Abstract
The past decade has witnessed a major transformation in our understanding of epilepsies presenting in the neonatal period. Old broad umbrella syndromes, such as Ohtahara syndrome and early myoclonic encephalopathy, are being parsed into new distinct genetic epilepsies based on the underlying etiology. Progress in understanding the genetic basis of these neonatal epilepsies has already led to novel therapeutic approaches resulting from (1) serendipitous clinical observation and (2) rationally devised directed therapy with in vitro validation. These advances have the potential to improve not only seizure control but perhaps also developmental trajectory. Yet implementation of a precision medicine paradigm in the nursery entails early recognition of genetic epilepsy and treatment based on etiology. Meeting this dual challenge will require moving beyond the monolithic concept of “neonatal seizures” and one-size-fits-all treatment protocols for seizures in neonates.
Keywords
diagnosis, genetic epilepsy, KCNQ2, KCNQ3, KCNT1, personalized treatment, precision medicine, recognition, SCN2A, targeted treatment
- •
Long-term video-EEG monitoring and advances in neuroimaging and genetic analysis have led to a major transformation in our understanding of epilepsies presenting in the neonatal period.
- •
Older terms, such as Ohtahara syndrome and early myoclonic encephalopathy, are being parsed into new distinct genetic epilepsies, making possible the first attempts at precision medicine on the basis of etiology.
- •
Benign familial epilepsies and severe sporadic epilepsies in neonates can be caused by different variants in the same genes (e.g., KCNQ2 and SCN2A).
- •
Seizures in neonatal epilepsies due to variants in KCNQ2/3 and SCN2A have been showed to respond to sodium channel blockers (e.g., carbamazepine).
- •
Targeted therapeutic approaches have identified drugs, such as quinidine, that may be able to target the channel dysfunction induced by a genetic mutation (such as variants in KCNT1 in epilepsy of infancy with migrating focal seizures).
- •
A precision medicine paradigm in the nursery entails early recognition of genetic epilepsy, as well as moving beyond the monolithic concept of “neonatal seizures” and one-size-fits-all treatment protocols for seizures in neonates.
A female neonate with unremarkable prenatal course was born at 38 weeks’ gestation by uncomplicated spontaneous vaginal delivery to a 38-year-old G2P2 mother. Apgar scores were 9 and 9 at 1 and 5 minutes, respectively. On the second day of life, the mother noticed three episodes of brief rightward head turning. The child was otherwise doing well and was already “breastfeeding like a champ.” A fourth event was witnessed by the nurse, who suspected it to be a seizure and initiated a transfer to the intensive care nursery. There, additional paroxysmal episodes characterized by asymmetric limb stiffening, accompanied by oxygen desaturation to the 70s, were noted. The child was given a load of 20 mg/kg intravenous phenobarbital and video-electroencephalography (EEG) was ordered. Laboratory studies performed on cerebrospinal fluid (CSF), serum, and urine revealed no abnormal metabolic findings and raised no concern for infection; the result of magnetic resonance imaging (MRI) of the brain was negative. Following a second 20 mg/kg load of phenobarbital, there was a temporary cessation of the events but the infant was notably more sedated and no longer able to breastfeed. While treated with maintenance phenobarbital, the episodes recurred the next day; the phenobarbital level was 42 µg/mL. Video-EEG confirmed the events to be electroclinical seizures, consisting of asymmetric tonic posturing of the upper and lower limbs to the left or to the right, often accompanied by cyanosis and at times followed by bilateral, asynchronous clonic limb jerking, each lasting 1 to 2 minutes. Electrographically, seizures were characterized by low-voltage fast activity followed by focal spikes and waves over the central and temporal regions. Seizure onset alternated between the left and right hemispheres, corresponding to the variable lateralization of the clinical events. The EEG was essentially normal between the frequent seizures. The electroclinical presentation led the consulting neurologist to revisit the family history. Initial family history had been negative for seizures, but asking again with the maternal grandmother present elicited a history of seizures in the mother during the neonatal period. A diagnosis of benign familial neonatal epilepsy was made, phenobarbital was discontinued, and oral carbamazepine was started at 10 mg/kg per day with immediate cessation of seizures. Feeding resumed over the next couple of days; as the sedating effects of phenobarbital wore off, seizures did not recur and the child was discharged home. An epilepsy gene panel returned a pathogenic stop-gain variant in the KCNQ2 gene 1 week later. Subsequent testing demonstrated the same variant in the child’s mother.
