The Effect of Levetiracetam on the Interictal EEG in Children with Refractory Epilepsy

Abstract

Background

Interictal epileptiform discharges (IED) frequency is the most commonly studied interictal measure of ASM efficacy in patients with epilepsy, and IED suppression in a repeat pediatric EEG is usually interpreted as a reduced tendency towards seizures.

Objective

To assess whether the amplitude and morphology of IED can serve as surrogate measures of levetiracetam (LEV) efficacy in children with drug-resistant epilepsy.

Methods

Patients with refractory epilepsy, who were treated with oral LEV during a 5-year period, and who had at least 24 h long-term video-EEG recording (LTVEEG) before LEV was initiated and after attaining maximal dosage, were retrospectively selected from the database of the pediatric epilepsy clinic and LTVEEG monitoring unit. IED kurtosis (degree of peakedness), skewness (degree of asymmetry), and below-curve-area (area of the IED) were estimated by a signal processing analysis program.

Results

IED analysis in 20 patients aged 3.6 months to 15 years at maximal LEV dosage revealed increased kurtosis, increased skewness positivity and decreased below-curve-area compared to pre-treatment values. Visual analysis revealed a mean IED amplitude decrease in 30% of the patients, mainly in patients with generalized seizures. Neither the changes in IED measures nor the decrease of the visually determined IED amplitude correlated with decreased seizure frequency.

Conclusions

IED amplitude, kurtosis, skewness and below-curve-area were not correlated with seizure outcome after LEV initiation. Further studies with larger samples and additional seizure outcome measures are warranted.

Keywords

drug-resistant epilepsy levetiracetam interictal epileptiform discharges EEG kurtosis

Introduction

Seizure frequency is currently the measure of antiseizure medications (ASMs) efficacy in pediatric patients with epilepsy. Usually it is derived from seizure diaries that are completed by the patients and their caregivers. However, seizure reports by parents are inaccurate, especially in young children, and certain types of seizures such as absence seizures and infantile spasms can be missed altogether.¹ Seizure frequency can be assessed more accurately by long-term video-EEG recording (LTVEEG), but high cost and prolonged waiting periods make this test less accessible.^1–3 Attempts were therefore made to identify measures of ASM efficacy in the ambulatory EEG.

Interictal epileptiform discharges (IEDs) can be demonstrated in 90% of patients with epilepsy in repeated EEG recordings.^4,5 IED frequency is the most commonly studied interictal measure of ASM efficacy in patients with epilepsy,^6–10 and IED suppression in a repeat pediatric EEG is usually interpreted as a reduced tendency towards seizures.⁶ A decrease in IED frequency was reported for several ASMs^7,8,10,11 but controversial results were reported for others,^6,8,11–13 and although IED suppression may improve cognitive functions and hence the daily lives of patients with epilepsy,¹⁴ its relation to ASM efficacy was questioned.^6,15 There are only scarce reports on the assessment of other possible surrogate markers of ASM efficacy, such as IED propagation, delimitation of the spatial extent of interictal spiking activity and EEG background spectral changes.^16–19

Previously, we observed IED amplitude reduction accompanied by significant clinical improvement following intravenous methylprednisolone in children with refractory epilepsy.²⁰ IED amplitude reduction was observed in several children with sleep- enhanced focal epileptiform activity who were considered to be optimally treated following levetiracetam (LEV) introduction,¹⁶ and resection of brain regions with high kurtosis were significantly correlated with 1-year seizure outcome in children with epilepsy.²¹ LEV, a broad-spectrum ASM, was able to selectively suppress neuronal spike synchronization in vitro,²² but to our best knowledge, its effect on IED measures in the human EEG was not studied before. In this study, we aimed to assess whether kurtosis, skewness and below-curve-area, IED morphological measures that were not studies so far in relation to ASM efficacy, can serve as surrogate measures of LEV efficacy in children with drug-resistant epilepsy.

Methods

Subjects

The study was approved by the Ethics Committee at our Medical Center.

