- Email: [email protected]
Topographic mapping and clinical analysis of benign childhood epilepsy with centrotemporal spikes
Brain & Development 20 (1998) 27–32
Original article
Topographic mapping and clinical analysis of benign childhood epilepsy with centrotemporal spikes Min-Lan Tsai a ,*, Kun-Long Hung b a
Department of Pediatrics, Show-Chwan Memorial Hospital, 542 Chung-Shan Road, Sec. 1, Changhua, Taiwan b Department of Pediatrics, Cathay General Hospital, Taipei, Taiwan Received 11 July 1997; revised version received 29 September 1997; accepted 13 October 1997
Abstract We studied the topographic mapping of the electroencephalography (EEG) of 47 children whose clinical history and course were compatible with typical benign childhood epilepsy with centrotemporal spikes (BCECT). Twenty-nine (62%) patients showed typical dipole fields, with a negative potential field in the centrotemporal region and a positive field in the frontal region. Eighteen children did not demonstrate the typical dipole field. Their non-dipole rolandic discharges were localized in small fields of centrotemporal region. The patients with dipole fields in BCECT had significantly less frequent seizures than patients without dipole fields. Twelve of the 47 patients with BCECT (26%) had more than one EEG focus. The clinical courses of patients with multiple foci were not worse than those of patients with a single focus. We conclude that EEG topographic mapping is helpful in identifying typical or atypical EEG topographic patterns in patients with clinically diagnosed BCECT. We also conclude that the presence of dipole field usually indicates a better clinical course of epilepsy and multiple foci do not mean a poor clinical course. 1998 Elsevier Science B.V. Keywords: Benign childhood epilepsy with centrotemporal spikes; Dipole; Benign rolandic epilepsy; Rolandic spikes
1. Introduction Benign childhood epilepsy with centrotemporal spikes (BCECT) is a common clinical syndrome of childhood, characterized clinically by easily controlled seizures affecting a normal child [1]. The electroencephalography (EEG) is characterized by rolandic spikes, which are distinct, monomorphic sharp waves with blunted peaks in the centrotemporal area or occasionally in other locations [2,3]. The diagnosis of BCECT is made on both clinical and electroencephalographic grounds: clinical criteria are the occurrence of seizures and the absence of neurologic deficits and the EEG criterion is the presence of rolandic spikes. Rolandic spikes in BCECT often show a bipolar field, with a negative potential field maximum at the midtemporocentral region with a simultaneous positive field at the * Corresponding author. Fax: +886 4 7276106; e-mail: [email protected]
0387-7604/98/$19.00 1998 Elsevier Science B.V. All rights reserved PII S0387-7604 (97 )0 0089-2
bifrontal region [4,5]. Topographic studies of rolandic spikes have reported this dipole topography to be a consistent finding of BCECT [6,7]. However, in a quantitative topographic study of rolandic spikes, some children classified as having BCECT showed this dipole, but other children with BCECT had a unipolar field, with a negative potential field in the centrotemporal region [8]. In the present study, we classified two groups of BCECT patients with or without topographic dipoles by their interictal EEGs. The main purpose of this study was to analyze the topography of these discharges and their distribution in order to evaluate the relationship between EEG topography and the clinical prognosis in children with BCECT. Patients who had an EEG temporal-frontal dipole of their rolandic spikes were compared to patients who did not have an EEG dipole. Their clinical course and presentations, response to antiepileptic drugs, frequency of seizures and topographic features were compared. The location and number of foci on their serial EEGs were also analyzed.
28
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
2. Patients and methods We reviewed the EEGs from 47 children (30 boys and 17 girls) with the following criteria of typical benign rolandic epilepsy of childhood: (i) history of typical seizures, (ii) age of onset seizures between 4 and 12 years, (iii) normal neurologic examination and intelligence, (iv) no evidence of brain damage, (v) good response to antiepileptic therapy, and (vi) EEG evidence of a uni- or bilateral focus in the midtemporal or rolandic area activated by sleep and normal background activities. The image studies of the brain (CT or MRI) of these patients were all normal. Children with neurologic deficits, learning disabilities or organic brain lesions were excluded. All the patients were followed at our clinics for more than 2 years (range 2–10 years, mean 4.2 years). The medical records of the 47 children who met the EEG and clinical criteria were then reviewed. Their seizure type, presence of nocturnal partial or secondary generalized seizures, seizure frequency and response to anticonvulsants were also recorded. All patients had routine waking and sleep EEG recordings using a 21-channel digital EEG machine (Biologic System, Mundelein, ILL). An international 10–20 electrode placement system was used. The recording time was at least 20 min. Data were recorded on common reference electrodes on the earlobes. No contamination was found in the data recorded and results were also confirmed by comparison with a bipolar montage. Records of EEG activity stored on optical discs served as a source for our patients. Analog to digital conversion using 8 bits precision at a 250 Hz sampling rate per set of 21 channels was carried out. The ‘Brain Atlas Reader’ (Biologic System Corp., version 2.5, model 594, 1993) software was used for brain mapping and topographic analysis.
