Antiepileptic drugs for the treatment of severe myoclonic epilepsy in infancy

  • Review
  • Intervention

Authors

  • Francesco Brigo,

    Corresponding author
    1. University of Verona, Department of Neurological, Neuropsychological, Morphological and Movement Sciences. Section of Clinical Neurology, Verona, Italy
    • Francesco Brigo, Department of Neurological, Neuropsychological, Morphological and Movement Sciences. Section of Clinical Neurology, University of Verona, P.le L.A. Scuro, 10, Verona, 37134, Italy. dr.francescobrigo@gmail.com.

    Search for more papers by this author
  • Monica Storti

    1. University of Verona, Department of Medicine, Verona, Italy
    Search for more papers by this author

Abstract

Background

Severe myoclonic epilepsy in infants (SMEI), also known as Dravet syndrome, is a rare, refractory form of epilepsy, for whose treatment stiripentol (STP) has been recently licensed for add-on use.

Objectives

To evaluate the efficacy and tolerability of STP and other antiepileptic drug treatments (including ketogenic diet) as therapy for patients with SMEI.

Search methods

We searched the Cochrane Epilepsy Group Specialised Register (15 May 2013), the Cochrane Central Register of Controlled Trials (CENTRAL, Issue 4 of 12, The Cochrane Library, April 2013), MEDLINE (1946 to May 2013) and SCOPUS (1823 to May 2013). The online trials registries ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform were systematically searched. The bibliographies of any identified study were searched for further references. We handsearched selected journals and conference proceedings. No language restrictions were imposed.

Selection criteria

Randomised controlled trials (RCTs) or quasi-randomised controlled trials; double- or single-blinded or unblinded trials; and parallel-group studies. Administration of at least one antiepileptic drug therapy given singly (monotherapy) or in combination (add-on therapy) compared with add-on placebo or no add-on treatment.

Data collection and analysis

Review authors independently selected trials for inclusion according to predefined criteria, extracted relevant data and evaluated the methodological quality of trials. The following outcomes were assessed: at least 50% seizure reduction, seizure freedom, adverse effects, proportion of dropouts and quality of life. Outcomes were assessed using a Mantel-Haenszel meta-analysis to calculate risk ratio (RR) with 95% confidence intervals (95% CIs).

Main results

No RCTs assessing drugs other than STP were found. Two RCTs evaluating the use of STP (total of 64 children) were included. Both studies were generally at unclear risk of bias. A significantly higher proportion of participants had 50% or greater reduction in seizure frequency in the STP group compared with the placebo group (22/33 vs 2/31; RR 10.40, 95% CI 2.64 to 40.87). A significantly higher proportion of participants achieved seizure freedom in the STP group compared with the placebo group (12/33 vs 1/31; RR 7.93, 95% CI 1.52 to 41.21). No significant difference in the proportion of dropouts was found in the STP group compared with the placebo group (2/33 vs 8/31; RR 0.24, 95% CI 0.06 to 1.03). Only one study explicitly reported the occurrence of side effects; higher proportions of participants were reported to experience side effects in the STP group compared with the placebo group (100% vs 25%; RR 3.73, 95% CI 1.81 to 7.67).

Authors' conclusions

Data derived from two small RCTs indicate that STP is significantly better than placebo with regards to 50% or greater reduction in seizure frequency and seizure freedom. Adverse effects occurred more frequently with STP. Further adequately powered studies with long-term follow-up should be conducted to unequivocally establish the long-term efficacy and tolerability of STP in the treatment of SMEI.

Résumé scientifique

Médicaments antiépileptiques pour le traitement d’épilepsie myoclonique sévère chez le nourrisson

Contexte

L’épilepsie myoclonique sévère du nourrisson (EMSN), également connue sous le nom de syndrome de Dravet, est une forme d’épilepsie rare et réfractaire aux médicaments, pour laquelle le stiripentol (STP) a été récemment homologué en traitement en association avec d'autres médicaments antiépileptiques.

Objectifs

Évaluer l'efficacité et la tolérance du STP et d’autres traitements antiépileptiques (y compris le régime cétogène) comme traitement pour les patients atteints d’EMSN.

Stratégie de recherche documentaire

Nous avons effectué des recherches dans le registre spécialisé du groupe Cochrane sur l'épilepsie (15 mai 2013), le registre Cochrane des essais contrôlés (CENTRAL, numéro 4 sur 12, La Bibliothèque Cochrane, Avril 2013), MEDLINE (de 1946 à mai 2013) et SCOPUS (de 1823 à mai 2013). Les registres d'essais en ligne ClinicalTrials.gov et le système d’enregistrement international WHO des essais cliniques ont été systématiquement consultés. Les références bibliographiques de toutes les études identifiées ont été consultées afin d'identifier d'autres références. Nous avons effectué une recherche manuelle dans des journaux et des actes de conférence. Aucune restriction de langue n'a été imposée.

Critères de sélection

Études contrôlées randomisées (RCT) ou quasi randomisées, études en double ou simple aveugle ou en ouvert, et études en groupes parallèles. Administration d'au moins un médicament antiépileptique en traitement administré seul (monothérapie) ou en association (traitement adjuvant) par rapport au placebo en traitement adjuvant ou à l’absence de traitement adjuvant.

Recueil et analyse des données

Des auteurs de la revue ont indépendamment sélectionné les essais à inclure sur la base de critères prédéfinis, extrait les données pertinentes et évalué la qualité méthodologique des essais. Les critères de jugement suivants ont été évalués : au moins 50% de réduction des crises, l'absence de crises, les effets indésirables, le nombre de sorties d'étude et la qualité de vie. Les critères de jugement étaient évalués à l'aide d'une méta-analyse de Mantel-Haenszel pour calculer le risque relatif (RR) avec des intervalles de confiance (IC à 95%).

