Neonatal Seizures and Syndromes (Epilepsia, 2002)

Neonatal Seizures and Syndromes

Barry R. Tharp

Department of Neurology, School of Medicine, and The M.I.N.D. Institute, University of California, Davis,Sacramento, California, U.S.A.

Summary: Neonatal seizures frequently accompany neonatalencephalopathies. Seizures occur in∼1.8–5/1,000 live births inthis country and are caused by virtually any condition thataffects neonatal brain function. This review provides a simpleclassification of seizures and emphasizes that many abnormalintermittent behaviors in this age group are not accompanied byictal EEG patterns. Additionally,#50% of neonatal seizuresare not associated with abnormal clinical behavior. This is acommon phenomenon, particularly after anticonvulsant treat-

ment in which the clinical seizures are suppressed but electro-graphic seizures continue unabated. Seizures also may becaused by genetic disorders, several of which are benign, fa-milial, and caused by channelopathies involving potassiumchannels. The review also discusses the epileptic syndromesseen in neonates, including early myoclonic encephalopathy,Ohtahara syndrome, pyridoxine dependency, and glucose trans-porter type 1 syndrome.Key Words: Neonate—Seizures—Syndrome—GLUT-1—Genetic—Myoclonus.

Seizures are a common manifestation of encephalop-athy in newborn infants including the very small prema-ture. They occur in∼0.15–3.5% of newborns. TheNational Collaborative Perinatal Project (1) reported anincidence of five per 1,000, whereas in a more recentpopulation-based study in a single county in Kentucky,the risk was 3.5/1,000 (2). The incidence in infants bornbetween 1992 and 1994 in Harris County, Texas, was1.8/1,000 live births (3).

In some studies the incidence was high in very smallprematures, decreased in older prematures, and increasedagain in term infants. In others, including the Texasstudy, the incidence decreased as birth weight increased.In all these studies, some or all of the infants were iden-tified with seizures solely on the basis of clinical events(in the Kentucky study, for example, only 22% of theinfants with presumed seizures had confirmatory ictalEEGs). It is now well known that many seizure-like epi-sodes in neonates are not associated with concomitantelectrographic abnormalities. Some may actually be sei-zures arising from deeper sites, such as the hippocampus,or even the brainstem or cerebellum, and not be recordedby distant scalp electrodes or be limited surface eventswith restricted fields on the scalp. Many represent ab-normal nonepileptic motor phenomena that arise from

brainstem centers as a consequence of the reduction ofneocortical inhibitory input secondary to a diffuse en-cephalopathy, so-called “brainstem release phenomena.”These nonepileptic events are quite common, forexample, in encephalopathic premature infants withmassive intraparenchymal hemorrhages and extensivewhite-matter damage and term infants with hypoxic–ischemic encephalopathy (HIE). They are probablycaused by the removal of the normal cortical inhibitoryinput as well as by a direct disturbance of brainstemfunction from increased intracranial pressure, temporallobe herniation, or subarachnoid blood.

Infants with severe encephalopathies often have subtleseizures (e.g., brief deviation of the eyes, cessation ofnormal motor activity, apnea, autonomic symptoms, andsubtle posturing), which may be missed by trained per-sonal. Additionally a significant number of infants,>50% in several studies, have electrographic seizuresonly. These seizures would be missed without continu-ous EEG monitoring. Electrographic-only seizures usu-ally occur in infants with diffuse encephalopathies, thosewho have received anticonvulsant drugs (AEDs), par-ticularly barbiturates, and in iatrogenically paralyzed in-fants. It is well known, for example, that infants withclinical status epilepticus often remain in electrographicstatus after the clinical events are eliminated by paren-teral phenobarbital (PB).

Neonatal seizures are usually self-limited and sponta-neously disappear in days to a few weeks. Seizures wanewhen the causative encephalopathy resolves (e.g., HIE,

Address correspondence and reprint requests to Dr. B. R. Tharp atDepartment of Neurology, School of Medicine, and The M.I.N.D. In-stitute, University of California, Davis, 4860 Y Street, Suite 3020,Sacramento, CA 95817, U.S.A. E-mail: [email protected]

Epilepsia,43(Suppl. 3):2–10, 2002Blackwell Publishing, Inc.© International League Against Epilepsy

2

meningitis, and hemorrhage). Many infants, particularlythose with static encephalopathies, will experience sei-zures later in life after a seizure-free period of months toyears. Seizures also may continue beyond the neonatalperiod without a seizure-free interval in some infants(Table 1).

ETIOLOGY

It is beyond the scope of this review to discuss thecause of neonatal seizures, which includes virtually ev-ery encephalopathy of infancy. Table 2 lists some of themost common etiologies.