A dipole spike was interpreted only if the discharge was of a negative polarity in the midtemporal region with simultaneous positive field in the ipsilateral frontal region and the field was present on at least 80% of EEG discharges [9]. The selected spikes represented the same focus and their distributions were all identical. At least 20 individual spikes were selected and averaged for the topographic mapping. Each record was analyzed by digital EEGs to identify: (i) number and location of foci, (ii) topographic mapping of the discharges, and (iii) the presence of horizontal dipoles. The sequential topographic mapping was also performed on a referential montage to the ipsilateral ear, including the ascending phase, the negative maximum and the descending phase, from 64 ms before to 56 ms after the maximum of the spike. These were analyzed by means of sequential topographic mapping with steps of 4 ms [10]. We also evaluated the EEG focus of each recording and their relation to the clinical history and course. Data were analyzed using Fisher’s exact test or the chisquare statistical method for comparison of each parameter between the dipole and non-dipole groups of patients according to their EEG mappings.
3. Results There were 47 patients, 30 boys (63.8%) and 17 girls (36.2%) with a mean age of 9.1 years (range 4 years 5 months to 13 years 6 months) at the time of diagnosis. The 47 children, who all met the electroclinical criteria of BCECT, were divided into two groups: (i) children with a typical dipole field on EEG (n = 29) and (ii) children without a dipole field on EEG (n = 18).
Fig. 1. EEG of referential and bipolar montage of a typical dipole in a 12-year-old boy with benign childhood epilepsy with centrotemporal spikes. The spikes at the right centrotemporal area are negative and simultaneous positivity is at the bifrontal and midline areas.
29
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
Fig. 2. Topographic voltage mapping of the typical dipole field, which demonstrates the positive field at the right centrotemporal area and negative at the frontal and midline areas in the same patient as in Fig. 1. Fig. 3. Topographic analysis by means of sequential mapping of the average rolandic spikes in the same patient as in Fig. 1. Fig. 4. Topographic mapping of the non-dipole discharge the negative field at the right centrotemporal area in a 9-year-old boy with a clinical diagnosis of benign childhood epilepsy with centrotemporal spikes.
3.1. Topographic mapping of rolandic spikes In patients with a stereotypic dipole field by voltage mapping, the dipoles were situated at the rolandic area, i.e., the centrotemporal strip, with the most negative field at the midtemporal or central region and the positive field anteriorly involving the bifrontal regions and spreading to the central midline (Figs. 1 and 2). Fig. 3 demonstrates the sequential mapping of rolandic spikes with dipoles. In
patients without a dipole field, only the most negative field located at the central or temporal area was without a positive field at frontal or other areas (Fig. 4). 3.2. Comparison of clinical features in both groups Clinical histories including atypical generalized seizure patterns, presence of nocturnal seizures, whether EEG discharges were activated by sleep, and family history of feb-
Table 1 Clinical presentation and family history between dipole (group 1) and non-dipole (group 2) groups
Group 1 (n = 29) Group 2 (n = 18)
Nocturnal seizures (n = 36)
EEG discharges activated by sleep (n = 38)
GS(+) (n = 17)
FC history (n = 7)
Family history of epilepsy (n = 8)
Family history of FC (n = 4)
24 (83%) 12 (67%)
24 (83%) 14 (78%)
6 (21%) 11 (61%)*
5 (20%) 2 (11%)
7 (28%) 1 (5%)
2 (8%) 2 (11%)
FC, febrile convulsions; GS, generalized seizures. *P , 0.05.
30
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
Table 2
Table 4
Numbers of EEG foci between dipole (group 1) and non-dipole (group 2) groups
Seizure frequency in dipole (group 1) and non-dipole groups (group 2)
Group 1 (n = 29) Group 2 (n = 18)
Single focus (n = 35)
Two foci (n = 11)
Three foci (n = 1)
19 (66%) 16 (89%)*
9 (31%) 2 (11%)
1 (3%) 0
*P = 0.09, Fisher’s exact test.
3.3. Significance of EEG findings and location of focus Of 47 children, 35 had a single focus and 11 patients had two independent foci (one from each hemisphere in most patients) and only one patient had three foci. In the nondipole group, 16 (89%) patients had a single EEG focus and two (11%) had more than one focus. The non-dipole group had a higher percentage of children with a single focus than the dipole group in which 19 (66%) patients had a single focus and 10 (34%) had more than one focus. The differences were not statistically significant (P = 0.09, Fisher’s exact test; Table 2). In 12 patients with two or more foci, eight (67%) cases had bilateral independent central or temporal foci, four (33%) had the second or the third focus beyond the centrotemporal area, i.e., two cases with foci at the parietal areas, one case with the focus at the inferior frontal-posterior temporal areas and one case with independent occipital spikes. In patients with more than one focus, their seizure frequency was not higher than patients with a single EEG focus (Table 3). Table 3 The relation of seizure frequency and number of foci Seizure frequency more than five times per year (n = 6)*
Seizure frequency less than five times per year (n = 41)
5 1 0
31 10 1
*P = 0.76, Fisher’s exact test.
