SEARCH

SEARCH BY CITATION

Keywords:

  • Typical absence seizures;
  • Childhood absence epilepsy;
  • Early onset absence epilepsy;
  • Antiepileptic drugs;
  • GLUT1 deficiency

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Appendix 1

Purpose

To investigate whether patients with typical absence seizures (TAS) starting in the first 3 years of life, conformed to Panayiotopoulos's definition of childhood absence epilepsy (CAE), show different electroclinical course than those not fulfilling CAE criteria.

Methods

In this multicenter retrospective study, we choose a fixed duration follow-up of 36 months to examine the electroclinical course of epilepsy in all children with TAS starting before 3 years of age. The probands who fulfilled Panayiotopoulos's criteria for CAE were classified as having pure early onset absence epilepsy (P-EOAE), whereas those who did not as nonpure EOAE (NP-EOAE). In addition, these two groups of patients were further stratified according to the number of antiepileptic drugs taken to obtain initial seizure control (mono-, bi-, and tritherapy).

Key Findings

Patients with P-EOAE (n = 111) showed earlier initial seizure control (p = 0.030) and better seizure-free survival curve (p = 0.004) than those with NP-EOAE (n = 77). No mutation in SLC2A1 gene or abnormal neuroimaging was observed in P-EOAE. Among patients with NP-EOAE, those receiving tritherapy showed increased risk of structural brain abnormalities (p = 0.001) or SLC2A1 mutations (p = 0.001) but fewer myoclonic features (p = 0.031) and worse seizure-free survival curve (p = 0.047) than those treated with mono- and bitherapy. Children with NP-EOAE had 2.134 the odds of having relapse during the follow-up compare to those with P-EOAE.

Significance

Children with early onset TAS who did meet Panayiotopoulos's criteria showed a favorable course of epilepsy, whereas patients not fulfilling Panayiotopoulos's criteria showed increased risk of relapse at long-term follow-up.

Typical absences seizures (TAS) are generalized seizures of sudden onset and termination, usually lasting for seconds up to minutes. The main clinical manifestation is impairment of consciousness with concomitant bilateral, regular, symmetric, and generalized 3–4 Hz spike-waves on electroencephalography (EEG) (Berg et al., 2010). These seizures usually occur in the context of idiopathic (genetic) generalized epilepsies (IGE), such as childhood absence epilepsy (CAE), epilepsy with myoclonic absences, juvenile absence epilepsy, and juvenile myoclonic epilepsy (Panayiotopoulos, 2008).

Onset of TAS is frequently limited between 4 and 9 years of age and the youngest age has been set at 3 years (Panayiotopoulos, 2008). Nevertheless, clinical series of children with TAS before 3 years of age have been reported (Shahar et al., 2007; Caraballo et al., 2011; Giordano et al., 2011; Verrotti et al., 2011; Agostinelli et al., 2013). In the first 3 years of life, TAS may occur in different epileptic syndromes, such as CAE of early onset, benign myoclonic epilepsy of infancy, eyelid myoclonia with absences, and epilepsy with myoclonic absence (Caraballo et al., 2011). Recently, we showed that the application of strict criteria for CAE, suggested by Panayiotopoulos, leads to a group of children showing homogeneous electroclinical features with response to therapy and prognosis, similar but not identical to CAE (Giordano et al., 2011; Verrotti et al., 2011; Agostinelli et al., 2013). In contrast, some authors, using broad inclusion criteria, observed a variable epilepsy outcome, ranging from complete control with first antiepileptic drug (AED) monotherapy to severe refractoriness despite polytherapy (Suls et al., 2009; Arsov et al., 2012). More importantly, mutations in the SLC2A1 gene encoding the glucose transporter type 1 (GLUT1), have been found in >10% of children (Suls et al., 2009; Arsov et al., 2012).

The purpose of this multicenter study was to investigate if patients with TAS starting in the first 3 years of life but otherwise conformed to CAE definition had a different electroclinical course than those not fulfilling CAE criteria. Occurrence of SLC2A1 mutations and comparisons between children with different AED initial responses were evaluated. Another goal was to analyze multiple electroclinical factors to identify patients who had the strongest risk of poor prognosis with seizure relapse during follow-up.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Appendix 1

Setting

In this retrospective study, we evaluated children with TAS starting before the age of 3 years who were screened for SLC2A1 gene mutations at the department of molecular genetics, Institute “G. Gaslini,” Genoa, Italy.

Patients were recruited and followed, between January 2001 and February 2013, at 16 different epilepsy and child neurology centers in Italy (Appendix 1).

Written informed consent was provided by all parents or guardians. The study was approved by the ethics committee of each institution.

Study design and patients

We chose a fixed duration follow-up of 36 months for all children (instead of a variable length) to facilitate outcome comparisons at a fixed interval and to examine the course of epilepsy of all children during the entire follow-up.

Clinical records were reviewed to obtain information including sex, age of TAS onset, additional seizure types (myoclonic attacks and/or generalized tonic–clonic seizures), previous febrile seizures (FS), first-degree family history of IGE, neurologic examination, EEG data, brain magnetic resonance imaging (MRI) data, SLC2A1 gene analysis, treatment, and outcome variables.