The diagnostic workup should include metabolic tests,particularly if a structural abnormality is not obvious(Table 3) The first four items in Table 3 should be doneexpeditiously. One should search for acute reversiblecauses that, if left untreated, could lead to significantinjury (e.g., hypoglycemia, sepsis, acidosis, hypocalce-mia, and hyperammonemia). These metabolic perturba-tions also may be markers of a more serious inborn errorof metabolism that, if left untreated, could have seriouslong-term consequences. A trial of pyridoxine and folinicacid should be considered if seizures persist and no eti-ology is apparent (4).

Classification of neonatal seizuresMany behavioral events previously called seizures

have been shown by long-term monitoring to lack anictal electrographic signature. Certain clinical events (fo-cal clonic and tonic and some myoclonic movements) areusually accompanied by ictal EEG patterns, whereas oth-ers (generalized tonic, oral–buccal, and motor automa-

tisms) are infrequently associated with a corticaldischarge. They can be suppressed by gentle restraint orrepositioning and precipitated by stimulation. The extentof the involvement, including the amplitude of the mus-cular contraction and the number of muscles involved,often increases with each stimulation (“recruiting re-sponse”), consistent with a brainstem release mecha-nism. Table 4 presents a clinical classification ofneonatal seizures. Seizure types underlined are morelikely to be accompanied by ictal EEG abnormalities.

Progressive epileptic syndromes in the first year oflife with onset in the neonatal period

Several well-defined syndromes have their onset in theneonatal period. In some individuals, the cause is known,whereas in others, with virtually identical symptoms,none is found.

1. Early (neonatal) myoclonic encephalopathy (EME)is characterized by erratic, fragmentary myoclonic jerksthat begin in the first month of life, often in the first weekof life. These are followed by partial seizures, massivemyoclonus, and infantile spasms and, infrequently, tonicseizures. The EEG shows a characteristic suppression-burst pattern (S-B; bursts of spikes, sharp waves, andslow activity lasting 5–6 s, alternating with 4- to 12-speriods of attenuation). The EEG often evolves into ahypsarhythmic pattern or a markedly abnormal back-

TABLE 1. Neonatal seizures: long-term outcome

High mortality (30%) and morbidity (50% of survivors)Approximately 30% of survivors develop epilepsyWorst outcome in infants with hypoxic–ischemic encephalopathy,

meningitis, and cerebral dysplasiaBetter outcome with transient neonatal hypocalcemia, idiopathic

and familial seizures, and strokeNeonatal EEG, neurologic examination, and imaging results are

best predictors of outcome

TABLE 2. Common etiology of neonatal seizures

Hypoxic–ischemic encephalopathy (prenatal and postnatal onset)Infection: meningitis/encephalitis (congenital and postnatal),

sepsis without meningitisVascular disease (stroke, venous thrombosis)Intracranial hemorrhage, particularly intraparenchymal and

subarachnoidMetabolic encephalopathy (transient metabolic disturbance and

inborn error of metabolism)Cerebral dysgenesis, migration disorders, and malformations,

including those associated with chromosomal disordersTrauma (delivery-related and, more commonly, nonaccidental)IdiopathicEpileptic syndromes, including familial epilepsies

TABLE 3. Diagnostic workup of neonatal seizures

Serum glucose, calcium, magnesium, ammonia, lactate, pH, anda complete chemistry panel

Cerebrospinal fluidCranial ultrasoundEEG with perfusion of pyridoxineToxicology screenUrine organic acids, serum and cerebrospinal fluid amino acidsMaternal and fetal titers for congenital infectionCT scan (hemorrhage and calcium) or MRI scan (cerebral

dysgenesis, stroke)

CT, computed tomography; MRI, magnetic resonance imaging.

TABLE 4. Clinical classification of neonatal seizures

ClonicFocal (unifocal, multifocal)HemiconvulsiveAxial

TonicFocal (limb, asymmetric truncal posturing, eye deviation)Generalized

MyoclonicFocalGeneralized (commonly no EEG concomitant)

Motor automatismsOral–buccal–lingualOcular (excluding tonic deviation)Movements of progression (swimming, bicycling)

Other uncommon manifestationsCessation of motor activityInfantile spasmsApnea (rarely a seizure when occurring alone)

NEONATAL SEIZURES AND SYNDROMES 3

Epilepsia, Vol. 43, Suppl. 3, 2002

ground with multifocal spikes and sharp waves. Theseinfants have an obvious encephalopathy from birth withmarkedly delayed milestones, symmetric hypotonia, andprogressive microcephaly from cerebral atrophy. The eti-ology is often not found, although some cases are famil-ial and due to an inborn metabolic disorder (e.g., glycineencephalopathy,D-glyceric acidemia, propionic acide-mia, methylmalonic acidemia). Infants with similar clini-cal courses but without the S-B EEG pattern also havebeen described. Intrauterine seizures (due to cerebraldysgenesis) (5), a pyridoxine-dependent syndrome, andhydranencephaly also have been reported with EME.

