Clinical aspects

Description

Although the majority of seizures in the neonatal period occur in the context of an acute illness, in some cases seizures may be the first manifestation of early infantile epilepsy (78). Early differentiation of provoked seizures from neonatal-onset epilepsies has important diagnostic, therapeutic, and prognostic implications because the evaluation and long-term management of neonatal epilepsies are distinct from those of provoked seizures (94). Neonatal-onset epilepsies follow a different time course and evolution than acquired symptomatic neonatal seizures. Acute symptomatic neonatal seizures are commonly due to hypoxic ischemic encephalopathy, intraventricular hemorrhage, infections, or metabolic derangements. Hence, they tend to abate after the first few days of life but can be followed by the onset of unprovoked seizures later in life (epilepsy). In a few patients, hypoxic ischemic encephalopathy or stroke-related seizures may continue as neonatal epilepsy. On the contrary, neonatal seizures caused by structural (genetically driven), metabolic, or genetic causes are usually self-sustained from the onset and are considered early-onset epilepsies (75). With the increased use of sequencing methods, these etiologies now represent 10% to 15% of all neonatal seizures (94).

Clinical recognition has been the cornerstone of the diagnosis and management of neonatal seizures. The Task Force on Neonatal Seizures of the International League Against Epilepsy (ILAE) has published a new classification system that has significant implications for the care of newborns suspected of seizures (59). The clinical approach to neonatal epilepsy is different than that for older children or adolescents (78). The neonatal classification framework by the International League Against Epilepsy emphasizes the role of EEG in the diagnosis of seizures in the neonate and includes a classification of seizure types relevant to this age group (78). Researchers have tried to analyze clinical characteristics of newborns with genetic epilepsies and have compared them to acute provoked seizures. Newborns with genetic epilepsies tend to have a positive family history, presence of neonatal epileptic status, normal neuroimaging, a longer time to first seizure, and tonic or myoclonic seizures (23; 69).

The next important step is the identification of the etiology of neonatal seizures to administer etiology-specific therapy (111; 81). This evaluation simultaneously accompanies the active control of seizures using the recommended antiepileptic drugs for neonatal seizures. The underlying etiology and the degree and distribution of brain injury are the most important determinants for long-term outcome (93). Although rare and infrequent, there is an increasing recognition of neonatal-onset epilepsy syndromes.

The term “epileptic encephalopathy” is used for conditions in which the abundant epileptiform activity and high number of epileptic seizures contribute to cognitive regression (15). The newer term “developmental and epileptic encephalopathy” was proposed in 2017 for conditions in which cognitive development and behavior are impaired independent of the onset of epilepsy, with a high frequency of seizures and abundant epileptiform abnormalities (89; 88). Most children with developmental and epileptic encephalopathy have an underlying genetic etiology that is responsible for both impaired cognition and severe epilepsy; control of the epilepsy may not completely improve cognitive outcomes. Both these terms should not be confused with the term “developmental encephalopathy,” which signifies a developmental delay or intellectual impairment due to a nonprogressive brain state with or without a coexisting epilepsy (90). However, the epilepsy in developmental encephalopathy is not so severe as to cause epileptic encephalopathy. Similarly, for some neonatal epilepsies, such as those associated with KCNQ2 or STXBP1 mutations, the genetic defect may directly lead to both severe epilepsy and profound intellectual disability as two independent dimensions of the phenotype (28). They are often termed “neonatal” or “early-onset developmental and epileptic encephalopathy.”

The previous classification of neonatal-onset electroclinical syndromes identified the following four syndromes: (1) benign familial neonatal seizures, (2) benign neonatal seizures (nonfamilial), (3) early myoclonic encephalopathy, and (4) Ohtahara syndrome. The newer classification of neonatal seizures and epilepsies combines the benign syndromes as “self-limited neonatal epilepsy” (previously benign familial neonatal seizures) and the catastrophic ones as “early infantile developmental and epileptic encephalopathy” (previously early myoclonic epilepsy and early infantile epileptic encephalopathy) (78). It has been proposed that the two syndromes of early myoclonic encephalopathy and Ohtahara syndrome may represent a continuum and that the prominence of tonic seizures in Ohtahara syndrome may be an indication of brainstem dysfunction, which may play an important role in the subsequent transition to infantile spasms (27). A brief discussion of these disorders is essential in understanding the management of neonatal epilepsies.