Patients with refractory epilepsy, who were treated with oral LEV during a 5-year period between January 2007 and January 2012, were retrospectively selected from the database of the pediatric epilepsy clinic and LTVEEG unit at our Medical Center. Of those, only patients who had IEDs on ≥24 h LTVEEG recording before LEV was initiated and after attaining maximal LEV dosage were included in the study. Based on these criteria, one patient without IEDs on the second LTVEEG was excluded.

The computerized medical records of the patients were reviewed and the data collected included age and duration of epilepsy at treatment initiation, gender, epilepsy syndrome, developmental delay, MRI abnormalities, family history of epilepsy, concomitant ASMs, highest LEV dosage (total and per weight), seizure frequency before initiation of LEV and at maximal dosage, changes in neurological and general function, reasons for LEV discontinuation, and side effects. Epilepsy syndromes were classified based on criteria of the International League Against Epilepsy.²³

LEV efficacy was defined as reduction of seizure frequency and was determined according to seizure frequency during the month prior to treatment initiation and the month after the maximal dosage of LEV was attained based on seizure diaries. Cognitive, behavioral or functional improvement was determined according to reports of the parents and the school teacher. Patients were regularly evaluated for seizure response, functional status and adverse events every month by phone, at LTVEEG admissions, and every 6 months during a visit at the Epilepsy Clinic.

Video-EEG

EEGs were recorded from 20 scalp electrodes placed according to the international 10–20 system, and four additional electrodes were used to monitor eye movement, ECG signal and EMG activity. The EEG was collected with a reference located between Cz and Pz, sampling rate of 256 Hz, and band pass of 0.1–100 Hz. The EEG signal was transmitted from the patients’ rooms to the EEG reading room through a telemetry system (Biologic Vision Netlink EEG collection system), where it was stored on hard disk. Montages consisted of 21 channels, with the possibility of different referential and bipolar reconstructions.

EEG Analysis

Visual analysis of continuous referential EEG was performed every 24 h by one of the authors (E.H), and several epochs of a total of 10 min of awake and sleep interictal EEG containing the highest voltage IEDs were saved from the first day of each recording for further analysis. IED was defined as a single spike, sharp wave, and spike/sharp wave-slow wave complex.

For each patient, both visual analysis and calculation of IED measures, performed by one of the authors (R.G.M), involved the EEG channel with the highest IED amplitude on the pre-LEV LTVEEG recordings. In patients with independent IEDs in two different foci, each was separately analyzed. For spike/sharp wave-slow wave complexes, the spike/sharp wave and the slow wave components were separately evaluated. The analyzed IEDs on the pre- and post LEV recordings were from the same eyes-open or eyes-closed condition in each patient.

Calculation of IED Measures

A single IED with the highest peak-to-peak amplitude was chosen from each EEG sample based on measurement with the amplitude scale cursor of the EEG program.

Additional IED measures were estimated by a signal processing analysis program, and included kurtosis, skewness and below-curve-area.

Kurtosis was defined as the degree of peakedness of a distribution relative to the normal distribution²⁴:

Kurtosis = μ^{4} / σ^{4}

Where: µ⁴ = ∑ _i (y_i-ŷ)⁴ / N; i = 1, n; σ⁴ = ∑ _i (y_i-ŷ)² / N ; i = 1, n; y_i – samples points.

Kurtosis value of 3 designated a normal-like curve, values higher than 3 indicated a relatively high peak (leptokurtic), and relatively flat curves (platykurtic) resulted in values lower than 3.

Skewness evaluated the degree of departure from symmetry compared to a normal distribution curve:

Skewness = μ^{3} / σ^{3}

Where: µ³ = (∑ _i (y_i-ŷ)³) / N ; i = 1, n; σ³ = (∑ _i (y_i-ŷ)^3/2) / N; i = 1, n; y_i – samples points.

Skewness was negative when the curve's tail was to the left of the maximum and positive when the tail was to the right. A symmetrical shape results in zero skewness.

