Abstract
Introduction: Frontal-predominant epileptiform discharges (EDs) include generalized spike-wave (GSW) and frontal spikes (FS). However, negative bi-frontal ED with simultaneous occipital positivity (BFOD) are rare, leading to questions regarding physiological generators. Methods: To determine the clinical significance of BFOD, electroclinical features of children with BFOD (n = 40) were compared to control patients with GSW (n = 102) and FS (n = 100). Results: Results are presented in the following order: BFOD, GSW, and FS. Epilepsy was prevalent among the groups: 95.0%, 90.2%, and 77.0%, respectively. The median age of seizure-onset did not significantly differ between groups: 3.00, 4.00, and 2.25 years, respectively. Regarding EEG background features, the BFOD group had more disorganized sleep architecture than other groups, P < .005. There was a significant difference in the proportion of developmental delay (DD) between the groups (P < .005). BFOD had much higher odds of DD compared to GSW and FS groups: odds ratio (OR) (confidence interval [CI]) 19.44 [5.64, 64.05] and 3.98 [1.16, 13.34]. Furthermore, BFOD had much higher odds of severe DD compared to GSW and FS groups: 9.60 [2.75, 33.45] and 2.73 [1.03, 7.27]. A Gross Motor Function Classification System (GMFCS) score of ≥ 4 was more prevalent in BFOD (22.5%), than GSW (0%) and FS groups (9%). On neuroimaging, BFOD had more structural (P < .005) and multilobar structural (P < .05) abnormalities than control groups. Conclusion: Children with BFOD had particularly severe significant DD, considerable motor deficit (GMFCS ≥ 4), and brain structural abnormalities, often multilobar. This suggests BFOD is a marker of severe underlying brain dysfunction and not benign when encountered on routine EEG review.
Keywords: EEG, pediatric, frontal sharp wave, generalized spike-wave, dipole
Introduction
Various epileptiform abnormalities with apparent frontal predominance arise on pediatric EEG including generalized spike wave discharges (GSW) and frontal spikes/sharp waves (FS). Furthermore, in children, various artifacts, normal variants, or sleep features can have a frontal predominance and need to be carefully distinguished from epileptiform discharges (EDs). GSW, generated by thalamocortical circuits, 1 are an electrographic hallmark of genetic generalized epilepsy (GGE), but not exclusive to it. FS have a strong association with epilepsy2,3 and may indicate a frontal epileptogenic focus; however, they also occurs in epilepsies of various localizations and etiologies. Therefore, for these findings, the clinical correlation and anticipated neurodevelopment is contingent on the underlying cerebral pathology that gives rise to seizures, thus requiring further clarification.
Rarely encountered on pediatric EEG are negative bi-frontal EDs with simultaneous occipital positivity, a so-called dipole field (BFOD). Usually, EEG spike generators are oriented perpendicularly with 1 pole detectable on the scalp, resulting in mainly surface-negative unipolar phenomena. Therefore, when encountering electrical fields with simultaneous surface-negative and surface-positive dipolar potentials at different scalp electrodes, such as BFOD, questions arise about the physiological and clinical significance of such an unusual tangentially orientated generator/field. In this study, clinical features of children with BFOD were compared to those of 2 control groups with anterior predominance: children with GSW and children with FS. The study objective was to describe the electro-clinical features and to determine the clinical significance of BFOD when encountered on routine EEG analysis.
Methods
At our institution, electroclinical data obtained by a technologist at each visit for all patients are entered into a database. Our institution is the only pediatric tertiary care center in the area, serving a population of 5 million people. BFOD was defined as bi-frontal negative discharges with simultaneous (ie, synchronous) bi-occipital positive discharges on scalp EEG (Figures 1 and 2). The database was queried for BFOD (all spike types were classified separately) from 2000 to 2021 in patients aged 2 months to 18 years. Patients in a comatose state, with other types of EDs or inadequate clinical information were excluded. BFOD had to have uniform morphology and voltage configuration in the EEG tracing to be included. EEGs were recorded for 25 to 45 min with Biologic or Natus machines, using the international 10 to 20 system with 256 Hz sampling rate. For EEG acquisition, a low-frequency filter of 0.5 Hz and a high-frequency filter of 70 Hz were used. On close inspection, included patients may have had a subtle lead-in or frontal-occipital lag, especially when viewed on faster display speed (Figure 3). However, our aim is to study this rare pattern of frontal negativity and occipital positivity, detected on visual inspection on routine clinical (not research) EEG analysis. Sixty-six patients in the database with BFOD were identified. All EEGs were re-reviewed by investigators using various montages to confirm the finding of BFOD and 40 patients met inclusion criteria.
Figure 1.
Examples of BFOD.(A)-(E) Ear reference montage. HFF = 100 Hz, LFF = 0.5 Hz; sensitivity: sensitivity; A, C, D and E = 30 µV/mm, B = 20 µV/mm; the distance between the bold vertical lines represents 1s. Abbreviations: BFOD, bi-frontal negative discharges and synchronous bi-occipital positive discharges; ED, epileptiform discharges; HFF, high frequency filter; LFF, low frequency filter.