2. Early infantile epileptic encephalopathy with S-BEEG pattern (EIEE; Ohtahara syndrome)is often con-fused with early myoclonic encephalopathy. The EEGabnormalities are very similar, and the etiologies mayoverlap. This disorder is characterized by intractabletonic seizures in the neonatal period or early infancy(often with an ictal EEG pattern similar to that accom-panying typical infantile spasms in older infants). Sei-zures are accompanied by a severe encephalopathy,which progresses to other seizure syndromes includinginfantile massive spasms (IMSs) with hypsarrhythmia,partial seizures with multifocal spike foci on the EEG,and the Lennox–Gastaut syndrome. In some instances,the seizures may disappear, despite the persistence of thesevere encephalopathy.

Many infants with EIEE have cerebral dysgenesis, in-cluding major brain malformation. Ohtahara (6), in oneof his earlier articles reporting 14 cases, described cere-bral dysgenesis in eight, porencephaly in two, nonspe-cific atrophy in one; three infants had no obviouspathology. Aicardi syndrome also has been associatedwith EIEE. Williams (7) described the first case of EIEEwith a neurometabolic cause: cytochrome oxidase defi-ciency (complex IV), diagnosed with a muscle biopsy.

Lombroso (8) described several infants with symp-toms and EEGs consistent with EME caused by ahypoxic–ischemic encephalopathy. The seizures, how-ever, stopped after several weeks, which is not typical ofEME. He found no significant differences in the EEGS-B pattern in these two syndromes but emphasized thattonic seizures were more common in EIEE (and mayoccur in series or clusters similar to IMS) and myoclonicseizures typified EME. He also recognized a more severecourse in EIEE, the rarity of neurometabolic etiologies inEIEE, and the occurrence of cerebral dysgenesis in both.He also pointed out that the S-B pattern may resemblethe pattern seen during sleep in some infants with typicalIMS, who do not have a hypsarrhythmic EEG patternduring wakefulness. He considered EIEE an early variantof IMS.

Infants with EIEE may rarely have a focal corticaldysplasia with asymmetric tonic seizures and S-B pat-tern. Some have been helped by resective surgery (9).

We recently saw an infant with EIEE whose magneticresonance imaging (MRI) showed typical hemimegalen-cephaly (Figs. 1–3). The child’s seizures were com-pletely eliminated by a hemispherectomy at age 7months.

Miller et al. (10) described an infant with EIEE and adiffuse neuronal migration disorder with cortical laminardisorganization, microdysgenesis, abnormal hippocampi,and immature neurons in both amygdalae and bilateralenlargement of the dentate and olivary nuclei. The MRIshowed only atrophy and a thin corpus callosum. Therealso was an absence ofg-aminobutyric acid (GABA) inthe cerebrospinal fluid (CSF), and the positron emissiontomography (PET) scan showed decreased metabolismin both frontal lobes and left anterior temporal lobe.

Cerebral dysgenesis is a well-recognized cause of in-fantile and neonatal seizures. It should be kept in mind,however, that cerebral dysgenesis may be caused by aneurometabolic disorder that disturbs early fetal braindevelopment. An MRI that shows a migrational disorder,therefore, should not stop the investigation for anotheretiology of the infant’s seizures. Table 5 lists some of themetabolic disorders that have been associated with cere-bral dysgenesis. Some have been reported with EME.

3. Glucose transporter type 1 syndrome (Glu-1 defi-ciency syndrome; Glut-1 DS)is probably more commonthan the small number of reported cases would suggest(>40 patients as of July 2000; D. DeVivo, personal cor-respondence). It is characterized by infantile seizures(complex partial, atypical absences, and generalizedmyoclonic and tonic–clonic), usually beginning betweenages 6 and 12 weeks, progressive postnatal microcepha-ly, and developmental delay including ataxia (11–17).The infants may initially appear normal with rare neo-natal seizures, but later in the first year of life, the seizurefrequency increases, and the delay in development be-comes evident. The encephalopathy is caused by a defi-ciency of GLUT-1 (positionally mapped to 1p35-p31.3),resulting in decreased transport of glucose at the blood–brain barrier and across astroglial cell membranes andmarkedly reduced brain and CSF glucose (CSF/bloodratio of <0.65; CSF glucose usually <40 mg/dl). CSFlactate is usually in the lower range of normal (elevatedlactate levels would suggest a mitochondrial disorder).DeVivo’s group (11) initially reported two distinctclasses of mutations as the molecular basis for the func-tional defect of glucose transport: hemizygosity ofGLUT-1 and nonsense mutation resulting in truncationof the GLUT-1 protein. Subsequent studies have dis-covered a variety of mutations including large-scaledeletions, missense mutations, deletions, insertions,splice-site mutations, as well as nonsense mutations (12).