Below-curve-area evaluated the area between the curve and a horizontal axis at the minimum value of the IED. Below-curve-area was calculated as:

Below the Curve Area (mVolt / cm) = \sum i (y_{i} – y_{min}); i = 1, n

y_i = amplitude values of curve; y_min= amplitude value of the lowest point of the curve when the curve was positive deflected and amplitude of the highest point of the curve when the curve was negative deflected.

Statistical Analysis

Statistical analysis was done by SPSS version 15, Chicago, IL. Pearson Chi-Square and Fisher's exact test were used to test categorical variables. Continuous variables were analyzed using t-test, and the Bonferroni correction was applied. Statistical significance was determined at p < .05.

Results

Patients

Demographic and clinical characteristics before LEV initiation are presented in Table 1. LEV starting dosage was 10 mg/kg/day given twice a day in two equal doses, and was increased every week up to a maximum of 18 to 76 mg/kg/day (mean- 45.9 ± 14.9 mg/kg/day), depending on clinical response and tolerability. Seizure frequency decreased in 13 (65%) patients, including 6(30%) with seizure freedom (Table 2). Side effects occurred in 7 (35%) patients and were mainly behavioral (n = 6), including irritability (n = 3), aggression (n = 2) and emotional instability (n = 1). LEV was discontinued in 4 (20%) patients due to inefficacy (n = 2) and side effects (n = 2) (Table 2).

Table 1.

Demographic and clinical characteristics before levetiracetam initiation.

Age (n = 20)	6.5 ± 4.6 yearsRange 3.6 Months- 15 Years
Gender boys/girls (n = 20)		15(75)/5(25)
Seizure type (n = 20)	Focal	14(70)
Seizure type (n = 20)	Generalized	6(30)
Developmental delay (n = 20)		13(65)
Family history of epilepsy (n = 20)		6(30)
Brain MRI (n = 18)	Abnormal	6(33)
Epilepsy duration (n = 20)	1.50 ± 1.43 years range 2.5 months – 5 years
Concomitant ASMs (n = 20)		18(90)

n (%).

ASMs – antiseizure medications.

Table 2.

Clinical Outcome at Maximal Levetiracetam Dosage.

Seizure outcome (n = 20)	Seizure frequency decreased	13 (65)
	Seizure freedom	6 (30)
	Seizure frequency unchanged	7 (35)
Functional improvement (n = 2)	Decreased fatigue	1 (5)
Functional improvement (n = 2)	Improved communication	1 (5)
Side effects (n = 7)	Irritability	3 (15)
	Aggression	2 (10)
	Emotional instability	1 (5)
	Increased appetite	1 (5)
	Headache	1 (5)
	Fatigue	1 (5)
LEV discontinuation (n = 4)	Inefficacy	2 (10)
LEV discontinuation (n = 4)	Side effects	2 (10)

n (%).

LEV- levetiracetam.

EEG

According to visual analysis, mean IED amplitude decreased in 6 (30%) patients. There was no amplitude change in 14 (70%). Overall, mean IED amplitude decreased from 379.5 ± 204.5 microvolts (range- 120-800) before LEV treatment to 333.5 ± 266.3 microvolts (range- 120-1000) at maximal LEV dosage. Decreased IED amplitude was found in 4/6 (66.7%) patients with generalized seizures and 2/14 (14.3%) with focal seizures (p = .037; after applying Bonferroni correction- p = .44), and was unaffected by gender (p = 1.0), age at epilepsy onset (p = .193), developmental delay (p = 0. 613), family history of epilepsy (p = 1.0), the presence or number of current ASMs (p = .757), brain MRI abnormalities (p = .342), age at LEV initiation (p = .365), LEV daily dosage- total (p = .4) and per weight (p = .931), and LEV-related functional improvement (p = .521). There was no correlation between the decrease of IED amplitude and decreased seizure frequency (p = 1.0).

