EEG in Children with Early-onset Benign Occipital Seizure Susceptibility Syndrome: Panayiotopoulos Syndrome

Transcription

1 Epilepsia, 44(3): , 2003 Blackwell Publishing, Inc International League Against Epilepsy EEG in Children with Early-onset Benign Occipital Seizure Susceptibility Syndrome: Panayiotopoulos Syndrome * Mayu Ohtsu, *Hirokazu Oguni, *Kitami Hayashi, *Makoto Funatsuka, *Kaoru Imai, and *Makiko Osawa *Department of Pediatrics, Tokyo Woman s Medical University, Tokyo, and Division of Pediatric Neurology, Yokohama Ryoikuen, Yokohama, Japan Summary: Purpose: We analyzed sequential changes in the localization of EEG foci along with age to identify a specific EEG pattern, and the relation between the clinical manifestations and the EEG pattern in patients with Panayiotopoulos syndrome (PS). Methods: The subjects were 76 children, who had been followed up >2 years with repeated EEG examinations at 6-month intervals. Analysis of EEG findings included the determination of localization of spike foci, as a function of age, by using cross-sectional data, and the identification of subgroups with homogeneous EEG patterns. Then we compared certain clinical features among these subgroups. Results: In the cross-sectional EEG study, the occipital EEG spike focus was most frequently seen between ages 2 and 5 years. Independent and synchronous frontopolar and occipital spikes (Fp-O spikes) and centroparietotemporal (CPT) EEG spike foci had increased incidences between ages 4 and 7 years, and between ages 6 and 10 years, respectively. We subclassified the 76 patients into the following five subgroups based on the evolutional changes in epileptic EEG foci, which frequently showed shifting, multiplications, and generalization: (a) persistent occipital focus group (O group), (b) Fp-O spikes group (Fp-O pattern group), (c) generalized EEG pattern group, (d) CPT foci group (CPT group), and (e) no epileptic EEG focus group. The Fp-O group showed the latest age at onset of epilepsy. The generalized EEG pattern group had the highest frequency of seizures as well as recurrences of status epilepticus (SE), as well as the longest active seizure period among the five groups. Conclusions: These results indicated that the EEG foci in most of patients with PS are frequently shifting location, multiplying, and propagating diffusely with age, rather than persistently localizing in the occipital region. In addition, the EEG patterns showed a certain trend and roughly corresponded to certain clinical characteristics. However, the prognosis of the seizures appeared to be favorable regardless of the EEG pattern. Key Words: Early-onset benign occipital seizure susceptibility syndrome Shifting EEG foci Frontopolar-occipital EEG foci Multifocal EEG foci Panayiotopoulos syndrome. Early-onset benign occipital seizure susceptibility syndrome (EBOSS), recently known as Panayiotopoulos syndrome, has received attention as a newly identified idiopathic focal epilepsy developing during early childhood (1 11). This syndrome has now been incorporated into the proposed International Classification of Epileptic Syndromes as Panayiotopoulos-type early-onset benign childhood occipital epilepsy, together with Gastauttype late-onset childhood occipital epilepsy (12). Recent studies have shown the incidence of this epileptic syndrome accounting for 6 10% of all childhood focal epilepsy, the most common idiopathic type after benign childhood epilepsy with centrotemporal foci (BECT), Accepted October 20, Address correspondence and reprint requests to Dr. H. Oguni at Department of Pediatrics, Tokyo Women s Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162, Japan. . hoguni@ped. twmu.ac.jp which accounted for % (6,13). Panayiotopoulos originally defined the following six criteria to characterize this syndrome: (a) brief or prolonged seizures, characterized by initial vomiting, followed by eye deviation with or without secondary generalization; (b) normal development before the onset of epilepsy; (c) absence of structural neuroimaging abnormalities; (d) the onset of epilepsy between ages 1 and 8 years; (e) occipital EEG foci, at times shifting to centrotemporal or frontal regions with age; and (f) ultimate remission before age 12 years (14). The age at onset of epilepsy and the EEG findings have recently been modified (15). Our recent study showed that the EEG foci in this syndrome tend either to shift their location, to become multifocal, or to spread diffusely with age (6,16). Thus we noted that the occipital EEG foci frequently identified in the early stage of this syndrome would be easily missed depending on the timing of the EEG recording (17). In his most recent 435

