- Email: [email protected]
High-frequency EEG activity in epileptic encephalopathy with suppression-burst
Brain & Development xxx (2014) xxx–xxx www.elsevier.com/locate/braindev
Original article
High-frequency EEG activity in epileptic encephalopathy with suppression-burst Yoshihiro Toda a,b, Katsuhiro Kobayashi a,⇑, Yumiko Hayashi a, Takushi Inoue a, Makio Oka a, Fumika Endo a, Harumi Yoshinaga a, Yoko Ohtsuka a,c a
Department of Child Neurology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences and Okayama University Hospital, Okayama, Japan b Department of Pediatrics, School of Medicine, University of Tokushima, Tokushima, Japan c Asahigawaso Rehabilitation and Medical Center, Okayama, Japan Received 30 November 2013; received in revised form 7 April 2014; accepted 8 April 2014
Abstract Objective: We explored high-frequency activity in the suppression-burst (SB) pattern of interictal electroencephalogram (EEG) in early infantile epileptic encephalopathy including Ohtahara syndrome (OS) and early myoclonic encephalopathy (EME) to investigate the pathophysiological characteristics of SB. Methods: Subjects included six patients with the SB EEG pattern related to OS or EME (Group SB). The results were evaluated in comparison to trace´ alternant (TA) observed during the neonatal period in nine patients to rule out possible nonspecific relationships between high-frequency activity and periodic EEG patterns (Group TA). EEG was digitally recorded with a sampling rate of 500 Hz and the analysis was performed in each of the particular bipolar channel-pairs. We visually selected 20 typical consecutive burst sections and 160 inter-burst sections for comparison from the sleep record of each patient and performed the time–frequency analysis. We investigated the maximum frequencies of power enhancement in each derivation in both groups. Results: In Group SB, a significant increase in power at a frequency of 80–150 Hz was observed in association with the bursts, particularly in the bilateral parieto-occipital derivations, in all patients. In Group TA, on the contrary, no significant increase in high-frequency power was found. The maximum frequencies of power enhancement were significantly higher in Group SB than in Group TA (p < 0.001 by repeated-measures ANOVA). Conclusion: Interictal high frequencies of up to 150 Hz were detected in the suppression-burst EEG patterns in epileptic encephalopathy in early infancy. Further studies will be necessary to identify the role of the interictal high-frequency activity in the pathophysiology of such early epileptic encephalopathy. Ó 2014 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.
Keywords: High-frequency oscillations; Ohtahara syndrome; Early myoclonic encephalopathy; Trace´ alternant; Time–frequency analysis
1. Introduction ⇑ Corresponding author. Address: Department of Child Neurology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences and Okayama University Hospital, 5-1 Shikatacho 2-chome, Kita-ku, Okayama 700-8558, Japan. Tel.: +81 86 235 7372; fax: +81 86 235 7377. E-mail address: [email protected] (K. Kobayashi).
Analysis of high-frequency oscillations (HFOs) beyond the gamma band has been enabled by the recent technical development of digital electroencephalography (EEG). Remarkably, HFOs are suggested to have a close relation with epileptogenicity and ictogenicity.
http://dx.doi.org/10.1016/j.braindev.2014.04.004 0387-7604/Ó 2014 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
2
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
Gamma activity in scalp EEGs has been reported in ictal EEGs associated with epileptic spasms (ESs) in West syndrome and with the interictal suppression-burst (SB) pattern in epileptic encephalopathy in early infancy [1–3]. Likewise, high-frequency activity >80 Hz has now been observed in scalp and cortical ictal EEGs of ESs [4–7], but it has not yet been clearly identified in association with the SB pattern. The SB pattern in EEG characterizes epileptic encephalopathy during the neonatal period and early infancy including Ohtahara syndrome (OS) and early myoclonic encephalopathy (EME). The SB pattern and ES in series both consist of recurring bursts of activity, and it has already been indicated that there is a mutual transition between the ictal activity of ES and the bursts in peri-ictal SB, and that gamma activity is augmented in association with the change in EEG pattern from SB to the ictal activity of ES [2]. Hence the SB pattern is suggested to have a close relation with the generative mechanisms of ES. We have therefore hypothesized that the bursts in SB, which are already known to be accompanied by activity occurring as high as the gamma band, may be associated with the generation of high-frequency activity >80 Hz that is not readily visible through ordinary analysis. To prove this hypothesis, we attempted to use a statistical time-frequency analysis method that we had developed for the analysis of high-frequency changes associated with intracranially recorded epileptic discharges [8,9]. In addition, we similarly analyzed the EEG pattern known as trace´ alternant (TA), which occurs during quiet sleep in neonates, for comparison in order to exclude the possibility that the generation of high frequencies may be a non-specific finding associated with any type of cyclic bursting pattern during the neonatal period and early infancy. 2. Subjects and methods 2.1. Patients Subjects with the SB pattern in EEG (Group SB) were a total of six patients at 44–55 weeks postconceptional age (the sum of gestational age and chronological age), comprising three with OS and three with EME. The onset of both OS and EME is very early, typically under three months of age, and occurs most commonly in the neonatal period. OS is characterized by (1) epileptic spasms (tonic spasms) with or without clustering, though additional partial seizures and rare myoclonus may occur in some patients, (2) consistent appearance of SB with regular periodicity in both waking and sleep EEGs, and (3) heterogeneous etiologies including brain malformations and gene mutations; EME is characterized by (1) fragmentary myoclonia as the main seizure type, (2) partial seizures, (3) later
appearance of massive myoclonia or epileptic spasms, (4) SB which may present only during sleep or be most prominent during sleep, and (5) unknown etiology though genetic or metabolic origins have been suggested [10–12]. Subjects whose TA in EEG was observed for comparison (Group TA) were nine newborns at 36–43 weeks of post-conceptional age who were being examined with EEG for various reasons including low birth weight, fetal hydrops, cerebral ventricular dilatation, neonatal convulsions, intraventricular hemorrhage, and agenesis of the corpus callosum. The patients with the SB pattern are indexed by numbers with the prefix S and those with TA by numbers with the prefix T (Table 1). We are aware that comparison between SB and TA is not ideal because the ages of the patients are different, but we undertook this comparison to achieve a methodological confirmation which would otherwise have been impossible. 2.2. EEG recording and analysis EEG was recorded with a sampling rate of 500 Hz using the Nihon-Kohden (Tokyo, Japan) Neurofax digital EEG system. The international 10–20 electrode placement system was used, and the analysis was performed in each of the following bipolar channel-pairs: F3-C3, F4-C4, P3-O1, and P4-O2. Both the SB pattern and the TA pattern were recorded during quiet/non-REM sleep. We then visually selected 20 typical artifact-free consecutive burst sections and 160 inter-burst sections from the sleep record of each patient for the time–frequency analysis described below. Each burst section was a 3 s segment of EEG data, as it included not only the 1 s onset part of the burst but also 2 s of the inter-burst period immediately before the burst. Every burst section was manually selected using the moment of the burst’s onset as a trigger. Each inter-burst section included a non-overlapping low-amplitude EEG epoch lasting 512 ms. A representative EEG trace showing the SB pattern recorded from Patient S2 with OS is depicted in Fig. 1A: bursts of bilaterally diffuse irregular high amplitude slow waves containing spikes with a duration of 1–2 s periodically repeated with intervals of low amplitude suppression phase lasting for 3–4 s each. A representative EEG trace of TA recorded from Patient T4 with ventricular dilatation observed in the prenatal period is shown in Fig. 1B: bursts of diffuse slow waves containing some sharp transients occurred periodically and lasted for a few seconds each with relatively low amplitude intervals. In TA, compared with the SB pattern, bursting slow waves were not very high in amplitude and lacked truly epileptic discharges, and activity during the relatively long inter-burst intervals was not flat.