IED measures (Table 3) were calculated in 13 patients with one focus of spike-slow complexes and in 7 patients with one focus of spikes. Two of the patients had additional independent spike-slow complexes (n = 1), and additional independent spikes (n = 1). After attaining maximal LEV dosage, there was no correlation between decrease in seizure frequency and the decrease in below-curve-area (independent spikes- p = 1.0, slow waves- p = .103), increased skewness positivity (spikes- p = 1.0, slow waves – p = .592), and increased kurtosis (independent spikes- p = 1.0, slow waves- p = 1.0).

Table 3.

Interictal Epileptiform Discharge Measures Before Levetiracetam and After Attaining Maximal Dosage.

		LEV Treatment
		Before	At Maximal Dosage
Kurtosis	Spike 1 (n = 20)	1.79 ± 0.27 (1.42, 2.38)	1.85 ± 0.33 (1.32, 2.55)
	Spike 2 (n = 2)	2.02 ± 0.57 (1.61,2.42)	2.04 ± 0.52 (1.67, 2.41)
	Slow wave 1 (n = 13)	2.25 ± 0.68 (1.53, 4.02)	2.48 ± 0.92 (1.42, 4.48)
	Slow wave 2 (n = 1)	1.71	2.23
Skewness	Spike 1 (n = 20)	−0.14 ± 0.34 (−0.77, 0.40)	−0.03 ± 0.35 (−0.44, 0.74)
	Spike 2 (n = 2)	−0.36 ± 0.21 (−0.51, −0.21)	0.02 ± 0.32 (−0.21, 0.24)
	Slow wave 1 (n = 13)	0.51 ± 0.4 (0.07, 1.32)	0.63 ± 0.51 (−0.28, 1.24)
	Slow wave 2 (n = 1)	0.07	0.88
Below-curve-area	Spike 1 (n = 20)	3701.9 ± 3075.4 (426.2, 11 683.2)	3310.1 ± 1960.7 (120.7, 7189.5)
	Spike 2 (n = 2)	6144.4 ± 6031.1 (1879.7, 10 409)	5792.6 ± 1772.2 (4539.5, 7045.7)
	Slow wave 1 (n = 13)	15 003.6 ± 7868.5 (3761.7, 25 217.2)	10 013.8 ± 7780.5 (2068, 30 132)
	Slow wave 2 (n = 1)	32 450.4	2534.8

Mean ± SD (Minimum, Maximum).

LEV- levetiracetam.

Discussion

In this study, treatment with LEV led to decreased seizure frequency in 65% of our patients, including seizure freedom in 30%. No correlations were found between seizure outcome and changes in IED amplitude, kurtosis, skewness and below-curve-area. IED amplitude decrease was more frequent among patients with generalized seizures. To our best knowledge, this is the first study that evaluated kurtosis, skewness and below-curve-area in relation to ASM influence.

The amplitude of the IEDs recorded at the scalp surface is a quantitative parameter that is influenced by voltage of the cortical discharge, the location of the dipole generators in relation to the cortical convolutions, the magnitude of the source area and the degree of synchrony. In this study we employed computed evaluation of IED morphological measures as it can supplement visual inspection.²⁵ Both IED amplitude and below-curve-area decreased in our patients after attaining maximal LEV dosage. A decrease of IED amplitude after the initiation of ASMs was reported by Larsson et al, who used electric source imaging to visualize the propagation of interictal epileptiform activity between different brain regions after LEV introduction.¹⁶ However, in their study, only IED propagation was considered an indicator of effective ASM treatment. Although LEV is a broad-spectrum ASM, there are previous reports on differing LEV effects on seizure frequency according to seizure type,^26,27 but none could be found regarding IED amplitude change, as observed in our patients.

We found increased skewness positivity (a steeper spike's up-going slope or first half-wave) of the spike and slow wave after LEV initiation, and the spike and slow wave kurtosis (degree of peakedness) increased. Previous studies on the morphology of spikes yielded controversial results. A steeper slope of the spike's first half-wave was considered by Gloor a feature that differentiated epileptiform discharges from other sharply contoured waves on the EEG.²⁸ A similar finding was described by Lemieux and Blume in half of the epileptiform potentials on electrocorticograms.²⁹ However, Gotman found that in 30% of the spikes and sharp waves that had asymmetrical slopes, the second half-wave slope was steeper than the first.³⁰ According to Blume and Lemieux,³¹ when a significant inter-half-wave difference was present for the top half of the spike, the second half wave was steeper than the first, while the opposite was observed for spikes with asymmetries of the whole slop. In this study, the second half-wave amplitude value exceeded the first half-wave amplitude in bisynchronous spikes.