2 436 M. OHTSU ET AL. publication, Panayiotopoulos (15) described the fundamental EEG features of this syndrome as being multifocal EEG foci rather than having an occipital EEG focus. He also stressed the importance of ictal vomiting or autonomic seizure symptoms as a major feature of this syndrome. Epilepsy syndromes are characterized by clusters of homogeneous clinical and EEG manifestations, including the onset age, seizure types, circadian rhythm of the seizures, and epileptic EEG abnormalities. As the specific EEG pattern is an important biologic marker to distinguish epileptic syndromes, together with clinical features, we must clarify the EEG characteristics of this syndrome. In this study, we analyzed the sequential changes in localization of EEG foci with age to confirm whether there is a specific EEG pattern, and the relation between the clinical manifestations and EEG pattern in this syndrome. SUBJECTS AND METHODS The subjects were 76 children who visited our clinic between 1968 and 1998, and fulfilled the following criteria: (a) normal development before the onset of epilepsy; (b) absence of structural abnormalities on either computed tomography (CT) or magnetic resonance imaging (MRI); (c) onset of epilepsy between ages 1 and 8 years; (d) presence of brief or prolonged attacks characterized by initial vomiting and eye deviations with or without subsequent secondary generalization; and (e) follow-up >2 years with repeated EEG examinations. In this study, the EEG findings were excluded from the criteria. The cases previously reported were included in this study (6,15). The EEGs were recorded with TC, , HF, Hz, and both monopolar and bipolar montages. The EEG examinations were performed at 6-month intervals because the principal policy to look after children with epilepsy in our hospital has been to check EEG examinations regularly at 6-month intervals from EEGs were recorded with 12 channels from 1968 to 1980, and thereafter with 16- to 21-channel machines. Natural or triclofos sodium induced sleep EEGs were obtained in all cases, but waking EEGs were not recorded in some of early cases. As for electrode placements, we used a special system for children younger than 6 years, and thereafter, the standard international system until We took this difference in the montage into consideration when analyzing the data. The details of interpolations and methods were described previously (18). We defined shifting of EEG foci as movement of a spike focus from one region to another, with disappearance of the original focus. This study included the following: (a) determination of localization of spike foci, as a function of age, using cross-sectional data; and (b) classifying the 76 patients into subgroups with homogeneous EEG patterns, based on the analysis of evolutional changes in localization of EEG spike foci at 6-month intervals in an individual patient. We analyzed the sleep EEGs only, because spike foci were highly activated during sleep but were infrequent during wakefulness. Then we compared the clinical features including sex, onset of epilepsy, family history of convulsive disorders up to third-degree relatives, a history of febrile convulsions (FCs), types of seizures, total frequency of attacks, incidence and frequency of status epilepticus (SE), the duration of the active seizure period, and the prognosis among these subgroups. We did not investigate the visual reactivity and fixation-off sensitivity of the occipital paroxysms (15), because, in part, the occipital paroxysms were rarely seen during wakefulness, and we do not routinely perform these procedures. Statistical analysis The 2 test or Fisher s exact probability test was used when n was less than 5, and analysis of variance (ANOVA) testing was used for comparisons. A value of p < 0.05 was considered significant. RESULTS The 76 