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
3
Table 1 Subjects. Patient
Postconceptional age at examination (w)
Gestational age at birth
Birth weight (g)
Disorder
Patients with suppression-burst (SB) pattern in EEG S1 55 38 w 2 d S2 52 40 w 3 d S3 46 38 w 4 d S4 50 39 w S5 44 40 w S6 52 38 w Average 50 38 w 3 d
2746 3306 2840 2790 2814 2706 2867
EME OS OS EME EME OS
Patients with trace´ alternant (TA) in EEG T1 39 T2 40 T3 41 T4 39
28 28 36 39
w w w w
d d d d
962 805 3136 2644
T5 T6 T7 T8
43 41 36 39
39 31 30 36
w w4d w4d w6d
3084 758 1905 1973
T9 Average
42 40
41 w 3 d 34 w 5 d
2936 2023
LBW LBW Hydrops fetalis Cerebral ventricular dilatation observed in prenatal period Neonatal convulsion LBW LBW, IVH I Cerebral ventricular dilatation observed in prenatal period Agenesis of corpus callosum
4 2 3 4
Postconceptional age: the sum of gestational age and chronological age. EME: early myoclonic encephalopathy; OS: Ohtahara syndrome. LBW: low birth weight; IVH I: intraventricular hemorrhage Grade I.
Our procedure for statistical time-frequency analysis is illustrated in Supplementary Fig. 1, which shows representative SB data of the left parieto-occipital derivation recorded from Patient 2 with OS (same patient as in Figs. 1 and 2A). The burst sections were selected in each patient, and the raw EEG data in these sections were subjected to time–frequency analysis to yield an average power spectrum. We used the Gabor Transform (windowed Fourier transform with a full width at half maximum 50 ms Gaussian window) for time-frequency analysis, and the frequency range was 20–150 Hz. Power increases were sometimes observed in association with the bursts, but these could not be evaluated with certainty in the original spectra. The power levels of the burst sections were statistically compared with those of control inter-burst data sections by means of an unpaired t-test to obtain the t value and the corresponding p value at each pixel of the time–frequency spectrum. The control spectral data were obtained through application of the Fourier transform to each of the 160 inter-burst sections. The resulting spectrum of t-values, however, had an enormous number of pixels, and was noisy and difficult to interpret with an ordinary significance level a (say, 0.05 or a 5% error rate) because too many pixels were declared active (false positivity or type I error). Therefore we adopted a statistical procedure for controlling false discovery rate (FDR) to avoid type I errors; the details of this statistical procedure have already been described by Kobayashi et al. [8] and are
briefly summarized here. The FDR is defined as the ratio of the number of false positive pixels to the number of pixels declared active, and the FDR bound q, which must be defined by the user, is the maximum tolerable FDR on average. In the present study, the two-tailed test was used with q = 0.025. Pixels with significantly above-average power are indicated in red, and pixels with t values that did not reach the limits are in green, in the FDR-controlled t-spectrum. Computation was performed using a program written in-house for Matlab (version 6.5.1; MathWorks Inc., Natick, MA, USA). 2.3. Statistical comparison of high-frequency power changes between SB and TA We performed repeated-measures analysis of variance (ANOVA) to compare the maximum frequencies of power enhancement between groups and between derivations with a Greenhouse-Geisser correction using SPSS Statistics Japanese version 17.0 (IBM Japan, Tokyo). Relationships were considered statistically significant if p < 0.05. 3. Results During SB, a significant increase in power >80 Hz and up to 150 Hz was observed in association with the bursts in the bilateral parieto-occipital derivations in all patients, as indicated by a representative spectrum
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
4
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
in the frequency band > 39 Hz in the parieto-occipital derivations or in the band > 64 Hz in the fronto-central derivations (Fig. 3, crosses). Repeated-measures ANOVA showed that the maximum frequencies of power enhancement were significantly higher in Group SB than in Group TA (p < 0.001), as shown in Table 2. When there was no increase in power > 20 Hz, as in the TA case depicted in Fig. 2B, the maximum frequency of power enhancement was assumed to be 20 Hz in this statistical analysis (Patients T1, T5, and T9 in the fronto-central derivations, and Patients T1, T3–T6, and T9 in the parieto-occipital derivations) (Fig. 3). Because the effects of derivations were not significant (p = 0.418), no post-hoc multiple comparisons between derivations were performed. 4. Discussion
Fig. 1. EEGs showing a suppression-burst (SB) pattern and trace´ alternant (TA). (A) A representative SB pattern recorded from Patient S2 with Ohtahara syndrome (OS). Bursts of bilaterally diffuse irregular high-amplitude slow waves containing spikes with a duration of 1–2 s occur periodically, separated by intervals of low amplitude suppression phase lasting for 3–4 s. (B) A representative TA pattern recorded from Patient T4 with ventricular dilatation observed in the prenatal period. Bursts of diffuse slow waves containing some sharp transients with a duration of a few seconds occur periodically, separated by intervals of relatively low amplitude. In TA, in contrast to the SB pattern, bursting slow waves did not always have a high amplitude and lacked truly epileptic discharges, and activity during the relatively long inter-burst intervals was not flat.