Résultats principaux

Aucun ECR évaluant des médicaments autres que STP n'a été trouvé. Deux ECR évaluant l'utilisation de STP (total de 64 enfants) ont été inclus. Les deux études possédaient généralement un risque de biais incertain. Le pourcentage de patients ayant obtenu une réduction d’au moins 50% de la fréquence des crises a été significativement plus élevé dans le groupe STP que dans le groupe placebo (22/33 versus 2/31 ; RR : 10,40, IC à 95% : 2,64 40,87). Le pourcentage de patients n’ayant pas eu de crises a été significativement plus élevé dans le groupe STP que dans le groupe placebo (12/33 versus 1/31 ; RR : 7,93, IC à 95% : 1,52 41,21). Il n’a pas été observé de différence significative en termes de taux d’abandons dans le groupe STP par rapport au groupe placebo (2/33 versus 8/31 ; RR : 0,24, IC à 95% : 0,06 1,03). L’incidence des effets indésirables n’était rapportée explicitement que dans une seule étude ; les pourcentages de patients ayant présenté des effets indésirables ont été plus élevés dans le groupe STP que dans le groupe placebo (100 % versus 25 % ; RR : 3,73, IC à 95% : 1,81 7,67)

Conclusions des auteurs

Les données issues de deux petites études randomisées contrôlées indiquent que le stiripentol est significativement plus efficace que le placebo en termes de réduction d’au moins 50% de la fréquence des crises et d’absence de crises. Les effets indésirables ont été plus fréquents avec le STP. D’autres études ayant la puissance adéquate avec un suivi à long terme doivent être menées afin d’établir clairement l’efficacité et la tolérance à long terme du stiripentol dans le traitement de l’EMSN.

Plain language summary

Antiepileptic drugs for the treatment of severe myoclonic epilepsy in infancy

Severe myoclonic epilepsy in infants (SMEI), also known as Dravet syndrome, is a rare, drug-resistant form of epilepsy, for whose treatment stiripentol (STP) has been recently licensed for use in combination with other antiepileptic drugs. In this review we evaluated the efficacy and tolerability of STP and other antiepileptic drug treatments in the treatment of SMEI. Upon systematically searching (15th May 2013) the medical literature, we found no randomised controlled trials (RCTs) assessing drugs other than STP. We found two RCTs evaluating the efficacy and tolerability of STP in 64 children with SMEI compared with placebo. Compared with patients given placebo, patients receiving STP were more likely to be seizure free or to have 50% or greater reduction in seizure frequency. Only one study explicitly reported the occurrence of side effects, which occurred more frequently in patients treated with STP. These data, derived from two small studies, indicate that STP is significantly better than placebo with regards to efficacy but is less well tolerated. Further studies should be conducted to unequivocally establish the long-term efficacy and tolerability of STP in the treatment of SMEI.

Résumé simplifié

Médicaments antiépileptiques pour le traitement d’épilepsie myoclonique sévère chez le nourrisson

L’épilepsie myoclonique sévère du nourrisson (EMSN), également connue sous le nom de syndrome de Dravet, est une forme d’épilepsie rare et réfractaire aux médicaments, pour laquelle le stiripentol (STP) a été récemment homologué en traitement en association avec d'autres médicaments antiépileptiques. Dans cette revue, nous avons évalué l'efficacité et la tolérance du STP et d’autres traitements antiépileptiques chez des patients présentant une EMSN. Suite à une recherche systématique (15 mai 2013) dans la littérature médicale, nous n'avons trouvé aucun essai contrôlé randomisé (ECR) évaluant des médicaments autres que le STP. Nous avons identifié deux ECR évaluant l'efficacité et la tolérance du STP chez 64 enfants atteints d’EMSN par rapport à un placebo. Comparés aux patients ayant reçu un placebo, les patients recevant le STP étaient plus susceptibles de ne pas faire de crise ou d'avoir une réduction de 50% ou plus dans la fréquence des crises. Seule une étude a explicitement rapporté la survenue d'effets secondaires, qui se produisait plus fréquemment chez les patients traités par STP. Ces données, issues de deux petites études, indiquent que le STP est significativement plus efficace que le placebo, mais est moins bien toléré. De nouvelles études devraient être menées pour établir clairement l'efficacité à long terme et la tolérance du STP dans le traitement d’EMSN.

Notes de traduction

Traduit par: French Cochrane Centre 1st November, 2013
Traduction financée par: Financeurs pour le Canada : Instituts de Recherche en Santé du Canada, Ministère de la Santé et des Services Sociaux du Québec, Fonds de recherche du Québec-Santé et Institut National d'Excellence en Santé et en Services Sociaux; pour la France : Ministère en charge de la Santé

Summary of findings(Explanation)

Summary of findings for the main comparison. 
STP compared with placebo for SMEI

Patient or population: children with SMEI

Settings:

Intervention: STP 50 mg/kg/d

Comparison: placebo

OutcomesRelative effect
(95% CI)
No of participants
(studies)

≥ 50% reduction in seizure frequency

2 months' follow-up

RR 10.40 (2.64 to 40.87)64
(2)

Seizure freedom

2 months' follow-up

RR 7.93 (1.52 to 41.21)64
(2)

Proportion of dropouts

2 months' follow-up

RR 0.24 (0.06 to 1.03)64
(2)
Adverse effects RR 3.73 (1.81 to 7.67)

41

(1)

Background

Description of the condition

Severe myoclonic epilepsy in infants (SMEI), also known as Dravet syndrome (Dravet 2005), is one of the most refractory forms of epilepsy, for whose treatment stiripentol (STP) has been recently licensed for add-on use. It is a rare disease, with an estimated incidence ranging from 1 per 20,000 (Yakoub 1992) to 1 per 40,000 children (Hurst 1990). De novo mutations in the sodium channel SCN1A are a major cause of this syndrome (Fujiwara 2003) and have been found in 40% to 70% of patients with typical SMEI (Nabbout 2003).

According to the International League Against Epilepsy classification (ILAE 1989), SMEI is defined by "febrile and afebrile generalized and unilateral, clonic or tonic–clonic, seizures, that occur in the first year of life in an otherwise normal infant and are later associated with myoclonus, atypical absences, and partial seizures. All seizure types are resistant to antiepileptic drugs (AEDs). Developmental delay becomes apparent within the second year of life and is followed by definite cognitive impairment and personality disorders."

In this nosologically and aetiologically homogeneous syndrome, seizures are refractory to conventional AEDs. Newer drugs, such as STP, have been proposed to achieve better seizure control in SMEI.