subjects included 37 boys and 39 girls. Family history of convulsive disorders was recognized in 21 cases, accounting for 27.6% of the 76 cases. A history of FCs was confirmed in 33 (43.4%) cases. The age at onset of epilepsy ranged from 14 to 118 months, with an average of 48 months. The duration of the active seizure period ranged from 0 to 139 months, with a median of 69.5 months. The total number of seizures in each case was from one to as many as 27, with a mean of 4.8. Forty-five children or 61.6% had SE >30 min at least once during the clinical course. A total of 493 EEG examinations, an average of 6.5 times in each patient, were performed. Localization of the epileptic EEG foci as a function of age The epileptic EEG discharges were identified at the first EEG examination in 59 (78%) of the 76 cases. The earliest age the epileptic EEG discharges was identified was 2 years and 0 months. The occipital EEG focus was most frequently seen between ages 2 and 5 years (Fig. 1). Thereafter, synchronous and independent occipital and frontopolar spike foci increased in incidence. The details of this EEG pattern, which we have termed the Fp-O pattern, were described elsewhere (19). The Fp-O pattern was most frequently seen in the age range from 4 to 7 years. After age 3 years, frontopolar (Fp) spikes appeared in several cases, and steadily increased in incidence thereafter. Centrotemporoparietal (CPT) spike foci

3 EEG STUDY OF EBOS 437 EEG-focus group, there was no epileptic EEG focus throughout the clinical course. Three patients were not able to be subclassified into the previously mentioned criteria. Persistent O group This group included 12 cases. The epileptic EEG discharge was recorded from the occipital region from as early as age 3 years (Figs. 2A and 3A). The occipital EEG spike was identified in 80% of them between ages 4 and 6 years, and in 60 70% between ages 7 and 11 years, and in 20% of all cases thereafter. FIG. 1. Distribution of the epileptic EEG foci as a function of age (n = 76). Occipital foci were most frequent at ages 2 6 years, whereas frontopolar foci were seen at the lowest frequency in this age group. CPT foci were most frequent at ages 7 to 12 years. Fp-O foci were maximal at ages 6 7 years and decreased in frequency thereafter. Generalized rhythmic and sustaining spikeand-wave discharges appeared from age 4 to 11 years. G, Generalized rhythmic and sustaining spike and wave discharges; C, central spikes; O, occipital EEG spikes; Fp-O, synchronous and independent frontopolar and occipital spikes; Fp, frontopolar spikes; CPT, centroparietotemporal spikes including central, centrotemporal, and parietal spikes; N/N, no paroxysmal EEG discharges. including central, centrotemporal, or parietal spike foci, first appeared as early as age 3 years. They increased in incidence between ages 5 and 10 years, accounting for 20 30% of all cases. The incidence of CPT spike foci decreased sharply after age 11 years. A few cases showed Fp or occipital EEG spikes after age 14 years. Classification of the EEG patterns based on the evolutionary changes in localization of epileptic EEG foci in individual patients We subclassified the 76 patients into five subgroups based on the evolutionary changes in epileptic EEG foci, which frequently shifted, multiplied, and were generalized. The persistent occipital EEG focus group (O group) was characterized by the persistent appearance of an occipital EEG focus without spreading to the other areas throughout the clinical course. The frontopolaroccipital EEG foci group (Fp-O pattern group) was characterized by the new appearance of Fp spikes on the EEG with an occipital EEG focus, resulting in synchronous and independent Fp and occipital EEG foci. Sometimes the EEG spikes also appeared in CPT regions, although they never became predominant. In the generalized EEG pattern group, the EEG findings were at first identical to the Fp-O pattern group, but the spike wave discharges subsequently became generalized and appeared in a rhythmic and sustained fashion, at times looking like the continuous spike wave during slow sleep (CSWS) (20). The CPT group was characterized by the shifting of the previous occipital EEG focus to CPT regions, but it never spread to the frontal regions. In