Figure 2.
Examples of BFOD with referential and bipolar montage; BFOD (A) Ipsilateral ear referential montage; (B) longitudinal bipolar montage; HFF = 100 Hz, LFF = 1 Hz; sensitivity: 15 µV/mm; the distance between the bold vertical lines represents 1 s. BFOD (C) Ipsilateral ear referential montage; (D) longitudinal bipolar montage; HFF = 100 Hz, LFF = 1 Hz; sensitivity: 15 µV/mm; the distance between the bold vertical lines represents 1 s. Abbreviation: BFOD, bi-frontal negative discharges and synchronous bi-occipital positive discharges.
Figure 3.
Examples of BFOD at 30 and 60 mm/s to demonstrate that the dipoles are heterogeneous and not bisynchronous. (A) Negativity at FP1 and FZ precedes negativity at FP2, followed by bi-central parietal occipital positivity. HFF = 70 Hz, LFF = 3 Hz; sensitivity 30 µV/mm; the distance between the bold vertical lines represents 1 s. (B) Negativity in the left anterior quadrant precedes right-frontal polar negativity and bi-central parietal occipital positivity. HFF = 70 Hz, LFF = 3 Hz; sensitivity 30 µV/mm; the distance between the bold vertical lines represents 1 s. (C) Negativity in the right anterior quadrant precedes left anterior quadrant negativity and bi-occipital positivity. HFF = 70 Hz, LFF = 3 Hz; sensitivity 50 µV/mm; the distance between the bold vertical lines represents 1 s. EKG reference for A. and C. is a noncephalic reference that is reliable for the period between active cardiac signal, that is, midway between EKG QRS complexes. Although it has limitations, it can be helpful when other noncephalic electrodes are unavailable to clarify topography of EDs. Abbreviation: BFOD, bi-frontal negative discharges and synchronous bi-occipital positive discharges
As our EEG lab is at the only tertiary care pediatric center in the province, more children with developmental delay and seizures may be referred to the EEG lab. To mitigate such referral bias and strengthen the results, 2 control groups, with anterior predominant EDs, were used. A GSW group consisted of patients with bilateral synchronous and symmetric spike-and-wave EDs, with an anterior bi-frontal predominance. This may have had shifting predominance based on asymmetry or asynchrony. The FS group consisted of patients with right, left, or bi-frontal spikes/sharp waves that were transient potentials, distinguishable from the background, with a sharp negative peak. A spike has a duration of 20 to 70 ms with a more rapid ascending than descending peak, whereas a sharp wave has a duration of 70 to 200 ms. 4 For the control groups, a random 150 patients were selected from the EEG database with GSW and FS from the same timeframe as BFOD EEGs. Detailed chart reviews were performed and similar to the study group, all patients in a comatose state, with other focal EDs or with inadequate clinical data were excluded.
EEG background features, coded in the EEG database, were also examined. A normal EEG background included evidence of a posterior dominant rhythm that was appropriate for age, during relaxed wakefulness. Background slowing was unvarying and unreactive slowing of the background in a vigilant patient, corresponding to mild generalized cerebral dysfunction. Dysrhythmia was defined as isolated episodic paroxysmal focal slow activity, or sharply contoured waveforms, that were not epileptiform. Disorganized sleep architecture was defined as the absence of well-formed sleep architecture, including vertex waves, sleep spindles and K-complexes during nonrapid eye movement (NREM) sleep. EEG suppression was defined as background activity with significantly lower-than-expected amplitudes.
Clinical data, including seizure-type, neurodevelopment, neuro-imaging, and epilepsy co-morbidities, such as attention-deficit/hyperactivity disorder (ADD/ADHD), autism spectrum disorder (ASD), and psychiatric diagnoses, were obtained from medical records. Developmental delay (DD) was defined as a delay in meeting at least one developmental milestone. If a patient's development was less than half of the expected chronological age, the patients were classified as having severe DD. This measure, although arbitrary, was of clinical utility and was applied to the study and control groups, based on quantitative descriptions by clinicians. Standardized neuropsychological evaluations were not available for most of the patients. A specialist formally diagnosed the children with ASD. The Gross Motor Function Classification System (GMFCS), is a 5-level classification that differentiates children with cerebral palsy based on the child's gross motor abilities, limitations in gross motor function, and need for assistive technology and wheeled mobility. 5 Specific notation was made of severe impairment, levels ≥ 4. Seizure type and etiology were classified according to 2017 International League Against Epilepsy. 6 Neuro-imaging abnormalities were classified as normal, nonspecific, focal, multifocal, hemispheric, or diffuse.