This case illustrates the challenges and potential rewards of parsing neonatal epilepsies from among more common acute symptomatic causes of seizures in the nursery. The role of video-EEG is highlighted, not only for determining the presence of seizures, but also for providing clues to the underlying etiology. Carbamazepine is not a medication typically used to treat seizures in neonates. The application of targeted treatment for genetic epilepsy stands in contrast to a protocol-driven one-size-fits-all approach to “neonatal seizures.”
The New Chapter of Neonatal Epilepsies
The past decade has witnessed a major transformation in our understanding of epilepsies presenting in the neonatal period as a result of two advances. First, the implementation of long-term video-EEG monitoring in the nursery provided an opportunity for neurologists to be more involved in the clinical evaluation of newborns with seizures. A result of this experience was the rejection of the notion that seizure phenomenology in neonates is essentially different from that in older children and adults. Second, advances in neuroimaging and metabolic testing, together with readily available sophisticated genetic analysis, has allowed for the identification of an increasing proportion of epilepsies presenting in the nursery from among the large number of newborns with seizures resulting from acute brain injury. The consequence of these developments has been an appreciation of the nuanced differences among patients previously combined under broad umbrella syndromes—Ohtahara syndrome and early myoclonic encephalopathy (EME)—thus allowing for the recognition of new distinct electroclinical phenotypes reflecting discrete etiologic entities (e.g., STXBP1 encephalopathy, KCNQ2 encephalopathy, glycine encephalopathy). Indeed, Ohtahara syndrome and EME were first described in 1976 and 1978, respectively, when neuroimaging was in its infancy and gene sequencing was being developed. Differentiating neonates with epilepsy from neonates with acute symptomatic seizures has prompted controversy about treatment: does one size fit all? 4
Parsing of the old syndromes into the neonatal epilepsies has made possible the first attempts at precision medicine on the basis of etiology. We provide an overview of the types of genetic epilepsies present in newborns and then discuss three genes in which there is some evidence for emerging targeted therapeutics.
Landscape of the Neonatal Epilepsies
Genetic epilepsies may be conceptualized in a number of ways, including by genetic mechanism, inheritance pattern, pathophysiology, associated features, and outcomes. Generally, neonatal epilepsies with known genetic etiologies affect either dominant or recessive genes and the pathogenic variants either are inherited or occur de novo, in which case the mutation may occur either before fertilization or during postzygotic development ( Fig. 8.1 ). This dichotomous scheme is useful, as the resultant groupings tend to share certain clinical features. For example, recessive genes associated with epilepsy often give rise to metabolic disorders in which the complete lack of an enzyme or critical cofactor results in a severe, often progressive, condition (e.g., sulfite oxidase deficiency). Heterozygous de novo mutation in dominant epilepsy genes tends to cause severe epilepsies characterized by intractable seizures and encephalopathy (e.g. STXBP1 ). Malformations of cortical development arise from similar de novo genetic variants (e.g., PAFAH1B1 , associated with lissencephaly) but can also result from postzygotic gene mutation affecting a subset of cortical cells (e.g., PIK3CA in hemimegalencephaly). Inherited variants in dominant genes are responsible for dominant familial epilepsies.