the no-epileptic- Fp-O pattern group This group comprised 14 cases. At age 2 years, eight of the 14 cases showed occipital EEG spikes (Figs. 2B and 3B). After age 3 years, either they had shifted to Fp regions, or Fp spikes newly appeared in >20% of cases. The Fp-O EEG pattern was most frequently observed between ages 4 and 7 years, when 20 40% of all cases showed this EEG pattern. Then this pattern gradually decreased in frequency after age 9 years. Finally, it was replaced by Fp spikes at age 15 years. Generalized EEG pattern group This group consisted of 19 cases. More than half of the cases had occipital EEG spikes between ages 2 and 3 years (Figs. 2C and 3C). After age 4 years, they decreased in incidence, and disappeared after age 10 years. Fp-O foci were observed in 50% of cases at an age ranging from 2 to 9 years. Rhythmic and sustaining generalized spike wave discharges developed in 10 30% of all cases as early as age 4 years. CSWS developed in one case during the follow-up. Generalized spike wave discharges were seen in 10 20% of all cases between ages 4 and 11 years. They were replaced by either CPT or Fp spike foci after age 12 years. CPT pattern group This group comprised 21 cases. The occipital EEG spikes were seen in >50% of all cases between ages 3 and 5 years (Figs. 2D and 3D). Then the incidence of the occipital EEG focus steadily decreased over the subsequent 6 years, and to zero after age 11 years. The CPT spikes developed as early as age 3 years and were discovered in >50% of all cases between ages 6 and 10 years. The CPT foci gradually decreased in incidence and disappeared after age 14 years. In two children, CSWS EEG patterns developed at some point in the clinical course. Although the children in the Fp-O pattern group showed Fp spikes in >10% of all cases after age 12 years, those in the CPT group showed CPT spikes in only a few cases after age 12 years. No paroxysmal EEG pattern group This EEG pattern was seen in seven cases demonstrating epileptic EEG abnormalities during the follow-up

4 438 M. OHTSU ET AL. FIG. 2. Distribution of the epileptic EEG foci as a function of age in the individual group patients. A: Persistent O group (n = 12). Occipital foci were frequent at age 4 11 years and decreased in frequency thereafter. B: Fp-O pattern group (n = 14) Fp-O foci appeared at age 4 years and were maximum at age 7 years. Note the frequent association of CPT foci and Fp spikes. C: Generalized EEG pattern group (n = 19). The occipital foci and Fp-O foci were first seen at age 2 3 years. Generalized rhythmic and sustaining spike-and-wave discharges were distributed at age 4 11 years. D: CPT pattern group (n = 21). The ratio between the occipital foci and CPT foci gradually changed with the former being replaced by the latter with age. Abbreviations are same as those of Fig. 1. ranging from 40 to 138 months with a mean of 79 months. However, one patient demonstrated parietooccipital EEG spikes only once, immediately after cessation of a complex partial seizure and did not show any epileptic abnormality thereafter. A comparison of the clinical features among the five groups There were no statistically significant differences in the clinical parameters among the five groups, except for the active seizure period (p < 0.05), although the EEG pattern roughly corresponded to certain clinical characteristics (Table 1). The history of FCs was highest in the persistent O group, and it was lowest in the Fp-O pattern group. The patients in the generalized EEG pattern group had more frequent seizures and a significantly longer duration of the active seizure period than did those in the other four groups. As for the CPT group, the mean duration of the active seizure period was 20.0 months and appeared shortest among the five groups. All patients in the no-paroxysmal-eeg-pattern group experienced SE at least once. With respect to the types of seizures, those in the persistent O group and Fp-O pattern group showed more frequent recurrences of complex partial seizures, whereas those in the other three groups experienced more frequent secondary generalizations. DISCUSSION Regarding the EEG findings of PS, we have stressed the variable evolutional changes in EEG foci with age, which may have weakened the