Clinical Neurophysiology 118 (2007) 1369–1376 www.elsevier.com/locate/clinph
Focal epileptiform activity described by a large computerised EEG database H. Aurlien
a,* a
, J.H. Aarseth b, I.O. Gjerde c, B. Karlsen a, H. Skeidsvoll a, N.E. Gilhus
d,c
Section of Clinical Neurophysiology, Department of Neurology, Haukeland University Hospital, Bergen, Norway b National MS Registry, Department of Neurology, Haukeland University Hospital, Bergen, Norway c Department of Neurology, Haukeland University Hospital, Bergen, Norway d Department of Clinical Medicine, University of Bergen, Bergen, Norway Accepted 21 February 2007 Available online 23 April 2007
Abstract Objective: To study the age-related topographical tendency of expressing epileptiform activity, and the effect of focal epileptiform activity (FEA) on the general cortical brain activity. Methods: 1647 consecutive routine EEGs containing FEA were visually assessed for FEA location and asymmetry. Background activity was compared with that in normal EEGs from 3268 drug-free outpatient controls. Results: FEA localisation was age-related (p < 0.0005) except for the temporal region (p = 0.22) where FEA was found equally often in the young and the old. The left hemisphere was more prone to FEA (p = 0.018). The left–right asymmetry varied by age (p = 0.013). FEA asymmetry occurred most frequently in EEGs from patients older than 80 years, and least frequent in the age-group 20–39 years. FEA was associated with lower alpha rhythm (AR) frequencies (p = 0.0041) and higher AR amplitudes (p = 0.0023), as well as higher general background activity (GBA) amplitude (p < 0.0005), while GBA frequencies were the same (p = 0.96). Conclusions: Topographical localisation of FEA was age-dependent. There was an overall left dominance, but the side asymmetry was modest and varied by age. FEA was associated with changes in AR and GBA. Significance: The results demonstrate that FEA is associated with cerebral cortical dysfunction also distant from the epileptic focus. Ó 2007 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. Keywords: Electroencephalography; Databases; Alpha rhythm; Background activity; Asymmetry; Topography
1. Introduction Epilepsy is a clinical diagnosis, but electroencephalography (EEG) plays a major role in evaluating epilepsy (Flink et al., 2002). EEG provides a convenient and inexpensive way to demonstrate the physiological manifestations of abnormal cortical excitability that underlie epilepsy (Smith, 2005). Epilepsy is the single most studied patient diagnosis in nearly all EEG laboratories, and is the area in which EEG is of greatest clinical value (Binnie and Stefan, * Corresponding author. Address: Section of Clinical Neurophysiology, Department of Neurology, Haukeland University Hospital, 5021 Bergen, Norway. Tel.: +47 55975000; fax: +47 55975164. E-mail address:
[email protected] (H. Aurlien).
1999). EEG differentiates between epileptic and non-epileptic seizures, seizure types, epilepsy syndromes, focal or generalised epilepsies, and symptomatic or idiopathic epilepsies. Thereby EEG also facilitates the choice of antiepileptic medication and prediction of prognosis. In previous papers, we have described a system for categorization of digital EEG data in a computerised database, thus achieving accessibility of routine EEGs for research and quality control (Aurlien et al., 1999, 2004). Knowledge about the age-dependency of commonly studied EEG parameters is essential to understand the significance of epileptiform activity both clinically and scientifically. There are, however, only a few published studies concerning age correlated topographical localisation of focal epileptiform activity (FEA). Gibbs and Gibbs
1388-2457/$32.00 Ó 2007 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2007.02.027
1370
H. Aurlien et al. / Clinical Neurophysiology 118 (2007) 1369–1376
(1952) analysed crude topographical distribution of FEA more than half a century ago, but not in age-related categories. Koufen and Gast (1981) reported age-related distribution of focal EEG abnormalities, but not specified for FEA. Several authors have reported the left cerebral hemisphere to be more prone to FEA than the right (Dean et al., 1997; Labar et al., 2001; Holmes et al., 2001; Gatzonis et al., 2002; Doherty et al., 2002) but only a few have studied such asymmetries related to age (Doherty et al., 2003a,b). Abnormal EEG background activity has been described in patients with focal epilepsies using quantitative EEG techniques (Miyauchi et al., 1991; Farkas et al., 1992; Diaz et al., 1998; Drake et al., 1998; Braga et al., 2000). Similar findings have to our knowledge not been published using visual EEG analysis, which is the main EEG method for clinical use (Nuwer, 1997; Flink et al., 2002; Stokes et al., 2004). EEG-studies concerning epileptiform activity are usually limited to clinically well-defined epilepsy entities or syndromes. This study, in contrast, focuses on the epileptiform activity itself as obtained from a large unselected population of patients referred to EEG examination. Good clinical information is essential for the assessment of the significance of epileptiform activity in clinical practice (Binnie and Stefan, 1999; Sam and So, 2001). Nevertheless, looking at certain important EEG phenomena from an EEGer’s point of view, and including a large unselected population, provides an overview that may easily be lost if studying selected clinical entities only. The aim of this study was to study age-related topographical aspects of expressing epileptiform activity, and also the effect of FEA on the general cortical brain activity. Does the expression of epileptiform activity in different brain regions change during life? Is the left hemisphere more prone to FEA? Is FEA asymmetry age-dependent? Does FEA alter the EEG background activity?
