JLE

Epileptic Disorders

MENU

Early-onset childhood absence epilepsy: is it a distinct entity? Volume 13, issue 4, December 2011

epd.2011.0465

Auteur(s) : Pue Farooque1, Jatinder Goraya2, Ignacio Valencia2, Karen S Carvalho2, H Huntley Hardison2, Agustin Legido2, Divya S Khurana2 divya.khurana@drexelmed.edu

1 Department of Neurology, Hahnemann University Hospital, 245 North 15th Street, Philadelphia, PA 19102, USA

2 Section of Child Neurology, St. Christopher's Hospital for Children, 3601 A Street, Philadelphia, PA 19134, USA and Department of Pediatrics, Drexel University of College of Medicine, Philadelphia, Pennsylvania, USA

Correspondence: D. S. Khurana Section of Child Neurology, St. Christopher's Hospital for Children, 3601 A Street, Philadelphia, PA 19134. USA

Childhood absence epilepsy (CAE) is an age-specific epilepsy syndrome classified as one of the idiopathic generalised epilepsies (Commission on Classification and Terminology of the International League Against Epilepsy, 1989). It is defined by episodes of unresponsiveness associated with bilaterally synchronous spike-and-wave discharges at 3 Hz with a normal background on EEG. CAE typically starts at four to seven years of age with a peak at around six or seven years, whereas juvenile absence epilepsy (JAE) presents at 11 to 12 years. The onset of absence epilepsy before the age of three years is rare and has not been previously reported in North America. Optimal treatment with antiepileptic drugs (AEDs) for this age group is also not known. Studies from Europe and South America have evaluated the significance of early-onset absence epilepsy, however, most of the data are contradictory (Cavazzuti et al., 1989; Aicardi, 1995; Darra et al., 1996; Covanis, 1998; Chaix et al., 2003; Fernandez-Torre et al., 2006; Shahar et al., 2007; Verrotti et al., 2011b; Caraballo et al., 2011; Giordano et al., 2011). The aim of this study was to assess the clinical relevance of early-onset absence epilepsy and to characterise seizure frequency, control, and prognosis in a cohort of patients with this condition.

Materials and methods

Approval was obtained from the Institutional Review Board of Drexel University College of Medicine, St. Christopher's Hospital for Children. A retrospective chart review was performed of the EEG database at St. Christopher's Hospital for Children, from January 2000 to June 2009, to identify patients with pure absence epilepsy that began before the age of three years. Absence seizures were defined as paroxysmal episodes of staring, unresponsiveness, or motor arrest with a concomitant electrographic 3-Hz spike-and-wave discharge on a normal EEG background. All EEGs were read and interpreted by an epileptologist. Clinical data of the patients identified were then reviewed and collected from the neurology clinic medical records. Data gathered included demographic variables such as age, gender, age at onset, diagnosis, and duration of follow-up. Information on seizure control, AEDs taken, response to treatment, side effects, and the presence or absence of developmental problems was also gathered for all the patients identified. Exclusion criteria included presence of other seizure types such as concomitant eyelid myoclonia, myoclonus, or generalised tonic-clonic seizures at initial presentation and during follow-up. This approach allowed us to identify patients with only pure absence epilepsy.

Results

Patient demographics, AEDs previously and currently taken, side effects, co-morbidities, and EEG findings are summarised in table 1.

Table 1 Patient demographics, AEDs and side effects, co-morbidities, and EEG findings.

Age at onset/Age at diagnosis First AED Second AED Current AED regimen Seizure control Recurrence Normalised EEG on Rx Neuro-development Co-morbidity Side effects Duration of therapy/Duration of follow-up
11 m/2 y ETX ETX No N/A No Language delay ADHD None Ongoing/2.3 y
1 y/18 m ETX VPA VPA Yes N/A No Normal None Rash (ETX) Ongoing/3 y
16 m/16 m TPM ETX None Yes No Yes Language delay Premature Mood (TPM) 2 y/3 y
16 m/19 m ETX VPA None Yes Yes (VPA tapered off) No Language delay ADHD None 8 y/9 y
1.5 y/2.5 y VPA VPA Yes Yes No Normal ADHD None Ongoing/6 y
2 y/7 y ETX LEV LEV, ETX No N/A No Language delay Febrile seizures None Ongoing/5 y
2 y/6 y ETX VPA LEV, LTG No N/A No Normal None None Ongoing/11 y
2 y/5 y ETX VPA (seizure free but SE) LEV Yes No No Normal None None Ongoing/3.5 y
2.5 y/3 y ETX None Yes Yes twice Photo-paroxysmal response Normal Febrile seizures None 6 y/7 y
2 y/2 y ETX VPA None Yes No Yes Normal None None 3.5 y/4 y
2 y/4 y ETX VPA VPA No N/A No Language delay Autism Rash (ETX) Ongoing/2 y
2 y/2 y LEV ETX None No N/A No Normal None None 4 m/parents stopped ETX

ETX: ethosuximide; VPA: valproate; LEV: levetiracetam; LTG: lamotrigine; TPM: topiramate; ADHD: attention deficit hyperactivity disorder; y: years; m: months; N/A: not applicable; SE: side effects.