specificity of this epilepsy syndrome (6,15). We must rely heavily on clinical findings to make a syndromic diagnosis if the EEG findings are nonspecific. Although the most important clinical finding is paroxysmal vomiting among the various characteristic clinical manifestations (4,5,11,17,21,22), we should reinvestigate how the EEG findings contribute to the diagnosis of this syndrome when patients have extraoccipital EEG foci. In his recent publication, Panayiotopoulos (15) stressed the importance of multifocal EEG foci as a fundamental EEG feature of PS, and the cloned-like repetitive multifocal spike wave complexes being pathognomonic of PS if they are seen in patients with a few seizures (15). However, as cloned-like repetitive multifocal spike wave complexes were seen in only 19% of PS, and he has not described when the multifocal spike foci develop and how they change with age, we must clarify the details of EEG findings in PS. Therefore we retrospectively studied the long-term sequential EEG changes recorded at 6-month intervals in the patients with PS fulfilling the criteria without the

5 EEG STUDY OF EBOS 439 FIG. 3. Illustrative EEG samples of the each EEG pattern group except for the no-paroxysmal-eeg-pattern group. A: Persistent O group. The patient was a 17-year-old-girl. She developed sudden vomiting followed by eye deviation and left unilateral seizures during sleep, which lasted for 45 min. The symptoms stopped after intravenous diazepam injection at our hospital. The second attack lasting a few minutes recurred 5 months later, but she has been seizure free until now. The EEGs constantly demonstrated occipital spike discharges. B: Fp-O pattern group. The patient was a 6-year, 2-month-old boy. The first EEG, performed incidentally at age 2 years 8 months because of minimal speech delay, showed active sharp-waves recorded independently from left posterior temporooccipital regions, and the right occipital region. He had only one brief seizure at age 4 years 1 month. An EEG, recorded 2 months later, showed independent and synchronous occipital and frontopolar spike discharges. His modified Binet IQ test was 100. He has been followed up without medication. C: Generalized EEG pattern group. The patient was an 8-year-old girl. She had nine seizures including one episode of status epilepticus. The EEG showed bilateral independent occipital spikes at age 3 years 6 months, which became generalized and appeared almost continuously during sleep at age 5 years 9 months. D: CPT pattern group. The patient was an 8-year-old girl. She had only once experienced status epilepticus at age 3 years 3 months. The EEG demonstrated parietooccipital EEG foci until age 4 years 10 months, which shifted to bilateral centrotemporal region at age 5 years 6 months. She has been seizure free without medication since age 5 years 9 months. EEG findings. As a result, the occipital EEG focus was most frequently seen between ages 2 and 5 years. However, the characteristic EEG pattern consisting of independent and synchronous Fp and O spikes increased in incidence and was observed in several to 10% of all cases from age 2 years and in 20% of cases between ages 4 and 7 years. The shifting (or multiplication) of EEG foci to the CPT regions was most frequently seen between ages 6 and 10 years, and reached a maximum at age 10 years, when 30% of all cases had shown this EEG pattern. However, the incidence of the epileptic EEG discharges sharply decreased after age 11 years. The localization of the final EEG foci tended to be clustered in either the Fp or occipital region. Gibbs and Gibbs (23,25,26) and Gibbs et al. (24) investigated the migration of epileptic EEG foci with age in children, demonstrating that occipital, centrotemporal, and frontal EEG foci were most predominant between ages 3 and 6 years, 7 and 10 years, and during adulthood, respectively, regardless of etiology, with the absence or presence of seizures. In 50% of patients with an occipital or centrotemporal EEG focus, the epileptic EEG discharges disappeared between ages 9 and 15 years. Subsequent studies largely supported the preferential localization of epileptic foci in childhood, although they largely disagreed on the concept of the migration of epileptic foci (18,27 30). The present results, regarding the cross-sectional analysis of the epileptic EEG foci in patients with PS, were identical to those of Gibbs and Gibbs (24,25). However, the evolutionary changes in the localization of epileptic EEG foci in individual patients with PS appeared diverse and complex, and different from those demonstrated by Gibbs, who showed that the centrotemporal EEG foci in 15% of all patients whose