GLUT-1 is multifunctional; in addition to transportinggalactose, water, and glycopeptides, it also transports de-hydroascorbic acid (DHAS), the oxidized form of vita-

B. R. THARP4


min C. This latter property formed the basis of an earlydiagnostic blood test that measured the transport ofDHAS into red blood cells (RBCs). In patients with glu-cose transporter protein syndrome (GTPS), there is areduced uptake into the RBCs (13). A recently developedRBC assay uses14C-labeled 3-O-methyl-D-glucose up-take and is the diagnostic test of choice (14).

The encephalopathy is treated with the ketogenic diet,which supplies the necessary energy to the brain in theabsence of glucose. Long-term outcome data are notavailable on the effectiveness of this diet. One of ourpatients with GTPS had an improvement of his enceph-alopathy with the diet but continued to receive PB for hisseizures. After Klepper et al. (15) reported that barbitu-rates may inhibit glucose transport mediated by GLUT-1and could theoretically worsen the encephalopathy ofGTPSs, we switched him to carbamazepine (CBZ) withseizure control. It also is advised that caffeine-containingbeverages and other methylxanthines, which have a simi-lar effect on the glucose transporter, be avoided by theseindividuals (15). De Vivo (personal correspondence) also

recommends thioctic (a-lipoic) acid as a dietary supple-ment (200–600 mg daily). Thioctic acid in vitro increasesthe expression of the Glut-1 gene.

Although initially thought to be caused by a sponta-neous mutation, the syndrome has recently been reportedin several families with what appears to be dominanttransmission (17). The phenotype was quite mild in someof the involved family members, suggesting that manycases may be missed if they are mildly mentally retardedand without seizures.

4. Neonatal myoclonus without EEG epileptiform ac-tivity. Myoclonus is not uncommon in the neonate andcan occur without concomitant epileptiform abnormali-ties in the following situations:

a. Severe neonatal encephalopathies with isoelectricor markedly abnormal EEG backgrounds, in whichthe myoclonus is probably of brainstem origin.

b. The EEG is performed without midline parasagittalelectrodes, which may miss the not uncommon fo-cal epileptiform activity in this region. Central ver-

FIG. 1. Ohtahara syndrome in a 2-week-old term infant with uncomplicated pregnancy and delivery. From birth there were brief periodsof increased tone, extension and elevation of the extremities, and upward eye deviation every few minutes. EEG showed an invariantmarkedly discontinuous pattern with high-voltage bursts of spikes and sharp and slow waves, with low-voltage interburst background(“suppression-burst”). Note the interhemispheric asymmetry due to the extremely depressed background on the right. No response totactile stimulation.



tex epileptiform activity temporally related to themyoclonic jerks is often stimulus sensitive and fre-quently overlooked or not sought by the technolo-gist.

c. Segmental/spinal myoclonus(1) HIE(2) Spinal cord tumor(3) Hyperglycorrhachia (18)(4) Meningitis(5) Spinal dysraphic syndromes(6) Hyperexplexia/stiff-baby syndrome

d. Jitteriness and drug-withdrawal syndromes thatmimic myoclonus

e. Benign neonatal sleep myoclonus. This is a rela-tively common disorder characterized by the onsetof focal, multifocal, or generalized myoclonus inthe first 3 weeks of life. The myoclonus, which maybe quite dramatic, occursonly during sleep, is mostabundant during non–rapid-eye-movement(NREM) sleep, may appear in clusters lasting up toan hour, and disappears immediately on awakening.It is often not stopped by gentle restraint and canoccasionally be provoked by touch or sound. TheEEG is normal, including during periods of myoc-lonus, and the neurologic examination and subse-

quent development are normal. The myoclonususually disappears over the first 6 months of life.There is usually no family history.

f. Normal infants, particularly prematures, duringREM sleep

5. Pyridoxine-dependent seizures (PDSs).The sei-zures in this autosomal recessive disorder usually havetheir onset between birth and age 3 months, althoughoccasional cases have been reported with onset at 3years. Seizures usually appear within 6 h ofbirth and notuncommonly are recognized in utero. Despite treatment,some children are mentally retarded, and some will de-velop MRI evidence of a leukodystrophy, particularly inthe frontal and occipital areas, and cerebral atrophy(19,20). It is a rare disorder with an estimated pointprevalence of one per 687,000 (21).

