- Email: [email protected]
Origin and dynamics of epileptic activity in a symptomatic case of Panayiotopoulos syndrome: Correlation with clinical manifestations
Clinical Neurophysiology 124 (2013) 20–26
Contents lists available at SciVerse ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Origin and dynamics of epileptic activity in a symptomatic case of Panayiotopoulos syndrome: Correlation with clinical manifestations Alberto J.R. Leal a,b,⇑, Ricardo Lopes c, José C. Ferreira d a
Department of Neurophysiology, Hospital Júlio de Matos, Lisbon, Portugal CIS-ISCTE, Lisbon University Institute (IUL), Lisbon, Portugal c Department of Psychology, University of Coimbra, Coimbra, Portugal d Department of Pediatric Neurology, Centro Hospitalar Lisboa Ocidental, Lisbon, Portugal b
a r t i c l e
i n f o
Article history: Accepted 1 June 2012 Available online 23 July 2012 Keywords: Epilepsy Panayiotopoulos syndrome EEG Parietal lobe
h i g h l i g h t s Panayiotopoulos syndrome (PS) is a frequent childhood epileptic syndrome with typical clinical features but unknown localisation of the epileptogenic area. We describe the first symptomatic case of PS where both clinical and EEG features converge to demonstrate that an epileptic focus in the inferior parietal lobe can originate the epileptic syndrome. Fast spread of epileptic activity through physiological networks involved in eye-movement control, gastrointestinal autonomic control and consciousness can explain the diversity of clinical manifestations in PS.
a b s t r a c t Objective: The aim of the study was to demonstrate the dynamics and structure of the epileptic network and provide a tentative correlation with the clinical manifestations, in a symptomatic case of Panayiotopoulos syndrome (PS). Methods: JP, 5-year-old girl. Gestational period and developmental milestones were normal. At age 4 years, two episodes of recurrent vomiting, tonic eye deviation and consciousness impairment lasting for about 30 min occurred. Multifocal spikes were apparent over frontal areas in the EEG and MRI demonstrated an inferior parietal lobe (IPL) lesion. Results: A long-term 35-channel scalp EEG was obtained, which was processed with a Blind Source Separation algorithm. The most significant components with a dipolar field were submitted to source analysis and the recovered generators used to build the nodes of a brain network associated with each spike type. Analysis of information flow supported epileptic propagation from the left parietal lobe to both frontal and temporal lobes around spike peak. The good spatial overlap with physiological networks controlling eye movements, autonomic functions and consciousness, provides a tentative explanation to the diverse clinical manifestations of PS. Conclusions: Spreading patterns of epileptic activity form an extended network in PS. Significance: An epileptic focus in an IPL can reproduce both neurophysiological and clinical features of PS. Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction The Panayiotopoulos syndrome (PS) is one of the most recently described and prevalent benign epilepsy syndromes of childhood. Despite the significant effort devoted to elucidate its clinical (Panayiotopoulos, 2002), neurophysiological (Caraballo et al., ⇑ Corresponding author. Address: Department of Neurophysiology, Hospital Júlio de Matos, Ava do Brasil nr 53, 1749-002 Lisbon, Portugal. Tel.: +351 969851734; fax: +351 217819809. E-mail address: [email protected] (A.J.R. Leal).
2007) and neuropsychological aspects (Specchio et al., 2010a), several problems remain that prevent a deeper insight into the aetiology of the syndrome: the multifocal interracial spike activity present in most cases precluded the unambiguous identification of a common epileptogenic area; the peculiar combination of ictal autonomic manifestations, eye deviation and long-lasting impairment of consciousness, typical of the condition, has not allowed the determination of the precise locus of the epileptogenic zone on clinical grounds; It remains uncertain whether a single, yet unknown, cortical focus is responsible for the stereotyped seizure manifestations or these can be produced by foci in diverse areas.
1388-2457/$36.00 Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinph.2012.06.007
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
The few neurophysiological recordings of ictal events in the literature have also failed to provide focal EEG patterns with clear clues to their origin (reviewed in Specchio et al., 2010b). Despite the close morphological similarities of the interictal spikes in PS with the ones of rolandic epilepsy, the localisation information of each type is very different, with rolandic spikes showing stable dipolar fields with consistent generators within the primary somato-sensory area (Leal et al., 2007) and PS spikes demonstrating variable scalp topographies and significant spread of epileptic activity (Leal et al., 2008; Yoshinaga et al., 2010; Kokkinos et al., 2010). Clues to the localisation of the epileptogenic area in PS could potentially be obtained from the analysis of the few reported symptomatic cases. Nevertheless, the analysis of such cases failed to provide a consistent hypothesis and has led the authors to postulate a fortuitous coincidence between the epileptic syndrome and the structural abnormalities (Panayiotopoulos, 2002; Yalçin et al., 2009). In none of the reported cases was a source analysis of the EEG epileptic activity performed, nor any type of processing besides visual evaluation. We report on a patient with important clues to the source of epileptic activity in PS. 2. Methods and patient data 2.1. Patient data JP, 5-year-old girl. In the gestational period, the mother was continuously medicated with warfarin. Vaginal delivery occurred at the 37th week and the birth weight was 2160 kg. The neonatal period was complicated with sepsis and transient insulin-dependent diabetes (up to the 53rd day of life) but developmental milestones were reached at appropriate age. In the first 3 years, three renal urinary infections were reported, with associated febrile seizures. At the age of 4 years, a night event occurred lasting about 30 min with vomiting, conjugate and sustained eye deviation upward and to the left side, with unresponsiveness. After regaining consciousness, there was a transient period of confusion and somnolence but ultimately she recovered completely. She was investigated at the emergency department of a local hospital and discharged after normal blood and serum parameters and also a normal head CT were obtained. Five months later, a similar nocturnal event was reported. An EEG revealed multifocal spike activity. A 1.5-T MRI scan demonstrated a cortical lesion in the left inferior parietal lobe (IPL), with imaging features compatible with cortical dysplasia (Fig. 1b). A long-duration (24 h) video-EEG monitoring could not record seizure events, but abundant interictal spikes were noticed, occurring independently over the central area of the right hemisphere and the frontal area of the left one (Fig. 1a). The cognitive development was evaluated with the Portuguese version of the Wechsler Preschool and Primary Scale of Intelligence-Revised (WPPSI-R). The neurological examination was normal. No genetic analyses were performed.