Data was summarized using descriptive statistics including medians and interquartile ranges (IQRs) for continuous variables and frequencies and percentages for categorical variables. To assess whether the relative proportions of clinically relevant risk factors were different between the study groups, chi-square tests were performed, with a P-value < .05 being significant. The presence of DD was studied using a logistic regression model, with the study group as a covariate. The results of this logistic regression model are presented as odds ratios (OR) with 95% confidence intervals and P-values. All analyses were performed using R Statistical Software (R Core Team 2022, Vienna Austria).
Results
From 2000 to 2021, with 43,061 records in the EEG database, 40 patients were identified with BFOD, comprising 0.09% of all patients referred for EEGs. For the control groups, after reviewing 150 random patients with GSW and FS, the control groups included 102 patients with GSW and 100 with FS. There was no difference between the groups in terms of the age at seizure-onset and the age of spike. The median [IQR] age at onset was 3 [1.00, 5.00] years in the BFOD group, 4.00 [2.00, 10.0] years in the GSW group, and 2.25 [0.500, 6.00] years in the FS group. The median [IQR] age in years of the EEG finding was 8.50 [6.00, 11.0] in the BFOD group, 9.45 [5.08, 13.9] in the GSW group, and 8 [3.00, 11.0] in the FS group. The results are presented in the following order: BFOD, GSW, and FS groups and are summarized in Table 1.
Table 1.
Clinical Features of BFOD and Control Groups.
| BFOD (N = 40) | GSW (N = 102) | FS (N = 100) | Significant Findings for BFOD vs Control Groups | |
|---|---|---|---|---|
| Age of seizure onset (year) | ||||
| Median [Q1, Q3] | 3.00 [1.00, 5.00] | 4.00 [2.00, 10.0] | 2.25 [0.500, 6.00] | |
| Mean (SD) | 3.23 (2.61) | 5.99 (4.73) | 3.96 (4.20) | |
| Age of spike (year) | ||||
| Median [Q1, Q3] | 8.50 [6.00, 11.0] | 9.45 [5.08, 13.9] | 8.00 [3.00, 11.0] | |
| Mean (SD) | 9.38 (3.71) | 9.61 (5.09) | 7.50 (5.12) | |
| History of epilepsy | 38 (95.0%) | 92 (90.2%) | 77 (77.0%) | |
| Seizure semiology | ||||
| Generalized | 7 (17.5%) | 82 (80.4%) | 30 (30.0%) | |
| Focal | 21 (52.5%) | 8 (7.8%) | 44 (44.0%) | |
| Both | 10 (25.0%) | 1 (1.0%) | 10 (10.0%) | |
| Unknown | 2 (5.0%) | 11 (10.8%) | 16 (16.0%) | |
| Seizure etiology | ||||
| Structural | 18 (45.0%) | 3 (2.9%) | 25 (25.0%) | P < .005 |
| Genetic | 9 (22.5%) | 12 (11.8%) | 20 (20.0%) | |
| Infectious | 0 (0%) | 1 (1.0%) | 4 (4.0%) | |
| Metabolic | 1 (2.5%) | 1 (1.0%) | 0 (0%) | |
| Autoimmune | 0 (0%) | 3 (2.9%) | 3 (3.0%) | |
| Genetic generalized | 0 (0%) | 57 (55.9%) | 9 (9.0%) | |
| Acute symptomatic | 0 (0%) | 5 (4.9%) | 3 (3.0%) | |
| Unknown | 11 (27.5%) | 17 (16.7%) | 30 (30.0%) | |
| Neuro-imaging | ||||
| Normal | 10 (25.0%) | 53 (52.0%) | 46 (46.0%) | |
| Nonspecific changes | 3 (7.5%) | 12 (11.8%) | 6 (6.0%) | |
| Focal abnormalities | 4 (10.0%) | 2 (2.0%) | 19 (19.0%) | |
| Multifocal | 7 (17.5%) | 0 (0%) | 3 (3.0%) | |
| Hemispheric | 2 (5.0%) | 0 (0%) | 0 (0%) | |
| Diffuse abnormalities | 8 (20.0%) | 4 (3.9%) | 15 (15.0%) | |
| Not performed or available | 6 (15.0%) | 31 (30.4%) | 11 (11.0%) | |
| Large structural: Diffuse, multifocal, and hemispheric | 19 (47.5%) | 4 (3.9%) | 25 (25.0%) | P < .005 |
| Developmental delay | 35 (87.5%) | 27 (26.5%) | 64 (64.0%) | P < .005 |
| Severe developmental delay | 15 (37.5%) | 6 (5.9%) | 18 (18.0%) | P < .005 |
| GMFCS >=4 | 9 (22.5%) | 0 (0%) | 9 (9.0%) | |
| ASD | 9 (22.5%) | 12 (11.8%) | 22 (22.0%) | |
| ADHD | 6 (15.0%) | 22 (21.6%) | 10 (10.0%) | |
| Psychiatric Diagnosis | 14 (35.0%) | 28 (27.5%) | 28 (28.0%) | |
| EEG Background | ||||
| Slowing | 17 (42.5%) | 0 (0%) | 31 (31%) | |
| Dysrhythmia | 17 (42.5%) | 1 (1%) | 49 (49%) | |
| Suppression | 2 (5%) | 0 (0%) | 4 (4%) | |
| Disorganized sleep architecture | 17 (42.5%) | 1 (1%) | 8 (8%) | P < 0.005 |
Abbreviations: ADHD, attention deficit hyperactivity disorder; ASD, autism spectrum disorder; BFOD, bi-frontal negative discharges and synchronous bi-occipital positive discharges; FS, frontal spikes and sharp waves; GMFCS, Gross Motor Function Classification System; GSW, generalized spike wave; [Q1, Q3]: interquartile range; SD, standard deviation.