6 440 M. OHTSU ET AL. TABLE 1. Comparisons of clinical features among the five EEG pattern groups Persistent 0 group Fp-0 pattern group Generalized EEG pattern group CPT group No paroxysmal EEG pattern group p Value Number of patients Sex (F/M) 9/3 6/8 8/11 10/11 3/4 >0.05 FH (Epi/FC) 5 (41.7%) 4 (28.6%) 4 (21.1%) 3 (14.3%) 2 (28.6%) >0.05 PH of FC 7 (58.3%) 3 (21.4%) 8 (42.1%) 11 (52.4%) 3 (42.9%) >0.05 Onset age of Epi (mo) 44.4 (6 92) 56.9 (34 71) 48.0 (28 75) 48.3 (14 118) 42.9 (10 73) >0.05 Number of Sz. 4.5 (2 13) 3.0 (1 8) 8.5 (1 27) 4.2 (1 10) 5.6 (1 14) >0.05 Incidence of SE 5 (41.7%) 7 (50%) 12 (63.2%) 14 (66.7%) 7 (100%) > (40%) 5 (71%) 4 (33%) 7 (50%) 3 (43%) (60%) 2 (29%) 6 (50%) 6 (42.9%) 4 (57%) 4< 0 (0%) 0 (0%) 2 (17%) 1 (7.1%) 0 (0%) Median duration of longest sz. in an individual patient (mo) 15 (5 120) 50 (5 120) 40 (10 360) 40 (2 190) 50 (30 60) Active sz. period (mo) 29.1 (3 94) 29.3 (0 139) 37.5 (4 104) 20.0 (0 47) 25.4 (0 64) <0.05 Sz type >0.05 CPS CPS evolving GTCS, HC F, female; M, male; FH, family history; PH, past history; FC, febrile convulsions; Sz, seizures; SE, status epilepticus; CPS, complex partial seizures; GTCS, generalized tonic clonic seizures; HC, hemiconvulsions; mo, months. occipital focus migrated to that region moved to the frontal region (24,25). Other studies, including our previous one, demonstrated that the EEG foci did not move regularly, as shown by Gibbs, and appeared as if they were unstable in terms of localization or moving in a random fashion (18,27). Our present results as to the variable evolutionary changes in EEG foci might simply reflect the maturational process of the developing brain because the ages at onset and active period of PS distribute during early childhood, when the EEG spike foci appear to be unstable (18). The EEG findings of PS include all the developmental EEG changes such as migration of epileptic EEG foci, multiplication and diffuse propagations of the epileptic EEG discharge, which are most frequently observed during preschool and school ages. However, we must keep in mind that these EEG characteristics were observed only in the sleeping state, when spike foci were highly activated. The visual reactivity or fixation-off sensitivity has been described as somewhat specific for occipital foci in PS, although we did not investigate this phenomenon (15). The presence of this phenomenon would strengthen the diagnosis of PS together with the evolutional EEG changes described here. We subclassified the EEG pattern into five subgroups and analyzed the clinical features according to differences in the EEG pattern. Although there were no statistically significant differences in the clinical parameters except for active seizure period, the EEG pattern showed some trends and roughly corresponded to certain clinical characteristics. The generalized EEG pattern group, demonstrating the most active and florid epileptic EEG abnormalities, had the highest frequency of seizures and significantly longest active seizure period among the five groups. Although this group was most resistant to treatment among the five groups, the patients finally became seizure free after age 11 years 6 months. In contrast, the CPT group patients had the shortest active seizure period. Although this group of patients tended to experience SE as a first seizure manifestation, they had few seizures thereafter, and appeared to be prototype cases, originally described by Panayiotopoulos (1,21). The epileptic EEG phenomena showing the EEG focus moving to the Fp from the occipital region and forming synchronous Fp-O spike discharges has received little attention in the past. The present study showed that the occipital EEG spikes appeared initially and then shifted to the Fp region or appeared at the same time as Fp spikes, forming an Fp-O EEG pattern. Some patients with the