Epilepsy

Meet the patient

A 16-year-old girl is brought to clinic because, three mornings in a row, she dropped her tea mug while getting ready for school. Her mother also describes "little jumps" of her arms first thing, and one morning last week she woke on the floor, confused, having bitten the side of her tongue. She is otherwise well, sleeps poorly before exams, and went to a strobe-lit party on Saturday.^[1]

Two exam questions are now live. What is the syndrome? — and which drug will you offer, and which will you refuse to prescribe? The dropped mugs and the early-morning jerks on waking, with photosensitivity, are the fingerprint of juvenile myoclonic epilepsy. Get the syndrome right and the drug chooses itself; get it wrong and you will make her worse.^[1]

Provoked versus unprovoked — the first fork

A seizure is the transient occurrence of signs or symptoms produced by abnormal excessive or synchronous neuronal activity. A provoked (acute symptomatic) seizure sits in close temporal relationship with a transient systemic or brain insult — hypoglycaemia, hyponatraemia, alcohol withdrawal, fever, head trauma, CNS infection, a high-dose stimulant — and does not, by itself, constitute epilepsy. An unprovoked seizure has no immediate precipitant.^[1]

The fork decides management. A provoked seizure — a convulsion in a hypoglycaemic, hyponatraemic or alcohol-withdrawing patient, or a febrile seizure in a child — is treated by correcting the precipitant, not by starting an AED. The error to avoid is conflating any convulsion with epilepsy.^[1]

The ILAE definition — "2, or 1-with-60%-over-10-years, or a syndrome"

The ILAE 2014 practical clinical definition is met by any one of three criteria:^[1]

• At least two unprovoked seizures more than 24 hours apart.
• A single unprovoked seizure with at least a 60 percent probability of further seizures over the next ten years — for example, after a single seizure with a structural lesion on imaging and epileptiform EEG, after a stroke, or after certain remote insults such as severe traumatic brain injury.
• The diagnosis of a recognised epilepsy syndrome.^[1]

The definition is deliberately operational — it guides the decision to treat, not an absolute pronouncement about identity. After a single truly unprovoked seizure with low recurrence risk, starting an AED is not mandatory; after a single seizure with high recurrence risk, or after two seizures, treatment is the norm.^[1]

ILAE 2017 — classify before you prescribe

Seizure classification is not an academic exercise — it dictates AED choice, prognosis, and the imaging and genetic work-up. The ILAE 2017 operational classification divides seizures by where in the brain they begin and by whether awareness is preserved.^[2]

Enlarge

^[2]

The 2017 framework also classifies the epilepsies themselves along three axes — seizure type, aetiology, and syndrome.^[3]

Enlarge

The six aetiologies — SAME-GI

Aetiology matters because structural and metabolic causes may be treatable, genetic causes carry counselling implications, and immune and infectious causes direct the work-up.^[3]

^[3]

How common, how lethal — and SUDEP

Epilepsy is one of the commonest serious neurological disorders, affecting roughly 1 in 100 to 1 in 200 people at any time — a lifetime prevalence around 1 percent and an incidence of about 50 to 120 per 100,000 per year. The age distribution is bimodal: a peak in early childhood (genetic and perinatal causes, febrile seizures, cerebral palsies) and a larger peak in older adults (stroke and neurodegeneration), with a trough in young adults.^[12]

^[12]

Sudden unexpected death in epilepsy (SUDEP) is the leading cause of epilepsy-related premature mortality in adults, with an incidence of roughly 1 per 1,000 adults per year overall — and much higher (up to 1 in 100 per year) in those with frequent generalised tonic-clonic seizures, nocturnal seizures, drug-resistant epilepsy, intellectual disability, or non-adherence. The mechanism is likely postictal central apnoea and cardiac arrhythmia after a generalised tonic-clonic seizure. The single most effective risk-reduction strategy is seizure freedom through adherence.^[12]