Patients were first evaluated in accordance with the Panayiotopoulos's definition of CAE (Panayiotopoulos, 2008), except for age criterion (i.e., onset between 4 and 10 years), namely: (1) age at onset within the first 3 years; (2) normal neurologic state and development; (3) brief (4–20 s) and frequent (many per day) absence seizures with abrupt and severe impairment of consciousness; and (4) EEG ictal discharges of generalized high-amplitude spike and double or maximum triple spike and slow-wave complexes, which are rhythmic at 3–4 Hz, with a gradual and regular slowdown from the initial to the terminal phase of the discharge. Exclusion criteria were the following: (1) other types of seizure, such as generalized tonic–clonic seizures, or myoclonic jerks prior to or during the active stage of absences; (2) eyelid myoclonia, perioral myoclonia, rhythmic massive limb jerking, and single or arrhythmic myoclonic jerks of the head, trunk, or limbs; (3) mild or no impairment of consciousness during the 3–4 Hz discharges; (4) very brief (<4 s) EEG 3–4 Hz spike-wave paroxysms, polyspikes (more than three) or ictal discharge fragmentations; and (5) visual (photic) and other sensory precipitation of clinical seizures.

The probands who fulfilled our modified Panayiotopoulos's criteria were classified as pure early onset absence epilepsy (P-EOAE), and those who did not as nonpure EOAE (NP-EOAE). In particular, children were considered having NP-EOAE if they presented at least one of the following inclusion criteria: abnormal neurologic state and development; very brief (<4 s) 3–4 Hz spike-waves, polyspikes (more than three); no loss of consciousness; other seizure type; myoclonic features; photic and other sensory precipitation. In addition, patients with P-EOAE and NP-EOAE were subdivided according to AED initial response, as follows: children who responded well to the first AED and became seizure-free (monotherapy group); children who became clinically seizure-free with add-on treatment of a second AED (bitherapy group); and children who stopped having seizures with association of three different AEDs (tritherapy group).

Measures and procedures

Neurologic examination was classified as abnormal if the following findings were identified: hypertonia, hypotonia, microcephaly, or motor symptoms such as dystonic movements and ataxia.

All patients underwent repeated awake-sleep standard EEG and/or video-EEG, which routinely included photic stimulation and hyperventilation, if the child cooperated. The EEG recordings were made using Ag-AgCl scalp electrodes positioned according to the International 10–20 System. Signals were acquired by computerized systems and recorded using montages with a common reference electrode. Both EEG recordings at onset of TAS and those with a time interval of 6 months were considered. EEG recordings were classified as normal or abnormal and scored for the presence of epileptiform abnormalities (i.e., generalized 3–4 Hz spike-waves; focal or generalized discharges of spikes, spike-waves, or polyspikes) by clinical epileptologists of each institution.

Brain MRI records were reexamined and categorized as normal, when all findings were normal or incidental (e.g., cysts; mild cortical atrophy; single small [<5 mm] subcortical focal hyperintensities), or definitely abnormal, when at least one finding was abnormal (e.g., abnormal white matter signal; focal cortical dysplasia; gray matter heterotopia; congenital midline abnormality; arachnoid cyst; focal decreased volume; enlarged lateral ventricle; cortical atrophy; periventricular leukomalacia; large [≥5 mm] or numerous subcortical and cortical focal hyperintensities; presence of structural brain abnormalities).

Mutation analysis of the SLC2A1 gene was performed as described (Striano et al., 2012). Briefly, genomic DNA was extracted from ethylenediaminetetraacetic acid (EDTA) anticoagulant blood samples and the coding (10 exons and exon/intron junctions) and promoter regions of SLC2A1 were sequenced using BigDye chemistry on ABI3730XL DNA analyzer (Applied Biosystems, Carlsbad, CA, U.S.A.). Multiplex ligation-dependent probe amplification (MLPA), using probemix P138-B1 (MRC Holland, Amsterdam, The Netherlands) was performed to detect deletions or duplications.

Electroclinical definitions

The followed electroclinical parameters were evaluated: time at absence control (time from the beginning of AED therapy to obtaining initial complete TAS control); number of AEDs to stop absence (number of AED necessary to achieve initial TAS control); seizure relapse (each seizure occurring after the initial stage of TAS control, with or without AED therapy); time at relapse; terminal remission (time from the last seizure to the end of follow-up); seizure freedom ≥24 months (complete absence of seizures with or without AEDs for ≥24 months at the final 36-month evaluation); duration therapy (cumulative time of AED treatment at the end of follow-up); failed withdrawal (recurrence of seizure during or immediately after the tapering or withdrawal of AED); ongoing therapy (patients who were still on therapy at the end of follow-up).

Statistical analysis

Data were analyzed using SPSS 17.0 (SPSS Inc., Chicago, IL, U.S.A.). Normal distribution of variables was verified with the Kolmogorov-Smirnov test. Descriptive statistics were expressed as means (medians and ranges) for continuous variables and as percentages for categorical variables. Comparisons of continuous data were performed with unpaired Student's t-test and Mann-Whitney U test, and those of categorical data with chi-square test. One-way analysis of variance (ANOVA) followed by Bonferroni's correction was used to compare continuous variables among the categories of NP-EOAE with different AED initial response (monotherapy, bitherapy, tritherapy groups). Differences between categories for the data that were not normally distributed variables were carried out with Kruskal-Wallis test. For the risk of relapse, Kaplan-Meier survival statistics were performed. Log-rank (Mantel-Cox) was used to test for equality of the survival distributions for the different groups. Binary logistic regression analysis was computed to evaluate the influence of each independent variable on the outcome measurement (seizure relapse). All variables significantly associated with seizure relapse at the bivariate level were entered into multivariate logistic regression (forced entry method), and those not significant were removed to obtain the most parsimonious model. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were calculated. Statistical significant was defined as a p-value < 0.05.