Fp-O EEG pattern showed subsequent sustained rhythmic generalized spike wave complexes thereafter (generalized EEG pattern group). Our previous study investigating the Fp-O EEG pattern demonstrated occipital EEG spikes constantly leading to Fp spikes, resulting in secondary occipitofrontopolar synchrony (20). We concluded that this phenomenon is an age-dependent nonspecific EEG pattern reflecting the maturational process of the brain. It contributes to the multiplication or diffusion of the preexisting localized spike discharges because this EEG phenomenon was observed not only in idiopathic, but also in symptomatic, focal epilepsy. As for the etiology of PS, genetic predisposition to seizure susceptibility has been considered to play an important role. Patients in the CPT group may have had a common underlying pathophysiology with those with BECT (4), because there is evidence for this [i.e., EEGs exhibiting both occipital and rolandic EEG foci in one patient with PS (6,7,15), the seizure manifestation changing from typical PS seizures to sylvian seizures with advancing age (31), and the presence of a family including patients with BECT and PS (11)]. The age at onset of epilepsy is generally older in BECT, ranging from 5 to 15 years with peak of 9 years, than in PS (32). The

7 EEG STUDY OF EBOS 441 common features shared by both BECT and PS are a tendency to develop seizures during sleep, high incidences of the previous FC, and excellent prognosis of the seizures and intelligence, with patients usually entering remission until puberty. This evidence can explain why the CPT group patients had a lower seizure frequency, as well as a shorter seizure period. Recently Doose (33) proposed the concept of hereditary impairment of brain maturation, which underlies childhood epilepsies with multifocal sharp waves, including benign childhood epilepsy with occipital paroxysms, BECT, CSWS, Landau Kleffner syndrome, and atypical benign partial epilepsy, suggesting that these syndromes are not independent, but represent one clinical and EEG spectrum of genetically determined childhood epilepsy. The present results as to the EEG evolutional process appear to be reasonable when we take this theory into account. In the present series, we identified a no paroxysmal EEG pattern group, which demonstrated no interictal EEG spike discharges throughout the clinical course. Their clinical findings were identical to those of the other groups except for the absence of the epileptic EEG abnormalities. However, as one patient in the noparoxysmal-eeg-pattern group demonstrated occipital foci only once immediately after SE, this patient group may represent a special type of the persistent O group. The present results indicated that the EEG foci in most of patients with PS easily shift location, multiply, and propagate diffusely with age, rather than persistently localizing in the occipital region. In addition, the EEG pattern showed certain trends and roughly corresponded to certain clinical characteristics. However, the prognosis of the seizures appeared to be favorable regardless of the EEG pattern, despite some differences in response to the treatment (16,17). We must carefully look after patients with PS to avoid unnecessary overdoses or polytherapy with AEDs, although some patients may have shown a resistant clinical course with florid active EEG abnormalities despite ultimate remission. REFERENCES 1. Panayiotopoulos CP. Benign nocturnal childhood occipital epilepsy; new syndrome with nocturnal seizures, tonic deviations of the eyes, and vomiting. J Child Neurol 1989;4: Panayiotopoulos CP. Benign childhood epilepsy with occipital paroxysms: a 15-year prospective study. Ann Neurol 1989;26: Panayiotopoulos CP. Benign childhood partial epilepsies: benign childhood seizure susceptibility syndromes. J Neurol Neurosurg Psychiatry 1993;56: Panayiotopoulos CP. Early onset benign childhood occipital seizures: Panayiotopoulos syndrome. In: Panayiotopoulos CP, ed. Benign childhood partial seizures and related epileptic syndromes. London: John Libbey, 1999: Ferrie CD, Beaumanoir A, Guerrini R, et al. Early-onset benign occipital seizure susceptibility syndrome. Epilepsia 1997;38: Oguni H, Hayashi K, Imai K, et al. Study on the early-onset variant of