Severe myoclonic epilepsy in infancy is one of the most severe epilepsies affecting the infants and deserves to be recognized as soon as possible in order to manage it appropriately. However, the diagnosis is not easy and every infant who presents with long and repeated febrile convulsions before the end of the first year of life must be carefully followed because he is at risk for this epilepsy. Up to now, no antiepileptic drug has allowed controlling the seizures completely and almost all the childern have an abnormal psychomotor development and cognitive impairment at the adult age. In the last years, the genetic origin of this epilepsy has been demonstrated, most of the patients being carriers of a de novo mutation in the SCN1A gene, but not all. Currently, many studies aim to discover either the other mutations that could be responsible or the function of the known mutations.

Severe myoclonic epilepsy in infants (SMEI) was first defined in 1982¹⁴ and recognized as a syndrome in the most recent international classification of the epilepsies.⁹ Since 1982, about 445 cases have been published in the literature. Most cases are reported from Southern Europe and Japan, but descriptions also come from other regions.^3,26,30 The name of the syndrome was used in the first description to underscore the ictal and interictal, myoclonic component, which allowed differentiating this syndrome from the Lennox-Gastaut syndrome. In the latter, the ictal hallmarks are occurring by generalized tonic seizures, mainly during sleep, and atypical absences. Later, it appeared that myoclonus could be absent in some patients. Seizures were polymorphic, and it was more difficult to find an appropriate semiologic designation.¹ For this reason the Commission on International Classification has proposed using the eponym “Dravet syndrome.”⁹ Since 2000, because of the discovery of a mutation in the sodium channel gene SCN1A in some patients,⁸ this syndrome has been the target of a number of genetic studies that have given new insights to its pathophysiology.

Severe myoclonic epilepsy in infants is characterized 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. Later there is with myoclonus, atypical absences, and partial seizures. All seizure types are resistant to antiepileptic drugs. Developmental delay becomes apparent within the second year of life and is followed by definite cognitive impairment and personality changes. The disorder has been categorized among “epilepsies and syndromes undetermined as to whether they are focal or generalized” because the syndrome shows both generalized and localized seizure types and electroencephalographic (EEG) paroxysms.⁹ In the most recent proposal it is placed among the “epileptogenic encephalopathies.”¹⁵

Many children have been reported with a similar picture but without myclonus, and this condition has been designated as “borderline SMEI” (SMEIB).³³ These patients can have different EEG features but have the same course and outcome as patients with myclonus, and they can be included in the same syndrome. Recent genetic studies have demonstrated of SCN1A mutations in SMEIB patients but with differences in the type and the site of these mutations.

SMEI is a rare condition. In the patient population of the Centre Saint-Paul, SMEI was diagnosed in 63 of 6,300 patients (1%) between 1970 and 1990. Hurst²⁶ estimated an incidence of at least 1 case in 40,000 children <7 years of age in the general population, and Yakoub et al.⁵⁰ gave a slightly higher figure (1/20,000 or 1/30,000). In children with seizure onset in the first year of life, the proportion varies from 3%² to 5%.⁵⁰ These findings show that this syndrome deserves to be recognized in infants with early convulsions. Mild male predominance has been reported.¹³

SMEI shares some characteristics with the idiopathic epilepsies (genetic predisposition, photosensitivity, generalized seizures, and generalized spikes-and-waves on EEG), but they are also associated with features of localization-related epilepsies (focal seizures, focal EEG anomalies, cognitive impairment). Moreover, the course of the disease—progressive during the first stage and always severe—could suggest a metabolic disease. However, none has been identified.

Until recently, magnetic resonance imaging (MRI) was said to be normal, even when repeated some years after the onset. However, in 2005, new data were published by Siegler et al.,⁴⁰ who performed the first systematic neuroimaging study with repeated MRI at different ages. They found hippocampal sclerosis (HS) in 10 of 14 patients with SMEI. In 6 of them, HS was not present on the first MRI, performed during the first 2 years in 5 patients and at 11 years in 1 patient, but developed later, between 14 months and 16 years. In the other 4 children, the first examination was not performed before the age of 4 years. In the last 4 children, MRI, performed between 6 and 17 years, remained normal. Because all the patients had a history of
numerous hemiconvulsions and generalized tonic–clonic seizures (GTCS) during the first year, these results support the hypothesis that HS is a consequence of prolonged febrile seizures, even if it is unclear why 4 of the patients did not show HS.

Neuropathologic findings reported by Renier and Renkawek³⁷ from an autopsied case included cerebellar microdysgenesis, irregular lamination of cerebral cortex, and threefold spinal cord canals. There are no other neuropathologic studies.

A mitochondrial cytopathy was suspected in three patients who presented with an extremely severe picture.^5,6,16 However, according to Dravet et al.¹³ and Giovanardi-Rossi et al.²⁰ no structural or histochemical abnormality was present in the biopsy samples from muscles in respectively five and nine patients.

A family history of epilepsy or febrile seizures is often present, ranging from 25% to 71%.¹³ Five families have been reported with two affected siblings in each.^13,33,47

Monozygotic twins have been reported by Fujiwara et al.,¹⁷ Musumeci et al.,³¹ and Ohki et al.³⁵ Recently, several publications described SMEI cases in the syndrome of generalized epilepsy with febrile seizures plus (GEFS+).^41,42,47 Thus, a genetic etiology could be suspected and was confirmed by the recent discovery of a de novo mutation in the sodium channel gene SCN1A.⁸ Several publications have reported analogous findings in a high proportion of patients with SMEI but not in all. The percentage varies from 35% among 93 patients,³² to 62% among 76 patients¹⁹ (in part already published in the Nabbout et al. study³²), to 71.4% among 12 patients,⁴⁴ to 82.7% among 29 patients.³⁶ Only two studies reported a GABRG2 mutation,^25,28 which was not found by Madia et al.,²⁹ in 53 patients without the SCN1A mutation. Different mutation types were reported (truncating, missense, nonsense) in different sites of the gene. Correlations of phenotypes to genotypes were variable in different studies. Mutations were observed also in SMEIB,¹⁸ but truncating mutations were largely predominant in the typical form.