21
for data acquisition, controlled by the SystemPlus software. The EEG data were exported to the Scan 4.3.1 software (Compumedics NeuroScan, Charlotte, NC, USA), where the trace was visually inspected by an experienced clinical neurophysiologist (AL), to identify spikes with independent topology. For each of the two recognised classes of spike topology, an amplitude threshold was applied at the channel with maximum spike amplitude to automatically detect spikes of this class and insert a trigger at each detection. The triggers were then used to epoch the EEG with a window of 200 to +500 ms, which were then concatenated. The final result was a continuous EEG file containing all the detected spikes for each topology class. The concatenated EEG record for each spike class was then imported to the EEGLAB 10.1 software (Delorme and Makeig, 2004) and decomposed using the Infomax (Bell and Sejnowski, 1995) independent component analysis (ICA) algorithm. The ICA spatial components were fitted with a dipole solution (using the standard boundary element method (BEM) (Montreal Neurological Institute model) included in EEGLAB) and only the ones with a residual variance (RV) lower than 10% were retained. These were then ordered by their relative contribution to the spike signal in the time range from 50 to +50 ms around the peak. The dipole solutions associated with the four ICA components with the largest contribution to each spike type were obtained with the CURRY 6 software (Neuroscan, Charlotte, NC, USA), using the sLORETA algorithm (PascualMarqui, 2002) in a standard BEM model of the head (Fig. 2a and c). Spectral analysis of the ICA components activity around the spike peak was done using the event-related spectral perturbation method (ERSP) (Makeig, 1993) as well as the inter-trial coherence (ITC) method (Tallon-Baudry et al., 1996), both implemented in the EEGLAB software. Statistical significance of the changes was obtained through a bootstrap procedure comparing changes in the ERSP with the spectral distribution in the prestimulus period. The functional connectivity analysis was performed in the averaged time course of the four ICA components with the largest contribution to the spike of each class, using the point of maximum score in the sLORETA inverse solution as the node of each component. We used the adaptive direct transfer function (ADTF) (Wilke et al., 2008) as a dynamical measure of information flow between the ICA component nodes, as implemented in the e-connectome software (He et al., 2011). The average amount of information flow in the band 4–30 Hz was calculated and statistical significance was obtained using a permutation test with phase randomisation. 3. Results 3.1. Electro-clinical data Our clinical case exhibits the main features of PS, with rare, predominantly autonomic seizures, long-lasting impairment of consciousness, sustained conjugate eye deviation and multifocal interictal spikes in a normal background EEG. The WPPSI-R test revealed a verbal IQ of 107 (score 68%), a performance IQ of 101 (score 53%) and a full-scale IQ of 103 (score 58%). These results support an average cognitive development.
2.2. EEG acquisition and processing 3.2. Dynamics of epileptic activity The EEG signal during the 24-h video-EEG monitoring was collected with 35 gold disc electrodes glued with Colodium to the scalp positions (Fp1/2, F3/4, C3/4, P3/4, O1/2, F7/8, T7/8, P7/8, Fz, Cz, Pz, F11/12, FT9/10, TP9/10, P11/12, FC5/6, FC1/2, CP5/6 and CP1/2). The sampling rate was 1000 Hz, the high- and low-pass filters were set at 0.5 and 70 Hz and the electrode impedances remained below 5 kX. A Micromed EEG recording system was used
The EEG revealed a normal background rhythm and multifocal spikes with consistent maxima over the frontal lobes of both hemispheres (Fig. 1a). No spikes could be identified in the left parietal– occipital electrodes, near the scalp projection of the cortical lesion. The mapping of the scalp potential at different times through the rising phase of the spikes revealed a changing topographical con-
22
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
Fig. 1. (a) Interictal EEG featuring a normal background activity and independent spikes with maxima over Fp1 and FC6. On the right, sequential maps of the scalp potential are shown, demonstrating changes in topology from spike onset to spike peak and (b) Structural MRI with abnormal signal over the left inferior parietal lobe, compatible with cortical dysplasia, (FLAIR sequence).
figuration (Fig. 1a, right), suggesting the involvement of several spatially distinct generators, with diverse temporal dynamics. The four ICA components with the largest (and statistically significant, Fig. 2) contribution to the spikes of each topology class were able to explain 99.1% and 99.0%, respectively of the left and right spikes, effectively preserving the main features of each epileptic event. The source analysis of these ICA components demonstrates that their generators are widely separated in space (Fig. 2a and c), supporting the contribution of several cortical areas to the genesis of each interictal spike class. The different timing of the peak of each spike class ICA components (Fig. 2a and b) further supports a sequential activation in time, likely associated with fast propagation of the epileptic activity. The sequential temporal
peaking of the ICA components demonstrates an early activation of the left parietal generators for both spike classes. Because the cortical lesion is located in the left parietal lobe, these data effectively support a dynamical model of interictal spike activity postulating an early activation of a generator in the anatomical lobe containing the lesion, followed by secondary propagation to the left temporal lobe and bilateral frontal lobe generators. 3.3. Functional connectivity analysis The functional connectivity analysis using the generators of the ICA components as nodes and their time courses to calculate the information flow in an epileptic network for each spike class re-
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
23
Fig. 2. Decomposition of the two types of epileptic spikes and selection of the four components (rows) with the largest contribution in the time range 50 to 50 ms (centred on spike peak). (a) Fixed spatial components are shown in the left column, the average time course on the middle one (horizontal axis 200 to +500 ms, vertical axis in lV) and the source generator of each component is represented in the right column, with a dipole on the maximum of the sLORETA scores, (b) the components spectral changes around the spike peak are shown, both for power (ERSP) and phase (ITC). Green colour represents non-significant changes in the bootstrap statistics used, (c) similar to (a), (d) similar to (b). The data demonstrates the contribution of several cortical generators for each spike type, widely separated in space and with diverse temporal dynamics. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
vealed: (a) from spike onset to spike peak, sources of outflow information were located in the left parietal lobe only (Fig. 3b–d and f) and (b) inflow sources were scattered between the frontal lobes and the left temporal one. These observations effectively
support a model with propagation of epileptic activity from the left parietal lobe to distinct frontal and temporal areas as the likely explanation for the diverse spike topography present in the raw scalp EEG.
24
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
Fig. 3. Analysis of information flow between the four generators of right frontal spikes (above) and the left frontal ones (below). The nodes of the networks obtained from previous source analysis (Fig. 2) are shown in (a) and (d), together with the average time course of each component (below). The average flow of information in the 4–30 Hz spectral band is shown just before (b and e) and at spike peak (c and f), together with the connection matrix (below). Only information flow from the left parietal lobe (component 4) to the remaining areas reached statistical significance.