Why a seizure happens — excitation versus inhibition

A seizure is fundamentally an imbalance between excitation and inhibition in cortical (and, for generalised seizures, thalamocortical) networks, tipping into abnormal, hypersynchronous neuronal firing. The two molecular players are glutamate — the dominant cortical excitatory neurotransmitter, acting on AMPA, kainate and NMDA ionotropic receptors — and GABA, the dominant inhibitory neurotransmitter, acting on GABA-A (chloride-channel, fast inhibition) and GABA-B (G-protein-coupled, slow inhibition) receptors.^[2]

In a focal seizure, a localised group of neurons (the epileptic focus) loses its normal restraint and fires in synchronised bursts. The cellular signature is the paroxysmal depolarising shift — a giant excitatory postsynaptic potential mediated by glutamatergic input through NMDA and AMPA channels. The discharge may stay localised, spread into adjacent cortex producing a Jacksonian march, or recruit the wider network and evolve into a focal-to-bilateral tonic-clonic seizure.^[2]

In a generalised seizure from onset, the abnormal activity engages bilaterally distributed thalamocortical networks essentially simultaneously. The classic example is typical absence epilepsy, where a circuit between cortical pyramidal neurons and thalamic relay neurons firing through T-type calcium channels generates the 3-Hz spike-and-wave on EEG and the brief blank-stare. T-type calcium channels are the molecular target of ethosuximide — a piece of mechanism that explains a piece of pharmacology.^[2]

Enlarge

The AED targets map onto the mechanism

Remembering which drug hits which target earns examiner credit.^[1]

^[1]

The postictal state — confusion, headache, sleepiness, sometimes focal deficit — reflects transient neuronal exhaustion, ion-gradient depletion and local neurotransmitter depletion. Postictal (Todd's) paralysis is a transient focal weakness lasting hours (occasionally a day or two) after a seizure, caused by transient cortical dysfunction in the motor strip. It is clinically indistinguishable from an acute stroke at first encounter — a witnessed seizure or a history of epilepsy is the discriminator. The mechanism is not infarction; the deficit resolves completely.^[1]

The semiology — what the seizure looks like

The semiology is the single most diagnostic piece of information, and a structured eyewitness account is the cornerstone of diagnosis.^[1]

Focal aware seizures (formerly simple partial) produce symptoms the patient remains conscious of and can describe — often an aura, which is in fact a small focal seizure: a rising epigastric sensation, déjà vu, an olfactory hallucination (classic for temporal lobe — the burnt-rubber smell of an uncinate fit), a visual distortion, or unprovoked fear. Motor focal seizures produce clonic movements that may march across a limb (the Jacksonian march, from hand to arm to face, reflecting orderly spread across the motor homunculus).^[1]

Focal impaired-awareness seizures (formerly complex partial) — typically temporal or frontal — produce an alteration of awareness with automatisms: lip-smacking, chewing, fumbling with buttons, picking at clothes, wandering, verbal automatisms. Frontal-lobe seizures tend to be brief, nocturnal, with bizarre hypermotor movements easily mistaken for a sleep disorder or a psychogenic episode.^[1]

Generalised tonic-clonic seizures follow a stereotyped four-phase sequence. A brief aura or warning (if focal onset with secondary generalisation) precedes the tonic phase — sudden loss of consciousness, a forced expiratory cry as thoracic muscles contract, falling, and tonic stiffening for 30 to 60 seconds, often with central cyanosis. The clonic phase follows: rhythmic symmetrical jerking of the limbs for one to two minutes, with the jerks slowing as the seizure ends. The postictal phase — flaccidity, deep unconsciousness, irregular breathing, tongue biting (typically the lateral border), and sometimes urinary or faecal incontinence — gives way to confusion, headache and sleep.^[1]

^[1]

Typical absence seizures (childhood absence epilepsy, age 4 to 12) are striking: a brief (5 to 10 seconds), abrupt-onset, abrupt-offset blank stare with arrest of activity, sometimes with subtle eyelid flickering, no aura, no fall, no postictal confusion. The child resumes the interrupted activity as if nothing happened and may have dozens per day, mistaken for day-dreaming at school. The EEG signature is pathognomonic: 3-Hz spike-and-wave discharges, generalised, provoked by hyperventilation.^[1]

