Clinical manifestations

Presentation and course

Self-limited infantile epilepsy typically presents at the age of 3 to 20 months with a peak at 6 months. Epilepsies that present earlier (in the first week and first few months) would likely be classified as the distinct syndrome of self-limited neonatal and neonatal-infantile epilepsies.

Preceding gestational, birth, and neonatal history is typically normal with normal head size and neurologic examination. Preceding and subsequent to the epilepsy, psychomotor development is normal.

Seizures present as focal seizures with behavioral arrest, automatisms, head and eye version, and clonic movements. Focal clonic seizures may alternate from one side to the other and progress to a bilateral tonic-clonic seizure.

Seizures are usually brief, lasting less than 3 minutes; longer seizures can occur but are rare. Seizures can be frequent at onset. Some patients may present with a single seizure 2 weeks before the frequent seizures start.

The seizures are classically quite responsive to pharmacologic treatment. However, in untreated cases, there may be isolated or brief clusters of seizures that will persist if treatment is withheld, often through the second year of life. Irrespective of pharmacologic treatment, all seizures will typically remit within 1 year of onset (by the age of 2). A minority of individuals may have epilepsy persisting into later life.

In familial cases first and second-degree relatives endorse history of infantile seizures that resolved, often with an autosomal dominant inheritance pattern. Testing for etiologic factors, including metabolic and neuroimaging studies, are unrevealing. Patients with PRRT2 pathogenic variants may develop paroxysmal kinesigenic dyskinesia/dystonia beginning from childhood to adult life. This diagnosis may often be missed as the symptoms of a movement disorder may be brief and last only seconds. These symptoms are easily controlled with carbamazepine.

In the rarer subtypes, there is often an additional clinical feature or symptom that emerges after the resolution of the epilepsy and each syndrome is so named. For example, patients with self-limited familial infantile seizures with choreoathetosis will develop choreoathetotic movements in infancy or later in childhood. They occur at rest or during exertion/anxiety and are dystonic (48; 18). Patients with self-limited familial infantile seizures with familial hemiplegic migraine go on to develop hemiplegic migraine attacks (49; 52; 35).

Seizures are associated with gastroenteritis in self-limited infantile seizures with mild gastroenteritis (11) and there is an electrographic signature of midline or vertex spikes during sleep in self-limited infantile focal seizures with midline spikes and wave during sleep (19).

Prognosis and complications

Children with self-limited infantile seizures generally have normal psychomotor development and do not manifest other forms of epilepsy later in life. Seizures remit within 1 year of onset. The EEGs performed at follow-up remain normal.

Of note, up to 10% to 30% can have seizure recurrence later in life, with one study noting this in more children with more seizures and longer duration of epilepsy (36; 27).

Clinical vignette

The following is an example of a familial case reported by F Vigevano as well as by Giordano and colleagues (22).

A 3-months 21-days-old girl presented with a cluster of seizures characterized by fixed gaze, bilateral clonic jerks, and loss of consciousness that lasted 2 minutes.

On admission, a seizure was observed that was characterized by slow deviation of the head and eyes to the left and then to the right, loss of consciousness, oral automatisms, diffuse hypertonia, and bilateral clonic jerks. All laboratory testing (organic acids, lactate, pyruvate, ammonia in plasma, and urine) and neuroradiological examinations (CT head and MRI brain) were negative. EEG during wakefulness and sleep were normal.

At that time, neurologic examination and psychomotor development were normal, and no antiepileptic drug therapy was prescribed at the parent’s request. The patient continued to have three to four seizures per month from 5 to 7 months of age. Only a single seizure was reported from 7 to 11 months of age. At 1 year of age, the patient became seizure free. Neurologic examination and developmental trajectory remained normal.

There was a strong family history of convulsions during infancy, including 18 of 35 members of the mother’s family. In particular, the father and all but one of the mother’s siblings (four sisters and six brothers) presented with similar seizures. In all 18 members of the affected family, the clinical characteristics of the seizures were similar to those described in the proband. In all 18 members, seizure onset was at 3 to 4 months of age and remission was achieved at 9 to 10 months of age. All affected members experienced seizure clusters at onset and subsequently isolated seizures, with frequency similar to that of the proband. The proband’s grandmother considered this type of convulsion to be a characteristic trait of the family and as such none of the 18 subjects received antiepileptic drug treatment.

All 18 affected family members had normal psychomotor development and attended high school. None of them experienced febrile convulsions, nor did they develop epilepsy.

Biological basis

Etiology and pathogenesis

The etiologies of the self-limited familial infantile epilepsy syndromes are genetic. The self-limited nonfamilial epilepsy syndromes are less well understood, and their etiologies are likely multifactorial. Autosomal dominant transmission is seen in self-limited familial infantile seizures with high penetrance.