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
3.4. Spatial relationship of the epileptic focus with symptomatogenic cortical areas The existence of a focus of interictal spikes in the pre-motor areas of the right hemisphere, in the neighbourhood of the frontal eye field (FEF) (Fig. 2a, third row) suggests that the tonic and conjugate eye deviation to the left side, reported on both clinical seizures, might be due to epileptic stimulation of this structure. The propagation of epileptic activity through the functional network responsible for the control of saccadic eye movements, involving the IPL, superior colliculus (supCol) and the FEF, could explain the selectivity of this motor manifestation. The dysplastic cortex seen in the MRI (Fig. 1b) is in the neighbourhood of a node of such a network and the associated epileptogenic focus is in a good position to spread the epileptic activity to the saccadic eye-movement system, which is supported by the significant flow of information from the left parietal lobe to the right FEF area (Fig. 3c). Another node of the epileptic network with interest to explain the vomiting typical of PS is located in the left temporal lobe, and the associated generator (Fig. 2a, fourth row) is compatible with insular cortex activation. The demonstration of significant flow of information from the IPL to such a node (Fig. 3b) provides support to the spread of epileptic activity from the IPL to the insula. The co-existence of a node of the default mode network (DMN) in the IPL (Raichle et al., 2001) and the recent studies establishing a relationship between the degree of DMN connectivity and consciousness (Vanhaudenhuyse et al., 2010) suggest that disturbances in such connectivity by an epileptic focus in the IPL could lead to the impairment of consciousness so typical of PS. 4. Discussion Our clinical case features the major manifestations of PS: rare episodes of recurrent vomiting, tonic eye deviation and long-lasting (30 min) impairment of consciousness in a cognitively welldeveloped child. The rare frequency of seizures and the typical multifocal EEG spike activity further support this syndrome classification. Despite the fact that most cases of PS have no structural lesion in the brain MRIs (Panayiotopoulos, 2002; Caraballo et al., 2007; Specchio et al., 2010a), symptomatic cases have been described (Yalçin et al., 2009; Panayiotopoulos, 2002). Case #60 of the series reported by Panayiotopoulos (2002) had a dysembryoblastic neuroepithelial tumour in the right temporo-parieto-occipital region and seizures with prominent autonomic features (vomiting and sweating) and impairment of consciousness. One of the seizures lasted for hours. Two of the three patients reported by Yalçin et al. (2009) had extensive structural lesions involving not only the occipital lobes but also surrounding parietal and temporal regions, and therefore preventing a precise anatomical localisation. In general, these cases did not provide data to suggest to the authors of the studies a candidate epileptogenic area that could explain both the clinical and neurophysiological features of the syndrome, because structural lesions were too diffuse and because they failed to demonstrate a good spatial co-localisation of cortical lesions and epileptic spikes. All the previous cases demonstrated, as far as it is possible to evaluate from the reported data, extensive structural lesions involving the occipital lobe but extending to neighbour lobes and therefore possibly overlapping with the IPL lesion present in the case JP. 4.1. Dynamics of epileptic spikes Several studies of interracial spikes in PS (Leal et al., 2007; Yoshinaga et al., 2010; Kokkinos et al., 2010) have documented the existence of significant topographical changes in the potential
25
maps from spike onset to spike peak, demonstrating the co-existence of several intracranial sources, widely separated in space. This observation supports the hypothesis that significant spread of epileptic activity takes place in interictal and, presumably, in ictal discharges. Because some of these sources of epileptic activity are of very low amplitude (Leal et al., 2008), they can remain unrecognised in clinical visual analysis, leading to wrong cortical localisations or to the illusion of multifocal spike activity if not analysed with sensitive enough tools. In the patient JP, the demonstration that the two apparently independent spike types are preceded in time by a small left parietal source (Fig. 2a and c, upper row), supports a model postulating a single epileptogenic focus with two different secondary spreading patterns versus an alternative model postulating independent foci in the two hemispheres. The fact that the primary source of epileptic activity is in the lobe containing the structural lesion strongly supports a non-fortuitous connection between the two. The ICA technique produces an unmixing of the different sources of activity that add in the scalp to produce the EEG. The method has previously demonstrated its capability to identify small components in the scalp EEG (Leal et al., 2008) and also to demonstrate epileptic propagation between brain lobes (Leal et al., 2007), both important features for a full characterisation of epileptic networks, despite the fact that current EEG source-analysis methods provide little information on the spatial extent of intracranial sources. In the case JP, several independent components could be recovered for the two types of spikes, with consistent phase relation (Fig. 2b and d) but different topographies and source localisation (Fig. 2a and c). In the patient JP, the analysis of the flow of information between these sources, (Fig. 3), produced convergent evidence for spread of epileptic activity from the left parietal lobe containing the structural lesion towards both the frontal lobes containing the interictal spikes visible in the surface EEG and also to the left temporal lobe. These results agree and complement the ones of the time to peak analysis of the ICA components composing the spikes (Fig. 2a and c, middle column), which show an earlier activation of parietal sources as compared with both frontal and temporal ones. The use of the ADTF in human epilepsy, for both interictal (Wilke et al., 2008; 2009) and ictal recordings (Mierlo et al., 2011; Wilke et al., 2012), has allowed the recovery of propagation patterns validated in successful surgery for epilepsy. 4.2. Electro-clinical correlations The demonstration that electrical stimulation of the IPL cortex produced conjugate eye movements was done long ago (Ferrier, 1875; Ferrie et al., 2007), but this effect is thought to be mediated through the strong functional connections of this area with the FEFs (Blanke and Seeck, 2003) and also the supCol. In our clinical case, the demonstration of spread of epileptic activity to the right frontal lobe (Fig. 3c), in the neighbourhood of the FEF, provides a tentative explanation for the reported conjugate eye deviation to the left during seizure events, an expected effect of the stimulation of the right FEF (Blanke and Seeck, 2003). Sekimoto et al. (2007) reported two cases of parietal lobe epilepsy with ictal vomiting and no impairment of consciousness, which presented interictal spikes detected in MEG but not in the EEG, suggesting foci in the depth of sulci. In line with the interpretation of our case, these authors suggest that the clinical manifestations result from epileptic propagation to the insula, a cortical area prominently associated with autonomic and visceral motor manifestations (Insnard et al., 2004). The suspicion of insular cortex involvement in the genesis of ictal vomiting became widely accepted ever since Penfield and Jaspers (1954) first reported visceral motor manifestations associated with insular electrical stimulation. In clinical epileptology, vomiting has been most often associ-
26
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
ated with temporal lobe epilepsy (Kramer et al., 1988), but a recent study using stereo-EEG (Catenoix et al., 2008) suggests that this is due to secondary propagation to the insula. The proposed spread of epileptic activity from the IPL to the insula also finds support in the strong functional and structural connectivity between the IPL and the insula demonstrated by Uddin et al. (2010), using fMRI resting-state analysis and diffusion tensor imaging (DTI) tractography. In patient JP, the information flow from the left parietal lobe towards the left temporal source before spike peak is compatible with epileptic spread to the insula, which is well known to produce surface source topographies undistinguishable from the ones produced by more superficial sources on the overlying temporal cortex. The impairment of consciousness in PS is a prominent and persistent ictal feature the origin of which is not known. Recent studies have implicated the default network as a functional system related to consciousness (Vanhaudenhuyse et al., 2010), and with disturbances of this function in several epilepsies (Danielson et al., 2011). The IPLs contain a node of the DMN and epileptic seizures involving this area are well positioned to produce a disturbance of its functional connectivity leading to consciousness impairment. Overall, our clinical case highlights the IPL as a brain region including important nodes of the networks controlling eye movement (IPL, FEF and supCol), visceral motor responses (IPL-Insula) and consciousness (default network), forming an extended epileptic network whose activation in seizure events can mimic some of the cardinal clinical manifestations of PS. We therefore suggest that the IPL is a cortical region with all the capabilites to act as the primary source of epileptic activity in our case of PS, and of reproducing the typical behavioural manifestations through spread of the epileptic activity along specific physiological networks. References Bell A, Sejnowski T. An information-maximization approach to blind source separation and blind deconvolution. Neural Comp 1995;7:1004–34. Blanke O, Seeck M. Direction of saccadic and smooth eye movements induced by electrical stimulation of the human frontal eye field: effect of orbital position. Exp Brain Res 2003;150:174–83. Caraballo R, Cersosimo R, Fejerman N. Panayiotopoulos syndrome: a prospective study of 192 patients. Epilepsia 2007;48:1054–61. Catenoix H, Isnard J, Guenot M, Petit J, Remy C, Mauguière F. The role of the anterior insular cortex in ictal vomiting: a stereotactic electroencephalography study. Epilep Behav 2008;13:560–3. Danielson N, Guo J, Blumenfeld H. The default mode network and altered consciousness in epilepsy. Behav Neurol 2011;24:55–65. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics. J Neurosc Meth 2004;134:9–21. Ferrie C, Caraballo R, Covanis A, Demirbilek D, Dervent A, Fejerman N, et al. Autonomic status epilepticus in Panayiotopoulos syndrome and other childhood and adult epilepsies: a consensus view. Epilepsia 2007;48:1165–72.