was examined (N = 341). The time point for the last epileptic seizure was not registered in the database, but these elective, drug-free outpatients were very unlikely to have had a seizure within the last 24 h before registration. As a control material all first EEGs from drug-free outpatients without EEG pathology from the study period were chosen (N = 3268). The EEGs were described by one of 6 EEGers irrespective of patient categories. Three of these were certified, experienced EEGers, the other three were trainees under supervision. The agreement between the different EEGers as studied previously using the same method showed minor to moderate differences in absolute values, but always with the same trends for all EEGers (Aurlien et al., 2004). The EEGs were recorded on the digital EEG system NervusÒ. 22 electrodes were placed according to the international 10–20 system (Fp1, Fp2, Fpz, F3, F4, F7, F8, Fz, A1, A2, T3, T4, T5, T6, C1, C2, Cz, P3, P4, Pz, O1, and O2).
2. Methods
2.3. EEG interpretation
2.1. EEG recordings
The interpretation was divided into three main sections: Alpha rhythm (AR), general background activity
All EEGs recorded at Haukeland University Hospital from 01.03.2000 to 31.12.2005 were visually evaluated and described. This included 17723 EEGs from 12511 patients. Haukeland University Hospital recruits patients from a population of about 500 000, and is the only provider of EEG within this area. Long-term registrations, EEGs during WADA tests and Tilt tests were not included in this study. The first EEG containing FEA from each patient was included. The total material consisted of 1647 EEGs from 852 females and 795 males. Postictal activity as well as many drugs acting in the central nervous system affect the EEG background activity (Salinsky et al., 2003; Blume, 2006). To study the effect of FEA per se on the general background activity (GBA), the subgroup of all drug-free outpatient subjects
2.2. EEG database The EEG interpretations were structured and categorized in an EEG database using software developed for the purpose (Aurlien et al., 1999). This software automatically collected patient demographic data and administrative test parameters such as name, patient ID, date of birth, patient address, referral doctor, test notes, patient notes, and medication from the hospital patient administrative system and the EEG administrative database. For each EEG, the EEGer set one or more ICD-10 diagnoses on basis of the doctor’s referral representing the reason for taking the EEG (Table 1). The system automatically registered all the patients’ ICD-10 clinical diagnoses from all hospital contacts. These were diagnoses set by the clinical doctors treating the patients.
Table 1 The 10 most common referral diagnoses for 1647 consecutive patients with FEA Diagnosis
N
% of patients
Epilepsy Encephalopathy Anoxic brain damage Syncope or collapse Encephalitis Somnolence, stupor or coma Brain tumour Stroke Headache Mental retardation
1513 40 35 25 20 19 18 12 12 11
91 2 2 2 1 1 1 1 1 1
Some patients had more than one diagnosis.