Demographics

Twelve patients were identified with absence seizures beginning earlier than three years of age, including seven boys and five girls. The mean patient age at seizure onset was 20.5 months (range: 11 months to two years) with mean age at diagnosis of 3.1 years (range: 16 months to seven years). Follow-up ranged from six months to 11 years (mean: 4.8 years).

Clinical and EEG findings

All patients came to clinical attention because of episodes of staring and unresponsiveness and an initial EEG showing a 3-Hz spike-and-wave discharge lasting from 3 to 22 seconds with no asymmetry or focal abnormalities. Sleep was recorded during initial EEG for three of 12 patients, for the other children sleep was captured on subsequent routine or 24-hour ambulatory EEGs. Background activity was normal for age as part of the inclusion criteria. Five patients had a photoparoxysmal response and three showed occipital intermittent rhythmic delta activity on EEG. Neurodevelopment, as assessed by the Denver Development Screening test, was normal in seven patients (58%) and five patients (42%) had language delay, of whom four had mild speech delay and one had a moderate-to-severe language delay. Two children had a history of febrile seizures and one child's mother had a history of absence epilepsy.

Antiepileptic drugs taken

Initial monotherapy included ethosuximide for nine children, valproate for one, topiramate for one, and levetiracetam for another. The second AED used was valproate for six children, ethosuximide for two, and levetiracetam for one. In all, ethosuximide was used for 11 patients, valproate for seven, levetiracetam for four, and lamotrigine and topiramate each for one patient. Ethosuximide is typically the first AED used for uncomplicated childhood absence epilepsy in North America.

Clinical course and response to AEDs

Seven of 12 patients (58%) were seizure-free. Four of these seven were no longer taking medication having completed a seizure-free period of greater than two years and remained seizure-free, although one child continued to show a photoparoxysmal response on EEG but did not have recurrence of any epileptic seizures for 18 months following AED taper. The remaining three children were seizure-free with a single AED (two are currently taking valproate and one levetiracetam) but all continued to show brief paroxysms of irregular generalised 3-Hz spike-and-wave discharges on EEG. Interestingly, one of these children had an allergic reaction to ethosuximide as initial AED at age 19 months. The parents elected not to treat the seizures, however, a year and a half later, because of an increase in frequency and duration of absences, valproate treatment was initiated and she has been seizure-free since. Three patients had seizure recurrence within three to four months after completing a two-year seizure-free period with a single AED. One of the three patients had two failed attempts to wean off ethosuximide after two years of seizure freedom but a third medication taper was successful and she has remained seizure-free for 18 months since. For the second patient, ethosuximide was successfully tapered and the third child remains on valproate treatment.

The remaining five patients (42%) continued to show clinical and/or electroencephalographic absence seizures; two on monotherapy with ethosuximide or valproate, one on a combination of ethosuximide and levetiracetam, and another on a combination of levetiracetam and lamotrigine. The fifth patient is currently on no medication although she was briefly taking levetiracetam with no response and ethosuximide with a good response, but the parents decided not to continue treatment and she still has seizures.

None of the twelve patients developed any other seizure types during follow-up.

Discussion

We report the first series from North America on the outcomes of children presenting with pure CAE before the age of three years. Several studies have been performed in Europe and South America attempting to classify early-onset absence seizures and determine whether they represent a distinct entity from traditional CAE and JAE. The results are variable and the outcomes are summarised in table 2 (Cavazzuti et al., 1989; Aicardi, 1995; Darra et al., 1996; Covanis, 1998; Chaix et al., 2003; Fernandez-Torre et al., 2006; Shahar et al., 2007; Caraballo et al., 2011; Verrotti et al., 2011b; Giordano et al., 2011).

Table 2 Outcomes associated with early-onset absence seizures based on the literature.