Ferrier D. Experiments on the brain of monkeys-No. 1. Proceedings of the Royal Society 1875;23:409-430. He B, Dai Y, Astolfi L, Babiloni F, Yuan H, Yang L. EConnectome: a MATLAB toolbox for mapping and imaging of brain. J Neurosci Meth 2011;195:261–9. Insnard J, Guenot M, Sindou M, Mauguière F. Clinical manifestations of insular lobe seizures: a stereo-electroencephalographic study. Epilepsia 2004;45:1079–90. Kokkinos V, Koutroumanidis M, Tsatsou K, Koupparis A, Tsiptsios D, Panayiotopoulos C. Multifocal spatiotemporal distribution of interictal spikes in Panayiotopoulos syndrome. Clin Neurophysiol 2010;121:859–69. Kramer R, Luders H, Goldstick L, Dinner D, Morris H, Lesser R, et al. Ictus emeticus: an electroclinical analysis. Neurology 1988;38:1048–52. Leal A, Nunes S, Dias A, Vieira J, Moreira A, Calado E. Analysis of the generators of epileptic activity in early-onset childhood benign occipital lobe epilepsy. Clin Neurophysiol 2007;118:1341–7. Leal A, Ferreira J, Dias A, Calado E. Origin of frontal lobe spikes in the early onset benign occipital lobe epilepsy (Panayiotopoulos syndrome). Clin Neurophysiol 2008;119:1985–91. Makeig S. Auditory event-related dynamics of the EEG spectrum and effects of exposure to tones. Electroencephalogr Clin Neurophysiol 1993;86:283–93. Mierlo P, Carrette E, Hallez H, Vonck K, Roost D, Boon P, et al. Accurate epileptogenic focus localization through time-variant functional connectivity analysis of intracranial electroencephalographic signals. Neuroimage 2011;56:1122–33. Panayiotopoulos CP. Panayiotopoulos syndrome: a common and benign childhood epileptic syndrome. London: John Libbey & Company; 2002. Pascual-Marqui R. Standardized low resolution electromagnetic tomography (sLORETA): technical details. Meth Find Exp Clin Pharmacol 2002;24D:5–12. Penfield W, Jaspers H. Alimentary seizures. In: Epilepsy and the functional anatomy of the brain. Boston: Little Brown; 1954. p. 412–37. Raichle M, MacLeod A, Snyder A, Powers W, Gusnard D, Shulman G. A default mode of brain function. PNAS 2001;98:676–82. Sekimoto M, Kato M, Kaneko Y, Onuma T. Ictal nausea with vomiting as the major symptom of simple partial seizures: electroencephalographic and magnetoencephalographic analysis. Epilepsy Behav 2007;11:582–7. Specchio N, Trivisano M, Ciommo V, Cappelletti S, Masciarelli G, Volkov J, et al. Panayiotopoulos syndrome: a clinical, EEG, and neuropsychological study of 93 consecutive patients. Epilepsia 2010a;51:2098–107. Specchio N, Trivisano M, Claps D, Battaglia D, Fusco L, Vigevano F. Documentation of autonomic seizures and autonomic status epilepticus with ictal EEG in Panayiotopoulos syndrome. Epilepsy Behav 2010b;19:383–93. Tallon-Baudry C, Bertrand O, Delpuech C, Pernier J. Stimulus specificity of phaselocked and non-phase-locked 40 Hz visual responses in human. J Neurosci 1996;16:4240–9. Yalçin A, Toydemir H, Çelebi L, Forta H. Panayiotopoulos syndrome with coincidental brain lesions. Epileptic Disord 2009;11:270–6. Yoshinaga H, Kobayashi K, Ohtsuka Y. Characteristics of the synchronous occipital and frontopolar spike phenomenon in Panayiotopoulos syndrome. Brain Dev 2010;32:603–8. Vanhaudenhuyse A, Noirhomme Q, Tshibanda L, Bruno M, Boveoux P, Schnakers C, et al. Default network connectivity reflects the level of consciousness in noncommunicative brain-damaged patients. Brain 2010;133:161–71. Wilke C, Ding L, He B. Estimation of time-varying connectivity patterns through the use of an adaptive directed transfer function. IEEE Trans Biomed Eng 2008;55:2557–64. Wilke C, Drongelen W, Kohrman M, He B. Identification of epileptogenic foci from causal analysis of ECoG interictal spike activity. Clin Neurophysiol 2009;120:1449–56. Wilke C, Drongelen W, Kohrman M, He B. Neocortical seizure foci localization by means of a directed transfer function method. Epilepsia 2012;51:564–72. Uddin L, Supekar K, Aminoácido H, Rykhlevskaia E, Nguyen D, Greicius M, et al. Dissociable connectivity within human angular gyrus and intraparietal sulcus: evidence from functional and structural connectivity. Cerebral Cortex 2010;20:2636–46.