In this study, the changes in IED measures and the decrease of the visually determined IED amplitude did not correlate with decreased seizure frequency. According to in vitro studies, ictal discharges were abolished by carbamazepine, phenytoin, phenobarbital and valproic acid at concentration that did not influence interictal discharges in the rat hippocampus.^32,33 Similar findings were observed with carbamazepine, valproic acid and topiramate in slices from rat entorhinal cortex and were related to activity-dependent characteristics (ie discharge duration) rather than GABA_A- receptor antagonism.¹⁵ In human studies, the length and number of EDs with duration > 30 s was significantly reduced in young patients (4-21 years) with generalized drug-resistant epilepsy after the initiation of lamotrigine, while single IEDs were not affected, and a ≥ 50% seizure reduction concomitant with the reduction of EDs on the EEG was observed in 42% of the patients.¹¹ In addition, the effect of a single ASM on IED frequency and morphology is not uniform. Compared to pre-LEV treatment, IED frequency and/or amplitude remained unchanged or even increased after the initiation of treatment with LEV in 6 of 17 patients with pediatric epilepsy, IED frequency or amplitude decreased in 5 patients and the epileptiform activity completely disappeared in 6 patients.¹⁶

Seizure outcome was better in our study compared to previous reports. Seizure freedom was found in 30% of our patients compared to reported rates of 5% to 22% and a decrease in seizure frequency was noted in 65% of our patients compared to 45% to 52% in previous studies.^26,34–36 We attribute these differences to the relatively short clinical follow-up in our patients. The rate of side effects in our patients (35%) was similar to rates reported by others (up to 33%),^26,34,35 and may reflect similar LEV titration rates. According to a meta-analysis that included 3174 children who were treated with LEV (median initial dose - 10 mg/kg/day, median final dose- 60 mg/kg/day),³⁷ the rate of side effects was higher (47%), the most common side effects were psychiatric and neurologic, and LEV was discontinued in 0.9% of the children on LEV monotherapy and 4.5% of those on polytherapy, mostly due to behavioral problems. Among the behavioral side effects, LEV-treated children had a higher risk of developing aggression, hostility and nervousness compared to those who were not using LEV.³⁸

Our study has several limitations. This was a retrospective study and the lack of a control group may have affected the control for confounding factors. The study group consisted of a typical patient population with refractory epilepsy in a tertiary pediatric epilepsy center and was small and heterogeneous regarding age, seizure type, epilepsy duration, and the presence of developmental delay. A greater baseline clinical and electrical similarity (i.e. one epilepsy syndrome) may have yielded a more uniform influence on IED measures. Seizure frequency assessment was based on seizure diaries and these may be inaccurate. Authors E.H and R.G.M were not blinded when collecting and analyzing the EEG data due to practical reasons. In addition, most of the patients in this retrospective study were treated with polytherapy and LEV was added to other ASMs. Since their dosage was stable throughout the pre-LEV and post-LEV video-EEG investigations, we attributed the changes in IED morphology only to LEV, although possible effect on IED morphology cannot be completely ruled out.

In conclusion, IED amplitude, kurtosis, skewness and below-curve-area were not correlated with seizure outcome after LEV initiation. Since the identification of a surrogate marker in the interictal EEG can possibly increase accuracy and reduce the associated resources of current measures used to determine ASM efficacy, EEG acquisition and analysis protocols should be standardized and prospective studies with larger samples, sub-analysis according to epilepsy type, IED location, and additional seizure outcome measures, e.g. seizure severity, are warranted.