The EEG abnormalities are relatively nonspecific andconsist of focal and multifocal epileptiform discharges,bursts of generalized spike-and-wave activity, photopar-oxysmal responses, stimulus-induced myoclonus, andoccasionally an S-B pattern in the neonatal period andhypsarrhythmia later in the first year. Mikati et al. (22)described an unusual 1- to 4-Hz ictal pattern in fourpatients, which they thought was highly suggestive of

FIG. 2. Same infant as Fig. 1,EEG during brief tonic event thathad the clinical characteristics ofan infantile spasm. A high-voltagesharp wave is followed by a com-plex slow wave and lower-voltagefast activity, much of the latterprobably representing artifact.

B. R. THARP6


this disorder. In our experience, this pattern may occur inother neonatal encephalopathies.

The intravenous (i.v.) pyridoxine test (50–100 mg i.v.)is usually positive, with prompt cessation of seizures anddisappearance of epileptiform activity over the next fewhours. The test should be repeated if the response isequivocal or transient. There are often transient re-sponses to parenteral B6 in infants who do not have thissyndrome. There also are reports of infants and childrenwho have a dramatic response to the injection, with ap-nea and hypotonia and, occasionally, coma, respiratoryarrest, and an isoelectric EEG. (It is speculated that thisdramatic effect to B6 is caused by the sudden increase inbrain GABA levels) (23,24).

Patients usually require life-long B6 therapy. Although50–100 mg daily usually controls the seizures, higherdoses may be needed. Very high doses (#2,000 mg)have been used but can cause a sensory neuropathy (25).Pyridoxine also may suppress seizures in children withother epileptic disorders, and some have even recom-mended it as an adjunct to standard AED therapy(26,27). These latter cases appear not to have pyridoxine-dependent seizures. The inclusion of such cases in epi-demiologic studies may lead to an overestimation of theincidence of this syndrome. If a child has complete sei-zure control with oral B6, all AEDs should be discontin-ued, and after 6–12 months, a trial of B6 reductionconsidered.

The biochemical cause of this syndrome is unknown.An inborn defect in pyridoxine metabolism has not beenfound. Reduced levels of GABA and increased glutamatehave been found in the CSF and the cerebral cortex (28).The latter may be normalized by pyridoxine (29).

Plecko et al. (30) reported significant elevationsof pipecolic acid concentrations in the plasma andCSF of two patients with B6-responsive seizures. Theyfound that the degree of pipecolic acid elevationwas inversely correlated with plasma pyridoxal phos-phate levels and speculated that there was an en-zyme defect in a pyridoxine-dependent step in pipecolicacid degradation, although they did not identify thatenzyme.

TABLE 5. Metabolic syndromes associated withCNS dysgenesis

Nonketotic hyperglycinemiaMitochondrial disorders (e.g., cytochrome oxidase deficiency)Glutaric aciduria type I and type II3-Hydroxyisobutyric aciduria3-Methylglutaconic aciduria3-Ketothiolase deficiencySulfite oxidase deficiencyPyruvate dehydrogenase deficiencyMenkes’ syndromeNeonatal adrenoleukodystrophyZellweger’s syndromeFumaric aciduria

CNS, central nervous system.

FIG. 3. Same infant as Figs. 1 and 2. T2-weighted magnetic resonance imaging scanshows an enlarged and markedly dysplasticright hemisphere with dilatation of the lateralventricle on the same side.



6. Fifth-day fits.These neonatal seizures of unknownetiology usually begin toward the end of the first week oflife. They were first reported by the French as a distinctclinical and EEG syndrome (31,32).

These authors estimated that 2–7% of neonatal sei-zures could be classified as “fifth-day fits.” Convulsionswere usually partial clonic, often with apnea, and oc-curred in 90% of infants between days 4 and 6 of life.Status epilepticus was not infrequent and could last sev-eral days. Interictal EEG abnormalities were nonspecific,except that 60% showed an unusual pattern with burstsof sharply contoured theta activity, particularly in thecentral regions (“theta pointu alternant”) (31,32). Thispattern has not been seen by other authors (33), and inour experience, can occur in a variety of encephalopa-thies including HIE without seizures. Very few cases ofthis syndrome have been reported since the early 1980s,and it has been suggested that the seizures may havecaused by a transient zinc deficiency or a virus.

7. Genetic neonatal epilepsy syndromesA. Benign fa-milial neonatal convulsions (BFNCs). These are autoso-mal dominant disorders with∼85% penetrance. Theseizures usually begin by the third day of life and areclonic, tonic, and partial, often with automatisms. Myo-clonic seizures are rare. The infants are neurologicallynormal, and most have a spontaneous remission;∼10–15% will have subsequent epilepsy, usually generalizedmotor seizures. A single case of infantile spasms hasbeen reported in a family with BFNC (34). Baxter andKandler (35) reported an infant whose seizures presum-ably began in utero. Several families have been reportedwith familial and nonfamilial neonatal seizures and be-nign rolandic epilepsy (benign epilepsy with central tem-poral spikes) (BECTS). Linkage of BECTS has not beendemonstrated on chromosomes associated with BFNC(36).