Description of the intervention

Stiripentol is a new AED developed in France and recently approved by the European Medicines Agency (EMEA) for the treatment of Dravet syndrome, as an adjunct treatment with valproate (VPA) and clobazam (CLB), with a promising effect (Chiron 2007).

Some adverse events associated with STP have been determined to be related to a significant increase in the plasma concentration of VPA and CLB after the addition of STP (Perez 1999). They include drowsiness, ataxia, nausea, abdominal pain and loss of appetite with weight loss. Asymptomatic neutropenia is also occasionally observed (Chiron 2007).

How the intervention might work

Stiripentol is structurally unrelated to any other marketed AED. A GABAergic effect of STP has been demonstrated in vitro (Quilichini 2006) and is probably due to an allosteric modulation of the GABA-A receptor (Fisher 2009). Therefore the efficacy of STP could be secondary to a potentiation of GABAergic inhibitory neurotransmission (Quilichini 2006) and enhancement of the action of benzodiazepines (Fisher 2009). In humans, STP also inhibits the cytochrome P450 enzymes (CYP) in the liver, resulting in an increased plasma concentration of concomitant AEDs metabolised by CYP (Chiron 2005). In Dravet patients, such a pharmacokinetic interaction particularly applies to clobazam (Giraud 2006).

Why it is important to do this review

To date, studies evaluating STP in SMEI have been uncontrolled and have included few participants, leading to potentially biased estimates of treatment effect.

Recently, a systematic review and meta-analysis of individual participant data was conducted to evaluate the efficacy of available treatments for SMEI (Kassai 2008). Twenty-three uncontrolled studies and two randomised controlled trials (RCTs) comparing STP with placebo were found. Overall, 64 children aged between 3 and 20 years were included in the two RCTs. The odds ratio of responding to STP relative to placebo was 32 (confidence Interval (CI) 6.2 to 161), and STP reduced the seizure rate by 70% (93% to 47%). However, this review did not consider the tolerability of STP. Moreover to date, no study has systematically reviewed the literature on the role of other AEDs in the treatment of SMEI.

In this systematic review, we aimed to assess and summarise existing evidence regarding the efficacy and adverse effects of STP and other AEDs (including ketogenic diet, although it is an antiepileptic treatment rather than a drug) for people with SMEI.

Objectives

To evaluate the efficacy and tolerability of STP and other antiepileptic drug treatments (including ketogenic diet) as therapy for patients with SMEI.

Methods

Criteria for considering studies for this review

Types of studies

We included studies that met all of the following criteria.

  1. Randomised controlled trials (RCTs) or quasi-randomised controlled trials (e.g. sequence generated by date of birth or by clinical record number).

  2. Double-blinded, single-blinded or unblinded trials.

  3. Parallel-group studies.

We excluded all other study designs, including cohort studies, case-control studies, outcomes research, case studies, case series and expert opinion.

We did not impose any language restrictions.

Types of participants

We considered people with a defined diagnosis of severe myoclonic epilepsy in infancy (SMEI)/Dravet syndrome made according to International League Against Epilepsy criteria (ILAE 1989), regardless of age, sex, ethnicity and prior therapy, including children with disabilities.

Types of interventions

  1. Any trial that compares at least one antiepileptic drug therapy against placebo.

  2. Any trial that compares at least one antiepileptic drug therapy against no therapy.

  3. Any trial that compares at least one antiepileptic drug therapy against another therapy or a different dose of the same therapy.

We also included trials evaluating ketogenic diet.

The therapies may be given singly (monotherapy) or in combination (add-on therapy).

Types of outcome measures

Primary outcomes
  1. Fifty per cent or greater reduction in seizure frequency: proportion of participants with 50% or greater reduction in seizure frequency at the end of the study compared with the pre-randomisation baseline period.

  2. Seizure freedom: proportion of participants achieving total cessation of seizures at the end of the trial. We will use the most current International League Against Epilepsy (ILAE) proposed definition of seizure freedom: no seizures of any type for either 12 months or 3 times the longest (pre-intervention) seizure-free interval, whichever is longest (Kwan 2010).

For each outcome, we made an intention-to-treat primary analysis to include all participants in the treatment group to which they were allocated, irrespective of the treatment they actually received. The analysis used intention-to-treat (ITT) data of all randomly assigned participants recorded during the entire treatment period, including both titration and evaluation phases.

Secondary outcomes

1. Adverse effects.

  • Proportion of participants experiencing at least one adverse effect.

  • Proportion of participants experiencing each separate adverse effect.

2. Proportion of dropouts/withdrawals due to side effects, lack of efficacy or other reasons. This is used as a measure of global effectiveness.

3. Improvement in quality of life as assessed by validated and reliable rating scales.

Search methods for identification of studies

Electronic searches

We searched the following databases on 15/05/2013.

  • The Cochrane Epilepsy Group Specialised Register using the search strategy set out in Appendix 1.

  • The Cochrane Central Register of Controlled Trials (CENTRAL, Issue 4, The Cochrane Library, April 2013) using the search strategy set out in Appendix 2.

  • MEDLINE (Ovid, 1946 to 15/05/2013) using the search strategy set out in Appendix 3.

  • SCOPUS (1823 to 15/05/2013) using the search strategy set out in Appendix 4.

  • ClinicalTrials.gov using the search strategy set out in Appendix 5.

  • The WHO International Clinical Trials Registry Platform (ICTRP) using the search strategy set out in Appendix 6.

We did not impose any language restrictions.

Searching other resources

We contacted experts in the field for information about any unpublished or ongoing studies. We reviewed the reference lists of articles retrieved by the electronic searches to check for other relevant reports. To find other relevant trials, we reviewed conference proceedings not included in the Cochrane Epilepsy Group Specialised Register after checking with the Trials Search Co-ordinator to avoid duplication of effort.

Data collection and analysis

Selection of studies

Two review authors (FB and MS) independently screened all titles and abstracts of publications identified by the searches to assess their eligibility. We excluded publications that did not meet the criteria at this stage. After screening, we assessed the full text of potentially eligible citations for inclusion. The review authors reached consensus on the selection of trials and the final list of studies. If we could not reach consensus, we consulted a third member of the review team (ADF).