Seizure frequency 1–5 times/year (n = 41)
1 (3%) 5 (28%)
28 (97%)* 13 (72%)
*P , 0.05.
rile convulsions and epilepsy in each group are listed in Table 1. Of 47 children with BCECT, 36 patients had only nocturnal seizures, 17 patients had generalized clonic seizure events, seven patients had a febrile convulsion history, eight patients had a family history of epilepsy and four had a family history of febrile convulsions. There were no statistically significant differences between dipole and nondipole groups as to presence of nocturnal seizures, whether the discharges were activated by sleep, previous history of febrile convulsions, family history of febrile convulsions and epilepsy. However, the non-dipole group had a statistically significant higher incidence of generalized seizure patterns compared with the dipole group (x2 = 6.21, P , 0.05).
Single focus (n = 35) Two foci (n = 11) Three foci (n = 1)
Group 1 (n = 29) Group 2 (n = 18)
Seizure frequency >5 times/year (n = 6)
3.4. Clinical course and seizure frequency between dipole and non-dipole groups Only one patient in the dipole group had seizures more than five times per year whereas five patients in the nondipole group had seizures more than five times a year. The dipole group had a statistically significant lower frequency of seizures than the non-dipole group (P , 0.05, Fisher’s exact test; Table 4).
4. Discussion The syndrome of BCECT was based originally on EEG features, especially the centrotemporal location of the discharges. Gibbs and Gibbs [4] described 120 children with a single spike focus in the midtemporal area. They concluded that midtemporal epilepsy was the most common focal epilepsy in children and that it had a good prognosis, unlike psychomotor epilepsy. The epileptiform discharges in this syndrome consist of stereotypic central and/or midtemporal spikes in a normal background [2,5,11]. The spikes can be unilateral or bilateral and independent. Most patients had a typical appearance of sharp waves activated by drowsiness and sleep. Rolandic spikes typically have a horizontal dipole [6,12]. Topographic EEG analysis with voltage mapping of rolandic spikes in this study demonstrated maximum negativity of spikes over the central or midtemporal electrodes and positivity over the frontal area often spreading to midline. Some patients with BCECT did not show dipoles in their EEG topography [10]. The precise location and mechanism of the generator of functional rolandic foci is still unknown [6]. Gregory and Wong assumed that a single generator oriented tangentially to the surface of the scalp and located in the lower rolandic region was the source of this dipole field [6,13]. The proposed site of origin of these potentials is either the base of the rolandic fissure or in the sylvian fissure (including the insula) [14]. The discharges may arise in the depths of sylvian fissure involving folded cortical areas. This occurs in such a way that the negative component of the discharges is concealed from the scalp electrodes. Thus, a relative positivity is recorded on the surface. This represents a dipole reversal of interictal discharges [14]. Quantitative study of sequential mapping of rolandic spikes
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
revealed two different topographic patterns, a pattern of stationary potential fields and a pattern of non-stationary potential fields [10]. In our series, 62% of the children presented with typical dipole fields in their EEG topographic mapping. The non-dipole group had a significantly higher incidence of secondary generalized seizures. No significant difference between the dipole and non-dipole groups was seen in their number of foci, presence of nocturnal seizures, previous history of febrile convulsions or family history of epilepsy and febrile convulsions. Children with BCECT may have spikes in areas other than the central and midtemporal regions [15]. Some investigators report that children with EEG foci located beyond the midtemporal or central area (such as the posterior temporal, or inferior frontal) had more frequent seizures than patients with typical locations on EEG topography [4]. However, Drury et al. reported that the EEG foci located exclusively in the centrotemporal area is not a criteria or prognostic factor in children with BCECT [3]. In this study, 28% of patients had two or more foci on their EEGs; they did not have an increase in seizure frequency during the course of followup. Multiple foci did not mean a poor prognosis in children with BCECT. This study demonstrated that patients with a typical dipole field usually have a better clinical course, and a lower frequency of seizures even though the two groups of patients both fit the clinical criteria of BCECT and the neurologic examinations were all normal. Temporal-frontal dipole discharges are associated with a lower incidence of clinical atypical seizure patterns than non-dipole rolandic spikes. These dipole discharges may present as a benign functional focus [9]. Many reports have studied patients with rolandic spikes, and some studies included ‘organic groups’ (e.g., cerebral palsy). Rolandic spikes also occur as an EEG phenomenon with a variety of other clinical syndromes [8,9]. In addition to the inheritance of an EEG trait for rolandic spikes and BCECT, there may be a hereditary susceptibility to epilepsy in the family of a neurologically normal child with rolandic spikes [8]. As in all idiopathic epilepsies, genetic factors may play a role in BCECT [16]. An autosomal dominant gene with age-dependent penetrance has been proposed [11,17,18]. The rolandic spike is found in 34% of siblings of children with BCECT, but only 15% of siblings suffer from seizures [19]. In this study, the percentage of positive family history of epilepsy was higher in dipole group but without statistical significance, indicating genetic factors may play a major role in patients with BCECT with an EEG dipole. In conclusion, we clarified the importance of EEG analysis in children with the diagnosis of BCECT. Topographic mapping can demonstrate the location of the dipole field and the location of foci, although it can also be recognized by conventional EEG recordings. Typical dipole fields on EEG studies usually indicate a good prognosis,
31
but multiple foci do not mean a poor clinical course in children with BCECT. Identification of this dipole may give us further information about children with a clinical diagnosis of BCECT.