In the past several years it has been discovered that a large percentage of patients affected by self-limited familial infantile seizures have a mutation in the proline-rich transmembrane protein 2 (PRRT2) gene located on chromosome 16p12-q12 (10; 60; 45; 42). In 2012, the largest study performed so far in families with “pure” self-limited familial infantile seizures showed the occurrence of a mutation of PRRT2 in 83% of them (41). Additionally, a large Italian collaborative study confirmed PRRT2 mutations in most of the families, but with a slightly lower incidence than in the previous study (70%) (62). This mutation is not seen in families with self-limited familial neonatal seizures (23).

Previously the PRRT2 gene was implicated in paroxysmal kinesigenic dyskinesia (13) and in fact paroxysmal kinesigenic dyskinesia is observed later in life in some patients with self-limited familial infantile seizures with PRRT2 mutations (26; 37). This entity will be discussed below.

The product of the PRRT2 gene is a proline-rich protein whose precise function has not yet been elucidated. Its role seems related to synaptic function because it interacts, both in vitro and in vivo, with synaptosomal-associated protein 25 kDa (SNAP 25), a synaptic protein (43; 29; 51). SNAP 25 plays a relevant role in neurotransmitter exocytosis, vesicle docking, and fusion to the plasma membrane. Indeed, it has been shown in vitro that PRRT2 plays an important role in specific steps of neurotransmitter release (51). Further work must be done to elucidate (1) whether the mutations found so far are related to a loss of function and to what extent, and (2) why the epileptic syndromes associated with PRRT2 mutations have a limited time window of phenotypic expression, resolving without sequelae. It is worth noting here that the expression levels of CNS PRRT2 seem to vary during development, and apparently with a region-specific pattern (13; 26); this might be at least in part in line with the age-specific/self-limiting presentation of seizures, as well as with the phenotypic variability among patients. The mechanism by which this mutation is related to the occurrence of this brief, self-limited epilepsy syndrome is not well understood.

Interestingly, the PRRT2 sequence is highly preserved throughout phylogenesis (29), and this might help in understanding its precise role and function through experimental studies in rodents.

Other pathogenic mutations have been suggested. Zara and colleagues suspected involvement of KCNQ2, KCNQ3, and SCN2A and found that only approximately 10% of self-limited familial infantile seizures families had none of the three mutations (62). Subsequent linkage studies confirming these results have not yet been undertaken.

Self-limited familial infantile seizures and choreoathetosis. As discussed above, PRRT2 mutations are frequently found in patients with isolated self-limited familial infantile seizures; however, they also occur in patients with self-limited familial infantile seizures who later develop choreoathetosis or paroxysmal kinesigenic dyskinesias.

In 1997 Szepetowski and colleagues demonstrated linkage to this pericentromeric region of chromosome 16 in four French families with the syndrome of self-limited familial infantile seizures and choreoathetosis (48). This finding was later replicated by several other groups studying families of varying ethnicities (32; 47; 10). The mutation is present in up to 80% of families with self-limited familial infantile seizures with choreoathetosis (29; 26; 37) and seems to be inherited in an autosomal dominant manner.

Additionally, a mutation of the SCN8A gene has been described in families of patients affected by both paroxysmal dyskinesia and self-limited familial infantile seizures. Notably, these families did not have a coexisting PRRT2 mutation (21).

Self-limited familial neonatal-infantile seizures. In 2002 Heron and colleagues reported the presence of a mutation in the sodium channel subunit gene SCN2A in two families with seizure onset between 1 and 3 months of life, which was intermediate between neonatal and infantile forms. One of the families also had mutations in KCNQ2 and KCNQ3 (25). Later, the same group found mutations in the sodium channel gene SCN2A in six other families (02). The authors coined the term “benign familial neonatal-infantile seizures,” referring to their patients whose epilepsy onset was between day 2 of life and 7 months of life. In 2006, Striano and colleagues reported a novel heterozygous mutation c3003 T → A in the SCN2A gene in a family with three affected individuals over three generations (44). All subjects in this series experienced clusters of focal seizures with or without secondary generalization, onset between 4 and 12 months of life, and had been diagnosed with self-limited familial infantile seizures. This report provided new evidence that suggested self-limited familial neonatal-infantile seizures and self-limited familial infantile seizures may have some overlapping clinical and genetic characteristics. However, it could not be excluded that the family described by Striano and colleagues (44) was affected by self-limited familial neonatal-infantile seizures, as suggested by Herlenius and colleagues (24).

Self-limited familial infantile seizures with familial hemiplegic migraine. There is a subset of patients with self-limited familial infantile seizures who go on to experience hemiplegic migraine later in life, long after the resolution of the epilepsy. A mutation in ATP1A2 has been identified in some of these patients (49; 52). The loss of function mutation is hypothesized to lead to neuronal hyperexcitability and facilitate cortical spreading depression, which is a mechanism believed to be involved in both epilepsy and migraine.

Self-limited nonfamilial infantile seizures. The data on gene(s) involved in the sporadic cases are scarce. It has been observed that PRRT2 mutations do occur in some nonfamilial cases, as a de novo mutation (40; 42).