of the bilateral parieto-occipital derivations of Patient S2 with OS (Fig. 2A). In five infants, this increase in power was observed in the bilateral fronto-central derivations as well, as indicated in Fig. 3 which shows the maximum frequency of significant power change in each derivation-pair (circles denote individual patients’ data; boxes and vertical bars indicate mean data and standard deviation, respectively). In the analysis of TA, to the contrary, as indicated by a representative spectrum of the bilateral parieto-occipital derivations of Patient T3 (Fig. 2B), no significant increase of power was observed
In the present study, interictal high frequencies up to 150 Hz were detected in the SB EEG patterns of epileptic encephalopathy in early infancy. No such high-frequency activity was detected in association with TA, a non-epileptic periodic EEG pattern observed in neonates. Thus it was suggested that pathological high-frequency activity is linked to the pathophysiology of epileptic encephalopathy in early infancy, and that high frequencies are unlikely to be generated non-specifically in association with all periodic EEG patterns. There appear to be no remarkable differences between OS and EME with respect to burst-associated high frequencies, although the number of subjects with each of these conditions was so small as to preclude any detailed comparison. The subjects in Group TA had some abnormalities and were therefore not ideal as controls. It cannot be denied that the conditions that caused them to undergo EEG examination could have some effect on their EEG activity, perhaps by decreasing the cerebral capability to generate high-frequency activity. EEG data of entirely normal neonates, however, were unavailable to us during the study period, so this question will have to be addressed in the future. The SB EEG pattern of epileptic encephalopathy shares common periodic characteristics with the burst suppression (BS) pattern, which is non-specifically observed in various neurological conditions such as anesthesia-induced deep coma. In BS, bursts of diffuse high-voltage slow waves are interrupted by isoelectric periods that reflect absence of cortical activity, but the BS pattern includes no epileptic discharges. Although the BS pattern appears to correspond to decreases in neuronal responsiveness associated with loss of consciousness due to anesthesia, it has been demonstrated in at least some coma patients that BS in fact results from a hyperexcitability in which bursting activity is triggered even by low-intensity stimuli in association
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
5
Fig. 2. Representative statistical time–frequency analysis. Spectra of t-values controlled for false discovery rate (FDR) in the EEG data of bilateral parieto-occipital derivations. (A) Analysis in SB recorded from Patient S2 with OS. Augmentation of power is observed in association with bursts in a frequency range up to 150 Hz. (B) analysis in TA recorded from Patient T4. Augmentation of power is not observed.
with suppressed inhibition [13–15]. There are no reports of HFOs in the BS EEG pattern to our knowledge, but these studies on non-epileptic BS may provide some hints regarding the pathophysiology of SB. Epileptic high-frequency activity has been reported in ictal scalp and intracranial EEGs of ESs [1,3,4,6,7]. It was suggested in the current study that high frequencies are generated not only transiently in association with seizures but also rather continuously during interictal
periods, at least in OS and EME, two representative types of epileptic encephalopathy in early infancy. Interictal HFOs occurred with a tight temporal relationship to interictal spikes in patients with benign childhood epilepsy with centrotemporal spikes and Panayiotopoulos syndrome, and the periods of active seizure occurrence were found to be closely related [16]. As the EEG pattern of continuous spikes and waves during slow sleep (CSWS), in which each spike is associated with intense
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
6
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
Fig. 3. Maximum frequency of significant power enhancement in each derivation. Each mark denotes an individual patient’s data (circle for SB and cross for TA); boxes and vertical bars indicate mean data and standard deviations, respectively. In almost all SB cases, power enhancement was observed at frequency P100 Hz, especially in the parieto-occipital derivation. In TA cases, in contrast, power enhancement was limited to <70 Hz and was even undetectable at frequency >20 Hz in three patients with fronto-central derivations and six patients with parieto-occipital derivations.
Table 2 Statistical analysis enhancement.
of
the
maximum
frequencies
of
power
Derivation/max frequency (Hz)
F3-C3
F4-C4
P3-O1
P4-O2
Group SB Average SD
123.5 33.7
135.0 17.6
143.0 17.1
146.3 9.0
Group TA Average SD
29.6 10.8
30.8 14.4
23.3 6.5
23.1 6.3
Repeated-measures ANOVA showed a statistically significant effect of groups (F = 598.3; p < 0.001) but no significant effects of derivations (F = 0.863; p = 0.418) nor of the interactions between groups and derivations (F = 3.063; p = 0.075).