Seizures
For the BFOD, GSW, and FS groups, respectively, the prevalence of epilepsy was high in all groups: 38 of 40 (95.0%), 92 of 102 (90.2%), and 77 of 100 (77.0%), possibly reflecting a referral bias at a tertiary pediatric epilepsy center. The classification of seizure-types was based on semiology. The GSW group had predominant generalized seizure semiology. However, the BFOD and FS groups had either focal or generalized semiology. In the BFOD, GSW, and FS groups, semiology was generalized [7 (17.5%), 82 (80.4%), 30 (30.0%)], focal [21 (52.5%), 8 (7.8%), 44 (44.0%)], both generalized and focal [10 (25.0%), 1 (0.9%), 10 (10.0%)], and unknown [2 (5.0%), 11 (10.8%), 16 (16.0%)].
The most common epilepsy etiologies in the BFOD group were structural 18 (45.0%), including hypoxic ischemic injury and brain malformations; unknown 11 (27.5%), and genetic 9 (22.5%), including Rett Syndrome. Table 2 provides the detailed seizure etiology for the BFOD group. In the GSW group, the common etiologies were GGE/Idiopathic generalized epilepsy (IGE) 57 (55.9%); unknown 17 (16.7%); and genetic 12 (11.8%), where known specific genetic mutations were identified, including SCN1A, ANKRD11, and 15q duplication. The prevalence of GGE/IGE was likely high, as patients with other EDs were excluded and patients with GGE/IGE are less likely to have focal EDs, than other epilepsy etiologies, such as genetic or metabolic. In the FS group, the most common etiologies were unknown 30 (30.0%), structural 25 (25.0%), and genetic 20 (20.0%). As this study spanned over 25 years, some patients with unknown etiology did not have advanced genetic testing or neuroimaging. For example, some patients with BFOD with unknown etiology had developmental epileptic encephalopathies, without whole-exome sequencing studies.
Table 2.
Etiology and Imaging of BFOD Group.
| # | Etiology | Imaging | # | Etiology | Imaging |
|---|---|---|---|---|---|
| 1 | Tuberous sclerosis complex | Multifocal tubers and subependymal nodules | 21 | Chromosome 8 rearrangement | Diffuse cortical atrophy, bilateral WM abnormality |
| 2 | PAX1 mutation | Vermian atrophy | 22 | Aqueductal stenosis; hydrocephalus; previous ESES | Diffuse cortical atrophy, hydrocephalus and shunt |
| 3 | Hypoglycemia with secondary multifocal infarcts | Infarcts R frontal, L parietal and L occipital regions | 23 | Unknown | Normal |
| 4 | Unknown | Normal | 24 | Neonatal hemorrhage; previous ESES | L thalamus and L PV WM injury, smaller L hemisphere |
| 5 | Unknown; DEE | Normal | 25 | Nonaccidental injury | Extensive cortical encephalomalacia |
| 6 | Arachnoid cyst | Arachnoid cyst anterior to R temporal tip | 26 | Chromosome 15q11.2 microdeletion | Normal |
| 7 | Unknown; pervious ESES | Normal | 27 | MECP2 mutation | Normal |
| 8 | Unknown | Not done | 28 | Unknown; Lennox-Gastaut syndrome | Not available |
| 9 | Premature, HIE | PVL and R hemorrhagic infarction and porencephaly | 29 | Unknown, DEE | Not available |
| 10 | Unknown | Normal | 30 | Unknown; previous ESES | Normal |
| 11 | 15q11.13 duplication | Diffuse gray and WM loss | 31 | HIE, diffuse | Multicystic WM with necrotic cavities in basal ganglia and thalamus |
| 12 | HIE | Not available | 32 | Mitochondrial disease | Global volume loss of cerebrum, cerebellum and brainstem. Incomplete myelination |
| 13 | Unknown | Nonspecific increased WM signal | 33 | R hemisphere brain malformation | R hemisphere lissencephaly |
| 14 | MECP2 mutation | Increased signal in peritrigonal WM | 34 | SCN1A | Normal |
| 15 | Antenatal infarct | R frontal infarct and gliosis | 35 | R hemisphere brain malformation | R hemisphere polymicrogyria and closed lip schizencephaly. Hyper-intensity left frontal WM |
| 16 | Unknown | Normal | 36 | Diffuse brain malformation | Diffuse pachygyria, abnormal sylvian fissures, decreased WM volume |
| 17 | Unknown | Punctate foci frontal WM, nonspecific | 37 | Tuberous sclerosis complex | Multifocal tubers |
| 18 | Hemispheric brain malformation | R hemisphere smaller, abnormal and WM | 38 | Unknown | Not available |
| 19 | POLG mutation | Normal | 39 | Heterozygous for SMAD4/Myhre syndrome | Periosteal thickening of skull, diffuse WM gliosis |
| 20 | HSV infection; previous ESES | L MTS, anteromedial temporal lobe gliosis, and calcification | 40 | N/A (no epilepsy) | Not done |
Abbreviations: #, Subject number; BFOD, bi-frontal negative discharges and synchronous bi-occipital positive discharges; DEE, developmental epileptic encephalopathy; ESES, electrical status epilepticus in slow wave sleep; HIE, hypoxic ischemic encephalopathy; L, left-sided; MTS, mesial temporal sclerosis; N/A, not-applicable; PVL, peri-ventricular leukomalacia; R, right-sided; SeLECTS, self-limited epilepsy with centrotemporal spikes; WM, white matter.