Acknowledgements We thank Ms. Chau Hsu and Ms. Chu-Ju Hwang for their technical assistance and Ms. Chiu-Kuei Nien, MS for her statistical analysis.
References [1] Lermin P, Kivity S. Benign focal epilepsy of childhood. A followup study of 100 recovered patients. Arch Neurol 1975;32:261– 264. [2] Lu¨ders H, Lesser RP, Dinners DS, Morris HH. Benign focal epilepsy of childhood. In: Lu¨ders H, Lesser RP, editors. Epilepsy. Electroclinical syndromes. Berlin: Springer-Verlag, 1987:303– 346. [3] Drury I, Beydoun A. Benign partial epilepsy of childhood with monomorphic sharp waves in centrotemporal and other locations. Epilepsia 1991;32:662–667. [4] Gibbs FA, Gibbs EL. Jacksonian seizure and focal convulsions: midtemporal epileptic foci. In: Gibbs FA, Gibbs EL, editors. Atlas of electroencephalography, vol. 2. Boston, MA: Addison-Wesley, 1952:214–222. [5] Blum W. ‘Rolandic’ spikes. In: Blum W, editor. Atlas of pediatric electroencephalography. New York: Raven Press, 1982:144–145. [6] Gregory DL, Wong PKH. Topographic analysis of the centrotemporal discharges in benign rolandic epilepsy of childhood. Epilepsia 1984;25:705–711. [7] Wong PKH, Gregory D, Farrell K. Comparison of spike topography in typical and atypical benign rolandic epilepsy of childhood. Electroenceph clin Neurophysiol 1985;61:47. [8] van der Meij W, van Huffelen AC, Willemse J, Schenk-Rootlieb AJF, Meiners LC. Rolandic spikes in the interictal EEG of children: contribution to diagnosis, classification and prognosis of epilepsy. Dev Med Child Neurol 1992;34:893–903. [9] Gregory DL, Wong PKH. Clinical relevance of a dipole field in rolandic spikes. Epilepsia 1992;33:36–44. [10] van der Meij W, van Huffelen AC, Wieneke GH, Willemse J. Sequential EEG mapping may differentiate ‘epileptic’ from ‘nonepileptic’ rolandic spikes. Electroenceph clin Neurophysiol 1992;82:408–414. [11] Blom S, Heijbel J, Bergfors PG. Benign epilepsy of children with centrotemporal spike foci. Prevalence and follow up study of 41 patients. Epilepsia 1972;13:609–619. [12] Yoshinaga H, Amano R, Oka E, Ohatara S. Dipole tracing in childhood epilepsy with special reference to rolandic epilepsy. Brain Topogr 1992;4:193–199. [13] Wong PKH, Gregory DL. Rolandic dipole discharges in children. Am J EEG Technol 1988;28:243–250. [14] Gutierrez AR, Brick JF, Bodensteiner J. Dipole reversal: an ictal feature of benign partial epilepsy with centrotemporal spikes. Epilepsia 1990;31:544–548. [15] Holms GL. Benign focal epilepsies of childhood. Epilepsia 1993;34:S49–S61. [16] Ottman R, Annerger JF, Risch N, Hauser A, Susser M. Relations of genetic and environmental factors in the etiology of epilepsy. Ann Neurol 1996;39:442–449.
32
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
[17] Bray PF, Wiser WC, Wood C, Pusey SB. Hereditary characteristics of familiar temporo-central focal epilepsy. Pediatrics 1965;36:207– 211. [18] Blom S, Heijbel J. Benign epilepsy of childhood with centrotemporal spikes. In: Beck-Mannagetta G, Anderson VE, Doose H,
Janz D, editors. Genetics of the epilepsies. Berlin: Springer, 1989:67–72. [19] Heijbel J, Blom S, Rasmuson M. Benign epilepsy of children with centro-temporal EEG foci: a genetic study. Epilepsia 1975;16:285– 293.