HFOs [17], is regarded as a sort of electrical status epilepticus, the EEG pattern of SB with active high frequencies might well be akin to electrical status epilepticus. The origin of epileptic HFOs is not yet determined, though animal experiments have suggested that pathological HFOs are generated by action potentials of synchronously bursting principal cells and reflect the activity of pathologically interconnected clusters of neurons [18]. Accumulating evidence indicates that spikes and HFOs do not share identical clinical meanings and that HFOs may be a better biomarker of epileptogenicity/ictogenicity than spikes are: HFOs are more
tightly linked to seizures than spikes are [19], and success in epilepsy surgery is more closely related to the resection of the cortical regions that generate HFOs than to the elimination of spikes [20]. In our study on the effects of intravenous (IV) injection of diazepam (DZP), an agonist for the gamma-aminobutyric acid (GABA) A receptor, on the EEG pattern of CSWS, the effects of IV DZP on the intensity of spike-associated HFOs lasted longer than its effects on the amplitudes of the spikes themselves during the recovery process following a temporary suppression of CSWS in the three patients [21]. Thus the GABAergic interneuron system may be more closely related to HFOs than to spikes. Pathological gamma and high-frequency activity are of particular interest in pediatric epileptic encephalopathies including OS and EME. In epileptic encephalopathy, cognitive and behavioral impairment can be caused by epileptic activity itself; especially given that physiological gamma and high frequencies are reported to be involved in higher brain functions such as language, memory, and cognition [22,23], any possible interaction or mutual interference between pathological and physiological high-frequency activity should be pursued in future studies. We speculate that the persistent active generation of pathological high frequencies during interictal periods might be harmful to cognitive functions in the developing brain. High-frequency activity in scalp EEGs may not have the same meaning as intracranially recorded high frequencies because high frequencies are generally attenuated over the scalp, probably due to the summation of polyphasic cortical activity with variable phase [24]. Scalp high frequencies may represent the most intense intracranial high frequencies, but this issue requires further investigation. As the current study used EEG data that had already been recorded with a sampling rate of 500 Hz, it was difficult to analyze HFOs > 150 Hz. Experimenting with a much higher sampling rate might provide us new information. We used the correction method for controlling FDR to avoid false positivity, but the results may be somewhat different depending on the correction methods used and their thresholds. The significance of the present findings of highfrequencies in SB with respect to the pathophysiology, particularly in early infancy, is still at the initial step of research. Further studies will be necessary to identify the role of the interictal high-frequency activity in the pathophysiology of such early epileptic encephalopathy. Funding sources Dr. Kobayashi is a recipient of the Health and Labour Sciences Research Grant: Research on catastrophic epilepsy in infancy and early childhood– epidemiology, diagnosis and treatment guide, and is also supported in part by a Research Grant (21B-5) for
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
Nervous and Mental Disorders from the Ministry of Health, Labour and Welfare. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/ 10.1016/j.braindev.2014.04.004. References [1] Kobayashi K, Oka M, Akiyama T, Inoue T, Abiru K, Ogino T, et al. Very fast rhythmic activity on scalp EEG associated with epileptic spasms. Epilepsia 2004;45:488–96. [2] Kobayashi K, Inoue T, Kikumoto K, Endoh F, Miya K, Oka M, et al. Relation of spasms and myoclonus to suppression-burst on EEG in epileptic encephalopathy in early infancy. Neuropediatrics 2007;38:244–50. [3] Inoue T, Kobayashi K, Oka M, Yoshinaga H, Ohtsuka Y. Spectral characteristics of EEG gamma rhythms associated with epileptic spasms. Brain Dev 2008;30:321–8. [4] Akiyama T, Otsubo H, Ochi A, Ishiguro T, Kadokura G, Ramachandrannair R, et al. Focal cortical high-frequency oscillations trigger epileptic spasms: confirmation by digital video subdural EEG. Clin Neurophysiol 2005;116:2819–25. [5] Ramachandrannair R, Ochi A, Imai K, Benifla M, Akiyama T, Holowka S, et al. Epileptic spasms in older pediatric patients: MEG and ictal high-frequency oscillations suggest focal-onset seizures in a subset of epileptic spasms. Epilepsy Res 2008;78:216–24. [6] Nariai H, Nagasawa T, Juha´sz C, Sood S, Chugani HT, Asano E. Statistical mapping of ictal high-frequency oscillations in epileptic spasms. Epilepsia 2011;52:63–74. [7] Iwatani Y, Kagitani-Shimono K, Tominaga K, Okinaga T, Kishima H, Kato A, et al. Ictal high-frequency oscillations on scalp EEG recordings in symptomatic West syndrome. Epilepsy Res 2012;102:60–70. [8] Kobayashi K, Jacobs J, Gotman J. Detection of changes of highfrequency activity by statistical time–frequency analysis in epileptic spikes. Clin Neurophysiol 2009;120:1070–7. [9] Jacobs J, Kobayashi K, Gotman J. High-frequency changes during interictal spikes detected by time–frequency analysis. Clin Neurophysiol 2011;122:32–42. [10] Commission on Classification and Terminology of the ILAE. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989;30:389–99.
7
[11] Ohtahara S, Yamatogi Y. Ohtahara syndrome: with special reference to its developmental aspects for differentiating from early myoclonic encephalopathy. Epilepsy Res 2006;70:S58–67. [12] Mizrahi EM, Milh M. Early severe neonatal and infantile epilepsies. In: Bureau M, Genton P, Dravet C, Delgado-Escueta A, Tassinari CA, Thomas P, Wolf P, editors. Epileptic syndromes in infancy, childhood and adolescence. Montrouge, France: John Libbey Eurotext; 2012. p. 89–98. [13] Kroeger D, Amzica F. Hypersensitivity of the anesthesia-induced comatose brain. J Neurosci 2007;27:10597–607. [14] Ferron JF, Kroeger D, Chever O, Amzica F. Cortical inhibition during burst suppression induced with isoflurane anesthesia. J Neurosci 2009;29:9850–60. [15] Amzica F, Kroeger D. Cellular mechanisms underlying EEG waveforms during coma. Epilepsia 2011;52:25–7. [16] Kobayashi K, Yoshinaga H, Toda Y, Inoue T, Oka M, Ohtsuka Y. High-frequency oscillations in idiopathic partial epilepsy of childhood. Epilepsia 2011;52:1812–9. [17] Kobayashi K, Watanabe Y, Inoue T, Oka M, Yoshinaga H, Ohtsuka Y. Scalp-recorded high-frequency oscillations in childhood sleep-induced electrical status epilepticus. Epilepsia 2010;51:2190–4. [18] Bragin A, Wilson CL, Engel Jr J. Chronic epileptogenesis requires development of a network of pathologically interconnected neuron clusters: a hypothesis. Epilepsia 2000;41:S144–52. [19] Zijlmans M, Jacobs J, Zelmann R, Dubeau F, Gotman J. Highfrequency oscillations mirror disease activity in patients with epilepsy. Neurology 2009;72:979–86. [20] Jacobs J, Zijlmans M, Zelmann R, Chatillon CE, Hall J, Olivier A, et al. High-frequency electroencephalographic oscillations correlate with outcome of epilepsy surgery. Ann Neurol 2010;67:209–20. [21] Toda Y, Kobayashi K, Hayashi Y, Inoue T, Oka M, Ohtsuka Y. Effects of intravenous diazepam on high-frequency oscillations in EEGs with CSWS. Brain Dev 2013;35:540–7. [22] Brown EC, Rothermel R, Nishida M, Juha´sz C, Muzik O, Hoechstetter K, et al. In vivo animation of auditory-languageinduced gamma-oscillations in children with intractable focal epilepsy. Neuroimage 2008;41:1120–31. [23] Lachaux JP, Axmacher N, Mormann F, Halgren E, Crone NE. High-frequency neural activity and human cognition: past, present and possible future of intracranial EEG research. Prog Neurobiol 2012;98:279–301. [24] Pfurtscheller G, Cooper R. Frequency dependence of the transmission of the EEG from cortex to scalp. Electroencephalogr Clin Neurophysiol 1975;38:93–6.