EEG
For the BFOD, GSW, and FS groups, respectively, the following EEG background abnormalities were present: slowing [17 (42.5%), 0, 31 (31.0%)], dysrhythmia [17 (42.5%), 1 (0.9%), 49 (49.0%)], suppression [2 (5.0%), 0, 4 (4.0%)], and disorganized sleep architecture [17 (42.5%), 1 (0.1%), 8 (8.0%)]. The GSW group had fewer background abnormalities than the other groups did. There were no differences in the background slowing, dysrhythmia or suppression between the BFOD and FS groups. However, a chi-square test showed that the BFOD group had a higher incidence of disorganized sleep architecture than the FS and GSW groups, P < .005. None of the patients had electrical status epilepticus during slow-wave sleep (ESES) on the study EEG tracing. However, 4 (10%) in the BFOD group and 2 (2%) in the FS group had a history of ESES.
Co-Morbidities and Development
The results off the chi-square test showed a difference in the relative proportion of DD between the groups (P < .005). The BFOD group had much higher odds of having DD than the GSW and FS groups, OR [CI] 19.44 [5.64, 64.05] and 3.98 [1.16, 13.34], respectively. Furthermore, a chi-square test showed a difference in the relative proportion of severe DD between the groups (P < .005). The BFOD group has much higher odds of having severe DD than the GSW 9.60 [2.75, 33.45] and FS 2.73 [1.03, 7.27] groups, respectively.
The BFOD group had a higher incidence of severe motor impairment. A GMFCS score of ≥ 4 was present in 9 (22.5%) of the BFOD group, and 6 of 9 (67%) had a score of 5. A score of ≥ 4 was present in 9 (9%) of the FS group, and 5 of 9 (5%) had a score of 5. None of the patients with GSW had a score of ≥ 4. This finding is reflective of the high incidence of cerebral palsy in the BFOD group (Table 2).
There were no clear differences in other epilepsy co-morbidities, including ASD, ADD/ADHD, and psychiatric diagnoses among the groups. However, in children with severe DD, psychiatric disorders or ADHD are more challenging to detect and diagnose and may be under-represented, particularly in the BFOD group, which had a higher incidence of severe DD.
Neuro-Imaging
Neuro-imaging findings were normal or nonspecific [(13 (32.5%), 65 (63.7%), 52 (52.0%)], focal [4 (10%), 2 (2%), 19 (19%)], diffuse [8 (20%), 4 (3.9%), 15 (15%)], multifocal [7 (18%), 0, 3 (3%)], and hemispheric [2 (5%), 0, 2 (1%)]. Refer to Table 2 for imaging abnormalities in BFOD group. The BFOD group had a significantly higher incidence of structural abnormalities than either control groups (P < .005). In addition, the BFOD group had significantly more widespread (diffuse, multifocal, or hemispheric) structural abnormalities on imaging than the control groups (P < .05).
Discussion
In this study, BFOD was associated with a higher risk of DD, severe DD, significant motor deficit (GMFCS ≥4), and large brain structural abnormalities, possibly representing an EEG marker of severe underlying brain dysfunction, and less favorable clinical outcome. Various EEG waveforms have been associated with clinical correlations. Specifically, examples of tangential dipoles on routine EEG with clinical significance include centrotemporal spikes in Self-limited epilepsy with Centrotemporal spikes (SeLECTS), 7 and small sharp spikes during sleep (SSS), a normal variant, with negative anterior temporal spikes and widespread positive field over the posterolateral contralateral hemisphere. 8 Occasionally, sleep spindles of infancy can have a negative sharp component over the centroparietal and a positive sharp component over the frontal regions. 9 Spikes with tangential dipole configurations can also occur in patients with altered anatomy, resulting from brain surgery, trauma, or skull abnormalities.