Original article
Topographic mapping and clinical analysis of benign childhood epilepsy with centrotemporal spikes Min-Lan Tsai a ,*, Kun-Long Hung b a
Department of Pediatrics, Show-Chwan Memorial Hospital, 542 Chung-Shan Road, Sec. 1, Changhua, Taiwan b Department of Pediatrics, Cathay General Hospital, Taipei, Taiwan Received 11 July 1997; revised version received 29 September 1997; accepted 13 October 1997
Abstract We studied the topographic mapping of the electroencephalography (EEG) of 47 children whose clinical history and course were compatible with typical benign childhood epilepsy with centrotemporal spikes (BCECT). Twenty-nine (62%) patients showed typical dipole fields, with a negative potential field in the centrotemporal region and a positive field in the frontal region. Eighteen children did not demonstrate the typical dipole field. Their non-dipole rolandic discharges were localized in small fields of centrotemporal region. The patients with dipole fields in BCECT had significantly less frequent seizures than patients without dipole fields. Twelve of the 47 patients with BCECT (26%) had more than one EEG focus. The clinical courses of patients with multiple foci were not worse than those of patients with a single focus. We conclude that EEG topographic mapping is helpful in identifying typical or atypical EEG topographic patterns in patients with clinically diagnosed BCECT. We also conclude that the presence of dipole field usually indicates a better clinical course of epilepsy and multiple foci do not mean a poor clinical course. 1998 Elsevier Science B.V. Keywords: Benign childhood epilepsy with centrotemporal spikes; Dipole; Benign rolandic epilepsy; Rolandic spikes
1. Introduction Benign childhood epilepsy with centrotemporal spikes (BCECT) is a common clinical syndrome of childhood, characterized clinically by easily controlled seizures affecting a normal child [1]. The electroencephalography (EEG) is characterized by rolandic spikes, which are distinct, monomorphic sharp waves with blunted peaks in the centrotemporal area or occasionally in other locations [2,3]. The diagnosis of BCECT is made on both clinical and electroencephalographic grounds: clinical criteria are the occurrence of seizures and the absence of neurologic deficits and the EEG criterion is the presence of rolandic spikes. Rolandic spikes in BCECT often show a bipolar field, with a negative potential field maximum at the midtemporocentral region with a simultaneous positive field at the * Corresponding author. Fax: +886 4 7276106; e-mail: [email protected]
0387-7604/98/$19.00 1998 Elsevier Science B.V. All rights reserved PII S0387-7604 (97 )0 0089-2
bifrontal region [4,5]. Topographic studies of rolandic spikes have reported this dipole topography to be a consistent finding of BCECT [6,7]. However, in a quantitative topographic study of rolandic spikes, some children classified as having BCECT showed this dipole, but other children with BCECT had a unipolar field, with a negative potential field in the centrotemporal region [8]. In the present study, we classified two groups of BCECT patients with or without topographic dipoles by their interictal EEGs. The main purpose of this study was to analyze the topography of these discharges and their distribution in order to evaluate the relationship between EEG topography and the clinical prognosis in children with BCECT. Patients who had an EEG temporal-frontal dipole of their rolandic spikes were compared to patients who did not have an EEG dipole. Their clinical course and presentations, response to antiepileptic drugs, frequency of seizures and topographic features were compared. The location and number of foci on their serial EEGs were also analyzed.
28
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
2. Patients and methods We reviewed the EEGs from 47 children (30 boys and 17 girls) with the following criteria of typical benign rolandic epilepsy of childhood: (i) history of typical seizures, (ii) age of onset seizures between 4 and 12 years, (iii) normal neurologic examination and intelligence, (iv) no evidence of brain damage, (v) good response to antiepileptic therapy, and (vi) EEG evidence of a uni- or bilateral focus in the midtemporal or rolandic area activated by sleep and normal background activities. The image studies of the brain (CT or MRI) of these patients were all normal. Children with neurologic deficits, learning disabilities or organic brain lesions were excluded. All the patients were followed at our clinics for more than 2 years (range 2–10 years, mean 4.2 years). The medical records of the 47 children who met the EEG and clinical criteria were then reviewed. Their seizure type, presence of nocturnal partial or secondary generalized seizures, seizure frequency and response to anticonvulsants were also recorded. All patients had routine waking and sleep EEG recordings using a 21-channel digital EEG machine (Biologic System, Mundelein, ILL). An international 10–20 electrode placement system was used. The recording time was at least 20 min. Data were recorded on common reference electrodes on the earlobes. No contamination was found in the data recorded and results were also confirmed by comparison with a bipolar montage. Records of EEG activity stored on optical discs served as a source for our patients. Analog to digital conversion using 8 bits precision at a 250 Hz sampling rate per set of 21 channels was carried out. The ‘Brain Atlas Reader’ (Biologic System Corp., version 2.5, model 594, 1993) software was used for brain mapping and topographic analysis.