Contents lists available at SciVerse ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Origin and dynamics of epileptic activity in a symptomatic case of Panayiotopoulos syndrome: Correlation with clinical manifestations Alberto J.R. Leal a,b,⇑, Ricardo Lopes c, José C. Ferreira d a
Department of Neurophysiology, Hospital Júlio de Matos, Lisbon, Portugal CIS-ISCTE, Lisbon University Institute (IUL), Lisbon, Portugal c Department of Psychology, University of Coimbra, Coimbra, Portugal d Department of Pediatric Neurology, Centro Hospitalar Lisboa Ocidental, Lisbon, Portugal b
a r t i c l e
i n f o
Article history: Accepted 1 June 2012 Available online 23 July 2012 Keywords: Epilepsy Panayiotopoulos syndrome EEG Parietal lobe
h i g h l i g h t s Panayiotopoulos syndrome (PS) is a frequent childhood epileptic syndrome with typical clinical features but unknown localisation of the epileptogenic area. We describe the first symptomatic case of PS where both clinical and EEG features converge to demonstrate that an epileptic focus in the inferior parietal lobe can originate the epileptic syndrome. Fast spread of epileptic activity through physiological networks involved in eye-movement control, gastrointestinal autonomic control and consciousness can explain the diversity of clinical manifestations in PS.
a b s t r a c t Objective: The aim of the study was to demonstrate the dynamics and structure of the epileptic network and provide a tentative correlation with the clinical manifestations, in a symptomatic case of Panayiotopoulos syndrome (PS). Methods: JP, 5-year-old girl. Gestational period and developmental milestones were normal. At age 4 years, two episodes of recurrent vomiting, tonic eye deviation and consciousness impairment lasting for about 30 min occurred. Multifocal spikes were apparent over frontal areas in the EEG and MRI demonstrated an inferior parietal lobe (IPL) lesion. Results: A long-term 35-channel scalp EEG was obtained, which was processed with a Blind Source Separation algorithm. The most significant components with a dipolar field were submitted to source analysis and the recovered generators used to build the nodes of a brain network associated with each spike type. Analysis of information flow supported epileptic propagation from the left parietal lobe to both frontal and temporal lobes around spike peak. The good spatial overlap with physiological networks controlling eye movements, autonomic functions and consciousness, provides a tentative explanation to the diverse clinical manifestations of PS. Conclusions: Spreading patterns of epileptic activity form an extended network in PS. Significance: An epileptic focus in an IPL can reproduce both neurophysiological and clinical features of PS. Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction The Panayiotopoulos syndrome (PS) is one of the most recently described and prevalent benign epilepsy syndromes of childhood. Despite the significant effort devoted to elucidate its clinical (Panayiotopoulos, 2002), neurophysiological (Caraballo et al., ⇑ Corresponding author. Address: Department of Neurophysiology, Hospital Júlio de Matos, Ava do Brasil nr 53, 1749-002 Lisbon, Portugal. Tel.: +351 969851734; fax: +351 217819809. E-mail address: [email protected] (A.J.R. Leal).
2007) and neuropsychological aspects (Specchio et al., 2010a), several problems remain that prevent a deeper insight into the aetiology of the syndrome: the multifocal interracial spike activity present in most cases precluded the unambiguous identification of a common epileptogenic area; the peculiar combination of ictal autonomic manifestations, eye deviation and long-lasting impairment of consciousness, typical of the condition, has not allowed the determination of the precise locus of the epileptogenic zone on clinical grounds; It remains uncertain whether a single, yet unknown, cortical focus is responsible for the stereotyped seizure manifestations or these can be produced by foci in diverse areas.
1388-2457/$36.00 Ó 2012 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinph.2012.06.007
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
The few neurophysiological recordings of ictal events in the literature have also failed to provide focal EEG patterns with clear clues to their origin (reviewed in Specchio et al., 2010b). Despite the close morphological similarities of the interictal spikes in PS with the ones of rolandic epilepsy, the localisation information of each type is very different, with rolandic spikes showing stable dipolar fields with consistent generators within the primary somato-sensory area (Leal et al., 2007) and PS spikes demonstrating variable scalp topographies and significant spread of epileptic activity (Leal et al., 2008; Yoshinaga et al., 2010; Kokkinos et al., 2010). Clues to the localisation of the epileptogenic area in PS could potentially be obtained from the analysis of the few reported symptomatic cases. Nevertheless, the analysis of such cases failed to provide a consistent hypothesis and has led the authors to postulate a fortuitous coincidence between the epileptic syndrome and the structural abnormalities (Panayiotopoulos, 2002; Yalçin et al., 2009). In none of the reported cases was a source analysis of the EEG epileptic activity performed, nor any type of processing besides visual evaluation. We report on a patient with important clues to the source of epileptic activity in PS. 2. Methods and patient data 2.1. Patient data JP, 5-year-old girl. In the gestational period, the mother was continuously medicated with warfarin. Vaginal delivery occurred at the 37th week and the birth weight was 2160 kg. The neonatal period was complicated with sepsis and transient insulin-dependent diabetes (up to the 53rd day of life) but developmental milestones were reached at appropriate age. In the first 3 years, three renal urinary infections were reported, with associated febrile seizures. At the age of 4 years, a night event occurred lasting about 30 min with vomiting, conjugate and sustained eye deviation upward and to the left side, with unresponsiveness. After regaining consciousness, there was a transient period of confusion and somnolence but ultimately she recovered completely. She was investigated at the emergency department of a local hospital and discharged after normal blood and serum parameters and also a normal head CT were obtained. Five months later, a similar nocturnal event was reported. An EEG revealed multifocal spike activity. A 1.5-T MRI scan demonstrated a cortical lesion in the left inferior parietal lobe (IPL), with imaging features compatible with cortical dysplasia (Fig. 1b). A long-duration (24 h) video-EEG monitoring could not record seizure events, but abundant interictal spikes were noticed, occurring independently over the central area of the right hemisphere and the frontal area of the left one (Fig. 1a). The cognitive development was evaluated with the Portuguese version of the Wechsler Preschool and Primary Scale of Intelligence-Revised (WPPSI-R). The neurological examination was normal. No genetic analyses were performed.