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Appendix 1

Entire cohort

A total of 188 patients (115 (61.2%) male) were analyzed. Mean age at onset of TAS was 23.7 months (median 25.5; range 8.0–36.0 months), and mean age at last contact was 59.7 months (median 61.5; range 44.0–72.0 months). Thirty-six patients (19.1%) had febrile seizures before onset of TAS. A family history of IGE was reported in first-degree relatives in 61 patients (32.4%). No absence status epilepticus was observed in the entire cohort. Brain MRI was abnormal in 21 patients (11.2%) and SLC2A1 gene was mutated in 4 (2.1%). All children received AEDs and achieved absence control in monotherapy (n = 116; 61.7%), bitherapy (n = 61; 32.4%), or tritherapy (n = 11; 5.9%). Valproate (VPA), ethosuximide (ESM), lamotrigine (LTG), and levetiracetam (LEV) were used, alone or in combination, in 120 (63.8%), 75 (39.9%), 41 (21.8%), and 36 (19.1%) patients, respectively. At follow-up, 117 children (62.2%) were still on therapy, and seizure freedom of ≥24 months was observed in 147 subjects (78.2%).

Comparisons between P-EOAE and NP-EOAE

One hundred eleven children (59.0%) met Panayiotopoulos criteria (P-EOAE) and 77 (41.0%) did not (NP-EOAE). Therefore, early onset absences appeared to be a syndrome in itself or to be part of different types of epilepsy. Demographic and electroclinical characteristics of the two groups are shown in Table 1. Age at onset in patients with P-EOAE was significantly higher than in those with NP-EOAE (p = 0.009). The age distribution of both groups is illustrated in Figure 1. At the initial stage of epilepsy, children with P-EOAE controlled absences with less AEDs (p = 0.010) and shorter time (p = 0.030) than those with NP-EOAE. No differences were observed with respect to daily AED doses of VPA (p = 0.121), ESM (p = 0.112), LTG (p = 0.085), and LEV (0.973). At 6 months, normal EEG recordings were present in 82 (73.9%) of patients with P-EOAE but only 34 (44.2%) of those with NP-EOAE (p < 0.001). SLC2A1 mutations (p = 0.015) and abnormal brain MRI (p < 0.001) were reported only in NP-EOAE patients (Table 1). The most common type of definite imaging abnormality was subcortical focal hyperintensity of the frontal lobe (n = 11 of 21, 52.4%). Next most common were focal cortical dysplasia (n = 5 of 21, 23.8%), gray matter heterotopia (n = 3 of 21; 14.3%), and abnormalities of white matter signal (n = 2 of 21; 9.5%).

Table 1. Comparisons between children with and without Panayiotopoulos's criteria
 P-EOAE (n = 111)NP-EOAE (n = 77)p-Value
  1. P-EOAE, pure early onset absence epilepsy; NP-EOAE, nonpure early onset absence epilepsy; FS, febrile seizure; IGE, idiopathic generalized epilepsy; AED, antiepileptic drug; VPA, valproate; ESM, ethosuximide; LTG, lamotrigine; LEV, levetiracetam; EEG, electroencephalography.

  2. Data are expressed as mean (median; range) for continuous variables and number (percentage) for categorical variables.

Demographic data   
Female-to-male ratio40/7133/440.345
Age at onset (month)25.0 (26.9; 8.9–36.0)21.8 (22.5; 8.0–36.0)0.009
Previous FS24 (21.6%)12 (15.6%)0.199
Family history of IGE38 (34.2%)23 (29.9%)0.530
Early electroclinical data   
Time at absence control (month)2.0 (2.1; 0.7–4.1)2.4 (2.3; 0.7–4.5)0.030
No of AED at absence control1.3 (1.0; 1.0–2.0)1.6 (1.0; 1.0–3.0)0.010
Daily AED dose (mg/kg)   
VPA25.6 (26.5; 13.6–43.8)29.3 (28.5; 17.8–44.6)0.121
ESM22.3 (22.8; 14.3–31.3)18.2 (21.4; 11.8–30.0)0.112
LTG7.3 (7.3; 6.8–7.8)5.0 (4.4; 2.7–8.6)0.085
LEV27.1 (27.1; 24.1–30.2)26.1 (25.5; 23.5–30.0)0.973
EEG at 6 months (%)   
Normal82 (73.9)34 (44.2)<0.001
Generalized 3–4 Hz spike-waves19 (17.1)21 (27.3)0.069
Generalized abnormalities14 (12.6)24 (31.2)0.003
Focal abnormalities7 (6.3)15 (19.5)0.010
Genetic and neuroimaging data (%)   
SLC2A1 mutation04 (5.2)0.015
Abnormal neuroimaging021 (27.3)<0.001
36-Month outcome data   
Relapse (%)21 (18.9)30 (39.0)0.002
Time at relapse (month)14.7 (14.2; 4.2–29.4)19.2 (20.0; 5.3–24.1)0.082
Terminal remission (month)21.4 (21.6; 6.6–31.8)16.5 (16.0; 1.0–30.7)0.001
Seizure freedom ≥24 months (%)96 (86.5)51 (66.2)0.001
Duration therapy (month)30.8 (31.2; 23.9–36.0)33.4 (36.0; 21.3–36.0)0.017
Failed withdrawal (%)7/63 (11.1)7/30 (24.1)0.106
Ongoing therapy (%)60 (54.1)57 (74.0)0.005
image

Figure 1. Distribution of age of seizure onset in patients with P-EOAE and NP-EOAE.