At our institution, spike location and the presence of tangential dipoles are coded specifically and systematically in a database spanning over 25 years. BFOD was identified in only 40 (0.09%) of patients referred who were for EEG. As simultaneous negativity and positivity on EEG with a BFOD configuration is uncommon, when encountered it raises questions regarding generators and clinical significance. Cortical pyramidal neurons are the predominant generators of EEG waveforms, typically resulting in a vertical generator configuration perpendicular to the scalp. Tangential generators originate from within cortical fissures, sulcal depths (comprising 70% of the cortical surface), 10 and the mesial frontal, parietal, occipital, and basal temporal cortices. However, tangential potentials usually cancel when present in the adjacent fissural walls, allowing radial potentials from the cortical gyri to predominate. In our BFOD cohort, many children had large brain malformations and cerebral palsy, and 1 patient had periosteal thickening of the skull, in addition to diffuse white matter gliosis, all of which may have contributed to the unusual dipole configuration. The high incidence of motor impairment in the BFOD group, suggests white matter or motor-cortex involvement in many of the children. As BFOD is present in children with various epilepsy etiologies, there is likely heterogeneity of the neural-generator for BFOD.
In children with BFOD, bi-synchrony of the frontal and occipital regions may occur by interhemispheric propagation through subcortical mesial structures.11–13 Subcortical structures, such as the corpus callosum, play a major role in bilateral synchronization of spike-and-wave, while thalamocortical circuits generate rhythmic sustained spike-and-wave burst. 14 Focal cortical lesions, including cortical dysplasia or infarcts, particularly located near frontal callosal fibers, could have resulted in bi-synchronous discharges of BFOD in some patients. 15
A bi-synchronous connection is also possible via cortico-cortical spread via long association fibers. Widely separated activity can give rise to distinct negative and positive maxima on the scalp electrodes. For example, bi-synchronous temporal spikes, such that the negative component on the left aligns with the positive component on the right, may appear to represent huge transverse dipoles 16 ; however, demonstration of spike asynchrony can prove the fields represent not the source and sink of a single dipole generator but rather 2 generators linked by cortico-cortical propagation. 17 In our study, EEG review was by visual analysis without computer-assisted timing methods to determine subtle asynchrony. Therefore, multiple distant generators with a propagation pathway are possible in some individuals with BFOD, especially as many individuals have multilobar or diffuse cortical abnormalities.
In contrast to BFOD, synchronous negative occipital and frontopolar spikes (FO spike) are relatively common, and are characteristic of Self-Limited Epilepsy with Autonomic Seizures, (SeLEAS), which has a favorable prognosis, but also in other types of childhood epilepsies. Studies evaluating negative FO spikes with sequential topographical mapping, including a SeLEAS study and a heterogeneous epilepsy study, both observed, that the occipital peak consistently preceded the frontal peak, with an interhemispheric latency <10 ms, indicative of a primary interhemispheric bilateral synchronous pattern.18,19 The long occipitofrontal association fibers, such as the uncinate and arcuate fasciculus, play a role in generating coherent or synchronous negative FO spike discharges and may represent a developmental EEG phenomena related to brain maturation.
However, unlike FO spikes in children, where the synchronizing spike phenomenon propagate from the occipital to the frontal regions, the primary focus of BFOD is likely not as uniform or well-localized. This is demonstrated in Figure 3, where a paper speed of 60 mm/s better displays the lack of consistent synchrony and variability of BFOD. It is well established that a generator orientated obliquely to the scalp (midway between radial and tangential), generates potential field which is harder to localize reliably for various reasons. Between the 2 maxima of opposite polarity (frontal negativity and occipital positivity in BFOD), is an iso-potential boundary where the generator signal is not detected. 20 In the BFOD group, the high incidence of DD, particularly severe, are suggestive of involvement of large areas of cortex. Furthermore, the children had heterogeneous epilepsy etiologies and neuro-imaging findings, in agreement with this line of thinking.