H. Aurlien et al. / Clinical Neurophysiology 118 (2007) 1369–1376
(GBA), and EEG events. AR was defined as the dominant posterior rhythm with frequency 8–13 Hz that was blocked or attenuated by eye opening (Chatrian et al., 1983). Alpha variant rhythm was defined as with AR characteristics but with frequency outside the 8–13 Hz alpha band (Chatrian et al., 1983). AR and alpha variant rhythm were analysed together and are referred to as AR. AR was evaluated for frequency, amplitude, reactivity, and asymmetry. Frequency and amplitude were measured for the occipital leads in a monopolar average montage in an EEG segment where AR was distinctly appearing. They were evaluated as a range with a minimum and a maximum value. The maximum value was used for AR frequency. AR amplitude was measured from peak to peak and calculated as the mean of the minimum and maximum measured value of the typical AR. Pathological AR was in addition described as a ‘pathological AR’ EEG event. Background activity was defined as any EEG activity representing the setting in which a given normal or abnormal pattern appeared and from which such a pattern was distinguished (Chatrian et al., 1983). AR was described separately. The remaining continuous general activities were defined as the GBA. Focal, asymmetric, and intermittent activities were not defined as part of the background activity. Such activities were described as EEG events. In addition examples of pathological AR and GBA were described as EEG events. The GBA was evaluated in a monopolar average montage without signs of drowsiness. To assure alertness the patients were instructed to open the eyes during parts of the test. If drowsy the patients were alerted with an auditive signal (knocking). In some cases, the patients were asked to perform calculations and repeat series of digits to assure alertness. GBA was described as one or more ranges of frequencies and amplitudes. The minimum value was used for GBA frequency as this value was assumed to be most relevant for EEG pathology (Gloor et al., 1977). GBA amplitude was measured from peak to peak and calculated as the mean of the upper and lower measured value of the typical GBA. Pathological GBA was defined as frequent or continuous generalized theta or delta activity with only moderate asymmetry. This was marked as an EEG event. EEG events were defined as any part of the EEG that the EEGer described separately from the frequency and amplitude measurement of the AR and GBA. An example of each type of EEG event was manually marked in the EEG editor with start and stop time. These events were automatically picked up by the description module, and the EEGer then classified them according to the American Standards for Testing and Materials (ASTM, 1994). The ASTM categories were divided hierarchically into four main branches: ‘epileptiform pathology’, ‘non-epileptiform pathology’, ‘normal findings/variants’, and ‘extra-cerebral activity’ (Westmoreland and Klass, 1990). Pathological EEGs were defined as having at least one EEG event from one of the first two branches.
1371
2.4. FEA The localisation of the EEG events was determined by clicking the traces where the activity occurred. The electrodes were divided into five regions; frontal, temporal, central, parietal, and occipital. Each event was automatically classified as general or focal. If three or more regions were affected at both sides and with not more than moderate asymmetry the event was classified as general. All other events were classified as focal. All EEGs with one or more focal events categorized as ‘epileptiform pathology’ were classified as having FEA, also if there were additional generalised events in the same EEG. Side asymmetry was automatically set for each event according to the electrodes involved. Values were set as 4/+4 (only left/right side), 3/+3 (marked left/right asymmetry), 2/+2 (moderate left/right asymmetry), 1/+1 (mild left/right asymmetry), or 0 (no asymmetry). These automatically set asymmetry values could be overruled by the EEGer. Asymmetric FEA was defined as FEA with amplitude or distribution predominance on either left or right side and was for each EEG calculated as the mean asymmetry value of all described FEA events. Side independent asymmetry was defined as more than mild asymmetry independent of lateralisation to the left or right side and was for each EEG calculated as the mean of the absolute values of asymmetry values of all described FEA events. 2.5. Statistics Chi-square test was used to test age dependency for the localisation of FEA in topographical regions as well as for FEA asymmetry and for association between the EEGers and asymmetry. Age-related amplitude and frequency variations were described using polynomial regression with age as the independent variable. By including nominal variables for gender, and other specific grouping parameters, we tested for differences in the respective groups. For each situation considered, we first fitted a polynomial model of order equal to the number of significant terms for age (p < 0.05). The individual nominal variables were entered into the model as well as significant interaction terms. To limit the number of figures presented, gender was not included in the figures of the estimated models even though it was a significant parameter in some of the models (Figs. 1 and 2). The polynomial model became unstable when including EEGs from patients older than 60 years because the number of registrations was too low. EEGs from patients above 60 years were therefore excluded from these analyses. To avoid the problem with multicollinearity, a matrix of orthonormal polynomials represented the basis for the polynomial regression. Several transformations were needed to achieve near normality for the error terms. The results are given for the back-transformed data. S-Plus 6.0 and SPSS 13.0 were used for the analyses.