Download figure to PowerPoint

At 36-month follow-up, children of P-EOAE group had greater terminal remission (p = 0.001) with higher percentage of seizure freedom ≥24 months (p = 0.001) than those of NP-EOAE. Shorter duration of therapy (p = 0.017) and fewer patients still on AEDs (p = 0.005) were observed in the P-EOAE group than in the NP-EOAE. More children with P-EOAE (56.8%) compared to children with NP-EOAE (39.0%) withdrew AED therapy during 36 months (p = 0.016), but no difference appeared with respect to failed withdrawals (p = 0.106).

The Kaplan-Meier survival curves for children with P-EOAE and those with NP-EOAE are shown in Figure 2. There was a significant difference between the two curves (log-rank [Mantel-Cox] χ2 = 8.3, d.f. = 1, p = 0.004). Time to relapse during the 36 months of follow-up did not differ between the two groups (p = 0.082).

image

Figure 2. Kaplan-Meier curves show the probability of seizure-free survival in patients with P-EOAE and NP-EOAE. Test of equality of survival distribution: log-rank (Mantel-Cox) χ2 = 8.3, d.f. = 1, p = 0.004.

Download figure to PowerPoint

Comparisons among different AED response groups in P-EOAE

In children with P-EOAE, 75 (67.5%) responded well and became seizure-free to AED monotherapy and 36 (32.4%) to AED bitherapy. Patients in monotherapy received VPA in 48 (64.0%), ESM in 17 (22.7%), LTG in 7 (9.3%), and LEV in 3 (4.0%); those in bitherapy received VPA in 30 (83.3%), ESM in 21 (58.3%), LTG in 13 (36.1%), and LEV in 8 (22.4%). No significant differences in demographic and electroclinical variables, except for time at initial absence control (p = 0.002) and daily LTG dose (p = 0.024) were observed in the two groups (Table 2).

Table 2. Comparisons between children with P-EOAE on AED monotherapy and bitherapy
 Monotherapy (n = 75)Bitherapy (n = 36)p-Value
  1. FS, febrile seizure; IGE, idiopathic generalized epilepsy; AED, antiepileptic drug; VPA, valproate; ESM, ethosuximide; LTG, lamotrigine; LEV, levetiracetam; EEG, electroencephalography.

  2. Data are expressed as mean (median; range) for continuous variables and number (percentage) for categorical variables.

Demographic data   
Female-to-male ratio26/4914/220.664
Age at onset (month)24.5 (26.5; 8.9–36.0)25.9 (27.8; 9.6–36.0)0.283
Previous FS (%)18 (24.0)6 (16.7)0.267
Family history of IGE (%)23 (30.7)15 (41.7)0.253
Early electroclinical data   
Time at absence control (month)1.9 (2.0; 0.7–4.1)2.4 (2.4; 1.5–3.4)0.002
Daily AED dose (mg/kg)   
VPA28.3 (27.8; 15.9–46.5)26.0 (24.3; 13.6–43.8)0.074
ESM24.2 (24.0; 18.9–29.4)22.3 (20.9; 14.3–31.3)0.294
LTG11.1 (10.3; 5.8–18.3)7.0 (7.5; 4.9–8.7)0.024
LEV28.2 (27.9; 26.7–30.1)27.2 (26.7; 21.9–32.7)0.776
EEG at 6 months (%)   
Normal57 (76.0)25 (69.4)0.462
Generalized 3–4 Hz spike-waves10 (13.3)9 (25.0)0.127
Generalized abnormalities11 (14.7)3 (8.3)0.347
Focal abnormalities4 (5.3)3 (8.3)0.543
36-Month outcome data   
Relapse (%)13 (17.3)8 (22.2)0.422
Time at relapse (month)17.4 (17.4; 4.2–29.4)10.4 (6.5; 4.6–23.0)0.104
Terminal remission (month)31.3 (34.0; 6.6–35.3)31.8 (33.5; 14.0–34.6)0.331
Seizure freedom ≥24 months (%)64 (85.3)32 (88.9)0.608
Duration therapy (month)28.1 (31.2; 13.2–36.0)29.3 (30.2; 14.3–36.0)0.405
Failed withdrawal (%)6/43 (14.0)1/20 (5.0)0.293
Ongoing therapy (%)40 (53.3)20 (55.6)0.826

Using the Kaplan-Meier survival analysis, the difference between curves was not significant (log-rank [Mantel-Cox] χ2 = 0.533, d.f. = 1, p = 0.465) (Fig. 3). In addition, time at relapse did not differ between children treated with monotherapy and those with bitherapy (p = 0.104).

image

Figure 3. Kaplan-Meier curves show the probability of seizure-free survival among patients with P-EOAE based on different initial AED response. Test of equality of survival distribution: log-rank (Mantel-Cox) χ2 = 0.5, d.f. = 1, p = 0.465.

Download figure to PowerPoint

Comparisons among different AED response groups in NP-EOAE

In the NP-EOAE group, 45 (53.2%), 25 (32.5%), and 11 (14.3%) patients achieved initial seizure control, respectively, with AED monotherapy (VPA = 17 [41.5%]; ESM = 7 [17.1%]; LTG = 6 [14.6%]; LEV = 12 [29.3%]), bitherapy (VPA = 16 [64.0%]; ESM = 19 [76.0%]; LTG = 9 [36.0%]; LEV = 6 [24.0%]), and tritherapy (VPA = 9 [81.8%]; ESM = 11 [100%]; LTG = 6 [54.5%]; LEV = 7 [63.6%]). Characteristics of these different AED response categories are listed in Table 3. Significant differences were detected in the following variables: age at onset (p = 000.7); time at initial absence control (p < 0.001); LTG dose (p = 0.001); normal EEG at 6 months (p < 0.001); focal EEG abnormalities at 6 months (p = 0.008); SLC2A1 mutation (p = 0.001); abnormal brain MRI (p = 0.001); terminal remission (p = 0.017); and abnormal neurologic/developmental state (p = 0.001). Subjects treated with tritherapy were more frequently associated with abnormal neurologic/developmental state, abnormal brain MRI, and SLC2A1 mutation than those with monotherapy or bitherapy (p = 0.001 for all between-group comparisons) (Table 3). Myoclonic features were present more frequently in monotherapy (65.9%) or bitherapy (72.0%) groups than in tritherapy group (27.3%) (p = 0.031 for all between-group comparisons).