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
Original article
High-frequency EEG activity in epileptic encephalopathy with suppression-burst Yoshihiro Toda a,b, Katsuhiro Kobayashi a,⇑, Yumiko Hayashi a, Takushi Inoue a, Makio Oka a, Fumika Endo a, Harumi Yoshinaga a, Yoko Ohtsuka a,c a
Department of Child Neurology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences and Okayama University Hospital, Okayama, Japan b Department of Pediatrics, School of Medicine, University of Tokushima, Tokushima, Japan c Asahigawaso Rehabilitation and Medical Center, Okayama, Japan Received 30 November 2013; received in revised form 7 April 2014; accepted 8 April 2014
Abstract Objective: We explored high-frequency activity in the suppression-burst (SB) pattern of interictal electroencephalogram (EEG) in early infantile epileptic encephalopathy including Ohtahara syndrome (OS) and early myoclonic encephalopathy (EME) to investigate the pathophysiological characteristics of SB. Methods: Subjects included six patients with the SB EEG pattern related to OS or EME (Group SB). The results were evaluated in comparison to trace´ alternant (TA) observed during the neonatal period in nine patients to rule out possible nonspecific relationships between high-frequency activity and periodic EEG patterns (Group TA). EEG was digitally recorded with a sampling rate of 500 Hz and the analysis was performed in each of the particular bipolar channel-pairs. We visually selected 20 typical consecutive burst sections and 160 inter-burst sections for comparison from the sleep record of each patient and performed the time–frequency analysis. We investigated the maximum frequencies of power enhancement in each derivation in both groups. Results: In Group SB, a significant increase in power at a frequency of 80–150 Hz was observed in association with the bursts, particularly in the bilateral parieto-occipital derivations, in all patients. In Group TA, on the contrary, no significant increase in high-frequency power was found. The maximum frequencies of power enhancement were significantly higher in Group SB than in Group TA (p < 0.001 by repeated-measures ANOVA). Conclusion: Interictal high frequencies of up to 150 Hz were detected in the suppression-burst EEG patterns in epileptic encephalopathy in early infancy. Further studies will be necessary to identify the role of the interictal high-frequency activity in the pathophysiology of such early epileptic encephalopathy. Ó 2014 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.
Keywords: High-frequency oscillations; Ohtahara syndrome; Early myoclonic encephalopathy; Trace´ alternant; Time–frequency analysis
1. Introduction ⇑ Corresponding author. Address: Department of Child Neurology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences and Okayama University Hospital, 5-1 Shikatacho 2-chome, Kita-ku, Okayama 700-8558, Japan. Tel.: +81 86 235 7372; fax: +81 86 235 7377. E-mail address: [email protected] (K. Kobayashi).
Analysis of high-frequency oscillations (HFOs) beyond the gamma band has been enabled by the recent technical development of digital electroencephalography (EEG). Remarkably, HFOs are suggested to have a close relation with epileptogenicity and ictogenicity.
http://dx.doi.org/10.1016/j.braindev.2014.04.004 0387-7604/Ó 2014 The Japanese Society of Child Neurology. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
2
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
Gamma activity in scalp EEGs has been reported in ictal EEGs associated with epileptic spasms (ESs) in West syndrome and with the interictal suppression-burst (SB) pattern in epileptic encephalopathy in early infancy [1–3]. Likewise, high-frequency activity >80 Hz has now been observed in scalp and cortical ictal EEGs of ESs [4–7], but it has not yet been clearly identified in association with the SB pattern. The SB pattern in EEG characterizes epileptic encephalopathy during the neonatal period and early infancy including Ohtahara syndrome (OS) and early myoclonic encephalopathy (EME). The SB pattern and ES in series both consist of recurring bursts of activity, and it has already been indicated that there is a mutual transition between the ictal activity of ES and the bursts in peri-ictal SB, and that gamma activity is augmented in association with the change in EEG pattern from SB to the ictal activity of ES [2]. Hence the SB pattern is suggested to have a close relation with the generative mechanisms of ES. We have therefore hypothesized that the bursts in SB, which are already known to be accompanied by activity occurring as high as the gamma band, may be associated with the generation of high-frequency activity >80 Hz that is not readily visible through ordinary analysis. To prove this hypothesis, we attempted to use a statistical time-frequency analysis method that we had developed for the analysis of high-frequency changes associated with intracranially recorded epileptic discharges [8,9]. In addition, we similarly analyzed the EEG pattern known as trace´ alternant (TA), which occurs during quiet sleep in neonates, for comparison in order to exclude the possibility that the generation of high frequencies may be a non-specific finding associated with any type of cyclic bursting pattern during the neonatal period and early infancy. 2. Subjects and methods 2.1. Patients Subjects with the SB pattern in EEG (Group SB) were a total of six patients at 44–55 weeks postconceptional age (the sum of gestational age and chronological age), comprising three with OS and three with EME. The onset of both OS and EME is very early, typically under three months of age, and occurs most commonly in the neonatal period. OS is characterized by (1) epileptic spasms (tonic spasms) with or without clustering, though additional partial seizures and rare myoclonus may occur in some patients, (2) consistent appearance of SB with regular periodicity in both waking and sleep EEGs, and (3) heterogeneous etiologies including brain malformations and gene mutations; EME is characterized by (1) fragmentary myoclonia as the main seizure type, (2) partial seizures, (3) later
appearance of massive myoclonia or epileptic spasms, (4) SB which may present only during sleep or be most prominent during sleep, and (5) unknown etiology though genetic or metabolic origins have been suggested [10–12]. Subjects whose TA in EEG was observed for comparison (Group TA) were nine newborns at 36–43 weeks of post-conceptional age who were being examined with EEG for various reasons including low birth weight, fetal hydrops, cerebral ventricular dilatation, neonatal convulsions, intraventricular hemorrhage, and agenesis of the corpus callosum. The patients with the SB pattern are indexed by numbers with the prefix S and those with TA by numbers with the prefix T (Table 1). We are aware that comparison between SB and TA is not ideal because the ages of the patients are different, but we undertook this comparison to achieve a methodological confirmation which would otherwise have been impossible. 2.2. EEG recording and analysis EEG was recorded with a sampling rate of 500 Hz using the Nihon-Kohden (Tokyo, Japan) Neurofax digital EEG system. The international 10–20 electrode placement system was used, and the analysis was performed in each of the following bipolar channel-pairs: F3-C3, F4-C4, P3-O1, and P4-O2. Both the SB pattern and the TA pattern were recorded during quiet/non-REM sleep. We then visually selected 20 typical artifact-free consecutive burst sections and 160 inter-burst sections from the sleep record of each patient for the time–frequency analysis described below. Each burst section was a 3 s segment of EEG data, as it included not only the 1 s onset part of the burst but also 2 s of the inter-burst period immediately before the burst. Every burst section was manually selected using the moment of the burst’s onset as a trigger. Each inter-burst section included a non-overlapping low-amplitude EEG epoch lasting 512 ms. A representative EEG trace showing the SB pattern recorded from Patient S2 with OS is depicted in Fig. 1A: bursts of bilaterally diffuse irregular high amplitude slow waves containing spikes with a duration of 1–2 s periodically repeated with intervals of low amplitude suppression phase lasting for 3–4 s each. A representative EEG trace of TA recorded from Patient T4 with ventricular dilatation observed in the prenatal period is shown in Fig. 1B: bursts of diffuse slow waves containing some sharp transients occurred periodically and lasted for a few seconds each with relatively low amplitude intervals. In TA, compared with the SB pattern, bursting slow waves were not very high in amplitude and lacked truly epileptic discharges, and activity during the relatively long inter-burst intervals was not flat.