A dipole spike was interpreted only if the discharge was of a negative polarity in the midtemporal region with simultaneous positive field in the ipsilateral frontal region and the field was present on at least 80% of EEG discharges [9]. The selected spikes represented the same focus and their distributions were all identical. At least 20 individual spikes were selected and averaged for the topographic mapping. Each record was analyzed by digital EEGs to identify: (i) number and location of foci, (ii) topographic mapping of the discharges, and (iii) the presence of horizontal dipoles. The sequential topographic mapping was also performed on a referential montage to the ipsilateral ear, including the ascending phase, the negative maximum and the descending phase, from 64 ms before to 56 ms after the maximum of the spike. These were analyzed by means of sequential topographic mapping with steps of 4 ms [10]. We also evaluated the EEG focus of each recording and their relation to the clinical history and course. Data were analyzed using Fisher’s exact test or the chisquare statistical method for comparison of each parameter between the dipole and non-dipole groups of patients according to their EEG mappings.
3. Results There were 47 patients, 30 boys (63.8%) and 17 girls (36.2%) with a mean age of 9.1 years (range 4 years 5 months to 13 years 6 months) at the time of diagnosis. The 47 children, who all met the electroclinical criteria of BCECT, were divided into two groups: (i) children with a typical dipole field on EEG (n = 29) and (ii) children without a dipole field on EEG (n = 18).
Fig. 1. EEG of referential and bipolar montage of a typical dipole in a 12-year-old boy with benign childhood epilepsy with centrotemporal spikes. The spikes at the right centrotemporal area are negative and simultaneous positivity is at the bifrontal and midline areas.
29
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
Fig. 2. Topographic voltage mapping of the typical dipole field, which demonstrates the positive field at the right centrotemporal area and negative at the frontal and midline areas in the same patient as in Fig. 1. Fig. 3. Topographic analysis by means of sequential mapping of the average rolandic spikes in the same patient as in Fig. 1. Fig. 4. Topographic mapping of the non-dipole discharge the negative field at the right centrotemporal area in a 9-year-old boy with a clinical diagnosis of benign childhood epilepsy with centrotemporal spikes.
3.1. Topographic mapping of rolandic spikes In patients with a stereotypic dipole field by voltage mapping, the dipoles were situated at the rolandic area, i.e., the centrotemporal strip, with the most negative field at the midtemporal or central region and the positive field anteriorly involving the bifrontal regions and spreading to the central midline (Figs. 1 and 2). Fig. 3 demonstrates the sequential mapping of rolandic spikes with dipoles. In
patients without a dipole field, only the most negative field located at the central or temporal area was without a positive field at frontal or other areas (Fig. 4). 3.2. Comparison of clinical features in both groups Clinical histories including atypical generalized seizure patterns, presence of nocturnal seizures, whether EEG discharges were activated by sleep, and family history of feb-
Table 1 Clinical presentation and family history between dipole (group 1) and non-dipole (group 2) groups
Group 1 (n = 29) Group 2 (n = 18)
Nocturnal seizures (n = 36)
EEG discharges activated by sleep (n = 38)
GS(+) (n = 17)
FC history (n = 7)
Family history of epilepsy (n = 8)
Family history of FC (n = 4)
24 (83%) 12 (67%)
24 (83%) 14 (78%)
6 (21%) 11 (61%)*
5 (20%) 2 (11%)
7 (28%) 1 (5%)
2 (8%) 2 (11%)
FC, febrile convulsions; GS, generalized seizures. *P , 0.05.
30
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
Table 2
Table 4
Numbers of EEG foci between dipole (group 1) and non-dipole (group 2) groups
Seizure frequency in dipole (group 1) and non-dipole groups (group 2)
Group 1 (n = 29) Group 2 (n = 18)
Single focus (n = 35)
Two foci (n = 11)
Three foci (n = 1)
19 (66%) 16 (89%)*
9 (31%) 2 (11%)
1 (3%) 0
*P = 0.09, Fisher’s exact test.
3.3. Significance of EEG findings and location of focus Of 47 children, 35 had a single focus and 11 patients had two independent foci (one from each hemisphere in most patients) and only one patient had three foci. In the nondipole group, 16 (89%) patients had a single EEG focus and two (11%) had more than one focus. The non-dipole group had a higher percentage of children with a single focus than the dipole group in which 19 (66%) patients had a single focus and 10 (34%) had more than one focus. The differences were not statistically significant (P = 0.09, Fisher’s exact test; Table 2). In 12 patients with two or more foci, eight (67%) cases had bilateral independent central or temporal foci, four (33%) had the second or the third focus beyond the centrotemporal area, i.e., two cases with foci at the parietal areas, one case with the focus at the inferior frontal-posterior temporal areas and one case with independent occipital spikes. In patients with more than one focus, their seizure frequency was not higher than patients with a single EEG focus (Table 3). Table 3 The relation of seizure frequency and number of foci Seizure frequency more than five times per year (n = 6)*
Seizure frequency less than five times per year (n = 41)
5 1 0
31 10 1
*P = 0.76, Fisher’s exact test.