KCNQ2/3 -Associated Neonatal Epilepsies
Benign familial neonatal epilepsy (BFNE) is a rare autosomal dominant condition with incomplete penetrance and good neurodevelopmental outcome that presents with clusters of seizures in the first days of life and remits after weeks or months. A recent multicenter prospective study suggested that BFNE cases may represent about 3% of neonates with seizures. The seizures are characteristic, consisting of asymmetric tonic posturing evolving to unilateral or asynchronous bilateral clonic jerking, often accompanied by apnea and desaturation ( Fig. 8.2 ). Seizures shift laterality variably and are short, typically lasting 1 to 2 minutes, but may occur as frequently as 20 to 30 times a day. A distinctive pattern can be seen on amplitude-integrated EEG ( Fig. 8.2B ). BFNE was termed “benign” to reflect the fact that in most affected individuals, seizures are limited to the first year of life and neurologic development is normal. Nevertheless, seizures may be difficult to control in the neonatal period and some patients present with status epilepticus. Recent studies have shown that up to 25% of patients with BFNE develop epilepsy later in life, the risk of which correlates with the number of seizures in the neonatal period.
In more than 80% of cases BFNE is associated with pathogenic variants in KCNQ2 or KCNQ3 , which encode K v 7.2 and K v 7.3, voltage-gated ion channel subunits mediating a subthreshold potassium current important in limiting neuronal excitability. While inherited alterations in KCNQ2 are responsible for the vast majority of BFNE cases, de novo variants in KCNQ2 may result in profound neonatal encephalopathy with severe, frequent, intractable seizures, termed KCNQ2 encephalopathy. Seizures in KCNQ2 encephalopathy are of the same type as those of BFNE ; however, in contrast to BFNE, the interictal EEG is severely abnormal and the neurologic examination is often notable for hypotonia, paucity of spontaneous movements, no visual fixation, and altered reactivity. Whereas mutations that lead to BFNE result in mild reductions in potassium current, de novo mutations responsible for KCNQ2 encephalopathy represent more profound loss-of-function alterations, acting through a dominant negative effect. Gain-of-function variants (R201H and R201C) have also been reported to cause KCNQ2 encephalopathy, but careful analysis of the electroclinical characteristics revealed a separate gain-of-function phenotype that lacks seizures in the neonatal period and is characterized instead by irregular breathing patterns, exaggerated startle responses, and nonepileptic myoclonus. The association of a given gene, such as KCNQ2 , with both benign and severe phenotypes is a recurring theme for channelopathies, epilepsies associated with genes encoding ion channels. Fig. 8.3 depicts the spectrum of KCNQ2 phenotypes by underlying genotype.
The possibility for a targeted therapeutic approach for KCNQ2 encephalopathy has been suggested for retigabine/ezogabine, a drug that specifically augments current through the K v 7 channels. According to parental report, there was some improvement in mental status and development in six patients, four of whom experienced a seizure reduction, although seizure freedom was not achieved. Serious adverse effects, including skin and retinal pigmentation leading to visual loss, were reported with long-term treatment, and the production of retigabine/ezogabine was discontinued as of June 2017. Interestingly, clinical observation in patients with KCNQ2 encephalopathy has suggested that sodium channel blockers are particularly effective. Seizure freedom was reported with carbamazepine (CBZ) in six patients and with phenytoin in five patients. Therefore sodium channel blockers appear to represent a very effective therapy for KCNQ -associated seizures, and this is now considered an established precision medicine treatment. Interestingly, although retigabine/ezogabine had a more pertinent mechanism of action, it seemed to be less effective in KCNQ2 encephalopathy than sodium channel blockers.
Recently a multicenter study of 19 cases described treatment responses in BFNE and demonstrated rapid seizure cessation in all 17 neonates treated with low-dose (10 mg/kg per day) oral CBZ, regardless of when in the course it was initiated. In contrast, response to other agents was poor (e.g., no response to phenobarbital in 12 of 14 cases). Early treatment with CBZ reduced the length of hospitalization ( Fig. 8.4 ). These reports on the use of CBZ for KCNQ2 -associated epilepsies constitute the best evidence for the response of a genetic epilepsy to a specific therapeutic agent ( Table 8.1 ). It is unclear why sodium channel blockade would be particularly effective in countering potassium channel loss of function, but sodium channels are bound together with K v 7 potassium channels at the axonal initial segment and nodes of Ranvier, locations critical for action potential generation and maintenance, suggesting localized functional interaction ( Fig. 8.5B ).