21
for data acquisition, controlled by the SystemPlus software. The EEG data were exported to the Scan 4.3.1 software (Compumedics NeuroScan, Charlotte, NC, USA), where the trace was visually inspected by an experienced clinical neurophysiologist (AL), to identify spikes with independent topology. For each of the two recognised classes of spike topology, an amplitude threshold was applied at the channel with maximum spike amplitude to automatically detect spikes of this class and insert a trigger at each detection. The triggers were then used to epoch the EEG with a window of 200 to +500 ms, which were then concatenated. The final result was a continuous EEG file containing all the detected spikes for each topology class. The concatenated EEG record for each spike class was then imported to the EEGLAB 10.1 software (Delorme and Makeig, 2004) and decomposed using the Infomax (Bell and Sejnowski, 1995) independent component analysis (ICA) algorithm. The ICA spatial components were fitted with a dipole solution (using the standard boundary element method (BEM) (Montreal Neurological Institute model) included in EEGLAB) and only the ones with a residual variance (RV) lower than 10% were retained. These were then ordered by their relative contribution to the spike signal in the time range from 50 to +50 ms around the peak. The dipole solutions associated with the four ICA components with the largest contribution to each spike type were obtained with the CURRY 6 software (Neuroscan, Charlotte, NC, USA), using the sLORETA algorithm (PascualMarqui, 2002) in a standard BEM model of the head (Fig. 2a and c). Spectral analysis of the ICA components activity around the spike peak was done using the event-related spectral perturbation method (ERSP) (Makeig, 1993) as well as the inter-trial coherence (ITC) method (Tallon-Baudry et al., 1996), both implemented in the EEGLAB software. Statistical significance of the changes was obtained through a bootstrap procedure comparing changes in the ERSP with the spectral distribution in the prestimulus period. The functional connectivity analysis was performed in the averaged time course of the four ICA components with the largest contribution to the spike of each class, using the point of maximum score in the sLORETA inverse solution as the node of each component. We used the adaptive direct transfer function (ADTF) (Wilke et al., 2008) as a dynamical measure of information flow between the ICA component nodes, as implemented in the e-connectome software (He et al., 2011). The average amount of information flow in the band 4–30 Hz was calculated and statistical significance was obtained using a permutation test with phase randomisation. 3. Results 3.1. Electro-clinical data Our clinical case exhibits the main features of PS, with rare, predominantly autonomic seizures, long-lasting impairment of consciousness, sustained conjugate eye deviation and multifocal interictal spikes in a normal background EEG. The WPPSI-R test revealed a verbal IQ of 107 (score 68%), a performance IQ of 101 (score 53%) and a full-scale IQ of 103 (score 58%). These results support an average cognitive development.
2.2. EEG acquisition and processing 3.2. Dynamics of epileptic activity The EEG signal during the 24-h video-EEG monitoring was collected with 35 gold disc electrodes glued with Colodium to the scalp positions (Fp1/2, F3/4, C3/4, P3/4, O1/2, F7/8, T7/8, P7/8, Fz, Cz, Pz, F11/12, FT9/10, TP9/10, P11/12, FC5/6, FC1/2, CP5/6 and CP1/2). The sampling rate was 1000 Hz, the high- and low-pass filters were set at 0.5 and 70 Hz and the electrode impedances remained below 5 kX. A Micromed EEG recording system was used
The EEG revealed a normal background rhythm and multifocal spikes with consistent maxima over the frontal lobes of both hemispheres (Fig. 1a). No spikes could be identified in the left parietal– occipital electrodes, near the scalp projection of the cortical lesion. The mapping of the scalp potential at different times through the rising phase of the spikes revealed a changing topographical con-
22
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
Fig. 1. (a) Interictal EEG featuring a normal background activity and independent spikes with maxima over Fp1 and FC6. On the right, sequential maps of the scalp potential are shown, demonstrating changes in topology from spike onset to spike peak and (b) Structural MRI with abnormal signal over the left inferior parietal lobe, compatible with cortical dysplasia, (FLAIR sequence).
figuration (Fig. 1a, right), suggesting the involvement of several spatially distinct generators, with diverse temporal dynamics. The four ICA components with the largest (and statistically significant, Fig. 2) contribution to the spikes of each topology class were able to explain 99.1% and 99.0%, respectively of the left and right spikes, effectively preserving the main features of each epileptic event. The source analysis of these ICA components demonstrates that their generators are widely separated in space (Fig. 2a and c), supporting the contribution of several cortical areas to the genesis of each interictal spike class. The different timing of the peak of each spike class ICA components (Fig. 2a and b) further supports a sequential activation in time, likely associated with fast propagation of the epileptic activity. The sequential temporal
peaking of the ICA components demonstrates an early activation of the left parietal generators for both spike classes. Because the cortical lesion is located in the left parietal lobe, these data effectively support a dynamical model of interictal spike activity postulating an early activation of a generator in the anatomical lobe containing the lesion, followed by secondary propagation to the left temporal lobe and bilateral frontal lobe generators. 3.3. Functional connectivity analysis The functional connectivity analysis using the generators of the ICA components as nodes and their time courses to calculate the information flow in an epileptic network for each spike class re-
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
23
Fig. 2. Decomposition of the two types of epileptic spikes and selection of the four components (rows) with the largest contribution in the time range 50 to 50 ms (centred on spike peak). (a) Fixed spatial components are shown in the left column, the average time course on the middle one (horizontal axis 200 to +500 ms, vertical axis in lV) and the source generator of each component is represented in the right column, with a dipole on the maximum of the sLORETA scores, (b) the components spectral changes around the spike peak are shown, both for power (ERSP) and phase (ITC). Green colour represents non-significant changes in the bootstrap statistics used, (c) similar to (a), (d) similar to (b). The data demonstrates the contribution of several cortical generators for each spike type, widely separated in space and with diverse temporal dynamics. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
vealed: (a) from spike onset to spike peak, sources of outflow information were located in the left parietal lobe only (Fig. 3b–d and f) and (b) inflow sources were scattered between the frontal lobes and the left temporal one. These observations effectively
support a model with propagation of epileptic activity from the left parietal lobe to distinct frontal and temporal areas as the likely explanation for the diverse spike topography present in the raw scalp EEG.
24
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
Fig. 3. Analysis of information flow between the four generators of right frontal spikes (above) and the left frontal ones (below). The nodes of the networks obtained from previous source analysis (Fig. 2) are shown in (a) and (d), together with the average time course of each component (below). The average flow of information in the 4–30 Hz spectral band is shown just before (b and e) and at spike peak (c and f), together with the connection matrix (below). Only information flow from the left parietal lobe (component 4) to the remaining areas reached statistical significance.