Table 3. Comparisons between children without Panayiotopoulos's criteria on different therapy
 Monotherapy (n = 41)Bitherapy (n = 25)Tritherapy (n = 11)p-Value
  1. FS, febrile seizure; IGE, idiopathic generalized epilepsy; AED, antiepileptic drug; VPA, valproate; ESM, ethosuximide; LTG, lamotrigine; LEV, levetiracetam; EEG, electroencephalography; CAE, childhood absence epilepsy.

  2. Data are expressed as mean (median; range) for continuous variables and number (percentage) for categorical variables.

  3. a

    Monotherapy versus bitherapy, p = n.s.; monotherapy versus tritherapy, p = 0.002; bitherapy versus tritherapy, p = 0.006.

  4. b

    Monotherapy versus bitherapy, p = 0.042; monotherapy versus tritherapy, p < 0.001; bitherapy versus tritherapy p = 0.001.

  5. c

    Monotherapy versus bitherapy, p = 0.003; monotherapy versus tritherapy, p = 0.002; bitherapy versus tritherapy, p = n.s.

  6. d

    Monotherapy versus bitherapy, p < 0.001; monotherapy versus tritherapy, p < 0.001; bitherapy versus tritherapy, p = n.s.

  7. e

    Monotherapy versus bitherapy, p = 0.023; monotherapy versus tritherapy, p = 0.002; bitherapy versus tritherapy, p = n.s.

  8. f

    Monotherapy versus bitherapy, p = n.s.; monotherapy versus tritherapy, p = 0.001; bitherapy versus tritherapy, p = 0.041.

  9. g

    Monotherapy versus bitherapy, p = n.s.; monotherapy versus tritherapy, p = 0.001; bitherapy versus tritherapy, p = 0.010.

  10. h

    Monotherapy versus bitherapy, p = n.s.; monotherapy versus tritherapy, p = 0.014; bitherapy versus tritherapy, p = n.s.

  11. i

    Monotherapy versus tritherapy, p = 0.023; monotherapy versus tritherapy, p = n.s.; bitherapy versus tritherapy, p = 0.023.

  12. j

    Monotherapy versus bitherapy, p = n.s.; monotherapy versus tritherapy, p = 0.021; bitherapy versus tritherapy, p = 0.012.

Demographic data    
Female-to-male ratio17/2415-October5-June0.694
Age at onset (month)a23.3 (23.4; 8.8–36.0)22.5 (24.3; 8.7–34.5)14.6 (13.5; 8.0–29.5)0.007
Previous FS (%)9 (22.0)2 (8.0)1 (9.1)0.258
Family history of IGE (%)10 (24.4)11 (44.0)2 (18.2)0.158
Early electroclinical data    
Time at absence control (month)b2.0 (2.0; 0.7–3.7)2.5 (2.6; 1.1–4.2)3.7 (3.9; 2.4–4.5)<0.001
Daily AED dose (mg/kg)    
VPA30.7 (30.1; 21.8–43.8)28.2 (26.0; 17.8–44.6)28.6 (25.8; 17.8–44.6)0.244
ESM22.4 (23.8; 20.3–30.2)20.9 (20.4; 12.7–33.1)19.0 (16.7; 11.8–30.0)0.245
LTGc13.6 (15.0; 7.1–18.5)5.5 (5.6; 2.9–8.7)3.2 (3.1; 2.7–3.8)0.001
LEV29.7 (30.0; 23.5–43.3)25.6 (25.2; 20.5–32.3)23.4 (24.5; 16.8–29.0)0.119
EEG at 6 months (%)    
Normald28 (68.3)6 (24.0)0<0.001
Generalized 3–4 Hz spike-waves8 (19.5)8 (32.0)5 (45.5)0.186
Generalized abnormalities8 (19.5)12 (48.0)4 (36.4)0.071
Focal abnormalitiese3 (7.3)7 (28.0)5 (45.5)0.008
Genetic and neuroimaging data    
SLC2A1 mutation (%)f01 (4.0)3 (27.3)0.001
Abnormal neuroimaging (%)g7 (17.1)6 (24.0)8 (72.7)0.001
36-Month outcome data    
Relapse (%)14 (43.1)9 (36.0)7 (63.6)0.191
Time at relapse (month)20.1 (22.9; 5.3–35.0)21.5 (22.1; 9.7–33.4)14.7 (15.3; 6.1–23.6)0.199
Terminal remission (month)h27.8 (33.7; 1.0–35.3)26.3 (32.6; 2.6–34.8)23.1 (22.8; 10.4–32.1)0.017
Seizure freedom ≥24 months (%)30 (73.2)16 (64.0)5 (45.5)0.216
Duration therapy (month)31.0 (36.0; 16.3–36.0)29.4 (34.1; 19.8–36.0)35.1 (36.0; 26.7–36.0)0.052
Failed withdrawal (%)5/17 (31.3)2/12 (16.7)00.57
Ongoing therapy (%)32 (78.0)15 (60.0)10 (90.9)0.104
Conditions do not meet CAE definition (%)    
Abnormal neuro/develop statei4 (9.8)6 (24.0)7 (63.6)0.001
Other seizure type23 (56.1)13 (52.0)6 (54.5)0.501
Myoclonic featuresj27 (65.9)18 (72.0)3 (27.3)0.031
Photic induction6 (14.6)1 (4.0)00.182
No loss of consciousness1 (2.4)3 (12.0)00.166
Brief (<4 s) 3–4 Hz spike-waves8 (19.5)3 (12.0)2 (18.2)0.726