1372
H. Aurlien et al. / Clinical Neurophysiology 118 (2007) 1369–1376 Controls FEA
60
10
Amplitude (V)
Frequency (Hz)
Controls FEA
8
50
40
6 30 0
10
20
30
40
50
60
0
10
20
Age (years) Controls FEA - Not medicated outpatients
40
50
60
Controls FEA - Not medicated outpatients
60
10
Amplitude (V)
Frequency (Hz)
30
Age (years)
8
50
40
6 30 0
10
20
30
40
50
60
0
10
20
30
40
50
60
Age (years)
Age (years)
Fig. 1. AR frequency (a) and AR amplitude (b) from all patients with FEA and from drug-free outpatient controls. AR frequency (c) and AR amplitude (d) from drug-free outpatients with FEA and from drug-free outpatient controls.
3. Results 3.1. Material description Mean age was 34.4 years, median 27.5 years. (Age distribution at test time for the patients with FEA is shown in electronic supplementary OnlineFig. 1.) The 10 most common referral diagnoses are listed in Table 1. For 1252 (76%) of the 1647 patients with FEA a clinical diagnosis of epilepsy was confirmed. Clinical diagnoses for the 395 patients with FEA but without a clinical diagnosis of epilepsy are listed in Table 2. For 17 patients (1%) no clinical diagnosis was available. 1055 patients (64%) received one or more drugs (electronic supplementary OnlineTable 1), whereas 498 (30%) were drug-free. For 94 individuals (6%) information regarding medication was missing. For 1035 patients with FEA (63%) and 2726 outpatient controls (83%) the alpha rhythm (AR) could be identified. In drug-free outpatients with FEA, AR was identified in 77% (263 of 341), while only 44% (238 of 542) of the medicated inpatients with FEA had identified AR. For 1632 patients with FEA (99%) and all outpatient control EEGs
the GBA could be identified. EEGs from patients over 60 years of age were excluded from the statistical analysis for AR and GBA because the numbers were too low for the analysis used (see Section 2.5). Thus the AR statistics included 829 patients with FEA of which 254 were drugfree outpatients, 100 were drug-free inpatients, and 2623 outpatient controls. GBA statistics included 1264 patients Table 2 The 10 most common brain-related clinical diagnoses in 395 patients with FEA, but without a clinically confirmed diagnosis of epilepsy Diagnosis
N
% of patients
Stroke Syncope or collapse Convulsions Dementia Depressive episode Coma Dizziness or giddiness Anoxic brain damage Brain tumour Headache
89 47 44 30 28 26 26 23 21 21
23 12 11 8 7 7 7 6 5 5
Some patients had more than one diagnosis.
H. Aurlien et al. / Clinical Neurophysiology 118 (2007) 1369–1376 Controls FEA
1373 Controls FEA
50
Amplitude (V)
Frequency (Hz)
8
6
4
2
40 30 20 10
0
0 0
10
20
30
40
50
60
0
10
20
30
40
50
60
Age (years)
Age (years)
Controls FEA - Not medicated outpatients
Controls FEA - Not medicated outpatients
50 40
Amplitude (V)
Frequency (Hz)
8
6
4
2
30 20 10 0
0 0
10
20
30
40
50
0
60
10
Age (years)
20
30
40
50
60
Age (years)
Fig. 2. GBA frequency (a) and GBA amplitude (b) from all patients with FEA and from drug-free outpatient controls. GBA frequency (c) and GBA amplitude (d) from drug-free outpatients with FEA and from drug-free outpatient controls.
70% Front Temp Par Centr Occ
100% 90% 80%
Left asymmetry No asymmetry Right asymmetry
60%
50%
70%
40%
60% 50%
30% 40% 30%
20%
20%
10%
10% 0% 0-9
10-19
20-29
30-39
40-49
50-59
60-69
70-79
>80
Age (years)
0% 00-04
05-10
10-14
15-19
20-39
40-59
60-79
>80
Age (Years)
Fig. 3. Age-related topographical distribution of FEA in 1647 consecutive patients with EEGs containing FEA. FEA could be located in more than one region.
Fig. 4. Side asymmetry of FEA in 1647 consecutive patients with EEGs containing FEA.