Data extraction and management

Two review authors (FB and MS) independently extracted the following characteristics of each included trial from the published reports, when possible. We used data extraction forms and resolved any disagreements by mutual agreement. We recorded the rawest form of the data, when possible. In cases of missing or incomplete data, we contacted the principal investigators of included trials to request additional information.

Participant factors
  1. Age

  2. Sex

  3. Epileptic seizure types

  4. Etiology of epilepsy

  5. Duration of epilepsy

  6. Family history of epilepsy

  7. Number of seizures or seizure frequency before randomisation

  8. Presence of status epilepticus

  9. Number and types of AEDs previously taken

  10. Concomitant AEDs

  11. Presence of neurological deficit/signs

  12. Neuropsychological status

  13. Electroencephalography (EEG) findings

  14. Neuroradiological findings (computed tomography (CT)/magnetic resonance imaging (MRI))

Trial design
  1. Criteria used to diagnose epilepsy

  2. Inclusion and exclusion criteria

  3. Method of randomisation

  4. Method of allocation concealment

  5. Method of blinding

  6. Stratification factors

  7. Number of participants allocated to each group

  8. Length of baseline period, titration period, treatment period

Intervention and control
  1. Intervention given to controls

  2. Dosage of AED

Follow-up data
  1. Duration of follow-up

  2. Reasons for incomplete outcome data

  3. Dropout or loss to follow-up rates

  4. Methods of analysis (e.g. intention-to-treat, modified intention-to-treat, per protocol, worst-case or best-case scenario analysis)

Assessment of risk of bias in included studies

Two review authors (FB and MS) assessed the risk of bias of each trial according to the approaches described in The Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We assessed the risk of bias as low, high or unclear risk of bias. We evaluated the following characteristics:

  1. Random sequence generation (selection bias).

  2. Allocation concealment (selection bias).

  3. Blinding of participants and personnel (performance bias).

  4. Blinding of outcome assessment (detection bias).

  5. Incomplete outcome data addressed.

  6. Selective reporting (reporting bias).

  7. Other bias, including outcome reporting bias (Kirkham 2010), assessed in accordance with The Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

Measures of treatment effect

We used statistical methods in accordance with The Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) to measure treatment effect.

We analysed dichotomous data by calculating risk ratios (RRs) for each trial, with the uncertainty in each trial expressed with the use of 95% CI. For dichotomous data, we calculated a weighted treatment effect across trials from the meta-analysis.

Unit of analysis issues

Given the insufficient number of studies, we did not undertake a meta-regression to explore the effect of trial duration on treatment effect.

Dealing with missing data

For individual missing data, such as information on dropout or loss to follow-up, we used an intention-to-treat analysis, using as the denominator the total number of people who underwent randomisation. Intention-to-treat analysis was used only for outcomes related to efficacy (fifty per cent or greater reduction in seizure frequency and seizure freedom).

Assessment of heterogeneity

Visual inspection of the forest plots was used to investigate the possibility of statistical heterogeneity. We evaluated homogeneity among trial results using a standard Chi2 test; the hypothesis of homogeneity was rejected if the P value was less than 0.10.

Assessment of statistical heterogeneity was supplemented using the I2 statistic, which provides an estimate of the percentage of variability due to heterogeneity rather than a sampling error (Higgins 2003).

Interpretation of I2 for heterogeneity is as follows.

  • 0% to 40%: may not be important.

  • 30% to 60%: represents moderate heterogeneity.

  • 50% to 90%: represents substantial heterogeneity.

  • 75% to 100%: represents considerable heterogeneity.

We combined trial outcomes to obtain a summary estimate of effect (and the corresponding confidence interval (CI)) using a fixed-effect model, unless significant heterogeneity was noted (i.e. I2 > 75%).

We tested the robustness of our conclusions by performing a sensitivity analysis, as described later.

Assessment of reporting biases

The number of studies was insufficient to permit assessment of reporting biases using a funnel plot.

Data synthesis

We used statistical methods in accordance with The Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) to measure treatment effect.

We synthesised the trial results in a meta-analysis using the Mantel-Haenszel method for analysing dichotomous data.

We performed separate analyses on each intervention (AED or ketogenic diet) given singly (monotherapy) or in combination (add-on therapy). Different control groups were analysed separately.

We synthesised data on all seizures. Because of lack of data, it was impossible to perform an analysis according to seizure type. We used Review Manager to combine trial data.

Sensitivity analysis

For participants with missing data (i.e. dropout participants), we used the worst-case and best-case scenarios for efficacy outcomes (50% or greater reduction in seizure frequency and seizure freedom).

Worst-case analysis

Participants randomly assigned but excluded from analysis (e.g. for not completing follow-up or with inadequate seizure data) were assumed non-responders in the active treatment group and responders in the placebo/no treatment group.

Best-case analysis

Participants randomly assigned but excluded from analysis (e.g. for not completing follow-up or with inadequate seizure data) were assumed responders in the active treatment group and non-responders in the placebo/no treatment group.

Results

Description of studies

Results of the search

The search strategy described earlier yielded 81 results (7 Epilepsy Specialised Register; 8 CENTRAL; 28 MEDLINE; 27 SCOPUS; 4 ClinicalTrials.gov; 7 ICTRP). After duplicate publications (27), reviews (27) and uncontrolled trials (13) were excluded, 14 studies with the potential to be included were provisionally selected. We excluded 11 studies after reading the full published papers: 9 excluded studies were uncontrolled clinical trials (Coppola 2002, JPRN-JapicCTI-101116 2010, JPRN-JMA-IIA00014 2007, NCT01533506 2012, NCT01607073 2012, Nguyen Thanh 2002, Oguni 1994, Striano 2007, Thanh 2002), and 2 were abstracts (Chiron 1999a, Chiron 1999b) of a study subsequently published "in extenso" (Chiron 2000). One systematic review (Kassai 2008) with meta-analysis of individual participant data from two RCTs (Chiron 2000, Guerrini 2002) was taken into consideration, although not formally included in the present review, to obtain further data on the primary studies. Thus, two RCTs contributed to this review: The earliest was published in 2000 (Chiron 2000) and the most recent in 2002 (Guerrini 2002).