Self-limited neonatal epilepsy (previously benign familial neonatal seizures)

This is a genetic condition characterized by the onset of seizures during the first weeks of life in an otherwise healthy child with a normal neurologic examination. It is commonly associated with pathogenic variants involving genes with autosomal dominant inheritance (coding for voltage-gated channel subunits), such as KCNQ2 and, less frequently, KCNQ3 and SCN2A. It is frequently expressed in more than one generation in the same family, and a familiar history of neonatal seizures should lead to suspicion of this disorder (99).

Mutations in the KCNQ2 gene encoding the voltage-gated potassium channel KV 7.2 are well-known to be the genetic cause of self-limited familial neonatal seizures. However, de novo KCNQ2 variants have also been recognized as causative mutations of early-onset epileptic encephalopathy of varying severities (58). The KCNQ channels play a key role in maintaining neuronal excitability. Any loss of their activity can lead to a loss of control over neuronal excitability in the brain, resulting in a seizure predisposition (55).

Early infantile developmental and epileptic encephalopathy (previously early myoclonic encephalopathy, early infantile epileptic encephalopathy, or Ohtahara syndrome)

Early myoclonic encephalopathy is a severe neonatal epileptic encephalopathy that is characterized by fragmentary myoclonus followed by erratic focal seizures, massive myoclonic jerks, or tonic spasms. This is associated with grossly abnormal EEG showing suppression-burst activity that may evolve into hypsarrhythmia. The cause is frequently familial and of metabolic etiology, although there is no specific genetic pattern. The course is severe with poor outcomes.

Ohtahara syndrome (also known as early infantile epileptic encephalopathy) is another severe neonatal epileptic encephalopathy syndrome characterized by tonic seizures within the first 3 months of life. The EEG is grossly abnormal and shows a suppression-burst pattern. Identification of the underlying structural, genetic, or metabolic cause is important for appropriate treatment. Treatable metabolic etiologies (pyridoxine and pyridoxal-5-phosphate disorders, biotinidase deficiency, etc.) should be excluded early.

The electroclinical syndromes of Ohtahara syndrome and early myoclonic encephalopathy were first described more than 50 years ago. With the increasing use of video-EEG to recognize neonatal seizures and genetic testing, several new, discrete, etiology-related genetic syndromes are being recognized under these umbrella terms. In a prospective, multicenter U.S. Neonatal Seizure Registry cohort of neonates with seizures, neonatal epileptic encephalopathy was found in 13% of patients, of which 83% had a genetic etiology; KCNQ2 variants were the most common (94). Using targeted sequencing, pathological variants were identified in 42.9% of 70 infants with neonatal-onset epileptic encephalopathy; the most common pathogenic variants were detected in the KCNQ2, STXBP1, and CDKL5 genes (66.7%) (65). In 48 neonates with epileptic encephalopathy with burst suppression, 64% had a genetic etiology, such as STXBP1 (27%), KCNQ2 (10%), and SCN2A (10%) (49). Genetic neonatal-onset epilepsies with monogenic variants (without significant structural brain anomalies) include channelopathies (eg, KCNQ2/3, KCNT1, SCN2A); cell signaling disorders, including developmental transcription factor and RNA processing disorders (eg, FOXG1, GNAO1, BRAT1, CDKL5); synaptopathies and synaptic transmission disorders (STXBP1, DNM1, SPTAN1); and mitochondrial disorders (eg, PDHA1, PDHB, PDHX, POLG1) (100; 105). X-linked neonatal-onset epileptic encephalopathy and severe developmental delay have been associated with a gain-of-function variant p.R660T in GRIA3, causing alteration of synaptic transmission (102). A few examples are discussed below (28; 46; 63; 98).

Etiology-specific syndromes

KCNQ2 encephalopathy. KCNQ2 encephalopathy may first appear in the neonatal period. This genetic etiology can present as either self-limited familial neonatal epilepsy or as a severe neonatal epileptic encephalopathy. Infants with loss-of-function KCNQ2 variants present in the first week of life with an abnormal neurologic examination (encephalopathy, hypotonia, and lack of visual attentiveness) and severe, treatment-resistant seizures (67). Neonates with gain-of-function KCNQ2 variants may have prominent nonepileptic startle-like myoclonus, in addition to encephalopathy and abnormal EEG (64). A distinctive ictal aEEG pattern of marked and prolonged suppression of the amplitude at the end of the seizure has been recognized. Early recognition of the electroclinical phenotype at the bedside by using aEEG may direct targeted treatment with sodium channel blockers (110). Although the seizures may resolve by the age of 3 years, affected children typically have severe global neurodevelopmental disabilities. It is important to identify this etiology as the seizures may respond to antiseizure medications that act on sodium channels (eg, oxcarbazepine, carbamazepine, phenytoin, eslicarbazepine) (67; 62), similar to carbamazepine responsiveness in infants with SCN2A variants (86; 26). Based on a systematic review to identify the best reported therapy for these patients, relating to phenotype, it was found that phenobarbital or carbamazepine appear to be the most effective antiseizure medications for children with a “benign” variant (31). On the contrary, polytherapy is often needed for developmental epileptic encephalopathy patients, even if it does not seem to improve neurologic outcomes (31).