Seizure frequency 1–5 times/year (n = 41)
1 (3%) 5 (28%)
28 (97%)* 13 (72%)
*P , 0.05.
rile convulsions and epilepsy in each group are listed in Table 1. Of 47 children with BCECT, 36 patients had only nocturnal seizures, 17 patients had generalized clonic seizure events, seven patients had a febrile convulsion history, eight patients had a family history of epilepsy and four had a family history of febrile convulsions. There were no statistically significant differences between dipole and nondipole groups as to presence of nocturnal seizures, whether the discharges were activated by sleep, previous history of febrile convulsions, family history of febrile convulsions and epilepsy. However, the non-dipole group had a statistically significant higher incidence of generalized seizure patterns compared with the dipole group (x2 = 6.21, P , 0.05).
Single focus (n = 35) Two foci (n = 11) Three foci (n = 1)
Group 1 (n = 29) Group 2 (n = 18)
Seizure frequency >5 times/year (n = 6)
3.4. Clinical course and seizure frequency between dipole and non-dipole groups Only one patient in the dipole group had seizures more than five times per year whereas five patients in the nondipole group had seizures more than five times a year. The dipole group had a statistically significant lower frequency of seizures than the non-dipole group (P , 0.05, Fisher’s exact test; Table 4).
4. Discussion The syndrome of BCECT was based originally on EEG features, especially the centrotemporal location of the discharges. Gibbs and Gibbs [4] described 120 children with a single spike focus in the midtemporal area. They concluded that midtemporal epilepsy was the most common focal epilepsy in children and that it had a good prognosis, unlike psychomotor epilepsy. The epileptiform discharges in this syndrome consist of stereotypic central and/or midtemporal spikes in a normal background [2,5,11]. The spikes can be unilateral or bilateral and independent. Most patients had a typical appearance of sharp waves activated by drowsiness and sleep. Rolandic spikes typically have a horizontal dipole [6,12]. Topographic EEG analysis with voltage mapping of rolandic spikes in this study demonstrated maximum negativity of spikes over the central or midtemporal electrodes and positivity over the frontal area often spreading to midline. Some patients with BCECT did not show dipoles in their EEG topography [10]. The precise location and mechanism of the generator of functional rolandic foci is still unknown [6]. Gregory and Wong assumed that a single generator oriented tangentially to the surface of the scalp and located in the lower rolandic region was the source of this dipole field [6,13]. The proposed site of origin of these potentials is either the base of the rolandic fissure or in the sylvian fissure (including the insula) [14]. The discharges may arise in the depths of sylvian fissure involving folded cortical areas. This occurs in such a way that the negative component of the discharges is concealed from the scalp electrodes. Thus, a relative positivity is recorded on the surface. This represents a dipole reversal of interictal discharges [14]. Quantitative study of sequential mapping of rolandic spikes
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
revealed two different topographic patterns, a pattern of stationary potential fields and a pattern of non-stationary potential fields [10]. In our series, 62% of the children presented with typical dipole fields in their EEG topographic mapping. The non-dipole group had a significantly higher incidence of secondary generalized seizures. No significant difference between the dipole and non-dipole groups was seen in their number of foci, presence of nocturnal seizures, previous history of febrile convulsions or family history of epilepsy and febrile convulsions. Children with BCECT may have spikes in areas other than the central and midtemporal regions [15]. Some investigators report that children with EEG foci located beyond the midtemporal or central area (such as the posterior temporal, or inferior frontal) had more frequent seizures than patients with typical locations on EEG topography [4]. However, Drury et al. reported that the EEG foci located exclusively in the centrotemporal area is not a criteria or prognostic factor in children with BCECT [3]. In this study, 28% of patients had two or more foci on their EEGs; they did not have an increase in seizure frequency during the course of followup. Multiple foci did not mean a poor prognosis in children with BCECT. This study demonstrated that patients with a typical dipole field usually have a better clinical course, and a lower frequency of seizures even though the two groups of patients both fit the clinical criteria of BCECT and the neurologic examinations were all normal. Temporal-frontal dipole discharges are associated with a lower incidence of clinical atypical seizure patterns than non-dipole rolandic spikes. These dipole discharges may present as a benign functional focus [9]. Many reports have studied patients with rolandic spikes, and some studies included ‘organic groups’ (e.g., cerebral palsy). Rolandic spikes also occur as an EEG phenomenon with a variety of other clinical syndromes [8,9]. In addition to the inheritance of an EEG trait for rolandic spikes and BCECT, there may be a hereditary susceptibility to epilepsy in the family of a neurologically normal child with rolandic spikes [8]. As in all idiopathic epilepsies, genetic factors may play a role in BCECT [16]. An autosomal dominant gene with age-dependent penetrance has been proposed [11,17,18]. The rolandic spike is found in 34% of siblings of children with BCECT, but only 15% of siblings suffer from seizures [19]. In this study, the percentage of positive family history of epilepsy was higher in dipole group but without statistical significance, indicating genetic factors may play a major role in patients with BCECT with an EEG dipole. In conclusion, we clarified the importance of EEG analysis in children with the diagnosis of BCECT. Topographic mapping can demonstrate the location of the dipole field and the location of foci, although it can also be recognized by conventional EEG recordings. Typical dipole fields on EEG studies usually indicate a good prognosis,
31
but multiple foci do not mean a poor clinical course in children with BCECT. Identification of this dipole may give us further information about children with a clinical diagnosis of BCECT.