As shown in Figure 4, the difference between the survival curves appeared significant at overall comparisons (log-rank [Mantel-Cox] χ2 = 6.104, d.f. = 2, p = 0.047); patients treated with three AEDs had the highest risk of relapse in the first 36 months. Time to relapse did not differ between patients with different AED response (p = 0.199).

image

Figure 4. Kaplan-Meier curves show the probability of seizure-free survival among patients with NP-EOAE based on different initial AED response. Test of equality of survival distributions (overall comparisons): log-rank (Mantel-Cox) χ2 = 6.1, d.f. = 2, p = 0.047. Test of equality of survival distributions (pairwise comparisons): log-rank (Mantel-Cox) χ2 = 0.2, p = 0.892 (monotherapy vs. bitherapy); log-rank (Mantel-Cox) χ2 = 5.3, p = 0.022 (monotherapy vs. tritherapy); log-rank (Mantel-Cox) χ2 = 3.9, p = 0.046 (bitherapy vs. tritherapy).

Download figure to PowerPoint

Risk factors

p-Values of variables and ORs for relapse for each value as compared with the reference value of that variable are reported in Table 4. Significant variables associated with increased odds of relapse were time at absence control, AED number to stop absences, abnormal EEG at 6 months, and NP-EOAE. Normal neuroimaging was associated with decreased odds of relapse (Table 4). In the final multivariate logistic regression analysis predicting relapse, only NP-EOAE remained significant (Table 5). Children with NP-EOAE had 2.134 times the odds of having relapse during the 36-month follow-up compared to patients with P-EOAE.

Table 4. Prognostic factors for relapse during 36-month follow-up
 OR95% CIp-Value
  1. FS, febrile seizure; IGE, idiopathic generalized epilepsy; AED, antiepileptic drug; EEG, electroencephalography; P-EOAE, pure early onset absence epilepsy; NP-EOAE, nonpure early onset absence epilepsy.

  2. Data are expressed as odds ratio (OR) with 95% confidence interval (CI).

Age at onset (month)0.9850.947–1.0250.458
Gender   
Male1  
Female1.0230.529–1.9760.947
History of FS   
Yes1  
No1.7530.840–3.6590.135
Family history of IGE   
Yes1  
No1.5760.767–3.2420.216
Neuroimaging   
Abnormal1  
Normal0.3580.142–0.9040.03
SLC2A1 mutation   
Yes1  
No0.1180.012–1.1580.067
Time at absence control (month)1.5471.038–2.3060.032
AED number to stop absence1.8151.084–3.0370.023
EEG at 6 months   
Normal1  
Abnormal2.0571.070–3.9550.03
Withdrawal AED   
Yes1  
No1.0250.539–1.9500.94
Panayiotopoulos's criteria   
P-EOAE1  
NP-EOAE2.7361.414–5.2920.003
Table 5. Multiple logistic regression analysis (in one single step) predicting relapse
 OR95% CIp-Value
  1. AED, antiepileptic drug; EEG, electroencephalography; P-EOAE, pure early onset absence epilepsy; NP-EOAE, nonpure early onset absence epilepsy.

  2. Data are expressed as odds ratio (OR) with 95% confidence interval (CI).

Neuroimaging   
Abnormal1  
Normal0.740.276–2.4960.74
Time at absence control (month)1.180.728–1.9140.501
AED number to stop absence1.2550.659–2.3890.489
EEG at 6 months   
Normal1  
Abnormal1.290.602–2.7680.513
Panayiotopoulos's criteria   
P-EOAE1  
NP-EOAE2.1341.010–4.5100.027

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Appendix 1

In this study, we used a strict definition of P-EOAE because children had onset of absence seizures within the first 3 years of life, but otherwise conformed to CAE criteria proposed by Panayiotopoulos (Panayiotopoulos, 2008). These criteria are different from those previously used in the context of GLUT-1 deficiency syndrome, which included patients with onset of absences before 4 years of age, generalized spike-waves (>2.5 Hz) on EEG, no evidence of secondary cause for epilepsy, and absence of atonic–tonic seizures (Suls et al., 2009; Arsov et al., 2012).

We found significant clinical differences between P-EOAE and NP-EOAE to justify a syndrome-based definition. In P-EOAE the absences started at an older age and were better controlled by AEDs with shorter time and less 6-month EEG abnormalities than in NP-EOAE. At last contact, longer remission with shorter duration of treatment and less frequent relapses were observed in P-EOAE compared to NP-EOAE. In contrast to the latter, children with P-EOAE did not show any other seizure types, abnormal brain MRI, and SLC2A1 mutations. Reasonably, in patients with newly suspected P-EOAE who are in accordance with our proposed criteria, it is not advisable to assess the SLC2A1 gene mutations.