Developmental delay, epilepsy, and neonatal diabetes (DEND) syndrome. DEND is a rare form of severe neonatal epilepsy caused by an activating mutation in the KCNJ11 gene, which encodes the Kir6.2 subunit of the potassium ion channel. Oral sulfonylurea therapy appears to be more effective than insulin in controlling hyperglycemia in this condition and can also lead to improved seizure control and psychomotor development (19; 50).

Neonatal developmental and epileptic encephalopathy associated with SLC13A5. Autosomal recessive pathogenic variants of the SLC13A5 gene have been associated with severe neonatal epilepsy and epileptic encephalopathy, developmental delay, and tooth hypoplasia and hypodontia (56). The prognosis is poor, with severe cognitive and motor impairment. There is no specific treatment.

Neonatal-onset syndromic developmental and epileptic encephalopathy associated with biallelic GAD1 variants. This new entity is characterized by developmental delay, epileptic encephalopathy, cleft palate, joint contractures, and omphalocele (21).

Vitamin-responsive epilepsies. Although inborn errors of metabolism are considered rare, neonatal seizures and epileptic encephalopathy are a common manifestation of some disorders that respond to a specific cofactor or vitamin supplementation. There are several metabolic epilepsies that manifest in the neonatal period, but three vitamin-responsive epilepsies are easy to treat and respond dramatically to proper supplementation: (1) pyridoxine- (vitamin B6) or pyridoxal 5'-phosphate– (the active form of pyridoxine) responsive seizures, (2) biotinidase deficiency, and (3) folinic acid–responsive seizures (81). Pyridoxine-dependent epilepsy due to antiquitin deficiency and the related disorder, pyridoxamine 5'-phosphate oxidase deficiency, are rare but treatable genetic causes of medically refractory neonatal seizures and neonatal epileptic encephalopathy. Two novel biomarkers have been proposed in pyridoxine-dependent epilepsy caused by ALDH7A1 deficiency: 6-hydroxy-2-aminocaproic acid (HACA) and an isomer of C₉H₁₁NO₄ by untargeted metabolomics (17).

Malignant migrating partial seizures of infancy. Malignant migrating partial seizures of infancy, or epilepsy of infancy with migrating focal seizures, is a rare, early infantile epileptic encephalopathy with poor prognosis. Recurrent seizures begin before the age of 6 months but commonly start as infrequent focal-onset seizures within a few weeks of birth, which may become migrating and in clusters later. These may be associated with autonomic features, such as flushing, drooling, pupillary changes, pallor, rash, apnea, and epiphora. The genetic cause of malignant migrating partial seizures of infancy is not fully known. It has been associated with pathogenic variations in KCNT1.

Structural brain malformations. Developmental defects of the cerebral cortex may account for up to 4% of all neonatal seizures (112; 38). The common causes include corpus callosum agenesis, lissencephaly, schizencephaly, polymicrogyria (isolated or in combination with pachygyria complex), hemimegalencephaly, subcortical heterotopia, focal cortical dysplasia, and tuberous sclerosis (118). Several of these malformations occur during the period of embryogenesis and are genetically determined (eg, PAFAH1B1, TSC1, TSC2, DCX, LIS1, ARX, DEPDC5). These may occur in isolation or as part of a genetic syndrome (eg, Miller-Dicker and neonatal Zellweger syndromes). The onset of seizures in the neonatal period can be variable. An early diagnosis is important as seizures can be treatment resistant, and genetic counseling is an important part of management. Surgical intervention for epilepsy is infrequently offered for neonates, often reserved only for those with catastrophic presentations. Surgical resection of focal cortical dysplasia in neonates with atypical neonatal presentation (refractory seizures from day one of life) of diseases such as tuberous sclerosis complex have led to resolution of refractory seizures (36). Similarly, “endovascular embolic hemispherectomy” has been tried for treatment of medically intractable seizures in neonates and young infants with hemimegalencephaly in whom surgical hemispherectomy is not a viable option (70).