Included studies

All included studies evaluated the use of STP in the treatment of SMEI.

The study of Chiron 2000 was a randomised placebo-controlled add-on trial conducted in 15 French centres. The study was conducted in children aged ≥ 3 years with SMEI (defined as onset of epilepsy in the first year of life with clonic (or tonic-clonic) generalised seizures but normal psychomotor development and normal electroencephalogram (EEG); appearance of myoclonia after 1 year of age; atypical absences; generalised spikes and waves on EEG; and mental delay) with at least four clonic (or tonic-clonic) generalised seizures/month, and with VPA and CLB as ongoing antiepileptic drugs. Patients receiving other drugs (except progabide) and those whose parents were unable to comply regularly with drug delivery and the daily seizure diary were excluded.

A total of 42 participants with SMEI (one lost to follow-up) were included in the trial. The STP group consisted of 21 participants (28.6% male, 71.4% female; mean age 9.4 years; range 3 to 16.7 years); the placebo group included 20 participants (55% male, 45% female; mean age 9.3 years, range 3.2 to 20.7 years). One participant allocated to STP could not be assessed (no further information was provided by the authors). Furthermore, five participants dropped out: 1 (5%) treated with STP and 4 (20%) given placebo. Reasons for dropping out were not explicitly reported. It was stated only that one dropout given placebo was withdrawn because of an adverse event (not further specified).

After a 1-month baseline, participants were randomly allocated to STP (50 mg/kg a day, twice or three times daily) or placebo as add-on therapy. This double-blind phase lasted 2 months and was followed by an open phase that lasted at least 1 month. Maximum dose of co-medication was limited to 30 mg/kg a day for VPA and 0.5 mg/kg a day for CLB. Doses could be decreased by 10 mg/kg daily for VPA in case of loss of appetite and by 25% for CLB in case of drowsiness or hyperexcitability. Participants were assessed every month during the double-blind period of 2 months and in subsequent open treatment for at least 1 month.

One Italian study (Guerrini 2002), published as an abstract, was a randomised placebo-controlled double-blind add-on trial conducted in 22 refractory participants with SMEI (11 male, 11 female; aged 9.1 ± 4.0 years, range 3.5 to 18.9 years; sex and age for each study group not explicitly reported) under VPA + CLB therapy. Five participants dropped out: 1 (0.9%) on STP and 4 (36%) on placebo. Reasons for dropping out were not explicitly reported. During a 1-month baseline period, VPA and CLB were administered at 20 and 0.5 mg/kg/24 h respectively. During the following 2-month double-blind period, participants were randomly assigned to receive either placebo (n = 11) or STP (n = 11) at 50 mg/kg/24 h as add-on therapy. Responders were defined as having greater than 50% reduction in the rate of clonic seizures during the second month of the double-blind period compared with baseline.

It is worth considering a discrepancy in the number of participants in this RCT as reported in the abstract that reported the data of this study and in the systematic review that included this study (Kassai 2008). Whereas the abstract specified that the study was conducted in 22 children (11 male, 11 female; aged 9.1 ± 4.0 years, range 3.5 to 18.9 years), the systematic review provided different data (23 children, 56.1% male, 43.9% female, mean age 8.95 years, range 3.47 to 18.9 years). Such a discrepancy may be easily explained by the inclusion of an additional participant after publication of the abstract that reported preliminary results of the study. It was impossible to include this additional participant in the sensitivity analysis, as it was specified neither whether the participant was male or female, nor to which group the participant was allocated.

Excluded studies

Nine excluded studies were uncontrolled clinical trials (Coppola 2002, JPRN-JapicCTI-101116 2010, JPRN-JMA-IIA00014 2007, NCT01533506 2012, NCT01607073 2012, Nguyen Thanh 2002, Oguni 1994, Striano 2007, Thanh 2002), and two were abstracts (Chiron 1999a, Chiron 1999b) of a study subsequently published "in extenso" (Chiron 2000). One systematic review with meta-analysis of individual participant data from two RCTs (Chiron 2000, Guerrini 2002) was taken into consideration, although not formally included in the present review, to obtain further data on the primary studies.

Risk of bias in included studies

See also Figure 1.

Figure 1.

Risk of bias graph: review authors' judgements about all risk of bias items presented as percentages across all included studies.

In the study of Chiron 2000, participants were randomly allocated with the use of a computer-generated list (low risk of selection bias). The trial was defined as "double-blind" without further information (unclear risk of performance and detection bias). No outcome data were missing (low risk of attrition bias), and, although the study protocol was not available, the published report included all expected outcomes, including those that were prespecified (low risk of reporting bias).

The study of Guerrini 2002 was published as abstract, hence very limited information was provided to permit judgement with regards to selection bias, performance and detection bias, attrition bias and reporting bias. It is worth reporting that this study was defined as "randomised" not in the abstract that provided data of this trial (Guerrini 2002), but in a subsequent systematic review (Kassai 2008) that included this study. Furthermore, in the abstract, the trial was simply defined as "double-blind" without further information on blinding.

Effects of interventions

See: Summary of findings for the main comparison

In the study of Chiron 2000, the frequency of participants with 50% or greater reduction in seizure frequency was greater among those on STP (15 of 21 participants, 71%) than on placebo (1 of 20 participants, 5%). One participant allocated to STP could not be assessed (no further information provided from the authors). Furthermore, five participants dropped out: 1 (5%) on STP and 4 (20%) on placebo. Reasons for dropping out were not explicitly reported. It was stated only that one dropout on placebo was withdrawn because of an adverse event (not further specified). Nine of 21 (43%) participants on STP but none on placebo became free of clonic (or tonic-clonic) seizures. In each group, one participant had status epilepticus. Absolute seizure frequency was significantly lower on STP than on placebo after a decrease of 69% on STP, but an increase of 7% on placebo. Drug-related adverse events were noticed in the 21 (100%) participants on STP versus 5 (25%) on placebo.