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
3
Table 1 Subjects. Patient
Postconceptional age at examination (w)
Gestational age at birth
Birth weight (g)
Disorder
Patients with suppression-burst (SB) pattern in EEG S1 55 38 w 2 d S2 52 40 w 3 d S3 46 38 w 4 d S4 50 39 w S5 44 40 w S6 52 38 w Average 50 38 w 3 d
2746 3306 2840 2790 2814 2706 2867
EME OS OS EME EME OS
Patients with trace´ alternant (TA) in EEG T1 39 T2 40 T3 41 T4 39
28 28 36 39
w w w w
d d d d
962 805 3136 2644
T5 T6 T7 T8
43 41 36 39
39 31 30 36
w w4d w4d w6d
3084 758 1905 1973
T9 Average
42 40
41 w 3 d 34 w 5 d
2936 2023
LBW LBW Hydrops fetalis Cerebral ventricular dilatation observed in prenatal period Neonatal convulsion LBW LBW, IVH I Cerebral ventricular dilatation observed in prenatal period Agenesis of corpus callosum
4 2 3 4
Postconceptional age: the sum of gestational age and chronological age. EME: early myoclonic encephalopathy; OS: Ohtahara syndrome. LBW: low birth weight; IVH I: intraventricular hemorrhage Grade I.
Our procedure for statistical time-frequency analysis is illustrated in Supplementary Fig. 1, which shows representative SB data of the left parieto-occipital derivation recorded from Patient 2 with OS (same patient as in Figs. 1 and 2A). The burst sections were selected in each patient, and the raw EEG data in these sections were subjected to time–frequency analysis to yield an average power spectrum. We used the Gabor Transform (windowed Fourier transform with a full width at half maximum 50 ms Gaussian window) for time-frequency analysis, and the frequency range was 20–150 Hz. Power increases were sometimes observed in association with the bursts, but these could not be evaluated with certainty in the original spectra. The power levels of the burst sections were statistically compared with those of control inter-burst data sections by means of an unpaired t-test to obtain the t value and the corresponding p value at each pixel of the time–frequency spectrum. The control spectral data were obtained through application of the Fourier transform to each of the 160 inter-burst sections. The resulting spectrum of t-values, however, had an enormous number of pixels, and was noisy and difficult to interpret with an ordinary significance level a (say, 0.05 or a 5% error rate) because too many pixels were declared active (false positivity or type I error). Therefore we adopted a statistical procedure for controlling false discovery rate (FDR) to avoid type I errors; the details of this statistical procedure have already been described by Kobayashi et al. [8] and are
briefly summarized here. The FDR is defined as the ratio of the number of false positive pixels to the number of pixels declared active, and the FDR bound q, which must be defined by the user, is the maximum tolerable FDR on average. In the present study, the two-tailed test was used with q = 0.025. Pixels with significantly above-average power are indicated in red, and pixels with t values that did not reach the limits are in green, in the FDR-controlled t-spectrum. Computation was performed using a program written in-house for Matlab (version 6.5.1; MathWorks Inc., Natick, MA, USA). 2.3. Statistical comparison of high-frequency power changes between SB and TA We performed repeated-measures analysis of variance (ANOVA) to compare the maximum frequencies of power enhancement between groups and between derivations with a Greenhouse-Geisser correction using SPSS Statistics Japanese version 17.0 (IBM Japan, Tokyo). Relationships were considered statistically significant if p < 0.05. 3. Results During SB, a significant increase in power >80 Hz and up to 150 Hz was observed in association with the bursts in the bilateral parieto-occipital derivations in all patients, as indicated by a representative spectrum
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
4
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
in the frequency band > 39 Hz in the parieto-occipital derivations or in the band > 64 Hz in the fronto-central derivations (Fig. 3, crosses). Repeated-measures ANOVA showed that the maximum frequencies of power enhancement were significantly higher in Group SB than in Group TA (p < 0.001), as shown in Table 2. When there was no increase in power > 20 Hz, as in the TA case depicted in Fig. 2B, the maximum frequency of power enhancement was assumed to be 20 Hz in this statistical analysis (Patients T1, T5, and T9 in the fronto-central derivations, and Patients T1, T3–T6, and T9 in the parieto-occipital derivations) (Fig. 3). Because the effects of derivations were not significant (p = 0.418), no post-hoc multiple comparisons between derivations were performed. 4. Discussion
Fig. 1. EEGs showing a suppression-burst (SB) pattern and trace´ alternant (TA). (A) A representative SB pattern recorded from Patient S2 with Ohtahara syndrome (OS). Bursts of bilaterally diffuse irregular high-amplitude slow waves containing spikes with a duration of 1–2 s occur periodically, separated by intervals of low amplitude suppression phase lasting for 3–4 s. (B) A representative TA pattern recorded from Patient T4 with ventricular dilatation observed in the prenatal period. Bursts of diffuse slow waves containing some sharp transients with a duration of a few seconds occur periodically, separated by intervals of relatively low amplitude. In TA, in contrast to the SB pattern, bursting slow waves did not always have a high amplitude and lacked truly epileptic discharges, and activity during the relatively long inter-burst intervals was not flat.