Study Number of patients Age at onset Drugs used Outcome Comments
Cavazzuti et al., 1989 1 6.5 m Nitrazepam Well controlled
Aicardi, 1995 1 <2 y VPA Well controlled
Darra et al., 1996 6 <3 y Not specified Well controlled
Covanis, 1998 7 <3 y VPA Well controlled
Chaix et al., 2003 10 <3 y VPA, ETX, CLB, LTG, TPM 5/10 with persistent seizures on polytherapy Tapering of an AED in one patient resulted in recurrence.
Fernandez-Torre et al., 2006 3 <3 y VPA, ETX Well controlled One patient had a myoclonic absence seizure.
Treatment with ETX resulted in better seizure control for one patient.
Shahar et al., 2007 8 <3 y VPA, LTG Well controlled Treatment with LTG resulted in seizure control for two patients.
Three patients relapsed after VPA withdrawal.
Caraballo et al., 2011 11 <3 y 9/11-no medication 9/11 seizure-free Two patients had persistent absence and generalised tonic-clonic seizures.
Verrrotti et al. 2011b; Caraballo et al., 2011 40 <3 y VPA, ETX 33/40 (82%) seizure-free
Giordano et al., 2011 33 <3 y VPA, ETX, LTG, LEV 28/33 (85%) seizure-free
This study 12 <3 y VPA, ETX, LTG, LEV 7/12 (58%) seizure-free Tapering of AEDs in three patients resulted in recurrence.

VPA: valproate; ETX: ethosuximide; LTG: lamotrigine; CLB: clobazam; TPM: topiramate; LEV: levetiracetam; y: years; m: months.

In the last year alone, there have been three large series of patients with “pure” early-onset CAE; two multicentre studies from Italy each with 40 and 33 children (Verrotti et al., 2011b; Giordano et al., 2011) and one from South America with 11 children (Caraballo et al., 2011). Our patient population is similar in that these were children who had only absence seizures, did not develop any other seizure type during follow-up, and had normal EEG background features. However, 9/11 (82%) patients in Caraballo's series, 33/40 (82%) in Verrotti's cohort and 28/33 (85%) patients in Giordano's paper were described as seizure-free compared to only 58% in our group of children. One reason for this discrepancy could be that both EEG and clinical status were used in our study to determine seizure freedom, i.e. seizures were classified as ongoing seizures if they were captured on EEG. The discrepancy could also be due to the relatively low use of valproate and high use of ethosuximide as first line AED in our population. In North America, there is some hesitation on the part of both physicians and patients to use valproate in young children because of the risk of hepatotoxicity. However, the success rate of valproate and ethosuximide for the treatment of CAE in North America is similar, ranging from 49% (Sato et al., 1982) to 53%-58% (Glauser et al., 2010). Therefore, a lower level of valproate treatment in our group can not be the only explanation for the difference reported by Caraballo et al. (2011), Giordano et al. (2011) and Verrotti et al. (2011b). Finally, it is possible that there is geographically-determined heterogeneity in the therapeutic response of CAE to valproate treatment, with a response of 89% in Europe (Mazurkiewicz-Beldzinska et al., 2010) and 74% in China (Huang et al., 2009), whereas in North America the response has consistently been less than 60% (Sato et al., 1982; Glauser et al., 2010).

Some rare conditions such as paroxysmal dyskinesia and glucose transporter (Glut-1) deficiency syndrome have been shown to be associated with early-onset absence seizures (Guerrini et al., 2002; Hirsch, 2004). We did not evaluate our patients for Glut-1 deficiency as the constellation of typical clinical features associated with Glut-1 deficiency, including developmental delay, ataxia, hypotonia, and infantile seizures, was not present in any of our patients.

Clinicians should be aware that absence epilepsy can occur in children below the age of three years. Including the present series, approximately 130 patients with this type of epilepsy have been reported in the literature, indicating the relative rarity of this epileptic disorder. Early-onset absence epilepsy is thought to represent less than 1% of epilepsies with onset at less than three years (Chaix et al., 2003; Caraballo et al., 2011).

While the majority of patients had good seizure control with conventional AEDs, 21 had poorly controlled disease (Chaix et al., 2003; Verrotti et al., 2011b; Giordano et al., 2011, and the present series), indicating that this is not a homogeneous condition. Our study also shows that there are some patients who do not have a uniformly good outcome with regards to seizure control and intellectual outcome. The unique vulnerability of the developing brain in the early years of childhood, as well as genetic heterogeneity and geographically-determined epigenetic differences, might account for the different clinical response to AEDs seen in different cohorts.

However, the debate over whether early-onset CAE is a distinct epilepsy syndrome continues. Whereas Verrotti et al. (2011a) believe that this is the case, Giordano et al. (2011) argue that early-onset CAE should be considered a continuum within the wide spectrum of idiopathic generalised epilepsies. In agreement with Verrotti et al. (2011a), we believe that early-onset CAE may be a distinct epilepsy syndrome with some features that overlap with those of typical CAE, as well as unique distinguishing features. Large prospective multicentre studies would be necessary to definitively resolve this matter.

Disclosure

None of the authors has any conflict of interest to disclose.