The RCT of Guerrini 2002 chose as the main outcome 50% or greater reduction in seizure frequency. Five participants dropped out: 1 (0.9%) on STP and 4 (36%) on placebo. Reasons for dropping out were not explicitly reported. During the double-blind period, 7/21 participants on STP (63.6%) reached a 50% or greater reduction in seizure frequency, and 3/21 were seizure-free; 1 participant on placebo (9.1%) reached seizure freedom. At the end of the double-blind period, the percentage reduction in seizure rate from baseline was significantly higher on STP (72%) than on placebo (22%) (P < 0.05). Authors stated that the mean clonic seizure rate was lower on STP than on placebo, but the difference did not reach statistical significance because of the small number of participants and the large interindividual variability. However, no statistical results were explicitly provided. Five participants (1 on STP and 4 on placebo) dropped out (reasons for dropping out were not explicitly reported).

Comparison STP versus placebo

Fifty per cent or greater reduction in seizure frequency

Two studies with 64 participants were identified. No significant statistical heterogeneity among trials was detected. A statistically significant higher proportion of participants had 50% or greater reduction in seizure frequency from the pre-randomisation baseline period in the STP group compared with the placebo group (22/33 vs 2/31 participants; RR 10.40, 95% CI 2.64 to 40.87).

Worst-case analysis

Two studies with 64 participants were identified (hence including the 2 participants missing in the STP group and 8 in the placebo group). In this analysis, we assumed the participant missing in the STP group as not experiencing 50% or greater reduction in seizure frequency and the participant missing in the placebo group as experiencing 50% or greater reduction in seizure frequency. No significant statistical heterogeneity among trials was detected. A statistically significant higher proportion of participants had 50% or greater reduction in seizure frequency from the pre-randomisation baseline period in the STP group compared with the placebo group (22/33 vs 10/31 participants; RR 2.08, 95% CI 1.18 to 3.68).

Best-case analysis

Two studies with 64 participants were identified (hence including 2 participants missing in the STP group and 8 in the placebo group). In this analysis, we assumed the participant missing in the STP group as experiencing 50% or greater reduction in seizure frequency and the participant missing in the placebo group as not experiencing 50% or greater reduction in seizure frequency. No significant statistical heterogeneity among trials was detected. A statistically significant higher proportion of participants with 50% or greater reduction in seizure frequency from the pre-randomisation baseline period was reported in the STP group compared with the placebo group (24/33 vs 2/31 participants; RR 11.35, 95% CI 2.91 to 44.31).

Seizure freedom

Two studies with 64 participants were identified. No significant statistical heterogeneity among trials was detected. A statistically significant higher proportion of participants with seizure freedom was reported in the STP group compared with the placebo group (12/33 vs 1/31 participants; RR 7.93, 95% CI 1.52 to 41.21).

Worst-case analysis

Two studies with 64 participants were identified (hence including 2 participants missing in the STP group and 8 in the placebo group). In this analysis, we assumed the participant missing in the STP group as not experiencing seizure freedom and the participant missing in the placebo group as experiencing seizure freedom. No significant statistical heterogeneity among trials was detected. No statistically significant difference in the proportion of participants with seizure freedom was reported in the STP group compared with the placebo group (12/33 vs 9/31 participants; RR 1.26, 95% CI 0.61 to 2.62).

Best-case analysis

Two studies with 64 participants were identified (hence including 2 participants missing in the STP group and 8 in the placebo group). In this analysis, we assumed the participant missing in the STP group as experiencing seizure freedom and the participant missing in the placebo group as not experiencing seizure freedom. No significant statistical heterogeneity among trials was detected. A statistically significant higher proportion of participants with seizure freedom was reported in the STP group compared with the placebo group (14/33 vs 1/31 participants; RR 9.21, 95% CI 1.80 to 47.01).

Adverse effects

Only one study (Chiron 2000) explicitly reported the occurrence of adverse effects, hence preventing us from pooling results in a meta-analysis.

In the study of Chiron 2000, drug-related adverse events were noticed in 21/21 (100%) participants on STP versus 5/20 (25%) on placebo (RR 3.73, 95% CI 1.81 to 7.67). The adverse events most frequently observed on STP were drowsiness (19/21, 90% participants; RR 18.10 95% CI 2.67 to 122.86) and loss of appetite (7/21, 33%; RR 14.32, 95% CI 0.87 to 235.36), which could result in loss of weight (6/21, 29%; RR 12.41, 95% CI 0.74 to 206.86), whereas weight gain was equally reported for STP (5/21, 24%) and placebo (4/20, 25%) (RR 1.19, 95% CI 0.37 to 3.81). Adverse events were regarded as severe in 5/21 (24%) participants on STP (drowsiness in three, loss of weight in two) and in one on placebo (drowsiness). Some events led to a decrease in the co-medication dose (in 17 participants), as planned by the protocol; the events then disappeared. Only one participant was withdrawn from the study because of an adverse event; he was on placebo.

The study of Guerrini 2002 did not explicitly report the occurrence of any adverse effect.

Proportion of dropouts

Two studies with 64 participants were identified. No significant statistical heterogeneity among trials was detected. No statistically significant difference in proportion of dropouts was reported in the STP group compared with the placebo group (2/33 vs 8/31 participants; RR 0.24, 95% CI 0.06 to 1.03).

Improvement in quality of life

No study measured this outcome.

Discussion

This systematic review of the literature with meta-analysis indicates that STP is significantly better than placebo with regards to 50% or greater reduction in seizure frequency and seizure freedom in the treatment of SMEI (also called Dravet syndrome). Data on 50% or greater reduction in seizure frequency remained unchanged in the sensitivity analyses with worst- and best-case scenarios (performed as an intention-to-treat analysis using imputation for participants with missing data), hence supporting a good efficacy profile of STP compared with placebo. No statistically significant difference in the proportion of dropouts was reported in the STP group compared with the placebo group. Only one study explicitly reported the occurrence of adverse effects, with a higher proportion of participants with side effects reported in the STP group compared with the placebo group. No study measured quality of life. Another relevant limitation of the included RCTs was the short-term follow-up duration (2 months in each study). The results of this meta-analysis should be read with cautiousness, mainly because of the limited number of included studies (each including a small number of participants).