of the bilateral parieto-occipital derivations of Patient S2 with OS (Fig. 2A). In five infants, this increase in power was observed in the bilateral fronto-central derivations as well, as indicated in Fig. 3 which shows the maximum frequency of significant power change in each derivation-pair (circles denote individual patients’ data; boxes and vertical bars indicate mean data and standard deviation, respectively). In the analysis of TA, to the contrary, as indicated by a representative spectrum of the bilateral parieto-occipital derivations of Patient T3 (Fig. 2B), no significant increase of power was observed
In the present study, interictal high frequencies up to 150 Hz were detected in the SB EEG patterns of epileptic encephalopathy in early infancy. No such high-frequency activity was detected in association with TA, a non-epileptic periodic EEG pattern observed in neonates. Thus it was suggested that pathological high-frequency activity is linked to the pathophysiology of epileptic encephalopathy in early infancy, and that high frequencies are unlikely to be generated non-specifically in association with all periodic EEG patterns. There appear to be no remarkable differences between OS and EME with respect to burst-associated high frequencies, although the number of subjects with each of these conditions was so small as to preclude any detailed comparison. The subjects in Group TA had some abnormalities and were therefore not ideal as controls. It cannot be denied that the conditions that caused them to undergo EEG examination could have some effect on their EEG activity, perhaps by decreasing the cerebral capability to generate high-frequency activity. EEG data of entirely normal neonates, however, were unavailable to us during the study period, so this question will have to be addressed in the future. The SB EEG pattern of epileptic encephalopathy shares common periodic characteristics with the burst suppression (BS) pattern, which is non-specifically observed in various neurological conditions such as anesthesia-induced deep coma. In BS, bursts of diffuse high-voltage slow waves are interrupted by isoelectric periods that reflect absence of cortical activity, but the BS pattern includes no epileptic discharges. Although the BS pattern appears to correspond to decreases in neuronal responsiveness associated with loss of consciousness due to anesthesia, it has been demonstrated in at least some coma patients that BS in fact results from a hyperexcitability in which bursting activity is triggered even by low-intensity stimuli in association
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
5
Fig. 2. Representative statistical time–frequency analysis. Spectra of t-values controlled for false discovery rate (FDR) in the EEG data of bilateral parieto-occipital derivations. (A) Analysis in SB recorded from Patient S2 with OS. Augmentation of power is observed in association with bursts in a frequency range up to 150 Hz. (B) analysis in TA recorded from Patient T4. Augmentation of power is not observed.
with suppressed inhibition [13–15]. There are no reports of HFOs in the BS EEG pattern to our knowledge, but these studies on non-epileptic BS may provide some hints regarding the pathophysiology of SB. Epileptic high-frequency activity has been reported in ictal scalp and intracranial EEGs of ESs [1,3,4,6,7]. It was suggested in the current study that high frequencies are generated not only transiently in association with seizures but also rather continuously during interictal
periods, at least in OS and EME, two representative types of epileptic encephalopathy in early infancy. Interictal HFOs occurred with a tight temporal relationship to interictal spikes in patients with benign childhood epilepsy with centrotemporal spikes and Panayiotopoulos syndrome, and the periods of active seizure occurrence were found to be closely related [16]. As the EEG pattern of continuous spikes and waves during slow sleep (CSWS), in which each spike is associated with intense
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
6
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
Fig. 3. Maximum frequency of significant power enhancement in each derivation. Each mark denotes an individual patient’s data (circle for SB and cross for TA); boxes and vertical bars indicate mean data and standard deviations, respectively. In almost all SB cases, power enhancement was observed at frequency P100 Hz, especially in the parieto-occipital derivation. In TA cases, in contrast, power enhancement was limited to <70 Hz and was even undetectable at frequency >20 Hz in three patients with fronto-central derivations and six patients with parieto-occipital derivations.
Table 2 Statistical analysis enhancement.
of
the
maximum
frequencies
of
power
Derivation/max frequency (Hz)
F3-C3
F4-C4
P3-O1
P4-O2
Group SB Average SD
123.5 33.7
135.0 17.6
143.0 17.1
146.3 9.0
Group TA Average SD
29.6 10.8
30.8 14.4
23.3 6.5
23.1 6.3
Repeated-measures ANOVA showed a statistically significant effect of groups (F = 598.3; p < 0.001) but no significant effects of derivations (F = 0.863; p = 0.418) nor of the interactions between groups and derivations (F = 3.063; p = 0.075).
HFOs [17], is regarded as a sort of electrical status epilepticus, the EEG pattern of SB with active high frequencies might well be akin to electrical status epilepticus. The origin of epileptic HFOs is not yet determined, though animal experiments have suggested that pathological HFOs are generated by action potentials of synchronously bursting principal cells and reflect the activity of pathologically interconnected clusters of neurons [18]. Accumulating evidence indicates that spikes and HFOs do not share identical clinical meanings and that HFOs may be a better biomarker of epileptogenicity/ictogenicity than spikes are: HFOs are more