Although TAS are considered the paradigm seizure type of IGE, they may occasionally be symptomatic, and associated with a known disorder of the central nervous system (Ferrie et al., 1995; Panayiotopoulos, 2001). These symptomatic TAS may be caused by focal or diffuse lesions (Ferrie et al., 1995). The mesial surfaces of the frontal lobe are the most likely brain locations to generate TAS in symptomatic cases (Panayiotopoulos, 2001). TAS has also been reported as the result of subependymal heterotopia (Raymond et al., 1994). In line with these observations, we observed in our series brain imaging abnormalities, such as subcortical focal hyperintensity of the frontal lobe, cortical dysplasia, gray matter heterotopia, and abnormalities of white matter signal. The fact that symptomatic (structural-metabolic) etiologies (i.e., brain abnormalities and GLUT1 deficiency) were observed only in children with NP-EOAE and not in those with P-EOAE was probably the cause of their worse prognosis and perhaps illustrates the natural course of epilepsy. In these patients a poor clinical evolution started after 12 months of follow-up and treatment was not effective in children received polytherapy (three AEDs) because of high rate of recurrence. The favorable and improving outcome in children with P-EOAE who became seizure-free with AED monotherapy (67.5%) or bitherapy (32.4%) could have been the result of treatments, although many patients withdrew without subsequent relapses. Considering these findings, treatment does not seem to have a major impact on the outcome of epilepsy. In accordance with hypothesis of Shinnar and Berg, (1994), the natural course of epilepsy and the underlying etiology better explain our results, with P-EOAE being short-lived and NP-EOAE having a designated poor outcome.

Application of our strict criteria for P-EOAE leads to a group of patients with homogeneous disease characteristics and prognosis, independent of fast AED response. In fact, when we compared children who responded to monotherapy and those who responded to bitherapy, similar demographic and electroclinical aspects were shown. As might be expected, patients treated with two AEDs, had seizures for a longer period and received smaller LTG doses than those treated with two AEDs. Difference in LTG dose values can be due to a pharmacodynamic beneficial effect of small LTG doses when added to VPA, which is the first-line AED most frequently used in our patients on bitherapy. These results are in line with those that we reported previously (Giordano et al., 2011; Verrotti et al., 2011; Agostinelli et al., 2013).

Conversely, in the group that did not meet Panayiotopoulos's criteria, more differences emerged between patients with initial response to monotherapy, bitherapy, and tritherapy. These differences indicate a heterogeneous group of epilepsies, including both patients who had a poor electroclinical prognosis as more associated with symptomatic etiologies (i.e., brain abnormalities and GLUT1 deficiency) and those who had an intermediate outcome between P-EOAE and symptomatic forms of NP-EOAE. Children with this intermediate outcome responded well to monotherapy or bitherapy, and showed myoclonic features during the active stage of epilepsy. This category may include some well-defined epileptic syndromes, such as the benign myoclonic epilepsy of infancy, eyelid myoclonia with absences, and myoclonic absence epilepsy. This is in line with the results provided by Caraballo et al. (2011).

At the end of follow-up, 62.2% of our cohort continued AEDs, not only those who relapsed, but also a considerable number of patients in remission. The proportion of subjects continuing treatment was lower in P-EOAE (54.1%) than in NP-EOAE (74.0%). Differences in etiology, length of initial absence control, and relapses during follow-up could explain this.

Although, brain MRI and EEG data and parameters of early seizure control were associated with outcome measurement, in the final multivariate analysis only Panayiotopoulos's criteria remained significant. Indeed, children without strict Panayiotopoulos's criteria had 2.134 times the odds of having relapse during the 36-month follow-up compared to patients who met Panayiotopoulos's criteria.

In conclusion, this study demonstrates that differences in electroclinical features and outcomes between patients with P-EOAE and NP-EOAE may be sufficient to delineate distinct subsyndromes within children presenting early onset absences. Patients who meet our modified Panayiotopoulos's criteria have a favorable course of epilepsy, whereas those not presenting them have a poor outcome with high risk of relapse during follow-up. Early onset absences not conforming to our strict diagnosis are part of epileptic disorders secondary to brain abnormalities and GLUT1 deficiency, as well as some well-defined epileptic syndromes, such as the benign myoclonic epilepsy of infancy, eyelid myoclonia with absences, and myoclonic absence epilepsy. NP-EOAE associated with symptomatic etiologies shows a poorer electroclinical prognosis than that reported in the context of well-defined epileptic syndromes.

Patients with EOAE are probably more frequent than we previously considered, and they should be better recognized and studied.