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
3.4. Spatial relationship of the epileptic focus with symptomatogenic cortical areas The existence of a focus of interictal spikes in the pre-motor areas of the right hemisphere, in the neighbourhood of the frontal eye field (FEF) (Fig. 2a, third row) suggests that the tonic and conjugate eye deviation to the left side, reported on both clinical seizures, might be due to epileptic stimulation of this structure. The propagation of epileptic activity through the functional network responsible for the control of saccadic eye movements, involving the IPL, superior colliculus (supCol) and the FEF, could explain the selectivity of this motor manifestation. The dysplastic cortex seen in the MRI (Fig. 1b) is in the neighbourhood of a node of such a network and the associated epileptogenic focus is in a good position to spread the epileptic activity to the saccadic eye-movement system, which is supported by the significant flow of information from the left parietal lobe to the right FEF area (Fig. 3c). Another node of the epileptic network with interest to explain the vomiting typical of PS is located in the left temporal lobe, and the associated generator (Fig. 2a, fourth row) is compatible with insular cortex activation. The demonstration of significant flow of information from the IPL to such a node (Fig. 3b) provides support to the spread of epileptic activity from the IPL to the insula. The co-existence of a node of the default mode network (DMN) in the IPL (Raichle et al., 2001) and the recent studies establishing a relationship between the degree of DMN connectivity and consciousness (Vanhaudenhuyse et al., 2010) suggest that disturbances in such connectivity by an epileptic focus in the IPL could lead to the impairment of consciousness so typical of PS. 4. Discussion Our clinical case features the major manifestations of PS: rare episodes of recurrent vomiting, tonic eye deviation and long-lasting (30 min) impairment of consciousness in a cognitively welldeveloped child. The rare frequency of seizures and the typical multifocal EEG spike activity further support this syndrome classification. Despite the fact that most cases of PS have no structural lesion in the brain MRIs (Panayiotopoulos, 2002; Caraballo et al., 2007; Specchio et al., 2010a), symptomatic cases have been described (Yalçin et al., 2009; Panayiotopoulos, 2002). Case #60 of the series reported by Panayiotopoulos (2002) had a dysembryoblastic neuroepithelial tumour in the right temporo-parieto-occipital region and seizures with prominent autonomic features (vomiting and sweating) and impairment of consciousness. One of the seizures lasted for hours. Two of the three patients reported by Yalçin et al. (2009) had extensive structural lesions involving not only the occipital lobes but also surrounding parietal and temporal regions, and therefore preventing a precise anatomical localisation. In general, these cases did not provide data to suggest to the authors of the studies a candidate epileptogenic area that could explain both the clinical and neurophysiological features of the syndrome, because structural lesions were too diffuse and because they failed to demonstrate a good spatial co-localisation of cortical lesions and epileptic spikes. All the previous cases demonstrated, as far as it is possible to evaluate from the reported data, extensive structural lesions involving the occipital lobe but extending to neighbour lobes and therefore possibly overlapping with the IPL lesion present in the case JP. 4.1. Dynamics of epileptic spikes Several studies of interracial spikes in PS (Leal et al., 2007; Yoshinaga et al., 2010; Kokkinos et al., 2010) have documented the existence of significant topographical changes in the potential
25
maps from spike onset to spike peak, demonstrating the co-existence of several intracranial sources, widely separated in space. This observation supports the hypothesis that significant spread of epileptic activity takes place in interictal and, presumably, in ictal discharges. Because some of these sources of epileptic activity are of very low amplitude (Leal et al., 2008), they can remain unrecognised in clinical visual analysis, leading to wrong cortical localisations or to the illusion of multifocal spike activity if not analysed with sensitive enough tools. In the patient JP, the demonstration that the two apparently independent spike types are preceded in time by a small left parietal source (Fig. 2a and c, upper row), supports a model postulating a single epileptogenic focus with two different secondary spreading patterns versus an alternative model postulating independent foci in the two hemispheres. The fact that the primary source of epileptic activity is in the lobe containing the structural lesion strongly supports a non-fortuitous connection between the two. The ICA technique produces an unmixing of the different sources of activity that add in the scalp to produce the EEG. The method has previously demonstrated its capability to identify small components in the scalp EEG (Leal et al., 2008) and also to demonstrate epileptic propagation between brain lobes (Leal et al., 2007), both important features for a full characterisation of epileptic networks, despite the fact that current EEG source-analysis methods provide little information on the spatial extent of intracranial sources. In the case JP, several independent components could be recovered for the two types of spikes, with consistent phase relation (Fig. 2b and d) but different topographies and source localisation (Fig. 2a and c). In the patient JP, the analysis of the flow of information between these sources, (Fig. 3), produced convergent evidence for spread of epileptic activity from the left parietal lobe containing the structural lesion towards both the frontal lobes containing the interictal spikes visible in the surface EEG and also to the left temporal lobe. These results agree and complement the ones of the time to peak analysis of the ICA components composing the spikes (Fig. 2a and c, middle column), which show an earlier activation of parietal sources as compared with both frontal and temporal ones. The use of the ADTF in human epilepsy, for both interictal (Wilke et al., 2008; 2009) and ictal recordings (Mierlo et al., 2011; Wilke et al., 2012), has allowed the recovery of propagation patterns validated in successful surgery for epilepsy. 4.2. Electro-clinical correlations The demonstration that electrical stimulation of the IPL cortex produced conjugate eye movements was done long ago (Ferrier, 1875; Ferrie et al., 2007), but this effect is thought to be mediated through the strong functional connections of this area with the FEFs (Blanke and Seeck, 2003) and also the supCol. In our clinical case, the demonstration of spread of epileptic activity to the right frontal lobe (Fig. 3c), in the neighbourhood of the FEF, provides a tentative explanation for the reported conjugate eye deviation to the left during seizure events, an expected effect of the stimulation of the right FEF (Blanke and Seeck, 2003). Sekimoto et al. (2007) reported two cases of parietal lobe epilepsy with ictal vomiting and no impairment of consciousness, which presented interictal spikes detected in MEG but not in the EEG, suggesting foci in the depth of sulci. In line with the interpretation of our case, these authors suggest that the clinical manifestations result from epileptic propagation to the insula, a cortical area prominently associated with autonomic and visceral motor manifestations (Insnard et al., 2004). The suspicion of insular cortex involvement in the genesis of ictal vomiting became widely accepted ever since Penfield and Jaspers (1954) first reported visceral motor manifestations associated with insular electrical stimulation. In clinical epileptology, vomiting has been most often associ-
26
A.J.R. Leal et al. / Clinical Neurophysiology 124 (2013) 20–26
ated with temporal lobe epilepsy (Kramer et al., 1988), but a recent study using stereo-EEG (Catenoix et al., 2008) suggests that this is due to secondary propagation to the insula. The proposed spread of epileptic activity from the IPL to the insula also finds support in the strong functional and structural connectivity between the IPL and the insula demonstrated by Uddin et al. (2010), using fMRI resting-state analysis and diffusion tensor imaging (DTI) tractography. In patient JP, the information flow from the left parietal lobe towards the left temporal source before spike peak is compatible with epileptic spread to the insula, which is well known to produce surface source topographies undistinguishable from the ones produced by more superficial sources on the overlying temporal cortex. The impairment of consciousness in PS is a prominent and persistent ictal feature the origin of which is not known. Recent studies have implicated the default network as a functional system related to consciousness (Vanhaudenhuyse et al., 2010), and with disturbances of this function in several epilepsies (Danielson et al., 2011). The IPLs contain a node of the DMN and epileptic seizures involving this area are well positioned to produce a disturbance of its functional connectivity leading to consciousness impairment. Overall, our clinical case highlights the IPL as a brain region including important nodes of the networks controlling eye movement (IPL, FEF and supCol), visceral motor responses (IPL-Insula) and consciousness (default network), forming an extended epileptic network whose activation in seizure events can mimic some of the cardinal clinical manifestations of PS. We therefore suggest that the IPL is a cortical region with all the capabilites to act as the primary source of epileptic activity in our case of PS, and of reproducing the typical behavioural manifestations through spread of the epileptic activity along specific physiological networks. References Bell A, Sejnowski T. An information-maximization approach to blind source separation and blind deconvolution. Neural Comp 1995;7:1004–34. Blanke O, Seeck M. Direction of saccadic and smooth eye movements induced by electrical stimulation of the human frontal eye field: effect of orbital position. Exp Brain Res 2003;150:174–83. Caraballo R, Cersosimo R, Fejerman N. Panayiotopoulos syndrome: a prospective study of 192 patients. Epilepsia 2007;48:1054–61. Catenoix H, Isnard J, Guenot M, Petit J, Remy C, Mauguière F. The role of the anterior insular cortex in ictal vomiting: a stereotactic electroencephalography study. Epilep Behav 2008;13:560–3. Danielson N, Guo J, Blumenfeld H. The default mode network and altered consciousness in epilepsy. Behav Neurol 2011;24:55–65. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics. J Neurosc Meth 2004;134:9–21. Ferrie C, Caraballo R, Covanis A, Demirbilek D, Dervent A, Fejerman N, et al. Autonomic status epilepticus in Panayiotopoulos syndrome and other childhood and adult epilepsies: a consensus view. Epilepsia 2007;48:1165–72.