1374
H. Aurlien et al. / Clinical Neurophysiology 118 (2007) 1369–1376
with FEA of which 331 were drug-free outpatients, 148 drug-free inpatients, and 3155 outpatient controls. 131 EEGs (8%) were recorded after sleep deprivation with the patient both awake and asleep. Children referred for sleep deprived EEG did not sleep from 3 a.m. the previous night, while adults did not sleep the whole night before registration. No patients received drugs to induce sedation. Standard duration for a routine EEG was 20 min, and 60 min for EEG after sleep deprivation. 3.2. FEA topographical distribution Topographical distribution of FEA was age-dependent for all brain regions (p < 0.0005) except for the temporal (p = 0.17). Frontal FEA was more frequent in adults than in children and adolescents (Fig. 3). The percentage of EEGs with frontal FEA varied from 65% in children aged 0–9 years to 88% at 60–69 years. The percentage of EEGs with temporal FEA varied from 75% at age 0–9 years to 86% at 50–59 years. The percentage of EEGs with central, parietal, or occipital FEA decreased from maximal values of 61%, 61%, and 47%, respectively, in children 0–9 years to minimum values of 12–18% at 50–59 years (Fig. 3). The FEA was recorded from more than one brain region in 1369 EEGs (83%). 3.3. FEA asymmetry FEA lateralised more often to the left side of the brain compared to the right; 565 (34%) vs. 487 (30%) (p = 0.018). 595 EEGs (36%) had no side asymmetry. The left–right result did not vary significantly between the 6 EEGers (p = 0.18). There was still more left than right FEA asymmetry when only EEGs with completely unilateral FEA were included; 444 vs. 381 (27% vs 23%) (p = 0.031). Left and right FEA asymmetry varied significantly between age groups (p = 0.013) (Fig. 4). Also FEA asymmetry independent of left or right side varied between age groups (p < 0.0005). The relative risk for asymmetric FEA was highest in patients over the age of 80 years (94 asymmetric vs. 7 symmetric), and lowest at age 20–39 years (171 asymmetric vs. 80 symmetric) (electronic supplementary OnlineFig. 2). 3.4. FEA and AR The total group of patients with FEA had lower AR frequency (p < 0.0005) and higher amplitude (p < 0.0005) compared to the drug-free outpatient controls (Fig. 1a and b). There were 5 significant terms for age in the polynomial model for frequency and 2 in the model for amplitude. The subgroup of drug-free outpatients with FEA had lower AR frequency (p = 0.0041) and higher amplitude (p = 0.0023) compared to the outpatient controls (Fig. 1c and d). There were 5 significant terms for age in the polynomial model for frequency and 2 in the model for amplitude. Drug-free inpatients with FEA had lower AR
frequency compared to drug-free outpatients with FEA (p = 0.0046), while AR amplitude was the same (p = 0.46). There were 3 significant terms for age in the polynomial model for frequency and 1 in the model for amplitude. Patients over 60 years were excluded from these analyses due to too few EEGs. 3.5. FEA and GBA The total group of patients with FEA had lower GBA frequency (p < 0.0005) and higher amplitude (p < 0.0005) compared to the drug-free outpatient controls (Fig. 2a and b). There were 3 significant terms for age in the polynomial model for frequency and 6 in the model for amplitude. The subgroup of drug-free outpatients with FEA had higher amplitude (p < 0.0005) compared to the outpatient controls, while GBA frequency was the same (p = 0.96) (Fig. 2c and d). There were 3 significant terms for age in the polynomial model for frequency and 5 in the model for amplitude. Drug-free inpatients with FEA had lower GBA frequency (p = 0.0026), and similar amplitude (p = 0.41) compared to the drug-free outpatients with FEA. There were 3 significant terms for age in the polynomial model for frequency and 5 in the model for amplitude. Patients over 60 years were excluded from these analyses due to too few EEGs. 3.6. Non-FEA events The EEG from 162 of the 1647 patients (10%) contained generalised epileptiform activity in addition to FEA. 963 EEGs (58%) contained non-epileptiform pathology in addition to FEA. This pathology was focal in 656 EEGs (40%), generalised in 161 (10%), and both generalised and focal in 146 (9%). 4. Discussion This study shows an age dependent topographical localisation of FEA. The proportion of frontal FEA was higher among adults compared to children and adolescents, while the opposite was seen for parietal, central, and occipital FEA. Age-related changes in brain plasticity influence the pattern of seizure onset, spread and propagation velocity, but the processes underlying such variability are poorly understood (Weissinger et al., 2005). Age-related epilepsies, such as occipital epilepsies and benign partial epilepsy in childhood explain some of the observed difference. The small size of the head in children with shorter distance between the electrodes results in more overlap from the different regions when registered from the scalp surface, and therefore partly explains a more equal distribution of FEA between the brain regions in children. The frontal lobe volume is 3.5 times larger than the temporal and occipital lobes, while the parietal lobe is two times larger (DeCarli et al., 2005). The high frequency of FEA in the