Acknowledgements We thank Ms. Chau Hsu and Ms. Chu-Ju Hwang for their technical assistance and Ms. Chiu-Kuei Nien, MS for her statistical analysis.
References [1] Lermin P, Kivity S. Benign focal epilepsy of childhood. A followup study of 100 recovered patients. Arch Neurol 1975;32:261– 264. [2] Lu¨ders H, Lesser RP, Dinners DS, Morris HH. Benign focal epilepsy of childhood. In: Lu¨ders H, Lesser RP, editors. Epilepsy. Electroclinical syndromes. Berlin: Springer-Verlag, 1987:303– 346. [3] Drury I, Beydoun A. Benign partial epilepsy of childhood with monomorphic sharp waves in centrotemporal and other locations. Epilepsia 1991;32:662–667. [4] Gibbs FA, Gibbs EL. Jacksonian seizure and focal convulsions: midtemporal epileptic foci. In: Gibbs FA, Gibbs EL, editors. Atlas of electroencephalography, vol. 2. Boston, MA: Addison-Wesley, 1952:214–222. [5] Blum W. ‘Rolandic’ spikes. In: Blum W, editor. Atlas of pediatric electroencephalography. New York: Raven Press, 1982:144–145. [6] Gregory DL, Wong PKH. Topographic analysis of the centrotemporal discharges in benign rolandic epilepsy of childhood. Epilepsia 1984;25:705–711. [7] Wong PKH, Gregory D, Farrell K. Comparison of spike topography in typical and atypical benign rolandic epilepsy of childhood. Electroenceph clin Neurophysiol 1985;61:47. [8] van der Meij W, van Huffelen AC, Willemse J, Schenk-Rootlieb AJF, Meiners LC. Rolandic spikes in the interictal EEG of children: contribution to diagnosis, classification and prognosis of epilepsy. Dev Med Child Neurol 1992;34:893–903. [9] Gregory DL, Wong PKH. Clinical relevance of a dipole field in rolandic spikes. Epilepsia 1992;33:36–44. [10] van der Meij W, van Huffelen AC, Wieneke GH, Willemse J. Sequential EEG mapping may differentiate ‘epileptic’ from ‘nonepileptic’ rolandic spikes. Electroenceph clin Neurophysiol 1992;82:408–414. [11] Blom S, Heijbel J, Bergfors PG. Benign epilepsy of children with centrotemporal spike foci. Prevalence and follow up study of 41 patients. Epilepsia 1972;13:609–619. [12] Yoshinaga H, Amano R, Oka E, Ohatara S. Dipole tracing in childhood epilepsy with special reference to rolandic epilepsy. Brain Topogr 1992;4:193–199. [13] Wong PKH, Gregory DL. Rolandic dipole discharges in children. Am J EEG Technol 1988;28:243–250. [14] Gutierrez AR, Brick JF, Bodensteiner J. Dipole reversal: an ictal feature of benign partial epilepsy with centrotemporal spikes. Epilepsia 1990;31:544–548. [15] Holms GL. Benign focal epilepsies of childhood. Epilepsia 1993;34:S49–S61. [16] Ottman R, Annerger JF, Risch N, Hauser A, Susser M. Relations of genetic and environmental factors in the etiology of epilepsy. Ann Neurol 1996;39:442–449.
32
M.-L. Tsai, K.-L. Hung / Brain & Development 20 (1998) 27–32
[17] Bray PF, Wiser WC, Wood C, Pusey SB. Hereditary characteristics of familiar temporo-central focal epilepsy. Pediatrics 1965;36:207– 211. [18] Blom S, Heijbel J. Benign epilepsy of childhood with centrotemporal spikes. In: Beck-Mannagetta G, Anderson VE, Doose H,
Janz D, editors. Genetics of the epilepsies. Berlin: Springer, 1989:67–72. [19] Heijbel J, Blom S, Rasmuson M. Benign epilepsy of children with centro-temporal EEG foci: a genetic study. Epilepsia 1975;16:285– 293.