Ferrier D. Experiments on the brain of monkeys-No. 1. Proceedings of the Royal Society 1875;23:409-430. He B, Dai Y, Astolfi L, Babiloni F, Yuan H, Yang L. EConnectome: a MATLAB toolbox for mapping and imaging of brain. J Neurosci Meth 2011;195:261–9. Insnard J, Guenot M, Sindou M, Mauguière F. Clinical manifestations of insular lobe seizures: a stereo-electroencephalographic study. Epilepsia 2004;45:1079–90. Kokkinos V, Koutroumanidis M, Tsatsou K, Koupparis A, Tsiptsios D, Panayiotopoulos C. Multifocal spatiotemporal distribution of interictal spikes in Panayiotopoulos syndrome. Clin Neurophysiol 2010;121:859–69. Kramer R, Luders H, Goldstick L, Dinner D, Morris H, Lesser R, et al. Ictus emeticus: an electroclinical analysis. Neurology 1988;38:1048–52. Leal A, Nunes S, Dias A, Vieira J, Moreira A, Calado E. Analysis of the generators of epileptic activity in early-onset childhood benign occipital lobe epilepsy. Clin Neurophysiol 2007;118:1341–7. Leal A, Ferreira J, Dias A, Calado E. Origin of frontal lobe spikes in the early onset benign occipital lobe epilepsy (Panayiotopoulos syndrome). Clin Neurophysiol 2008;119:1985–91. Makeig S. Auditory event-related dynamics of the EEG spectrum and effects of exposure to tones. Electroencephalogr Clin Neurophysiol 1993;86:283–93. Mierlo P, Carrette E, Hallez H, Vonck K, Roost D, Boon P, et al. Accurate epileptogenic focus localization through time-variant functional connectivity analysis of intracranial electroencephalographic signals. Neuroimage 2011;56:1122–33. Panayiotopoulos CP. Panayiotopoulos syndrome: a common and benign childhood epileptic syndrome. London: John Libbey & Company; 2002. Pascual-Marqui R. Standardized low resolution electromagnetic tomography (sLORETA): technical details. Meth Find Exp Clin Pharmacol 2002;24D:5–12. Penfield W, Jaspers H. Alimentary seizures. In: Epilepsy and the functional anatomy of the brain. Boston: Little Brown; 1954. p. 412–37. Raichle M, MacLeod A, Snyder A, Powers W, Gusnard D, Shulman G. A default mode of brain function. PNAS 2001;98:676–82. Sekimoto M, Kato M, Kaneko Y, Onuma T. Ictal nausea with vomiting as the major symptom of simple partial seizures: electroencephalographic and magnetoencephalographic analysis. Epilepsy Behav 2007;11:582–7. Specchio N, Trivisano M, Ciommo V, Cappelletti S, Masciarelli G, Volkov J, et al. Panayiotopoulos syndrome: a clinical, EEG, and neuropsychological study of 93 consecutive patients. Epilepsia 2010a;51:2098–107. Specchio N, Trivisano M, Claps D, Battaglia D, Fusco L, Vigevano F. Documentation of autonomic seizures and autonomic status epilepticus with ictal EEG in Panayiotopoulos syndrome. Epilepsy Behav 2010b;19:383–93. Tallon-Baudry C, Bertrand O, Delpuech C, Pernier J. Stimulus specificity of phaselocked and non-phase-locked 40 Hz visual responses in human. J Neurosci 1996;16:4240–9. Yalçin A, Toydemir H, Çelebi L, Forta H. Panayiotopoulos syndrome with coincidental brain lesions. Epileptic Disord 2009;11:270–6. Yoshinaga H, Kobayashi K, Ohtsuka Y. Characteristics of the synchronous occipital and frontopolar spike phenomenon in Panayiotopoulos syndrome. Brain Dev 2010;32:603–8. Vanhaudenhuyse A, Noirhomme Q, Tshibanda L, Bruno M, Boveoux P, Schnakers C, et al. Default network connectivity reflects the level of consciousness in noncommunicative brain-damaged patients. Brain 2010;133:161–71. Wilke C, Ding L, He B. Estimation of time-varying connectivity patterns through the use of an adaptive directed transfer function. IEEE Trans Biomed Eng 2008;55:2557–64. Wilke C, Drongelen W, Kohrman M, He B. Identification of epileptogenic foci from causal analysis of ECoG interictal spike activity. Clin Neurophysiol 2009;120:1449–56. Wilke C, Drongelen W, Kohrman M, He B. Neocortical seizure foci localization by means of a directed transfer function method. Epilepsia 2012;51:564–72. Uddin L, Supekar K, Aminoácido H, Rykhlevskaia E, Nguyen D, Greicius M, et al. Dissociable connectivity within human angular gyrus and intraparietal sulcus: evidence from functional and structural connectivity. Cerebral Cortex 2010;20:2636–46.