The gene for this disorder has been localized to twochromosomes in many, but not all, of the families:

1. BFNC 1, locus on chromosome 20q13.3 (voltage-gated potassium channel gene, KCNQ2)

2. BFNC 2, chromosome 8q24 (potassium channelgene, KCNQ3)

BFNC is one of three idiopathic epileptic syndromescaused by a single gene mutation. The others areautosomal dominant nocturnal frontal lobe epilepsy(ADNFLE; caused by a mutation of thea4 subunit of theneuronal nicotinic acetylcholine receptor) and general-ized epilepsy with febrile seizures plus (GEFC+; a mu-tation in a voltage-gated sodium channel SCN1B).

Ronen et al. (37) reported the clinical and electro-graphic features of a large kindred (69 affected individu-als) in Newfoundland, Canada, with BFNC 1. Forty-twopercent had seizure onset on days 2–5, and 68% stoppedseizing during the first 6 weeks of life. Interictal EEGs

were usually normal, whereas ictal recordings showed apattern that has been described by others and thought tobe quite characteristic of BFNC (37,38). Sixteen percentof their patients had subsequent epilepsy (median age atonset, 8 years), and half of them continued having sei-zures into adulthood. They suggested that the seizureshad a “subcortical, possibly brainstem origin” because ofthe initial ictal EEG feature of generalized suppression ofthe background and the high incidence of ictal tonic,autonomic, and facial symptoms in their population. Ini-tial reports on this syndrome suggested that only thechromosome 20 families had later epilepsy; however,chromosome 8q24 patients also may have seizures laterin childhood (39).

How do these K+ channel mutations lead to epilepsy?KCNQ3 gains its K+ channel function as a heterote-tramer with KCNQ2. The two genes act synergisticallyin the formation of the native M-current, a slowly acti-vating and deactivating K+ conductance that regulatesthe excitability of neurons. In the inner ear, KCNQ3forms a heterotetramer with KCNQ4, and its mutationresults in autosomal dominant sensorineural deafness.Mutations of KCNQ1 produce the long-QT syndromewith deafness (Lange–Nielsen).

The mutation in KCNQ2 (BFNC 1) truncates the last300 amino acids and leads to a nonfunctional channelprotein. This causes a hyperexcitable neuron through areduced potassium-dependent repolarization of the neu-ronal membrane, similar to mutations in KCNQ1 in thelong-QT syndrome.

In the Italian family studied by Lerche et al. (40), anovel mutation within the distal, unconserved C-terminaldomain of KCNQ2 caused a 1-bp deletion, 2513delG, incodon 838, predicting substitution of the last sevenamino acids and the addition of another 56 amino acids.They expressed the mutation inXenopusoocytes andcaused a reduction of the potassium current to 5% of thewild-type current (impaired repolarization), but voltagesensitivity and kinetics were not significantly changed. Asecond artificial mutation yielded a stop codon at posi-tion 838, and the K+ current increased twofold, indicat-ing that the pathologic amino acid extension producedthe phenotype of BFNC and suggested an important roleof the distal, unconserved C-terminal domain of thischannel.

B. Benign familial infantile convulsions. These syn-dromes typically begin later in the first year of life, but insome families, neonatal seizures are seen.

1. Watanabe et al. (41) described a group of infantswith the onset of partial seizures at 3–10 months.Seizures tended to occur in clusters and were char-acterized by motion arrest, decreased responsive-ness, staring or “blank eyes” with simpleautomatisms and mild convulsive movements. Ictal

B. R. THARP8


EEGs showed focal paroxysmal discharges. Fourof their patients had a family history of benigninfantile convulsions. The interictal EEG and de-velopment were normal.

2. Singh et al. (42) described nine families with adominantly inherited epileptic disorder associatedwith febrile seizures. The most common phenotypewas febrile seizures plus (FS+; febrile seizures thatcontinued beyond 6 years or were associated withafebrile generalized tonic–clonic seizures; GTCSs).These were usually gone by adolescence but couldcontinue into adulthood. Additionally the pheno-types included typical FSs, as well as other seizuretypes (absences, myoclonic, or atonic seizures) andmyoclonic–astatic epilepsy (MAE). The disorderwas linked to chromosome 2q23-31. The mean ageat onset of the FS+ was 1.4 years (range, 6 weeksto 4 years), and for MAE, 1.5 years (range, 6 weeksto 4.5 years). All but one child had an onset withfever, and four had an explosive onset with myo-clonic, absence, and GTCSs. In five cases, therewere atonic head nods. All had falls due to myo-clonic seizures, or tonic or atonic seizures. EEGsshowed irregular 2.5- to 4-Hz generalized spike-and-wave or polyspike–wave discharges. Two ormore phenotypes within the GEFS+ spectrum wereobserved in each family.