tightly linked to seizures than spikes are [19], and success in epilepsy surgery is more closely related to the resection of the cortical regions that generate HFOs than to the elimination of spikes [20]. In our study on the effects of intravenous (IV) injection of diazepam (DZP), an agonist for the gamma-aminobutyric acid (GABA) A receptor, on the EEG pattern of CSWS, the effects of IV DZP on the intensity of spike-associated HFOs lasted longer than its effects on the amplitudes of the spikes themselves during the recovery process following a temporary suppression of CSWS in the three patients [21]. Thus the GABAergic interneuron system may be more closely related to HFOs than to spikes. Pathological gamma and high-frequency activity are of particular interest in pediatric epileptic encephalopathies including OS and EME. In epileptic encephalopathy, cognitive and behavioral impairment can be caused by epileptic activity itself; especially given that physiological gamma and high frequencies are reported to be involved in higher brain functions such as language, memory, and cognition [22,23], any possible interaction or mutual interference between pathological and physiological high-frequency activity should be pursued in future studies. We speculate that the persistent active generation of pathological high frequencies during interictal periods might be harmful to cognitive functions in the developing brain. High-frequency activity in scalp EEGs may not have the same meaning as intracranially recorded high frequencies because high frequencies are generally attenuated over the scalp, probably due to the summation of polyphasic cortical activity with variable phase [24]. Scalp high frequencies may represent the most intense intracranial high frequencies, but this issue requires further investigation. As the current study used EEG data that had already been recorded with a sampling rate of 500 Hz, it was difficult to analyze HFOs > 150 Hz. Experimenting with a much higher sampling rate might provide us new information. We used the correction method for controlling FDR to avoid false positivity, but the results may be somewhat different depending on the correction methods used and their thresholds. The significance of the present findings of highfrequencies in SB with respect to the pathophysiology, particularly in early infancy, is still at the initial step of research. Further studies will be necessary to identify the role of the interictal high-frequency activity in the pathophysiology of such early epileptic encephalopathy. Funding sources Dr. Kobayashi is a recipient of the Health and Labour Sciences Research Grant: Research on catastrophic epilepsy in infancy and early childhood– epidemiology, diagnosis and treatment guide, and is also supported in part by a Research Grant (21B-5) for
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004
Y. Toda et al. / Brain & Development xxx (2014) xxx–xxx
Nervous and Mental Disorders from the Ministry of Health, Labour and Welfare. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/ 10.1016/j.braindev.2014.04.004. References [1] Kobayashi K, Oka M, Akiyama T, Inoue T, Abiru K, Ogino T, et al. Very fast rhythmic activity on scalp EEG associated with epileptic spasms. Epilepsia 2004;45:488–96. [2] Kobayashi K, Inoue T, Kikumoto K, Endoh F, Miya K, Oka M, et al. Relation of spasms and myoclonus to suppression-burst on EEG in epileptic encephalopathy in early infancy. Neuropediatrics 2007;38:244–50. [3] Inoue T, Kobayashi K, Oka M, Yoshinaga H, Ohtsuka Y. Spectral characteristics of EEG gamma rhythms associated with epileptic spasms. Brain Dev 2008;30:321–8. [4] Akiyama T, Otsubo H, Ochi A, Ishiguro T, Kadokura G, Ramachandrannair R, et al. Focal cortical high-frequency oscillations trigger epileptic spasms: confirmation by digital video subdural EEG. Clin Neurophysiol 2005;116:2819–25. [5] Ramachandrannair R, Ochi A, Imai K, Benifla M, Akiyama T, Holowka S, et al. Epileptic spasms in older pediatric patients: MEG and ictal high-frequency oscillations suggest focal-onset seizures in a subset of epileptic spasms. Epilepsy Res 2008;78:216–24. [6] Nariai H, Nagasawa T, Juha´sz C, Sood S, Chugani HT, Asano E. Statistical mapping of ictal high-frequency oscillations in epileptic spasms. Epilepsia 2011;52:63–74. [7] Iwatani Y, Kagitani-Shimono K, Tominaga K, Okinaga T, Kishima H, Kato A, et al. Ictal high-frequency oscillations on scalp EEG recordings in symptomatic West syndrome. Epilepsy Res 2012;102:60–70. [8] Kobayashi K, Jacobs J, Gotman J. Detection of changes of highfrequency activity by statistical time–frequency analysis in epileptic spikes. Clin Neurophysiol 2009;120:1070–7. [9] Jacobs J, Kobayashi K, Gotman J. High-frequency changes during interictal spikes detected by time–frequency analysis. Clin Neurophysiol 2011;122:32–42. [10] Commission on Classification and Terminology of the ILAE. Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia 1989;30:389–99.
7
[11] Ohtahara S, Yamatogi Y. Ohtahara syndrome: with special reference to its developmental aspects for differentiating from early myoclonic encephalopathy. Epilepsy Res 2006;70:S58–67. [12] Mizrahi EM, Milh M. Early severe neonatal and infantile epilepsies. In: Bureau M, Genton P, Dravet C, Delgado-Escueta A, Tassinari CA, Thomas P, Wolf P, editors. Epileptic syndromes in infancy, childhood and adolescence. Montrouge, France: John Libbey Eurotext; 2012. p. 89–98. [13] Kroeger D, Amzica F. Hypersensitivity of the anesthesia-induced comatose brain. J Neurosci 2007;27:10597–607. [14] Ferron JF, Kroeger D, Chever O, Amzica F. Cortical inhibition during burst suppression induced with isoflurane anesthesia. J Neurosci 2009;29:9850–60. [15] Amzica F, Kroeger D. Cellular mechanisms underlying EEG waveforms during coma. Epilepsia 2011;52:25–7. [16] Kobayashi K, Yoshinaga H, Toda Y, Inoue T, Oka M, Ohtsuka Y. High-frequency oscillations in idiopathic partial epilepsy of childhood. Epilepsia 2011;52:1812–9. [17] Kobayashi K, Watanabe Y, Inoue T, Oka M, Yoshinaga H, Ohtsuka Y. Scalp-recorded high-frequency oscillations in childhood sleep-induced electrical status epilepticus. Epilepsia 2010;51:2190–4. [18] Bragin A, Wilson CL, Engel Jr J. Chronic epileptogenesis requires development of a network of pathologically interconnected neuron clusters: a hypothesis. Epilepsia 2000;41:S144–52. [19] Zijlmans M, Jacobs J, Zelmann R, Dubeau F, Gotman J. Highfrequency oscillations mirror disease activity in patients with epilepsy. Neurology 2009;72:979–86. [20] Jacobs J, Zijlmans M, Zelmann R, Chatillon CE, Hall J, Olivier A, et al. High-frequency electroencephalographic oscillations correlate with outcome of epilepsy surgery. Ann Neurol 2010;67:209–20. [21] Toda Y, Kobayashi K, Hayashi Y, Inoue T, Oka M, Ohtsuka Y. Effects of intravenous diazepam on high-frequency oscillations in EEGs with CSWS. Brain Dev 2013;35:540–7. [22] Brown EC, Rothermel R, Nishida M, Juha´sz C, Muzik O, Hoechstetter K, et al. In vivo animation of auditory-languageinduced gamma-oscillations in children with intractable focal epilepsy. Neuroimage 2008;41:1120–31. [23] Lachaux JP, Axmacher N, Mormann F, Halgren E, Crone NE. High-frequency neural activity and human cognition: past, present and possible future of intracranial EEG research. Prog Neurobiol 2012;98:279–301. [24] Pfurtscheller G, Cooper R. Frequency dependence of the transmission of the EEG from cortex to scalp. Electroencephalogr Clin Neurophysiol 1975;38:93–6.
Please cite this article in press as: Toda Y et al. High-frequency EEG activity in epileptic encephalopathy with suppression-burst. Brain Dev (2014), http://dx.doi.org/10.1016/j.braindev.2014.04.004