Drugs used in neonatal epilepsies

There is a lack of evidence on the use of antiseizure medications for neonatal epilepsies. There are no FDA-approved drugs for this age group. WHO guidelines indicate that phenobarbital should be used as the first-line agent for the treatment of neonatal seizures and should be made readily available in all settings (116). A Cochrane analysis of 18 trials (1342 infants) concluded that phenobarbital as a first-line antiseizure medication is probably more effective than levetiracetam in achieving seizure control after the first loading dose and after the maximal loading dose of antiseizure medication (moderate-certainty evidence) (01). The review also concluded that limited data and very low-certainty evidence precludes any reasonable conclusion on the effect of using one antiseizure medication versus another on other short- and long-term outcomes in term and late preterm infants. In neonates who continue to have seizures despite administering the maximum tolerated dose of phenobarbital, either a benzodiazepine, phenytoin, or lidocaine may be used as the second-line agent for the control of seizures. The following is a discussion of drugs considered for use in neonatal epilepsies.

Self-limited familial neonatal seizures. There is no consensus treatment for this entity. The seizures usually remit spontaneously without medication. Prolonged seizures may need to be terminated with benzodiazepines, phenobarbitone, or phenytoin. Hence, strategies to increase the activity of KCNQ2/3 represent a novel means to restore the control of neuronal excitability in epileptic patients.

Self-limited neonatal seizures (nonfamilial) (fifth-day seizures). These seizures also commonly remit without antiepileptic drugs. If medications are used for the termination of prolonged seizures, they are discontinued soon after the seizures subside. Seizures do not recur, and the neurodevelopmental outcome is good.

Structural focal epilepsies. The seizures are usually treatment resistant. Specific therapeutic options exist for tuberous sclerosis, including vigabatrin, adrenocorticotrophin hormone, and everolimus. For several focal malformations, surgical options may be tried based on the resectable ictal zone.

Malignant migrating partial seizures of infancy. The seizures do not respond well to treatment. Although several antiseizure medications and adrenocorticotrophin hormone have been tried, beginning in the neonatal period, the overall prognosis is poor. Affected individuals may develop normally at first, but neurodevelopment stagnates and soon starts to decline. KCNT1-associated migrating partial seizures may be treated with quinidine and ketogenic diet.

Neonatal epileptic encephalopathy / developmental and epileptic encephalopathy. Response to treatment depends on the underlying etiology. Some of the genetic developmental and epileptic encephalopathies, like KCNQ2 and SCN2A, may respond to sodium channel blockers. Documented structural focal epilepsies may need a sequential trial of antiseizure medications depending on the response. In resistant epilepsies of unknown etiologies, sequential therapeutic trials of pyridoxine (100 mg intravenous injection with continuous EEG and cardiorespiratory monitoring or 15 to 30 mg/kg/day orally in three divided doses for at least 3 to 5 days) and pyridoxal 5'-phosphate should be given (24). In nonresponsive cases, leucovorin (folinic acid) may also be administered (3 to 5 mg/kg/day in divided doses) as some cases of pyridoxine-dependent epilepsy may respond better to folinic acid. The EEG response is often dramatic, but neonates should undergo further biochemical evaluation (measurement of urinary alpha-aminoadipic semialdehyde and plasma pipecolic acid) and genetic confirmation (mutational analysis of the ALDH7A1 or PNPO gene). Long-term management includes lifelong supplementation of these vitamins and lysine-restricted formula or diet (24). Biotinidase deficiency due to pathogenic variations in the BTD gene also result in drug-refractory neonatal or early infantile seizures that are responsive to oral biotin supplementation (25). In regions where universal newborn screening is not yet available, a trial of pyridoxine or pyridoxal 5'-phosphate, folinic acid, and biotin supplementation may be considered in drug-refractory neonatal seizures.