Juvenile myoclonic epilepsy (JME) is the examinable classic of adolescence — onset 12 to 18 years with early-morning myoclonic jerks of the arms (the patient drops the breakfast mug), infrequent generalised tonic-clonic seizures also on waking, and often photosensitivity. EEG shows polyspike-and-wave discharges. Valproate is first-line and the response is excellent — but JME is lifelong and seizures recur in the great majority if the AED is withdrawn, even after years of freedom.^[1]

Febrile seizures — provoked, not epilepsy

Febrile seizures occur in children 6 months to 5 years. A simple febrile seizure is brief (under 15 minutes), generalised, with fever, no CNS infection, occurring once in 24 hours. A complex febrile seizure has at least one of: duration 15 minutes or more, focal features, or recurrence within 24 hours. Simple febrile seizures carry only a slightly elevated risk of later epilepsy; complex febrile seizures, febrile status epilepticus, a family history of epilepsy, and pre-existing neurodevelopmental abnormality all predict a higher risk. Febrile seizures are provoked — they are not epilepsy and are not usually treated with chronic AEDs.^[11]

The mimics — not every paroxysm is a seizure

The epilepsy clinic is filled with patients mislabelled as having epilepsy who in fact have syncope, a movement disorder, or a psychiatric mimic. The clinical cost of misdiagnosis is high — patients are exposed to teratogenic and side-effect-laden drugs unnecessarily, deprived of driving licences, and stigmatised.^[2]

^[2]

The decisive questions an examiner expects you to ask: Was there a prodrome? (syncope, migraine yes; seizure often no). What was the colour? (pallor and sweating — syncope). Was consciousness truly lost, and for how long? (syncope seconds, seizure minutes, PNES variable). Was there postictal confusion, tongue bite, incontinence? (seizure yes; syncope and PNES no). Did it occur on waking in the morning? (JME). Was it triggered by exertion, fright, or noise? (long-QT).^[1]

The single most useful question — ask the eyewitness

Investigations — EEG, MRI, bloods, and the ECG you must not forget

The work-up of a first seizure or new-onset epilepsy has three purposes: to confirm the event was epileptic, to classify the seizure type and syndrome, and to find the cause.^[1]

Electroencephalography (EEG) is the first-line investigation to classify seizures. An interictal EEG detects epileptiform discharges in roughly half of patients on a single recording, rising to over 90 percent with repeated recordings and provocative manoeuvres — sleep deprivation, hyperventilation, photic stimulation. A single normal EEG never excludes epilepsy. Video-EEG telemetry (prolonged inpatient recording with concurrent video) is the gold standard for definitive classification and for distinguishing epileptic seizures from PNES.^[1]

Brain imaging is required for any adult with new-onset focal epilepsy, any adult with new-onset seizures at all, and any child with focal features or an abnormal examination. MRI brain with a dedicated 3-Tesla epilepsy protocol (thin coronal slices through the hippocampus, FLAIR and T2) is the modality of choice — it detects mesial temporal sclerosis, focal cortical dysplasia, low-grade tumours (ganglioglioma, DNET), vascular malformations, and post-traumatic or post-stroke scarring. A non-contrast CT brain is the modality of choice in the acute or emergency setting.^[1]

Blood tests on a first seizure or in the ED fitting patient: capillary glucose (always, first), sodium, potassium, calcium, magnesium, urea and creatinine, liver function, full blood count, toxicology (especially in the young adult), AED levels if on treatment, inflammatory markers and blood cultures if febrile. Prolactin, drawn within 10 to 20 minutes of a suspected generalised tonic-clonic seizure, may be modestly elevated; it is not reliable enough to be routine, but a normal prolactin after a convulsive event of uncertain nature supports PNES over epilepsy.^[1]