Linkage to chromosomes 19q (43), SCN1B, a Na+-channelb1 subunit gene, and 2q24-q33 (44) also hasbeen reported. The inheritance in these families is auto-somal dominant with 54–89% penetrance in differentfamilies.

Escayg et al. (45) describe two mutations in SCN1A intwo families with GEFS+2, which probably cause a de-crease in the rate of channel inactivation resulting inincreased Na+ influx, increased excitability of neurons,and increased seizure susceptibility.

REFERENCES

1. Holden K, Mellits ED, Freeman JM. Neonatal seizures, 1: corre-lation of prenatal and perinatal events and outcomes.Pediatrics1982;70:165–76.

2. Lanska M, Lanska DJ, Baumann RJ, et al. A population-basedstudy of neonatal seizures in Fayette County, Kentucky.Neurology1995;45:724–32.

3. Saliba R, Annegars FJ, Waller DK, et al. Incidence of neonatalseizures in Harris County, Texas, 1992–1994.Am J Epidemiol1999;150:763–9.

4. Torres OA, Miller VS, Buist NM, et al. Folinic acid-responsiveneonatal seizures.J Child Neurol1999;14:529–32.

5. du Plessis AJ, Kaufmann WE, Kupsky WJ. Intrauterine-onsetmyoclonic encephalopathy associated with cerebral dysgenesis.JChild Neurol1993;8:164–70.

6. Ohtahara S, Ohtsuke Y, Yamatogi Y, et al. The early-infantileepileptic encephalopathy with suppression-burst: developmentalaspects.Brain Dev1987;9:371–6.

7. Williams A, Gray RG, Poulton K, et al. A case of Ohtahara syn-

drome with cytochrome oxidase deficiency.Dev Med Child Neurol1998;40:568–70.

8. Lombroso C. Early myoclonic encephalopathy, early infantile epi-leptic encephalopathy, and benign and severe infantile myoclonicepilepsies: a critical review and personal contributions. J Clin Neu-rophysiol1990;7:380–408.

9. Komaki H, Sugai K, Sasaki M, et al. Surgical treatment of a caseof early infantile epileptic encephalopathy with suppression-burstsassociated with focal cortical dysplasia.Epilepsia1999;40:365–9.

10. Miller SP, Dilenge ME, Meagher-Villemure K, et al. Infantileepileptic encephalopathy (Ohtahara syndrome) and migrationaldisorder.Pediatr Neurol1998;19:50–4.

11. Seidner G, Alvarez MG, Yeh JI, et al. GLUT-1 deficiency syn-drome caused by haploinsufficiency of the blood-brain barrier hex-ose carrier.Nat Genet1998;18:188–91.

12. Wang D, Kranz-Eble P, De Vivo DC. Mutational analysis ofGLUT1 (SLC2A1) in Glut-1 deficiency syndrome.Hum Mutat2000;16:224–31.

13. Klepper J, Vera JC, DeVivo DC. Deficient transport of dehydro-ascorbic acid in the glucose transporter protein syndrome.AnnNeurol 1998;44:286–7.

14. Klepper J, Garcia-Alverez M, O’Driscoll KR, et al. Erythrocyte3-O-methyl-D-glucose uptake assay for diagnosis of glucose-transporter-protein syndrome.J Clin Lab Anal1999;13:116–21.

15. Klepper J, Fischbarg J, Vera JC, et al. GLUT1-deficiency: barbi-turates potentiate haploinsufficiency in vitro.Pediatr Res1999;46:677–83.

16. Yuan-Yuan H, Hong Y, Klepper J, et al. Glucose transporter type1 deficiency syndrome (Glut1DS): methylxanthines potentiateGLUT1 haploinsufficiency in vitro.Pediatr Res2001;50:254–60.

17. Brockmann K, Wang D, Korenke CG, et al. Autosomal dominantGlut-1 deficiency syndrome and familial epilepsy.Ann Neurol2001;50:476–85.

18. Bass W, Lewis D. Neonatal segmental myoclonus associated withhyperglycorrhachia.Pediatr Neurol1995;13:77–9.

19. Gospe S, Hecht S. Longitudinal MRI finding in pyridoxine-dependent seizures.Neurology1998;51:74–8.

20. Jardim L, Pires R, Martins C, et al. Pyridoxine-dependent seizuresassociated with white matter abnormalities.Neuropediatrics1994;25:259–61.

21. Baxter P. Epidemiology of pyridoxine dependent and pyridoxineresponsive seizures in the UK.Arch Dis Child1999;81:431–3.

22. Mikati M, Trevathan E, Krishnamoorthy K, et al. Pyridoxine-dependent epilepsy: EEG investigations and long-term follow-up.EEG Clin Neurophysiol1991;78:215–21.