Disclosure

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Appendix 1

None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Appendix 1
  • Agostinelli S, Traverso M, Accorsi P, Beccaria F, Belcastro V, Capovilla G, Cappanera S, Coppola A, Dalla Bernardina B, Darra F, Ferretti M, Elia M, Galeone D, Giordano L, Gobbi G, Nicita F, Parisi P, Pezzella M, Spalice A, Striano S, Tozzi E, Vignoli A, Minetti C, Zara F, Striano P, Verrotti A. (2013) Early-onset absence epilepsy: SLC2A1 gene analysis and treatment evolution. Eur J Neurol 20:856859.
  • Arsov T, Mullen SA, Damiano JA, Lawrence KM, Huh LL, Nolan M, Young H, Thouin A, Dahl HH, Berkovic SF, Crompton DE, Sadleir LG, Scheffer IE. (2012) Early onset absence epilepsy: 1 in 10 cases is caused by GLUT1 deficiency. Epilepsia 53:e204e207.
  • Berg AT, Berkovic SF, Brodie MJ, Buchhalter J, Cross JH, van Emde Boas W, Engel J, French J, Glauser TA, Mathern GW, Moshé SL, Nordli D, Plouin P, Scheffer IE. (2010) Revised terminology and concepts for organization of seizures and epilepsies: report of the ILAE Commission on Classification and Terminology, 2005–2009. Epilepsia 51:676685.
  • Caraballo RH, Darra F, Fontana E, Garcia R, Monese E, Dalla Bernardina B. (2011) Absence seizures in the first 3 years of life: an electroclinical study of 46 cases. Epilepsia 52:393400.
  • Ferrie CD, Giannakodimos S, Robinson RO. (1995) Symptomatic typical absence seizures. In Duncan JS, Panayiotopoulos CP (Eds) Typical absences and related epileptic syndromes. Churchill Communications Europe, London, pp. 241252.
  • Giordano L, Vignoli A, Accorsi P, Galli J, Pezzella M, Traverso M, Battaglia S, Baglietto MG, Beccaria F, Cerminara C, Gambara S, Del Giudice E, Crichiutti G, Bisulli F, Pinci M, Tinuper P, Briatore E, Calzolari S, Coppola A, Canevini MP, Capovilla G, Striano S, Zara F, Minetti C, Striano P. (2011) A clinical and genetic study of 33 new cases with early-onset absence epilepsy. Epilepsy Res 95:221226.
  • Panayiotopoulos CP. (2001) Treatment of typical absence seizures and related epileptic syndromes. Paediatr Drugs 3:379403.
  • Panayiotopoulos CP. (2008) Typical absence seizures and related epileptic syndromes: assessment of current state and directions for future research. Epilepsia 49:21312139.
  • Raymond AA, Fish DR, Stevens JM, Sisodiya SM, Alsanjari N, Shorvon SD. (1994) Subependymal heterotopia: a distinct neuronal migration disorder associated with epilepsy. J Neurol Neurosurg Psychiatry 57:11951202.
  • Shahar E, Genizi J, Nevo Y, Kaufman R, Cabot S, Zelnik N. (2007) Typical absence epilepsy presenting prior to age of 3 years: an uncommon form of idiopathic generalized epilepsy. Eur J Paediatr Neurol 11:346352.
  • Shinnar S, Berg AT. (1994) Does antiepileptic drug therapy alter the prognosis of childhood seizures and prevent the development of chronic epilepsy? Semin Pediatr Neurol 1:111117.
  • Striano P, Weber YG, Toliat MR, Schubert J, Leu C, Chaimana R, Baulac S, Guerrero R, LeGuern E, Lehesjoki AE, Polvi A, Robbiano A, Serratosa JM, Guerrini R, Nürnberg P, Sander T, Zara F, Lerche H, Marini C; EPICURE Consortium. (2012) GLUT1 mutations are a rare cause of familial idiopathic generalized epilepsy. Neurology 78:557562.
  • Suls A, Mullen SA, Weber YG, Verhaert K, Ceulemans B, Guerrini R, Wuttke TV, Salvo-Vargas A, Deprez L, Claes LR, Jordanova A, Berkovic SF, Lerche H, De Jonghe P, Scheffer IE. (2009) Early-onset absence epilepsy caused by mutations in the glucose transporter GLUT1. Ann Neurol 66:415419.
  • Verrotti A, Olivieri C, Agostinelli S, Coppola G, Parisi P, Grosso S, Spalice A, Zamponi N, Franzoni E, Iannetti P, Chiarelli F, Curatolo P. (2011) Long term outcome in children affected by absence epilepsy with onset before the age of three years. Epilepsy Behav 20:366369.

Appendix 1

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Appendix 1

Appendix List of the involved centers and clinical investigators

Table 6. 
CentersInvestigators
Child Epilepsy Center, Department of Pediatrics, University of Chieti, ChietiAlberto Verrotti, Sergio Agostinelli
Pediatric Neurology and Muscular diseases Unit, Institute “G. Gaslini,” GenoaPasquale Striano, Monica Traverso, Federico Zara, Marianna Pezzella, Stella Vari, Carlo Minetti
Child Neuropsychiatry Unit, Hospital “Civile,” BresciaPatrizia Accorsi, Lucio Giordano
Department of Child Neuropsychiatry, Hospital “C. Poma,” MantovaGiuseppe Capovilla, Francesca Beccaria
Department of Neurosciences, Hospital “Sant'Anna,” ComoVincenzo Belcastro
Department of Pediatric Neurology, Hospital “Ospedali Riuniti,” AnconaSilvia Cappanera, Nelia Zamponi
Department of Neurological Sciences, Federico II University, NaplesSalvatore Striano, Luigi Del Gaudio
Unit of Child Neuropsychiatry, University of Verona, VeronaBernardo Dalla Bernardina, Francesca Darra
Department of Neurology, Institute “Oasi,” TroinaMaurizio Elia
Child Neurology, Children's Hospital “Giovanni XXIII,” BariMichela Sesta
Department of Child Neuropsychiatry, Hospital “Maggiore,” BolognaGiuseppe Gobbi
Department of Pediatrics, Hospital “V. Emanuele,” CataniaRaffaele Falsaperla, Piero Pavone
Child Neurology, II Faculty of Medicine, La Sapienza University, RomePasquale Parisi
Child Neurology Unit, Department of Pediatrics, La Sapienza University, RomeFrancesco Nicita, Alberto Spalice
Department of Child Neuropsychiatry, University of L'AquilaElisabetta Tozzi
Epilepsy Center, University of Milan, MilanAglaia Vignoli, Maria Paola Canevini