It is of paramount importance to identify epileptic syndromes that may be studies in RCTs assessing the efficacy of a new AED. To identify these syndromes, the European Medicine Agency (EMA) recently recommended the following two-step procedure: (1) an exploratory prospective-observational trial including a large variety of paediatric epilepsy syndromes and (2) a subsequent RCT to be performed in each of those that disclose a signal for possible efficacy (Chiron 2013). A recent study aimed to determine a minimal threshold for this signal for efficacy, estimating the number of participants to be included in such exploratory prospective-observational trials. The threshold was defined as the lowest percentage of responders observed in a prospective-observational trial with a positive corresponding RCT. The minimal number of participants to include for each syndrome for a prospective-observational trial with a new AED was estimated to reach at least this threshold of responders with a 95% confidence interval. The minimal responder threshold was found to be 25%. Among paediatric epileptic syndromes, SMEI was found to reach this threshold, with a minimal sample needed of 32 participants (Chiron 2013).

It is worth considering that only one of the included studies in the present meta-analysis had a sample size of more than 32 participants (Chiron 2000), whereas the meta-analysis assessed 64 participants, hence leading to increased statistical power. This highlights the importance of conducting systematic reviews with pooled analyses (meta-analyses) of data from the literature to allow an increase in statistical power and an improvement in precision.

Summary of main results

Data derived from two small RCTs indicate that STP is significantly better than placebo with regards to 50% or greater reduction in seizure frequency and seizure freedom. Adverse effects occurred more frequently with STP.

Quality of the evidence

Both included studies were conducted in small populations. Only one study explicitly reported the occurrence of adverse effects. No study measured quality of life. Both included studies had a short-term follow-up duration (2 months). Although data derived from RCTs in small populations suggest that STP is significantly better than placebo in achieving seizure control, further adequately powered studies with long-term follow-up duration should be conducted to unequivocally establish efficacy and tolerability of this drug in patients with SMEI.

Agreements and disagreements with other studies or reviews

Results of the present systematic review are consistent with results of a previous systematic review with meta-analysis of individual participant data (Kassai 2008), which included the results of the two studies identified in the present review (Chiron 2000, Guerrini 2002). In this systematic review, the number of participants from the Italian study was different (23 children, 56.1% male, 43.9% female) from that given in the published abstract reporting the results of the study (22 children, 11 male, 11 female) Guerrini 2002). Such a discrepancy may be easily explained by the inclusion of an additional participant after publication of the abstract reporting preliminary results of the study. However, it was impossible to include this additional participant in the sensitivity analysis, as it was specified neither whether the participant was male or female nor to which group the participant was allocated.Overall, 64 children aged between 3 and 20 years were included in the analysis of the treatment effect on seizure rate. STP resulted in an overall OR of response of 32 (95% CI 6.2 to 161): 47 (5.1 to 438) in the Chiron 2000 and 20 (1.85 to 216) in the Italian study (Guerrini 2002). No heterogeneity was detected between these studies (P = 0.63). The overall seizure rate was reduced by 70% (95% CI 47% to 93%): by 62% (P < 0.05) in the French study, and by 74% in the Italian study (P < 0.05). Authors concluded that the two RCTs included were performed with the same objectives and design and showed that seizure frequency is greatly reduced by STP in children with SMEI after two months of treatment.

Authors' conclusions

Implications for practice

Data derived from two small RCTs indicate that STP is significantly better than placebo with regards to 50% or greater reduction in seizure frequency and seizure freedom in the treatment of Dravet syndrome. Adverse effects occurred more frequently with STP.

Implications for research

Further adequate RCTs with long-term follow-up duration (at least 6 months) should be conducted to unequivocally establish the long-term efficacy and tolerability of STP in the treatment of Dravet syndrome.

Data and analyses

Download statistical data

Comparison 1. STP versus placebo
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 ≥ 50% reduction in seizure frequency264Risk Ratio (M-H, Fixed, 95% CI)10.40 [2.64, 40.87]
2 Seizure freedom264Risk Ratio (M-H, Fixed, 95% CI)7.93 [1.52, 41.21]
3 Proportion of dropouts264Risk Ratio (M-H, Fixed, 95% CI)0.24 [0.06, 1.03]
4 ≥ 50% reduction in seizure frequency, worst-case analysis264Risk Ratio (M-H, Fixed, 95% CI)2.08 [1.18, 3.68]
5 ≥ 50% reduction in seizure frequency, best-case analysis264Risk Ratio (M-H, Fixed, 95% CI)11.35 [2.91, 44.31]
6 Seizure freedom, worst-case analysis264Risk Ratio (M-H, Fixed, 95% CI)1.26 [0.61, 2.62]
7 Seizure freedom, best-case analysis264Risk Ratio (M-H, Fixed, 95% CI)9.21 [1.80, 47.01]
8 Patients with adverse effects (any type)141Risk Ratio (M-H, Fixed, 95% CI)3.73 [1.81, 7.67]
9 Drowsiness141Risk Ratio (M-H, Fixed, 95% CI)18.10 [2.67, 122.86]
10 Loss of appetite141Risk Ratio (M-H, Fixed, 95% CI)14.32 [0.87, 235.36]
11 Loss of weight141Risk Ratio (M-H, Fixed, 95% CI)12.41 [0.74, 206.86]
12 Weight gain141Risk Ratio (M-H, Fixed, 95% CI)1.19 [0.37, 3.81]
Analysis 1.1.

Comparison 1 STP versus placebo, Outcome 1 ≥ 50% reduction in seizure frequency.

Analysis 1.2.

Comparison 1 STP versus placebo, Outcome 2 Seizure freedom.

Analysis 1.3.

Comparison 1 STP versus placebo, Outcome 3 Proportion of dropouts.

Analysis 1.4.

Comparison 1 STP versus placebo, Outcome 4 ≥ 50% reduction in seizure frequency, worst-case analysis.

Analysis 1.5.

Comparison 1 STP versus placebo, Outcome 5 ≥ 50% reduction in seizure frequency, best-case analysis.

Analysis 1.6.

Comparison 1 STP versus placebo, Outcome 6 Seizure freedom, worst-case analysis.