Furthermore, a quarter (23%) of the children with BFOD had significant motor deficit, with a GMFCS score of ≥ 4, suggesting white matter or frontal involvement, which can involve deep brain structures. Specifically, deeper cortical generators can lead to a smaller surface potential and a more widespread field relative to the surface maximum, thus making identification of the source based on the potential difference between any scalp electrodes challenging.21,22 In addition, multiple superficial sources can overshadow a deep one, distorting or hiding it. Thus, a given electrode has a view of the nearby generators, such that dipoles that combine to reinforce each other will have a large net effect, whereas those that cancel will produce a smaller or null potential. 23 All study EEGs were re-reviewed in multiple montages, aiming for the most uninvolved referential montage; however, an adequate vantage point may be impossible with surface electrodes for deep generators. In this study, children with BFOD had various white matter abnormalities, ranging from periventricular leukomalacia to nonspecific white matter volume loss and gliosis (Table 2). Specifically, in studies of children with cerebral palsy, regardless of etiology or imaging pattern, heterogeneous EEG abnormalities occur, reflecting inconsistent cortical generators with white matter involvement. 24
Notably, on EEG, although the BFOD and FS group had a similar incidence of background slowing, dysrhythmia, and suppression, almost half of the BFOD group had disorganized sleep architecture, suggesting disruption of thalamocortical connections. Frequent sleep EDs can disrupt sleep architecture. However, none of the patients had ESES or had frequent EDs during sleep. Notably, 4 (10%) children in the BFOD group and 2 (2%) of children in the FS group had a history of ESES. Many children in the BFOD group had large cortical/subcortical structural abnormalities (ie, secondary to hypoxic ischemic encephalopathy), or functional abnormalities (ie, developmental epileptic encephalopathies, or genetic conditions) which could have resulted in disorganized sleep features on EEG. Furthermore, the current hypotheses suggest that the structural properties of the white matter tracts affect the synchronization of brain networks to produce slow oscillations during NREM sleep. Therefore, a high degree of white matter involvement may also contribute to the disorganized sleep architecture in almost half of the patients with BFOD.25,26 In addition, both structural and functional brain connectivity is closely linked to white matter. The axon conduction speed is variable, determined largely by myelination. 27 EEG and high-resolution computer-aided EEG could provide complementary measures of functional connectivity and further delineate frontal-occipital connectivity in BFOD.
The study results allow delineation of the clinical features of BFOD detected on routine EEG analysis. BFOD was associated with severe significant DD, motor deficit, and brain structural abnormalities, often multilobar, indicative that BFOD is not a favorable EEG finding when encountered. The limitations of this study include retrospective data collection without standard data collection forms. However, our EEG lab has systematically documented the presence of EEG tangential dipoles for 25 + years. Multicenter prospective studies with standardized neuropsychological testing would further characterize these EEG findings. Another important limitation to the study is the selection of optimal control groups. Control groups comprising EDs of anterior predominance were selected to create a benchmark for comparison of clinical features and to mitigate referral bias, as complex patients with higher rates of epilepsy are often referred to tertiary care centers. However, it is understood that clinical and electrophysiological differences exist between BFOD and the control groups, which could have contributed to bias in estimates of differences. Although, the focus of this study was visual analysis of routine EEG for the examination of BFOD, further studies with dipole localization software, sequential topographical mapping, and MEG, would provide further insight into cortical generators and networks. BFOD likely results from variable cortical generators and networks, as it is present in children with various etiologies of epilepsy. In conclusion, BFOD was associated with an increased risk of DD, particularly severe, significant motor deficit (GMFCS ≥4) and large brain structural abnormalities, suggesting that BFOD is a marker of severe underlying brain dysfunction and not a benign entity when encountered in routine EEG review.
Acknowledgments
Jeffrey Zhi helped to identify patients from the EEG database, and Ash Sandhu provided assistance with statistics. Dr Peter KH Wong created the EEG database, which was crucial to identify patients. He started systematically coding for tangential dipoles in the database 20 + years ago. He provided invaluable guidance in study design and analysis. He reviewed and edited the manuscript for important intellectual content.
Footnotes
Author Contributions: AND developed the original concept and design of the manuscript. She obtained clinical data by performing a detailed chart review and reviewing all EEG tracings. She helped to analyze the data, drafted the manuscript, and then reviewed, and edited it for intellectual content. JC reviewed all EEG tracings and prepared figures. She also helped to analyze the data and edited the manuscript for intellectual content. CM identified eligible patients from the EEG database. She reviewed all the EEG tracings. She reviewed, and edited the manuscript for intellectual content. All authors gave approval to the final version of the manuscript to be submitted and all authors are in agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethics Approval: Approved by University of British Columbia Ethics Board H20-01041-A009.
Funding: The authors received no financial support for the research, authorship, and/or publication of this article.