H. Aurlien et al. / Clinical Neurophysiology 118 (2007) 1369–1376
small temporal lobe illustrates the marked tendency for epileptiform activity in this brain region. The amplitude maximum of spikes recorded from scalp EEG does not necessarily correspond directly to the underlying focus in the cerebral cortex due to neuronal propagation from distant areas as well as to peculiarities in the brain coverings such as skull holes (Torre et al., 1999). This has to be taken into consideration when applying scalp EEG recordings to clinical practice. In this study, we used standard scalp electrodes according to the 10–20 system, and without any extra ‘‘true temporal’’ electrodes. The benefit of such electrodes has been debated, but they might have increased the sensitivity for deep medial temporal discharges (Sperling and Guina, 2003; Blume, 2003). The left cerebral hemisphere was more prone to FEA, the difference, however, smaller than in previous studies (Dean et al., 1997; Labar et al., 2001; Gatzonis et al., 2002; Doherty et al., 2002). No right-sided asymmetry of FEA in children less than 5 years old was present, in contrast to a previous report (Doherty et al., 2003b). Reviewing all 36 papers published in 2005 classifying focal nonlesional epilepsies according to a left or right hemisphere focus, we counted more patients with a left than with a right focus; 734 vs. 629 (p = 0.005). For references, see electronic supplementary reference list. (English language papers only, patients operated for epilepsy and case reports excluded.) Our data showed significant variation in side asymmetry at different ages. The age distribution of the tested populations will therefore influence the result. This study recruited patients from the total population within our region, as there were no other EEG-providers, and therefore reflects the actual distribution of FEA. As focal semiology may trigger referral for EEG testing, dominant or non-dominant semiologies could influence who is evaluated (Doherty et al., 2003b). Such a potential bias should, however, not depend on the patients’ age in the way observed in our study. Holmes et al. (2001) found that only left-handed patients had more frequent FEA in their left hemisphere. We have no information about handedness in our patients. Other authors have addressed the possible bias that EEGers might be more aware of epileptiform activity in one of the hemispheres (Dean et al., 1997), but concluded that this was unlikely because this asymmetry was not found for non-epileptiform pathology (Doherty et al., 2002). Koufen and Gast (1981) did, however, find considerable left-sided asymmetry also for nonepileptiform pathology. Their data also showed significant difference between EEGers in assessment of left vs. right EEG pathology showing that EEGer bias can occur. In the present study, there was no difference in asymmetry assessment for the six different EEGers. Furthermore, some age groups had more right-sided FEA while others had more on the left. This contradicts EEGer bias. The effect of FEA on AR frequency and amplitude and GBA amplitude was not due to medication, as it occurred also in drug-free patients. Complicating medical disorders could influence AR and GBA. Drug-free inpatients and
1375
drug-free outpatients were therefore compared, as the inpatients were assumed to have more complicating disorders. Inpatients were found to have more affected AR and GBA frequency than outpatients. There was, however, still an association between FEA and AR frequencies, AR amplitude, and GBA amplitude, as observed in the clean group of drug-free outpatients with FEA. The association between FEA and EEG background parameters does, however, not prove a causal effect. These findings are consistent with previous studies using quantitative EEG (Miyauchi et al., 1991; Diaz et al., 1998; Braga et al., 2000), but have previously not been documented using visual EEG analysis. FEA influence on GBA frequency could not be confirmed in the present study after the effects of medication and the general condition of the patients were excluded. The visual impression of the GBA is a combination of frequencies and amplitudes. In a previous study we found an amplitude limit of 31 lV for the 4–7 Hz theta GBA between EEGs evaluated as normal and pathological (Aurlien et al., 2004). EEGs with higher amplitudes will be judged as slow compared to an EEG with the same frequency, but lower amplitude. Diffuse GBA slowing is