Specific drugs used in neonatal epilepsies

Phenobarbital (loading dose 20 to 40 mg/kg intramuscularly, intravenously, or orally; maintenance dose 3 to 5 mg/kg once daily commencing 12 to 14 hours after the loading dose). Although considered as safe and efficacious as other antiepileptic drugs in neonates, the long-term neurodevelopmental outcomes in neonates treated with phenobarbital for neonatal seizures is lacking (04). Initial studies showed that 75% to 98% of neonates with seizures were initially treated with phenobarbitone (02; 38). Subsequent research suggested that seizure control may be achieved in only 40% to 50% of neonates treated with phenobarbitone. Although still widely used as first-line treatment, a multicenter study of 6245 infants from 47 U.S. institutions from 2007 to 2016 showed a decrease in both phenobarbital initiation within the neonatal period (96.9% to 91.3%, p=0.015) and continuation at discharge (90.6% to 68.6%, p< 0.001) in neonatal intensive care units (48). At the same time, levetiracetam initiation within the neonatal period (7.9% to 39.6%, p< 0.001) and continuation at discharge (9.4% to 49.8%, p< 0.001) increased (48). There are concerns over its proapoptotic effects on neurons and impairment of normal neonatal synaptic maturation in preclinical trials (35). In clinical studies it has been linked to impaired cognition and motor and academic underachievement in reading in children (101). This proapoptotic effect appears to be potentiated with polypharmacy. It is particularly difficult to assess these outcomes in neonatal epilepsies, which might affect neurodevelopment significantly (48).

A randomized, double-blind, parallel-group, phase III study (NCT03602118) with the aim of evaluating the efficacy of phenobarbital sodium injections in participants who have suffered from clinical seizures was reported on ClinicalTrials.gov but was later withdrawn (72). The study was designed to demonstrate the effectiveness of phenobarbital for the prevention of subsequent seizures and to demonstrate improved efficacy when used at a higher dose (40 mg/kg) compared with a lower dose (20 mg/kg). Study participants included infants born at 35 or more weeks + 0/7 weeks’ gestational age and infants aged 0 to 28 days after birth with evidence of an electrographic seizure lasting at least 10 seconds and undergoing continuous EEG monitoring.

Phenytoin (loading dose 15 to 20 mg/kg intravenously, maintenance dose 5 to 8 mg/kg/d 24 hours after the loading dose). In addition to the treatment of neonatal seizures and neonatal status epilepticus, phenytoin is used as a first-line treatment in confirmed or suspected SCN2A-related epileptic encephalopathies due to its sodium channel-blocking option (114).

Carbamazepine and oxcarbazepine. Carbamazepine and oxcarbazepine have both been used in neonatal epilepsies associated with channelopathies. Low-dose carbamazepine has been proposed as a first-line treatment for benign familial neonatal epilepsy, even in cases of status epilepticus (86). Because carbamazepine prevents repetitive firing of action potentials in depolarized neurons by blocking voltage-dependent sodium channels, it has infrequently been used in neonatal seizures via nasogastric tub in low doses (97; 106). Use in neonates is limited due to the lack of a parenteral formulation, low activity of isoenzyme CYP3A4 and epoxide hydrolase enzymes at birth, and reduced renal elimination. The drug has excellent therapeutic and side-effect profiles in older children. The first report of the long-term use of carbamazepine in 10 neonates with two or more seizures due to hypoxic ischemic encephalopathy used 10 mg/kg of carbamazepine as a loading dose via nasogastric tube (97). This preliminary study showed that carbamazepine may be an effective anticonvulsant for neonatal seizures. Another study evaluated the role of carbamazepine in 19 neonates with benign familial neonatal epilepsy (86). Earlier initiation of carbamazepine was associated with shorter hospitalization (p< 0.01), with no reported side effects. All patients had normal development and remained seizure-free at a mean follow-up period of 7.8 years (6 months to 16 years). Both studies provide evidence that carbamazepine is safe and rapidly effective in neonates with seizures and epilepsy. A retrospective study of five neonates with KCNQ2/3 pathogenic gene variants and drug-refractory neonatal seizures showed effective treatment with carbamazepine in four patients and control of neonatal seizures within the first week of life with three children showing normal development and good seizure control with carbamazepine later in life (109).