Footnotes

Acknowledgements

We thank Ilana Gelernter, M.A., from the statistical laboratory in the School of Mathematics, Tel-Aviv University, for the statistical analysis.

ORCID iD

Revital Gandelman-Marton

Ethics Approval

The study was approved by the local Ethics Committee at our medical center.

Consent

Waived by the Ethics Committee due to the retrospective nature of the study.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability

The authors confirm that the data supporting the findings of this study are available within the paper.

References

Akman

Montenegero

Jacob

Eck

Chiriboga

Gilliam

. Seizure frequency in children with epilepsy: Factors influencing accuracy and parental awareness. Seizure. 2009;18(7):524-529.

Modur

Rigdon

. Diagnostic yield of sequential routine EEG and extended outpatient video-EEG monitoring. Clin Neurophysiol. 2008;119(1):190-196.

McGonigal

Russell

Mallik

Oto

Duncan

. Use of short term video EEG in the diagnosis of attack disorders. J Neurol Neurosurg Psychiatry. 2004;75(5):771-772.

Worrell

Lagerlund

Buchhalter

. Role and limitations of routine and ambulatory scalp electroencephalography in diagnosing and managing seizures. Mayo Clin Proc. 2002;77(9):991-998.

Pillai

Sperling

. Interictal EEG and the diagnosis of epilepsy. Epilepsia. 2006;47(Suppl 1):14-22.

Libenson

Caravale

. Do antiepileptic drugs differ in suppressing interictal epileptiform activity in children? Pediatr Neurol. 2001;24(3):214-218.

Placidi

Tombini

Romigi

, et al. Topiramate: Effect on EEG interictal abnormalities and background activity in patients affected by focal epilepsy. Epilepsy Res. 2004;58(1):43-52.

Marciani

Spanedda

Bassetti

, et al. Effect of lamotrigine on EEG paroxysmal abnormalities and background activity: A computerized analysis. Br J Clin Oharmacol. 1996;42(5):621-627.

Stodieck

Steinhoff

Kolmsee

Van Rijckevorsel

. Effect of levetiracetam in patients with epilepsy and interictal epileptiform discharges. Seizure. 2001;10(8):583-587.

10.

Zupanc

Roell Werner

Schwabe

, et al. Efficacy of felbamate in the treatment of intractable pediatric epilepsy. Pediatr Neurol. 2010;42(6):396-403.

11.

Eriksson

Knutsson

Nergårdh

. The effect of lamotrigine on epileptiform discharges in young patients with drug-resistant epilepsy. Epilepsia. 2001;42(2):230-236.

12.

Duncan

Smith

Forster

Shorvon

Trimble

. Effects of the removal of phenytoin, carbamazepine, and valproate on the electroencephalogram. Epilepsia. 1989;30(5):590-596.

13.

Duncan

. Antiepileptic drugs and the electroencephalogram. Epilepsia. 1987;28(3):259-266.

14.

Binnie

. Cognitive impairment during epileptiform discharges: Is it ever justifiable to treat the EEG? Lancet Neurol. 2003;2(12):725-730.

15.

D'Antunono

Köhling

RIcalzone

Gotman

Biagini

Avoli

. Antiepileptic drugs abolish ictal but not interictal epileptiform discharges in vitro. Epilepsia. 2010;51(3):423-431.

16.

Larsson

Eeg-Olofsson

Michel

Seeck

Lantz

. Decrease in propagation of interictal epileptiform activity after introduction of levetiracetam visualized with electric source imaging. Brain Topogr. 2010;23(3):269-278.

17.

Mattia

Spanedda

Bassetti

, et al. Gabapentin as add-on therapy in focal epilepsy: A computerized EEG study. Clin Neurophysiol. 2000;111(2):311-317.

18.

Kim

Lim

Kim

, et al. Changes in background electroencephalography and regional cerebral glucose metabolism in focal epilepsy patients after 1-month administration of levetiracetam. Neuropsychiatr Dis Treat. 2015;11:215-223.

19.

