Federico Vigevano and Romina Moavero
Early-onset seizures have always been regarded with great caution and suspicion due to the possibility of long-term neurological and neurodevelopment sequelae. This was particularly true for focal forms of epilepsies, because a symptomatic etiology was always presumed. However, since 1963, this conception has progressively changed, thanks to the first description of infantile seizures with a benign evolution by Fukuyama (1963). Nowadays, benign infantile seizures are a well-described entity, also recognized by the International League against Epilepsy (ILAE) classification (Engel and International League against Epilepsy 2001). However, the appropriateness of the term ‘benign’ has been the object of long debates in the past few decades. A common thought, in fact, is that this term could only be appropriately used retrospectively, after a long-term follow-up witnessing seizure freedom and normal neurodevelopment. The word ‘benign’ literally means “to do good” or “gifted of natural goodness”, and this term is widely used in the medical literature, clearly opposed to “malignant” or “severe”, to define clinical entities in which the prognosis has a favorable outcome (Capovilla et al. 2009). The first ILAE explanation of the meaning of ‘benign’ comes in the 2001 classification, defining a “benign epilepsy syndrome” as “a syndrome characterized by epileptic seizures that are easily treated or require no treatment, and remit without sequelae” (Engel and International League against Epilepsy 2001). However, the meaning of the terms “no sequelae” and “seizure remission” was not clearly established. In the past few decades, most of the attention in infantile seizures was paid to the epilepsy outcome, with less data regarding the neurodevelopment outcome. However, recent studies using neuropsychological evaluations have demonstrated that in some ‘benign’ syndromes, a spectrum of neuropsychological sequelae can be present (Capovilla et al. 2009). Although it is not always possible nor simple, establishing since its onset whether an epilepsy could be considered as ‘benign’ is of crucial importance to make a definite plan for adequate management. The main goal in using the term ‘benign’ is to try to give early (at seizure onset or after a short follow-up) reliable information on the prognosis of the disease, thus reassuring and decreasing anxiety for patients and caregivers (Capovilla et al. 2009). Furthermore, by knowing a priori the possibility of a benign evolution, overtreatment and unnecessary procedures might be avoided. An Expert Meeting held in Italy in 2008 proposed that an epilepsy can be considered benign when characterized by clinical and/or EEG features that predict, with low risk of error, at onset or soon after, remission of seizures (with or without treatment) and without significant, permanent impact on the patient’s potential (Capovilla et al. 2009).
In the group of epilepsies with onset during infancy, the term ‘benign’ is only used for idiopathic/genetic epilepsy syndromes, both with focal seizures, such as the so-called “benign focal infantile seizures,” and with generalized seizures, such as the so-called benign myoclonic epilepsy. However, this latter form is actually known as “myoclonic epilepsy in infancy” because the term ‘benign’ was no more considered appropriate.
Benign focal infantile seizures
Benign familial and non-familial infantile seizures
This epileptic syndrome is characterized by the onset of seizures in otherwise healthy infants between 3 and 20 months of age. The benign familial infantile seizures (BFIS) present an autosomal dominant transmission with a slightly earlier onset than sporadic forms, usually between the fourth and the eighth month of life with a peak around the sixth month (Vigevano et al. 1992). In BFIS, first- or second-degree relatives presented convulsions in infancy without subsequent development of other seizure types. The first series of five infants affected by this epileptic syndrome was described more than 20 years ago by Vigevano et al. (1992). The authors stressed that none of the children or their relatives presented seizures in the neonatal age or after the eighth month of life. However, a similar form with seizure onset occurring between neonatal and infantile ages was reported by Kaplan and Lacey (1983). In this form, age at seizure onset ranged from 2 days to 3.5 months, thus partially overlapping with BFIS. The authors proposed to use the term “benign familial neonatal–infantile seizure” (BFNIS), and nearly 20 years later, two other families with onset between 1.9 and 3.8 months of life and a clear autosomal dominant inheritance have been described (Heron et al. 2002). Similar cases without a familial history of seizures, called benign non-familial infantile seizures (BNFIS) have been described.
In both the familial and non-familial cases, seizures are focal and often present in clusters of 8–10 seizures per day, with the cluster lasting 1 to 3 days. When examined in the interictal phase, infants appear to be completely normal, without any neurological deficits (Vigevano et al. 2012). Seizure semiology does not appear to differ between sporadic and familial forms. It is usually characterized by motor arrest, impaired consciousness, staring, head and eye deviation, and convulsions (Vigevano et al. 2012). The side of head and eye deviation might change from seizure to seizure in the same infant (Vigevano et al. 1992). Watanabe underlined the presence of limb and/or oral automatisms in the cases described as “benign partial epilepsy with complex partial seizures” and of prompt generalization with tonic–clonic manifestations in the so-called benign partial epilepsy with secondarily generalized seizures (Watanabe et al. 1990; 1993). Untreated children might present the recurrence of clusters of seizures, while a prompt antiepileptic treatment (i.e. with carbamazepine, valproate, phenobarbital, or zonisamide) is usually able to make seizures cease, and in almost all cases, no more seizures are observed throughout the life course. The interictal EEG is usually normal; lateralized slow waves and spikes in the occipito-parietal areas in the interictal EEG performed during a cluster of seizures have been reported (Vigevano et al. 2012).
Infantile convulsions with choreoathetosis
In 1997, four French families with autosomal inherited benign infantile convulsions and paroxysmal choreoathetosis were described, thus leading to the identification of a new syndrome called infantile convulsions and choreoathetosis (Szepetowski et al. 1997). Several subsequent reports confirmed the existence of this association, with paroxysmal choreoathetosis usually appearing at a later stage than seizures (Lee et al. 1998; Sadamatsu et al. 1999; Tomita et al. 1999; Bennett et al. 2000; Hattori et al. 2000; Swoboda et al. 2000; Thiriaux et al. 2002; Kato et al. 2006; Striano et al. 2006b; Rochette et al. 2010). Seizure characteristics widely overlap with the familial form of benign infantile seizures; therefore, this syndrome is considered as its variant, being characterized by the association of other neurological symptoms.
In familial cases of infantile seizures, autosomal dominant transmission appears to be evident; therefore, different studies tried to find a genetic substrate for this hereditary epilepsy syndrome. A linkage has been demonstrated on chromosome 19q (Guipponi et al. 1997) but also on chromosome 16p (Szepetowski et al. 1997; Caraballo et al. 2001) and 2q (Malacarne et al. 2001). In 1998, genetic mutations KCNQ2 and KCNQ3 were described in benign seizures with onset in the neonatal period, thus leading to the consideration of benign neonatal seizures as a channelopathy (Charlier et al. 1998; Singh et al. 1998), and later this concept also extended to BFIS and BFNIS. In fact, in 2002, the two families described by Heron et al. (2002) with BFNIS were found to present a missense mutation of SCN2A gene, which was later confirmed in other families with the same phenotype (Berkovic et al. 2004). Missense mutations of SCN2A have later been described even in classic cases of BFIS (Striano et al. 2006a), underlying that BNFIS and BFIS might share both clinical and genetic features.
More recently, PRRT2 gene mutations have been described in BFIS (Schubert et al. 2012; Specchio et al. 2013). Mutations in this gene have been recognized to be involved in paroxysmal choreoathetosis, with or without benign infantile seizures (Heron et al. 2012; Specchio et al. 2013).
LATER COGNITION AND BEHAVIOR