related to cognitive dysfunction and mental deterioration, and is regarded as strongly suggestive of CNS pathology (Dustman et al., 1993). It is also a predictor of drug-resistant epilepsy in children (Ko and Holmes, 1999). Our study shows that FEA is not an isolated phenomenon, but part of a more widespread dysfunction of the brain cortex. A recent study supports this by showing altered cortical excitability distant to the epileptogenic zone in patients with focal epilepsy (Hamer et al., 2005). The stringent dichotomy between focal and generalised epilepsies has been a debated issue (Leutmezer et al., 2002). Focal epilepsies with a rapid generalisation, especially in frontal epilepsies, can appear as generalised in EEG. On the other hand, patients with well documented idiopathic epilepsies can have clear focal traits in EEG. The diagnosis of epilepsy and the further specific ILAE classification of epilepsies and epileptic syndromes are clinically based. This study included all the clinical ICD10 diagnoses set by the clinical physicians from these patients’ contacts with the hospital. These ICD10 diagnoses were, however, not sufficient to classify the patients according to the ILAE classification system. This study was primarily an EEG-study that focused on the epileptiform activity as such, and as obtained from a large unselected population of patients referred to EEG examination. We chose to include EEGs from all patients with FEA even if the EEG in addition contained generalised epileptiform activity. These patients could theoretically belong to either group; idiopathic epilepsy, remote symptomatic epilepsy or cryptogenic epilepsy; the diagnosis being clinically determined. After the introduction of quantitative EEG, research studies often favour this technology. However, parameters from quantitative EEG can usually not be directly compared with those obtained from visual analysis. Studies
1376
H. Aurlien et al. / Clinical Neurophysiology 118 (2007) 1369–1376
using visual EEG analysis are necessary to strengthen the scientific basis of this main EEG method for evaluation of patients with epilepsy in clinical practice. This study has shown that the tendency of expressing epileptiform activity for different brain regions is age-related, and that FEA is associated with altered AR and GBA suggesting cerebral cortical dysfunction distant from the FEA. Such findings should in our opinion be of interest both for educational purposes, and for the understanding of basic functions of the brain. Acknowledgement We are indebted to Geir Egil Eide, The Regional Competence Centre for Clinical Research, Bergen, Norway, for guidance in the statistical analysis. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.clinph. 2007.02.027. References ASTM. Table X2.5 EEG waveform/activity identifiers. 1994:123 Aurlien H, Gjerde IO, Aarseth JH, Eldoen G, Karlsen B, Skeidsvoll H, et al. EEG background activity described by a large computerized database. Clin Neurophysiol 2004;115:665–73. Aurlien H, Gjerde IO, Gilhus NE, Hovstad OG, Karlsen B, Skeidsvoll H. A new way of building a database of EEG findings. Clin Neurophysiol 1999;110:986–95. Binnie CD, Stefan H. Modern electroencephalography: its role in epilepsy management. Clin Neurophysiol 1999;110:1671–97. Blume WT. The necessity for sphenoidal electrodes in the presurgical evaluation of temporal lobe epilepsy – Con position. J Clin Neurophysiol 2003;20:305–10. Blume WT. Drug effects on EEG. J Clin Neurophysiol 2006;23:306–11. Braga NI, Manzano GM, Nobrega JA. Quantitative analysis of EEG background activity in patients with rolandic spikes. Clin Neurophysiol 2000;111:1643–5. Chatrian GE, Bergamini L, Dondey M, Klass DW, Lennox-Buchtal M, Peterse´n I. A glossary of terms most commonly used by clinical electroencephalographers. 1983;11–26. Dean AC, Solomon G, Harden C, Papakostas G, Labar DR. Left hemispheric dominance of epileptiform discharges. Epilepsia 1997;38:503–5. DeCarli C, Massaro J, Harvey D, Hald J, Tullberg M, Au R, et al. Measures of brain morphology and infarction in the Framingham heart study: establishing what is normal. Neurobiol Aging 2005;26:491–510. Diaz GF, Virues T, San Martin M, Ruiz M, Galan L, Paz L, et al. Generalized background qEEG abnormalities in localized symptomatic epilepsy. Electroencephalogr Clin Neurophysiol 1998;106:501–7. Doherty MJ, Jayadev S, Miller JW, Farrell DF, Holmes MD, Dodrill CB. Age at focal epilepsy onset varies by sex and hemispheric lateralization. Neurology 2003a;60:1473–7. Doherty MJ, Simon E, De Menezes MS, Kuratani JD, Saneto RP, Homles MD, et al. When might hemispheric favouring of epileptiform discharges begin? Seizure 2003b;12:595–8. Doherty MJ, Walting PJ, Morita DC, Peterson RA, Miller JW, Holmes MD, et al. Do nonspecific focal EEG slowing and epileptiform abnormalities favor one hemisphere? Epilepsia 2002;43:1593–5.