Lin

Ouyang

Chiang

Yang

. Early prediction of medication refractoriness in children with idiopathic epilepsy based on scalp EEG analysis. Int J Neural Syst. 2014;24(7):1450023.

20.

Heyman

Lahat

Gandelman-Marton

. Interictal encephalography can influence patient selection for methylprednisolone therapy in pediatric refractory epilepsy. J Child Neurol. 2012;27(2):162-167.

21.

Akiyama

Osada

Isowa

, et al. High kurtosis of intracranial electroencephalogram as a marker of ictogenicity in pediatric epilepsy surgery. Clin Neurophysiol. 2012;123(1):93-99.

22.

Klitgaard

Matagne

Grimee

Vanneste-Goemaere

Margineanu

. Electrophysiological, neurochemical and regional effects of levetiracetam in the rat pilocarpine model of temporal lobe epilepsy. Seizure. 2003;12(2):92-100.

23.

Berg

Berkovic

Brodie

, et al. Revised terminology and concepts for organization of seizures and epilepsies: Report of the ILAE commission on classification and terminology, 2005-2009. Epilepsia. 2010;51(4):676-685.

24.

DeCarlo

. On the meaning and use of kurtosis. Psychol Methods. 1997;2(3):292-307.

25.

Santiago-Rodriguez

Harmony

Cardenas-Morales

Hernandez

Fernandez-Bouzas

. Analysis of background EEG activity in patients with juvenile myoclonic epilepsy. Seizure. 2008;17(5):437-445.

26.

Lee

Kang

Kim

Lee

. Efficacy and safety of adjunctive levetiracetam therapy in pediatric intractable epilepsy. Pediatr Neurol. 2010;42(2):86-92.

27.

Glauser

Dulac

. Preliminary efficacy of levetiracetam in children. Epileptic Disord. 2003;5(Suppl 1):S45-S50.

28.

Gloor

. The EEG and differential diagnosis of epilepsy. In: Van Duijn

Donker

DNJ

Van Huffelen

, eds. Current concepts in clinical neurophysiology. Drukkerij Trio; 1977:9-21.

29.

Lemieux

Blume

. Automated morphological analysis of spikes and sharp waves in human electrocorticograms. Electroencephal Clin Neurophysiol. 1983;55(1):45-50.

30.

Gotman

. Quantitative measurements of epileptic spike morphology in the human EEG. Electroencephal Clin Neurophysiol. 1980;48(5):551-557.

31.

Blume

Lemieux

. Morphology of spikes in spike-and-wave complexes. Electroencephal Clin Neurophysiol. 1988;69(6):508-515.

32.

Fueta

Avoli

. Effects of antiepileptic drugs on 4-aminopyridine-induced epileptiform activity in young and adult rat hippocampus. Epilepsy Res. 1992;12(3):207-215.

33.

Brückner

Heinemann

. Effects of standard anticonvulsant drugs on different patterns of epileptiform discharges induced by 4-aminopyridine in combined entorhinal cortex-hippocampal slices. Brain Res. 2000;859(1):15-20.

34.

Glauser

Pellock

Bebin

, et al. Efficacy and safety of levetiracetam in children with partial seizures: An open-label trial. Epilepsia. 2002;43(5):518-524.

35.

Lagae

Buyse

Deconinck

Ceulemans

. Effect of levetiracetam in refractory childhood epilepsy syndromes. Eur J Paed Neurol. 2003;7(3):123-128.

36.

Tonekaboni

Ghazavi

Karimzadeh

Mahvelati

Ghofrani

. Efficacy of levetiracetam in children with refractory epilepsy as an add-on trial. Epilepsy Res. 2010;90(3):273-277.

37.

Egunsola

Choonara

Sammons

. Safety of levetiracetam in paediatrics: A systematic review. PLoS One. 2016;11(3):e0149686.

38.

Halma

de Louw

Klinkenberg

Aldenkamp

IJff

Majoie

. Behavioral side-effects of levetiracetam in children with epilepsy: A systematic review. Seizure. 2014;23(9):685-691.