How Dravet syndrome became a model for studying childhood genetic epilepsies.

In this issue of Brain, Brunklaus et al. report a study of 241 patients with Dravet syndrome carrying a SCN1A mutation, with the aim of identifying predictors of developmental outcome and determining specific clinical and demographic characteristics. The electroclinical features of a large cohort were collected and analysed prior to genetic testing. A total of 355 patients were diagnosed with Dravet syndrome, but the authors selected only the 241 (68%) who were positive for a mutation of SCN1A. Patients who were excluded had not been tested for mutations in other genes that could potentially cause overlapping phenotypes, such as protocadherin 19 (Depienne et al., 2009). Dravet syndrome was first described briefly in French as severe myoclonic epilepsy in infancy (Dravet, 1978). In 1981, the same series was presented at the XIIIth Epilepsy International Congress and published 1 year later. Over the last 20 years, numerous series of patients have been reported, confirming the initial description, but also underlining variability of the clinical picture. The classic syndrome is defined by onset in an otherwise healthy infant during the first year of life, presenting febrile or afebrile and generalized or unilateral clonic or tonic clonic seizures, often long lasting, later associated with myoclonic, atypical absence and focal seizures. These are all resistant to drug treatment, and accompanied by slowing of developmental skills, as well as motor and behavioural disturbances. The EEG shows generalized and multifocal abnormalities whereas neuroimaging is normal. It quickly became apparent that a number of patients do not present with the complete semiology, mainly because the myoclonic component may be missing, and typical and atypical or ‘borderline’ forms are recognized, also showing an unfavourable prognosis. With time, the syndrome appeared no longer necessarily to be characterized by myoclonic epilepsy or confined to infancy. The aetiology remained unknown but the initial suggestion of a genetic cause (Benlounis et al., 2001) was soon confirmed by Claes et al. (2001) in seven probands. Noting the precipitation of seizures by fever, these authors screened the SCN1A gene because they knew it can cause the syndrome of generalized epilepsy with febrile seizures plus; and de novo mutations were found in all their probands. This finding paved the way for a growing body of research that showed these mutations to be present in 70–80% of patients with Dravet syndrome with 2–3% of mutation-negative patients harbouring exonic deletions or small chromosomal rearrangements involving SCN1A (Marini et al., 2009). Two animal models were generated using different techniques: deletion of a major exon (Yu et al., 2006) or knock-in of a stop codon (Ogiwara et al., 2007) in the SCN1A gene. These mice each showed the main characteristics of Dravet syndrome in humans and allowed more detailed studies of its complex pathogenic mechanisms. It is now clear that both missense and truncating mutations are found in Dravet syndrome. Missense mutations result either in loss or gain but most often loss of function (Catterall et al., 2008), a finding that was initially difficult to reconcile with cortical hyperexcitability. However, the subsequent demonstration that the -1 subunit of the sodium channel (SCN1A) is expressed in -aminobutyric acid (GABA)ergic interneurons and its loss of function decreases their inhibitory properties on pyramidal neurons (Yu et al., 2006; Ogiwara et al., 2007) shed light on epileptogenesis in the syndrome. Impaired function of cerebellar Purkinje cells, which are GABAergic and inhibitory, could account for the ataxia and cognitive impairment often observed in Dravet syndrome (Kalume et al., 2007). However, attempts to establish genotype/phenotype correlations in Dravet syndrome have proved disappointing. The different mutation types are not associated with well-defined phenotypes and these are found both in typical and atypical forms (Oguni, 2005). The most striking example is that of inherited mutations that result in various degrees of severity in the affected siblings. That evidence supports the concept of a unique syndrome, with more or less severe expression; the term ‘mild’ or ‘incomplete’ being perhaps more appropriate than ‘borderline’ for the less severe forms (Dravet and Guerrini, 2011). In keeping with this concept, Brunklaus et al. grouped their mutation-positive patients without differentiating severe and mild forms. Their results largely confirm the literature on several aspects of the syndrome. Their calculated incidence of 1/40 900 children affected in the UK is similar to that observed in previous Brain 2012: 135; 2309–2313 | 2309