23. Kroll J. Pyridoxine for neonatal seizures: an unexpected danger.Dev Med Child Neurol1985;27:369–82.

24. Bass N, Wyllie E, Cohen B, et al. Pyridoxine-dependent epilepsy:the need for repeated trials and the risk of electrocerebral silencewith intravenous infusion.J Child Neurol1996;11:422–4.

25. McLachlan R, Brown W. Pyridoxine dependent epilepsy with iat-rogenic sensory neuropathy.Can J Neurol Sci1995;22:50–1.

26. Jiao F, Gao D, Takuma Y, et al. Randomized, controlled trial ofhigh-dose intravenous pyridoxine in the treatment of recurrent sei-zures in children.Pediatr Neurol1997;17:54–7.

27. Pietz J, Benninger C, Schafer H, et al. Treatment of infantilespasms with high-dosage vitamin B6. Epilepsia1993;34:757–63.

28. Lott I, Coulombe T, DiPaolo R, et al. Vitamin B6-dependent sei-zures: pathology and chemical findings in brain.Neurology1978;28:47–54.

29. Baumeister F, Gsell W, Shin Y, et al. Glutamate in pyridoxine-dependent epilepsy: neurotoxic glutamate concentration in the ce-rebrospinal fluid and its normalization by pyridoxine.Pediatrics1994;94:318–21.

30. Plecko B, Stockler-Ipsiroglu S, Paschke E, et al. Pipecolic acidelevation in plasma and cerebrospinal fluid of two patients withpyridoxine-dependent epilepsy.Ann Neurol2000;48:121–5.

31. Dehan M, Quilleron D, Navelet Y, et al. Les convulsions ducinquième jour de vie: un nouveau syndrome?Arch Fr Pediatr1977;34:730–42.

32. Plouin P. Benign idiopathic neonatal convulsions (familial andnonfamilial). In: Roger J, Bureau M, Dravet F, et al., eds.Epileptic



syndromes in infancy, childhood and adolescence.2nd ed. London:John Libbey, 1992:3–11.

33. Ortibus E, Sum JM, Hahn J. Predictive value of EEG for outcomeand epilepsy following neonatal seizures.EEG Clin Neurophysiol1996;98:175–85.

34. Mori K, Yano I, Hashimoto T. Infantile spasms in one member ofa family with benign familial neonatal convulsions.Epilepsia1993;34:621–6.

35. Baxter P, Kandler R. Benign familial neonatal convulsions: abnor-mal intrauterine movements, provocation by feeding and ictalEEG.Seizure1997;6:485–6.

36. Neubauer B, Moises H, Lassker U, et al. Benign childhood epi-lepsy with centrotemporal spikes and electroencephalography traitare not linked to EBN1 and EBN2 of benign neonatal familialconvulsions.Epilepsia1997;38:782–7.

37. Ronen G, Rosales TO, Connolly M, et al. Seizure characteristics inchromosome 20 benign familial neonatal convulsions.Neurology1993;43:1355–60.

38. Andrews P, Stafsrom C. Ictal EEG findings in an infant withbenign familial neonatal convulsions.J Epilepsy1993;6:174–9.

39. Hirose S, Zenri F, Akiyoshi H, et al. A novel mutation of KCNQ3(c.925T→C) in a Japanese family with benign familial neonatalconvulsions.Ann Neurol2000;47:822–6.

40. Lerche H, Biervert C, Alekov AK, et al. A reduced K+ current dueto a novel mutation in KCNQ2 causes neonatal convulsions.AnnNeurol 1999;46:305–12.

41. Watanabe K, Yamoto N, Negoro T, et al. Benign complex partialepilepsies in infancy.Pediatr Neurol1987;3:208–11.

42. Singh, R, Sheffer IE, Crossland K, et al. Generalized epilepsy withfebrile seizures plus: a common childhood-onset genetic epilepsysyndrome.Ann Neurol1999;45:75–81.

43. Wallace R, Wand DW, Singh R, et al. Febrile seizures and gen-eralized epilepsy type 1 associated with a mutation in the Na+-channel beta-1 subunit gene SCN1B.Nat Genet1998;19:366–70.

44. Moulard B, Guippone M, Chaigne D, et al. Identification of a newlocus for generalized epilepsy with febrile seizures plus (GEF+2)on chromosome 2q24-q33.Am J Hum Genet1999;65:1396–4000.

45. Escayg A, MacDonald BT, Meisler MH, et al. Mutations ofSCN1A, encoding a neuronal sodium channel, in two families withGEFS+2.Nat Genet2000;24:343–5.

B. R. THARP10


Neonatal Seizures and Syndromes (Epilepsia, 2002)

Documents

Transcript of Neonatal Seizures and Syndromes (Epilepsia, 2002)