ORCID iD: Anita N Datta https://orcid.org/0000-0001-7620-4868
References
- 1.Snead OC, 3rd. Basic mechanisms of generalized absence seizures. Ann Neurol. 1995;37(2):146–157. [DOI] [PubMed] [Google Scholar]
- 2.Kanemura H, Sano F, Ohyama T, et al. EEG characteristics predict subsequent epilepsy in children with their first unprovoked seizure. Epilepsy Res. 2015;115:58‐62. [DOI] [PubMed] [Google Scholar]
- 3.Kuang YQ, Kong B, Yang T, et al. Epileptiform discharges and frontal paroxysmal EEG abnormality act as predictive marker for subsequent epilepsy in children with complex febrile seizures. Clin EEG Neurosci. 2014;45(4):299‐303. [DOI] [PubMed] [Google Scholar]
- 4.Noachtar S, Binnie C, Ebersole J, et al. A glossary of terms most commonly used by clinical electroencephalographers and proposal for the report form for the EEG findings. The International Federation of Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol. Suppl 1999;52:21‐41. [PubMed] [Google Scholar]
- 5.Palisano R, Rosenbaum P, Walter S, et al. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol. 1997;39(4):214‐223. [DOI] [PubMed] [Google Scholar]
- 6.Fisher RS, Cross JH, French JA, et al. Operational classification of seizure types by the International League Against Epilepsy: position paper of the ILAE commission for classification and terminology. Epilepsia. 2017;58(4):522‐530. [DOI] [PubMed] [Google Scholar]
- 7.Gregory DL, Wong PK. Topographical analysis of the centrotemporal discharges in benign rolandic epilepsy of childhood. Epilepsia. 1984;25(6):705‐711. [DOI] [PubMed] [Google Scholar]
- 8.Wennberg R, Tarazi A, Zumsteg D. Which small sharp spikes are benign epileptiform transients of sleep? Clin Neurophysiol. 2018;129(11):2492‐2494. [DOI] [PubMed] [Google Scholar]
- 9.Niedermeyer E. The EEG signal: polarity and field determination. In: Niedermeyer EL, ed. Electroencephalography. Williams and Wilkins; 1993:125‐130. [Google Scholar]
- 10.Carpenter MB. Core text of neuroanatomy. Baltimore, MD: Williams and Wilkins, 1991. [Google Scholar]
- 11.Gotman J. Interhemispheric relations during bilateral spike-and-wave activity. Epilepsia. 1981;22(4):453‐466. [DOI] [PubMed] [Google Scholar]
- 12.Gotman J. Measurement of small time differences between EEG channels: method and application to epileptic seizure propagation. Electroencephalogr Clin Neurophysiol. 1983;56(5):501‐514. [DOI] [PubMed] [Google Scholar]
- 13.Kobayashi K, Ohtsuka Y, Oka E, et al. Primary and secondary bilateral synchrony in epilepsy: differentiation by estimation of interhemispheric small time differences during short spike-wave activity. Electroencephalogr Clin Neurophysiol. 1992;83(2):93‐103. [DOI] [PubMed] [Google Scholar]
- 14.Blumenfeld H. Consciousness and epilepsy: why are patients with absence seizures absent? Prog Brain Res. 2005;150:271‐286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Tukel K, Jasper H. The electroencephalogram in parasagittal lesions. Electroencephalogr Clin Neurophysiol. 1952;4(4):481‐494. [DOI] [PubMed] [Google Scholar]
- 16.Jayakar P, Duchowny M, Resnick TJ, et al. Localization of seizure foci: pitfalls and caveats. J Clin Neurophysiol. 1991;8(4):414‐431. [DOI] [PubMed] [Google Scholar]
- 17.Ebersole JS. Intracranial EEG validation of single versus dual dipolar sources for temporal spikes in presurgical candidates. Epilepsia. 1990;31:621. [Google Scholar]
- 18.Ueno M, Oguni H, Yasuda K, et al. Neurophysiological study of secondary synchronous occipito-frontopolar spikes in childhood. Clin Neurophysiol. 2001;112(11):2106‐2112. [DOI] [PubMed] [Google Scholar]
- 19.Yoshinaga H, Kobayashi K, Ohtsuka Y. Characteristics of the synchronous occipital and frontopolar spike phenomenon in panayiotopoulos syndrome. Brain Dev. 2010;32(8):603‐608. [DOI] [PubMed] [Google Scholar]
- 20.Gloor P. Neuronal generators and the problem of localization in electroencephalography: application of volume conductor theory to electroencephalography. J Clin Neurophysiol. 1985;2(4):327‐354. [DOI] [PubMed] [Google Scholar]
- 21.Ludwig BI, Marsan CA. Clinical ictal patterns in epileptic patients with occipital electroencephalographic foci. Neurology. 1975;25(5):463‐471. [DOI] [PubMed] [Google Scholar]
- 22.Olivier A, Gloor P, Andermann F, et al. Occipitotemporal epilepsy studied with stereotaxically implanted depth electrodes and successfully treated by temporal resection. Ann Neurol. 1982;11(4):428‐432. [DOI] [PubMed] [Google Scholar]
- 23.Gregory DL. WP. Clinical and EEG features of tripole spike discharges in children. Epilepsia. 1986;27(5):605. [Google Scholar]
- 24.Al-Sulaiman A. Electroencephalographic findings in children with cerebral palsy: a study of 151 patients. Funct Neurol. 2001;16(4):325‐328. [PubMed] [Google Scholar]
- 25.Buchmann A, Kurth S, Ringli M, et al. Anatomical markers of sleep slow wave activity derived from structural magnetic resonance images. J Sleep Res. 2011;20(4):506‐513. [DOI] [PubMed] [Google Scholar]
- 26.Piantoni G, Poil SS, Linkenkaer-Hansen K, et al. Individual differences in white matter diffusion affect sleep oscillations. J Neurosci. 2013;33(1):227‐233. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gong J, Luo C, Chang X, et al. White matter connectivity pattern associate with characteristics of scalp EEG signals. Brain Topogr. 2017;30(6):797‐809. [DOI] [PubMed] [Google Scholar]