Self-limited infantile seizures with mild gastroenteritis. There are reports of sporadic cases of self-limited infantile convulsions associated with prolonged episodes of diarrhea caused by rotavirus infections (14). However, clear cause-effect relation is still to be demonstrated as other groups have found this entity to occur in infants with mild gastroenteritis that is rotavirus negative (11). Several large series of these patients have been described (15; 12). There has been some documentation of the lack of PRRT2 mutations in the few patients with convulsions with mild gastroenteritis assessed thus far (40; 28).

Self-limited focal epilepsy in infancy with midline spikes and waves during sleep. Capovilla and colleagues described a series of 12 patients who shared the clinical presentation of self-limited nonfamilial infantile seizures; however, they had interictal EEG abnormalities unlike the previously described cases. In all cases, characteristic spike and wave discharges were present over the vertex during sleep whereas the interictal EEG during wakefulness remained normal. On the basis of these uniform electroclinical features, the group proposed the creation of a new syndrome (06). The patients appear to have a self-limited outcome with normal psychomotor development and quick resolution of the epilepsy, and as such it has been suggested that they do not simply represent an early onset of other previously described epilepsy syndromes (07; 19).

Epidemiology

Self-limited (familial) infantile epilepsy accounts for 7% to 9% of all epilepsies beginning prior to 2 years of age. The incidence is estimated at 14.2 per 100,000 live births.

Due to the self-limited and sometimes brief nature of these epilepsy syndromes, their incidence and prevalence are difficult to estimate. Many families who possess the condition consider it a “family trait” and may not seek care for affected infants.

There are several historical worldwide reports of the familial and nonfamilial forms (31; 34; 16; 08; 04; 18; 39; 41; 62); however, review articles with reliable epidemiological data are more recent (38).

Differential diagnosis

In familial forms where the history is available, the diagnosis of self-limited familial infantile seizures can be made early. However, in many cases other etiologies such as infection, intracranial injury, metabolic disorder (including pyridoxine dependency/deficiency epilepsy), and electrolyte disturbance should be ruled out. Other epilepsy syndromes of infancy that are not self-limited must also be considered at onset as neurodevelopmental outcome will be uncertain.

Confusing conditions

Diagnostic workup

In the setting of an appropriate family history, diagnostic testing may be minimal if self-limited familial infantile seizures are suspected.

In the absence of such a history, diagnostic workup will often be directed at ruling out other causes of the seizures. This evaluation will begin with collecting historical information on prenatal, perinatal, or postnatal etiologic factors, data on psychomotor development, clinical condition prior to seizure onset, and family history when available. Normal neurologic examination must be confirmed.

Diagnostic testing may include blood and urine tests to exclude infectious or metabolic disorders (complete blood count, comprehensive metabolic panel, organic acids, lactate, pyruvate, ammonia, cultures of blood/urine/CSF as indicated). Video EEG is advisable to characterize the suspected episodes as seizures and document the interictal record, which will be normal in most cases. Although focal slowing may occur postictally, a structural brain abnormality should be considered if there is persistent focal slowing in one area (53). Diffuse, persistent slowing in the background would suggest a different syndrome. A variant with midline spikes during slow sleep has been described (03; 19).

The ictal recording is characterized by focal discharges, which is often onset in temporal or posterior head regions and may spread to both hemispheres (53). The seizure onset may vary from lobe to lobe or from hemisphere to hemisphere in different seizures in the same patient. However, the ictal pattern within the same seizure does not show a migrating pattern. This is consistent with the alternating semiology described by Vigevano and associates (55; 56). Subsequently, several groups have failed to demonstrate significant differences in region of onset or EEG characteristics between patients with familial and nonfamilial benign self-limited infantile seizures (33; 09).

Neuroimaging is normal.

A genetic etiology can be identified in up to 80% of cases, and PRRT2 is the gene most commonly implicated (05; 26). Other genes associated with this syndrome include SCN8A, SCN2A, KCNQ2, KCNQ3.

Exclusionary data would include the following (63):

• Age of onset younger than 1 month or older than 36 months
• Significant neurodevelopmental delay or neurocognitive regression
• Other seizure types aside from focal/focal to bilateral tonic clonic
• EEG findings of persistent focal or background slowing or hypsarrhythmia
• MRI lesion

Management

In general, these patients’ seizures are successfully treated with medications such as carbamazepine, phenobarbital, valproate, or zonisamide, which may be efficacious even at low doses (61). Sodium channel blocking agents are often considered first line when self-limited familial epilepsy is suspected (30; 36). Levetiracetam seems to be less efficacious (50).

Treatment in general is continued until the expected time of remission of the epilepsy, which is approximately 2 years.

Treatment is often withheld in benign familial infantile seizures cases, as many families are familiar with the self-limited outcome of the syndrome and as such may question the risk to benefit ratio of the antiseizure medications that are recommended.

Outcomes

The epilepsy remits usually by the second year of life if not before, and neurodevelopmental outcome is normal.

Of note, up to 10% to 30% can have seizure recurrence later in life, with one study noting this in more children with more seizures and longer duration of epilepsy (36; 27).