Although not used in neonates, oxcarbazepine has been used as an adjunctive therapy for focal-onset seizures in infants older than 1 month of age (74; 45; 117). In an open-label study on the safety, tolerability, and pharmacokinetics of oxcarbazepine (as monotherapy or adjunctive therapy) in infants and young children with focal seizures, oxcarbazepine was titrated from 10 to 60 mg/kg/day in a 30-day treatment phase, and its metabolite, the 10-monohydroxy derivative, was measured in the blood (66). The authors concluded that oxcarbazepine was safe and well tolerated in infants and young children. Another retrospective study assessed the efficacy and safety of oxcarbazepine oral suspension in 20 Asian infants and 38 children with various types of epilepsy (113). The drug suspension dose was titrated from 7.5 to 30 mg/kg/day within 1 month in all cases. The authors concluded that there were no statistically significant differences between the children and infants in efficacy (75% vs. 82%, p=0.7) and adverse effects (30% vs. 21%, p=0.5). Patients with focal-onset epilepsy responded better to oral suspension as monotherapy than those with multifocal epilepsy (113).

A retrospective analysis of 50 children with focal epilepsy onset within 3 months of age (median age of epilepsy onset of 11.5, IQR 2-42 days) and treated with oxcarbazepine in a tertiary pediatric epilepsy center in China showed that oxcarbazepine was effective in 76% cases (more in patients with sodium/ potassium channel gene variants), with seizure freedom in 56% patients, and it was well tolerated (54). The study population comprised of 32 cases of developmental epileptic encephalopathy, 10 cases of self-limited neonatal or neonatal-infantile epilepsy, and eight cases of focal epilepsy that could not be classified as epileptic syndrome. The median age of application of oxcarbazepine was 47 (IQR 31-66) days. The median follow-up time was 16.5 (IQR 10-25) months, with seven deaths (54).

Valproic acid. Valproic acid is an effective antiseizure medication for recurrent seizures and status epilepticus. Although it may be seen as maintenance antiseizure therapy in neonatal seizures (11), the risk of liver failure is highest in the neonatal period. Valproic acid has been proposed to cause developmental damage in the fetal and neonatal periods through this excitotoxicity mechanism (37). Prenatal exposure of valproic acid in mice significantly reduces synaptic δ-catenin protein and AMPA receptor levels, and impairs dendritic branching, thus impairing communication, which may probably underlie valproic acid-induced autism spectrum disorder in early childhood (83). It should be used with extreme caution in neonates, especially with polytherapy and if an underlying inborn error of metabolism is suspected.

Clobazam. Clobazam is not approved for use in neonates.

Levetiracetam. Levetiracetam is FDA-approved for seizures in neonates 1 month of age and older. Although there are some studies on the use of levetiracetam in acute neonatal seizures and in older children with pathogenic variants of STXBP1 (52), there are no good data specific to neonatal epilepsies.

Brivaracetam. Brivaracetam has not been approved for use in neonates.

Vigabatrin. Vigabatrin is FDA approved as monotherapy for pediatric patients 1 month to 2 years of age with infantile spasms and for whom the potential benefits outweigh the potential risk of vision loss. The drug has not been used in neonatal epilepsies.

Lacosamide. Lacosamide has been shown to reduce neocortical seizure-like activity in neonatal mice in vitro and in vivo without an acute increase in apoptosis, thus, supporting its use in treating neonatal seizures (47). Lacosamide has anecdotally shown benefits as a sodium channel blocker in SCN2A-related intractable neonatal and infantile seizures (33). A retrospective study on 28 neonates with seizures found no adverse effects attributed to lacosamide in the neonates, and it was well tolerated (34). A study to evaluate the efficacy, safety, and pharmacokinetics of lacosamide in neonates with repeated electroencephalographic neonatal seizures is currently registered and is recruiting (NCT04519645).

Topiramate. The clinical off-label use of topiramate in three cases of refractory neonatal seizures of unclear origin with no response to conventional antiepileptic drugs has been reported (82). All patients became seizure-free during hospitalization and were followed for approximately 1 year after hospital discharge, with monotherapy with topiramate. A multicenter 75-neonate study demonstrated that topiramate is effective and well-tolerated in neonates with refractory seizures and that the development of necrotizing enterocolitis after treatment with topiramate is rare (108). The role of topiramate cotreatment in preventing subsequent epilepsy deserves further study.

Zonisamide. Zonisamide has not been FDA-approved for use in neonatal epilepsy.

Perampanel. Perampanel is FDA-approved for use as adjunctive therapy for the treatment of partial-onset seizures with or without secondarily generalized seizures in patients with epilepsy 12 years of age and older.

