Pharmacol. 24 (Suppl. C), 91—96.
B. Clemens et al. Pascual-Marqui, R.D., Esslen, M., Kochi, K., Lehmann, D., 2002b. Functional imaging with low-resolution brain electromagnetic tomography (LORETA): review, new comparisons, and new validation. Jpn. J. Clin. Neurophysiol. 30, 81—94. Pizzagalli, D.A., Nitschke, J.B., Oakes, T.R., Hendrick, A.M., Horras, K.A., Larson, C.L., Abercrombie, H.C., Schaefer, S.M., Koger, J.V., Benca, R.M., Pascual-Marqui, R.D., Davidson, R.J., 2002. Brain electrical tomography in depression: the importance of symptom severity, anxiety, and melancholic features. Biol. Psychiatry 52, 73—85. Raichle, M.E., Snyder, A.Z., 2007. A default mode of brain function: a brief history of an evolving idea. NeuroImage 37, 1083—1090. Rodin, E., Ancheta, O., 1986. Cerebral electric fields during petit mal absences. Electroenceph. Clin. Neurophysiol. 66, 457—466. Rodin, E., 1999. Decomposition and mapping of generalized spikewave complexes. Clin. Neurophysiol. 110, 1868—1875. Saletu, B., Anderer, P., Saletu-Zyhlarz, G.M., Arnold, O., Pascual-Marqui, R.D., 2002. Classification and evaluation of the pharmacodynamics of psychotropic drugs by single-lead pharmaco-EEG, EEG mapping and tomography (LORETA). Methods Find Exp. Clin. Pharmacol. 24 (Suppl. C), 97—120. Sarthein, J., Morel, A., von Stein, A., Jeanmonod, D., 2003. Thalamic field potentials and EEG: high thalamocortical coherence in patients with neurogenic pain, epilepsy, and movement disorders. Thalamus Relat. Syst. 2, 231—238. Seeck, M., Lazeyras, F., Michel, C.M., Blanke, O., Gericke, C.A., Ives, J., Delavelle, J., Golay, X., Haenggeli, C.A., de Tribolet, N., Landis, T., 1998. Non-invasive epileptic focus localization using EEG-triggered functional MRI and electromagnetic tomography. Electroenceph. Clin. Neurophysiol. 106, 508—512. Simister, R.J., McLean, M.A., Barker, G.J., Duncan, J.S., 2003. A proton magnetic resonance spectroscopy study of metabolites in the occipital lobes in epilepsy. Epilepsia 44, 550—558. Steriade, M., Amzica, F., 1994. Dynamic coupling among neocortical neurons during evoked and spontaneous spike-wave seizure activity. J. Neurophysiol. 72, 2051—2069. Steriade, M., Contreras, D., 1995. Relation between cortical and thalamic cellular events during transition from sleep patterns to paroxysmal activity. J. Neurosci. 15, 623—642. Steriade, M., 2001. Impact of network activities on neuronal properties in corticothalamic systems. Am. Physiol. Soc. 86, 1—39. Sveinbjornsdottir, S., Duncan, J.S., 1993. Parietal and occipital lobe epilepsy: a review. Epilepsia 34, 493—521. Talairach, J., Tournoux, P., 1988. Co-planar Stereotaxic Atlas of the Human Brain: Three-Dimensional Proportional System. G. Thieme, Stuttgart. Thatcher, R.W., North, D., Biver, C., 2005. Evaluation and validity of a LORETA normative EEG database. Clin. EEG Neurosci. 36, 116—122. Vogt, B.A., Laureys, S., 2005. Posterior cingulate, precuneal and retrosplenial cortices: cytology and components of the neural network correlates of consciousness. Prog. Brain Res. 150, 205—217. Williamson, P.D., Spencer, S.S., 1986. Clinical and EEG features of complex partial seizures of extratemporal origin. Epilepsia 27 (Suppl. 2), S46—S63. Worrell, G.A., Lagerlund, T.D., Sharbrough, F.W., Brinkmann, B.H., Busacker, N.E., Cicora, K.M., O’Brien, T.J., 2000. Localization of the epileptic focus by low resolution electromagnetic tomography in patients with a lesion demonstrated by MRI. Brain Topogr. 12, 273—282.