Analysis 1.7.

Comparison 1 STP versus placebo, Outcome 7 Seizure freedom, best-case analysis.

Analysis 1.8.

Comparison 1 STP versus placebo, Outcome 8 Patients with adverse effects (any type).

Analysis 1.9.

Comparison 1 STP versus placebo, Outcome 9 Drowsiness.

Analysis 1.10.

Comparison 1 STP versus placebo, Outcome 10 Loss of appetite.

Analysis 1.11.

Comparison 1 STP versus placebo, Outcome 11 Loss of weight.

Analysis 1.12.

Comparison 1 STP versus placebo, Outcome 12 Weight gain.

Appendices

Appendix 1. Cochrane Epilepsy Group Specialised Register search strategy

"severe myoclonic epilepsy" OR SMEI OR Dravet*

Appendix 2. Cochrane Central Register of Controlled Trials search strategy

#1¬¬¬¬¬¬¬¬ "severe myoclonic epilepsy":ti,ab,kw¬ (Word variations have been searched)
#2¬¬¬¬¬¬¬¬ "SMEI":ti,ab,kw¬ (Word variations have been searched)
#3¬¬¬¬¬¬¬¬ Dravet*:ti,ab,kw¬ (Word variations have been searched)
#4¬¬¬¬¬¬¬¬ #1 or #2 or #3 in Trials

Appendix 3. MEDLINE search strategy

This search strategy is based on the Cochrane Highly Sensitive Search Strategy for identifying randomised trials (Lefebvre 2011).

1. (randomized controlled trial or controlled clinical trial).pt. or (randomized or placebo or randomly).ab.

2. clinical trials as topic.sh.

3. trial.ti.

4. 1 or 2 or 3

5. exp animals/ not humans.sh.

6. 4 not 5

7. severe myoclonic epilepsy.tw.

8. SMEI.tw.

9. Dravet$ syndrome.tw.

10. 7 or 8 or 9

11. 6 and 10

Appendix 4. SCOPUS search strategy

(TITLE((randomiz* OR randomis* OR controlled OR placebo OR blind* OR unblind* OR "parallel group" OR crossover OR "cross over" OR cluster OR "head to head") PRE/2 (trial OR method OR procedure OR study)) OR ABS((randomiz* OR randomis* OR controlled OR placebo OR blind* OR unblind* OR "parallel group" OR crossover OR "cross over" OR cluster OR "head to head") PRE/2 (trial OR method OR procedure OR study))) AND (TITLE-ABS-KEY("severe myoclonic epilepsy" OR SMEI OR Dravet*))

Appendix 5. ClinicalTrials.gov search strategy

"severe myoclonic epilepsy" OR SMEI OR Dravet*

Appendix 6. WHO International Clinical Trials Registry Platform ICTRP search strategy

severe myoclonic epilepsy OR SMEI OR Dravet*

Contributions of authors

Francesco Brigo and Monica Storti extracted the data, performed the meta-analysis and wrote the manuscript.

Declarations of interest

None known.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Chiron 2000

Methods

Multicenter randomised placebo-controlled double-blind add-on trial

1 month baseline period

2 months double-blind treatment period

Participants

42 participants with SMEI, 1 lost to follow-up

STP group: 6 M, 15 F; age 9.4 years (mean) (range 3 to 16.7 years)

Placebo group: 11 M, 9 F; age 9.3 years (mean) (range 3.2 to 20.7 years)

InterventionsSTP 50 mg/kg/d versus placebo
Outcomes
  • Fifty per cent or greater reduction in clonic (or tonic-clonic) seizure frequency the second month of the double-blind period compared with baseline

  • Count of clonic (or tonic-clonic) seizures during the second month of the double-blind period

  • Percentage of change from baseline

NotesMulticenter study conducted in France
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskParticipants were randomly allocated with the use of a computer-generated list
Allocation concealment (selection bias)Unclear riskInsufficient information to permit judgement
Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskInsufficient information to permit judgement. The trial is defined as "double-blind" without further information
Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskInsufficient information to permit judgement. The trial is defined as "double-blind" without further information
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data
Selective reporting (reporting bias)Low riskThe study protocol is not available, but the published report includes all expected outcomes, including those that were prespecified

Guerrini 2002

Methods

Multicenter randomised placebo-controlled double-blind add-on trial

1 month baseline period

2 months double-blind treatment period

Participants22 refractory participants with SME: 11 M, 11 F; age 9.1 years (mean), range 3.5 to 18.9 years under valproate + clobazam therapy
InterventionsSTP at 50 mg/kg/24 h versus placebo
OutcomesFifty per cent or greater reduction in seizure frequency
Notes

Published as abstract. Further information provided in Kassai 2008

Multicenter study conducted in Italy

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskInsufficient information about the sequence generation process to permit judgement. The study was defined as "randomised" not in the abstract reporting data of this trial, but in a systematic review (Kassai 2008)
Allocation concealment (selection bias)Unclear riskInsufficient information to permit judgement
Blinding of participants and personnel (performance bias)
All outcomes
Unclear riskInsufficient information to permit judgement. The trial is defined as "double-blind" without further information
Blinding of outcome assessment (detection bias)
All outcomes
Unclear riskInsufficient information to permit judgement. The trial is defined as "double-blind" without further information
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskInsufficient information to permit judgement
Selective reporting (reporting bias)Unclear riskInsufficient information to permit judgement

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Chiron 1999aAbstract of a study subsequently published "in extenso" (Chiron 2000)
Chiron 1999bAbstract of a study subsequently published "in extenso" (Chiron 2000)
Coppola 2002Uncontrolled trial
JPRN-JapicCTI-101116 2010Uncontrolled trial
JPRN-JMA-IIA00014 2007Uncontrolled trial
Kassai 2008Systematic review of the literature, not a randomised controlled clinical trial
NCT01533506 2012Uncontrolled trial
NCT01607073 2012Uncontrolled trial
Nguyen Thanh 2002Uncontrolled trial
Oguni 1994Uncontrolled trial
Striano 2007Uncontrolled trial
Thanh 2002Uncontrolled trial

Ancillary