Ezogabine. Ezogabine is a positive allosteric modulator of KCNQ2–5 (Kv7.2–7.5) ion channels. Thus, it functions as a neuronal potassium channel opener for the treatment of epilepsy. These channels are predominantly expressed in the neurons and are key determinants of cellular excitability (39). Ezogabine was used in 11 infants and children with loss-of-function KCNQ2 variants, and it improved refractory seizure activity without severe side effects (58). However, the youngest patient in this cohort, who was 2 months of age, showed no improvement. The drug was withdrawn in 2017 due to FDA safety concerns related to the risks of retinal abnormalities and blue discoloration of the skin, nails, and mucous membrane.

Fenfluramine. There are no data on its use in neonatal epilepsies.

Other therapies

Ketogenic diet. A study evaluated evidence on the use of the ketogenic diet and its variants in children, particularly in neonates and infants aged 0 to 23 months (29; 30; 104). Its role has been specifically evaluated in KCNT1-related developmental and epileptic encephalopathy. Although there are a few case reports of the use of ketogenic diet in neonates (103; 107), concerns for safety, efficacy, and tolerability in this age group remain. Although the ketogenic diet and its variants are safe and effective in children with drug-resistant refractory epilepsy, only one of 21 studies described its administration in neonates and infants (84). In a retrospective series of nine newborns and children younger than 4 months of age with refractory epilepsy in Columbia, ketogenic diet was started as early as day 9 of life, with the average age of initiation being 24 days (92). The therapy was successful in controlling seizure burden without considerable adverse effects. Successful use of the classic ketogenic diet has been reported in two identical twin premature neonates (at the conceptual age of 35 weeks) with SCN2A-related developmental and epileptic encephalopathy with drug-refractory epilepsy (73). Seizure frequency was significantly reduced (> 90% reduction in both patients), and the side effects were tolerable and correctable. Further studies on the use of the ketogenic diet in neonates and infants, ie, more clinical trials, are mandatory before its use is justified in the neonatal period. A retrospective study evaluated the effectiveness, safety, and survival in 19 neonates and infants younger than 3 months of age who started ketogenic dietary therapy for drug-resistant developmental epileptic encephalopathy and completed minimum 1 month of follow-up and concluded that ketogenic dietary therapy was safe and effective in newborns and very young infants (05). The authors found that 74% achieved greater than 50% response at 1 month on diet, 37% achieved a greater than 75% seizure reduction, and 10.5% became seizure-free. At 3 months, a greater than 50% decrease in seizure frequency was observed in 72%; 16% had a greater than 75% reduction; 21% became seizure-free. Overall survival was 76% at 1 year on diet. Incidence of acute and late adverse effects was low, and most adverse effects were asymptomatic and manageable. It has been suggested that the starting keto ratio should be lower in neonates and young infants, and the keto ratio titration period should be longer than for children older than 2 years (40).

Steroids. There are no clear data on the utility of steroids in complex neonatal epilepsies. Steroids are mentioned as a useful therapeutic option for neonatal-onset epileptic encephalopathies like Ohtahara syndrome in some of the classical monographs (13). However, the exact therapeutic protocols in the neonatal age group are not available.

Quinidine. Quinidine has been used in the treatment of KCNT1-related epilepsy (18). It has a variable efficacy (20% to 45%) and potential cardiotoxicity (long QTc) (68; 32).

Sulfonylureas. DEND syndrome, which is characterized by developmental delay, epilepsy, and neonatal diabetes, is mainly caused by gain-of-channel function mutations in the KCNJ11 gene (11p15.1) encoding a subunit of the ATP-sensitive potassium channel. Sulfonylurea and glinide drugs close the channel through a pathway independent of ATP and are now the primary therapy this disorder.

Ketamine. There is no clear evidence on the use of ketamine in neonatal epilepsies. A retrospective single-center cohort study of 69 children (n = 13, 19% neonates, median age of onset of refractory status epilepticus 0.7 years old) under intensive care showed seizure termination in 46%, reduction in 28%, and no change in 26% of cases (41). The study provided class IV evidence on the therapeutic utility of ketamine for the treatment of refractory status epilepticus in children and neonates.

Amitriptyline. Although loss-of-function variants in KCNQ2 are associated with a spectrum of neonatal-onset epilepsies, gain-of-function variants may cause a more complex phenotype. Bayat and colleagues describe three patients with de novo KCNQ2 “gain-of-function” variants presenting with mild global developmental delay, prominent language deficits, strong activation of interictal epileptic activity during sleep, and absence of epileptic seizures, who showed improvement in motor, verbal, social, sensory, and adaptive behavior skills with the Kv7 channel blocker antidepressant drug amitriptyline used over 2 years (Bayat 2023). Although no direct antiseizure effect was demonstrated, the authors propose that amitriptyline might represent a potential targeted treatment for patients with KCNQ2 “gain of function” variants. However, its use in neonatal age group is not yet approved.