benign childhood epilepsy with occipital paroxysms otherwise described as early-onset benign occipital seizure susceptibility syndrome. Epilepsia 1999;40: Caraballo R, Cersosimo R, Medina C, et al. Panayiotopoulos-type benign childhood occipital epilepsy: a prospective study. Neurology 2000;55: Kivity S, Ephraim T, Weitz R, et al. Childhood epilepsy with occipital paroxysms: clinical variants in 134 patients. Epilepsia 2000;41: Verrotti A, Domizio S, Guerra M, et al. Childhood epilepsy with occipital paroxysms and benign nocturnal childhood occipital epilepsy. J Child Neurol 2000;15: Tsai ML, Lo HY, Chaou WT. Clinical and electroencephalographic findings in early and late onset benign childhood epilepsy with occipital paroxysms. Brain Dev 2001;23: Ferrie CD, Grunewald RA. Panayiotopoulos syndrome: a common and benign childhood epilepsy. Lancet 2001;357: Engel J. A proposed diagnostic scheme for people with epileptic seizures and with epilepsy: report of the ILAE Task Force on Classification and Terminology. Epilepsia 2001;42: Panayiotopoulos CP. Epidemiology. In: Panayiotopoulos CP, ed. Panayiotopoulos syndrome: a common and benign childhood epileptic syndrome. London: John Libbey, 2002: Panayiotopoulos CP. Benign childhood epilepsy with occipital paroxysms. In: Andermann F, Beaumanior A, Mira L, Tassinary CA, eds. Occipital seizures and epilepsies in children. London: John Libbey, 1993: Panayiotopoulos CP. Electroencephlography. In: Panayiotopoulos CP, ed. Panayiotopoulos syndrome: a common and benign childhood epileptic syndrome. London: John Libbey, 2002: Oguni H, Hayashi K, Funatsuka M, et al. Study on early-onset benign occipital seizure susceptibility syndrome. Pediatr Neurol 2001;25: Oguni H. Panayiotopoulos syndrome. Lancet 2001:358: Oguni H, Hayashi K, Osawa M. Migration of epileptic foci in children. In: Stefan H, Andermann F, Chauvel P, et al., eds. Advances in Neurology. Vol 81. Plasticity in epilepsy: dynamic aspects of brain function. Philadelphia: Lippincott Williams & Wilkins, 1999: Ueno M, Oguni H, Yasuda K, et al. Neurophysiological study of secondary synchronous occipito-frontopolar spikes in childhood. Clin Neurophysiol 2001;112: Tassinari CA, Bureau M, Dravet C, et al. Epilepsy with continuous spikes and waves during slow sleep-otherwise described as ESES. In: Roger J, Dravet C, Bureau M, et al., eds. Epileptic syndromes in infancy, childhood and adolescence. 2nd ed. London: John Libbey Eurotext, 1992: Panayiotopoulos CP. Vomiting as an ictal manifestation of epileptic seizures and syndrome. J Neurol Neurosurg Psychiatry 1988; 51: Panayiotopoulos CP. Clinical manifestations. In: Panayiotopoulos CP, ed. Panayiotopoulos syndrome: a common and benign childhood epileptic syndrome. London: John Libbey, 2002: Gibbs FA, Gibbs EL. Changes in epileptic foci with age. Electroencephalogr Clin Neurophysiol 1953;(suppl 4): Gibbs EL, Gillen HW, Gibbs FA. Disappearance and migration of epileptic foci in children. Am J Dis Child 1954;88: Gibbs FA, Gibbs EL. Good prognosis of mid-temporal epilepsy. Epilepsia 1960;1: Gibbs FA, Gibbs EL. Age factor in epilepsy. N Engl J Med 1963; 269: Trojaborg W. Focal spike discharges in children, a longitudinal study. Acta Pediatr Scand Suppl 1966;168: Kellaway P. The incidences, significance and natural history of

8 442 M. OHTSU ET AL. spike foci in children. In: Henry CE, ed. Current clinical neurophysiology. Amsterdam: Elsevier:North-Holland, 1980: Hughes JR. Long-term clinical and EEG changes in patients with epilepsy. Arch Neurol 1985;42: Konishi T, Naganuma Y, Hongou K, et al. Changes in EEG foci with age in childhood partial epilepsies. Clin Electroencephalogr 1994;25: Parmeggiani L, Guerrini R. Idiopathic partial epilepsy: electronic demonstration of a prolonged seizure with sequential rolandic and occipital involvement: seizure spread due to regional susceptibility? Epileptic Disord 1999;1: Lerman P. Benign partial epilepsy with centro-temporal spikes. In: Roger J, Dravet C, Bureau M, et al., eds. Epileptic syndromes in infancy, childhood and adolescence. 2nd ed. London: John Libbey Eurotext, 1992: Doose H, Baier WK. Benign partial epilepsy and related conditions: multifactorial pathogenesis with hereditary impairment of brain maturation. Eur J Pediatr 1989;149:152 8.