Drake ME, Padamadan H, Newell SA. Interictal quantitative EEG in epilepsy. Seizure 1998;7:39–42. Dustman RE, Shearer DE, Emmerson RY. EEG and event-related potentials in normal aging. Prog Neurobiol 1993;41:369–401. Farkas V, Kohlheb O, Benninger C, Matthis P. Comparison of the EEG background activity of epileptic children and children with migraine. Epilepsy Res Suppl 1992;6:199–205. Flink R, Pedersen B, Guekht AB, Malmgren K, Michelucci R, Neville B, et al. Guidelines for the use of EEG methodology in the diagnosis of epilepsy. International League Against Epilepsy: Commission report. Commission on European Affairs: Subcommission on European Guidelines. Acta Neurol Scand 2002;106:1–7. Gatzonis SD, Roupakiotis S, Kambayianni E, Politi A, Triantafyllou N, Mantouvalos V, et al. Hemispheric predominance of abnormal findings in electroencephalogram (EEG). Seizure 2002;11:442–4. Gibbs FA, Gibbs EL. Diagnostic entities – Material and methods of classification. In: Gibbs FA, Gibbs EL, editors. Atlas of encephalography. Epilepsy 1952;vol. 2. Massachusetts: Addison-Wesley Press, Inc; 1952. p. 20–3. Gloor P, Ball G, Schaul N. Brain lesions that produce delta waves in EEG. Neurology 1977;27:326–33. Hamer HM, Reis J, Mueller HH, Knake S, Overhof M, Oertel WH, et al. Motor cortex excitability in focal epilepsies not including the primary motor area – a TMS study. Brain 2005;128:811–8. Holmes MD, Dodrill CB, Kutsy RL, Ojemann GA, Miller JW. Is the left cerebral hemisphere more prone to epileptogenesis than the right? Epileptic Disord 2001;3:137–41. Ko TS, Holmes GL. EEG and clinical predictors of medically intractable childhood epilepsy. Clin Neurophysiol 1999;110:1245–51. Koufen H, Gast C. Zur Frage der Alters- und Diagnoseabha¨ngigkeit der Links-Lateralisation und Lokalisation von EEG-Herden. Arch Psychiatr Nervenkr 1981;229:227–37. Labar D, Dilone L, Solomon G, Harden C. Epileptogenesis: left or right hemisphere dominance? Preliminary findings in a hospital-based population. Seizure 2001;10:570–2. Leutmezer F, Lurger S, Baumgartner C. Focal features in patients with idiopathic generalized epilepsy. Epilepsy Res 2002;50:293–300. Miyauchi T, Endo K, Yamaguchi T, Hagimoto H. Computerized analysis of EEG background activity in epileptic patients. Epilepsia 1991;32:870–81. Nuwer M. Assessment of digital EEG, quantitative EEG, and EEG brain mapping: report of the American Academy of Neurology and the American Clinical Neurophysiology Society. Neurology 1997;49:277–92. Salinsky TC, Oken BS, Storzbach T, Dodrill TB. Assessment of CNS effects of antiepileptic drugs by using quantitative EEG measures. Epilepsia 2003;44:1042–50. Sam MC, So EL. Significance of epileptiform discharges in patients without epilepsy in the community. Epilepsia 2001;42:1273–8. Smith SJM. EEG in the diagnosis, classification, and management of patients with epilepsy. J Neurol Neurosurg Psychiatry 2005;76:2–7. Sperling MR, Guina L. The necessity for Sphenoidal electrodes in the presurgical evaluation of temporal lobe epilepsy – Pro position. J Clin Neurophysiol 2003;20:299–304. Stokes T, Shaw EJ, Guarez-Garcia A, Camosso-Stefinovic J, Baker R. Clinical guidelines and evidence review for the epilepsies: diagnosis and management in adults and children in primary and secondary care. London: Royal College of General Practitioner; 2004. Torre JLF, Alarcon G, Binnie CD, Seoane JJG, Juler J, Guy CN, et al. Generation of scalp discharges in temporal lobe epilepsy as suggested by intraoperative electrocorticographic recordings. J Neurol Neurosurg Psychiatry 1999;67:51–8. Weissinger F, Buchheim K, Siegmund H, Meierkord H. Seizure spread through the life cycle: optical imaging in combined brain slices from immature, adult, and senile rats in vitro. Neurobiol Dis 2005;19:84–95. Westmoreland BF, Klass DW. Unusual EEG patterns. J Clin Neurophysiol 1990;7:209–28.