Indications

None of these drugs are FDA-approved for use in neonatal epilepsies.

Outcomes

The developing brain is exquisitely sensitive to drug-induced disturbances in the chemical homeostasis of neurons. Exposure to antiseizure medications in the critical period of development can cause neurologic and psychiatric abnormalities. There may be potential detrimental effects of antiseizure medications, but one must balance these with the treatment of the developmental and epileptic encephalopathies.

In addition, sleep macrostructure is generally preserved in self-limited genetic epilepsy in the first 6 months of age. In infants with epilepsies in the developmental and epileptic encephalopathy spectrum, sleep architecture is significantly impacted. It has been suggested that infants with abnormal sleep architecture may also experience compromised developmental outcomes, thereby hinting at the need to timely identify and treat epilepsies early in life (42).

Adverse effects

The majority of these drugs are not FDA-approved for use in neonates. The adverse effect reporting is limited. The data on effects on the nervous system and other systems are limited to children and adults. Cognitive outcomes are of particular concern in children who may be at an increased risk of cognitive adverse effects of early antiseizure medication treatment (16). The results of the Neurodevelopmental Effects of Antiepileptic Drugs (NEAD) study, a prospective observational study on fetal exposure to antiepileptic drug monotherapy that was performed in 25 centers in the United Kingdom and the United States, showed that valproic acid is associated with continued deficits in IQ and adaptive functioning as well as an elevated risk for ADHD at 6 years of age (22; 57). A systematic review of 35 studies on neurodevelopmental outcomes in children exposed to newer antiseizure medications highlighted that the current data on the safety of newer drugs, such as eslicarbazepine, lacosamide, perampanel, or zonisamide, are too limited (44).

Scientific basis

Early-onset epilepsy before the age of 1 year is a significant risk factor for developmental, behavioral, and cognitive problems (79). Hence, correct etiology and aggressive treatment are needed to control the seizures. It may not be possible to wait for several months to judge the efficacy of a treatment, as with older children or adults, because the neonatal period is a critical period of brain development. Some neonates rapidly progress to epileptic encephalopathy and EEG worsening, with developmental stagnation that continues into infancy (09). The long-term effects of antiepileptic drugs on the neurocognitive profile of neonates are not completely known (08). The belief that early-in-life seizures, especially status epilepticus, cause more pronounced brain injury and mesial temporal sclerosis is probably overstated, and the injury may occur in genetically predisposed individuals (60). The treatment outcomes for neonatal epilepsy syndromes depend on the underling etiology and responsiveness to available agents.

Precision medicine focuses on disease treatment and prevention tailored to the individual variability in genes, environment, and lifestyle for each patient. Genetic testing is advisable in all early-onset epilepsies, especially developmental and epileptic encephalopathy in the neonatal age group. A gene panel targeting 46 epilepsy genes was used on a cohort of 216 patients with a range of different epilepsies, from benign neonatal seizures to epileptic encephalopathies (61). The overall yield of the study in identifying a presumed disease-causing variant was 23%, and neonatal-onset epilepsies had the highest rate of positive findings (57%). The overall yield for patients with epileptic encephalopathies was 32%, compared to 17% among patients with generalized epilepsies and 16% in patients with focal or multifocal epilepsies. This highlights the enormous utility of genetic testing for therapeutic decision-making.

Neonatal-onset SCN2A- and SCN8A-related epilepsies due to gain-of-function mutations benefit from sodium channel blockers, such as phenytoin and carbamazepine (115). On the contrary, sodium channel blockers are avoided in loss-of-function mutations in SCN1A. The recommended drugs are stiripentol, valproic acid, and benzodiazepines. The ketogenic diet is recommended in loss-of-function mutations in SLC2A1-related epileptic encephalopathy. Quinidine has been suggested as a possible precision medicine treatment for KCNT1-associated epileptic encephalopathy. Sodium channel blockers such as phenytoin and carbamazepine are recommended as possible precision medicine treatments for KCNQ2-related epileptic encephalopathy (58). mTOR inhibitors (everolimus) are recommended as precision therapy in mTOR-signaling pathway–related epileptic encephalopathy, such as tuberous sclerosis and DEPDC5-related epilepsy (51).