Electrocardiography (ECG) is mandatory and the single most under-performed test in this setting. Long-QT syndrome and other arrhythmias can produce syncope with secondary anoxic convulsive movements and be mislabelled as epilepsy for years. An ECG reading QTc over 440 ms in men or 460 ms in women warrants cardiology assessment.^[1]

Lumbar puncture is indicated when CNS infection is suspected — a febrile fitting patient, febrile status epilepticus, or the immunocompromised patient with new seizures — and only after neuroimaging has excluded raised intracranial pressure or a mass. Autoimmune encephalitis antibody panels on CSF and serum are indicated when limbic encephalitis is suspected.^[1]

Enlarge

Syndrome-specific AED choice — the heart of the topic

Definitive management rests on antiepileptic drugs (AEDs) and, for the drug-resistant minority, surgery, neuromodulation or diet. The guiding principle is monotherapy, syndrome-appropriate, lowest effective dose, titrated to response. The single most important step is matching the drug to the syndrome.^[1]

Focal epilepsy: consider lamotrigine or levetiracetam as first-line monotherapy.^[15] SANAD-II failed to show levetiracetam non-inferior to lamotrigine for 12-month remission and concluded that lamotrigine should remain first-line; the original SANAD found lamotrigine better than carbamazepine for time to treatment failure.^[4][5] If first-line monotherapy fails, carbamazepine, oxcarbazepine or zonisamide are the second-line monotherapy options.^[15] In women of childbearing potential, lamotrigine or levetiracetam are preferred over enzyme-inducing drugs and over valproate.^[15]

Idiopathic (genetic) generalised epilepsy: valproate was the most effective drug in SANAD arm B, beating lamotrigine and topiramate on time to treatment failure in idiopathic generalised epilepsy.^[14] It is also the best documented teratogen: major malformations occurred in 9.2 percent of valproate-exposed pregnancies versus 2.1 percent with lamotrigine, and fetal exposure lowered age-6 IQ in a dose-dependent fashion (NEAD: mean IQ 97 vs 105 to 108 on comparators).^[17][7] Avoid valproate in women and girls of childbearing potential unless other options are unsuitable and a Pregnancy Prevention Programme is in place; there, lamotrigine or levetiracetam are the usual alternatives.^[16][15] JME usually needs long-term therapy: in one long-term cohort, about half the seizure-free patients who stopped drugs relapsed.^[18]

Childhood absence epilepsy first-line: ethosuximide. At 16 weeks, freedom from treatment failure was 53 percent with ethosuximide and 58 percent with valproate versus 29 percent with lamotrigine, and attentional dysfunction affected 33 percent on ethosuximide versus 49 percent on valproate.^[19] Twelve-month follow-up confirmed ethosuximide as the optimal initial monotherapy, titrated to response up to a maximum of 60 mg/kg/day or 2000 mg daily.^[20] The mechanism matches the EEG: ethosuximide reduces thalamic T-type calcium currents, the generator of the 3-Hz spike-and-wave.^[22] It treats absence only, so a child who also has generalised tonic-clonic seizures is moved to valproate or lamotrigine, which cover both.^[15] Most children remit: complete remission was reached in about three-quarters of a community cohort treated initially with ethosuximide.^[21]

The wrong-drug trap — make it a named pitfall

Key AED adverse effects and pitfalls

^[1]

The status epilepticus ladder — when a seizure will not stop

A seizure becomes status epilepticus — a continuous neurological emergency — when it lasts 5 minutes or more, or when recurrent seizures occur without recovery in between. Most self-limiting seizures stop within 2 minutes; seizures that have not stopped by 5 minutes are statistically unlikely to stop spontaneously, and benzodiazepine receptors internalise over time, making the drugs less effective the longer treatment is delayed. Status is a clinical diagnosis at the bedside — do not wait for EEG.^[1]

^{[8] [9] [8] [9]}

The ESETT trial established that levetiracetam, fosphenytoin and valproate were equally effective as second-line agents for established convulsive status epilepticus, ending the default use of phenytoin.^[6] Pre-hospital, the RAMPART trial showed IM midazolam at least as effective as IV lorazepam when IV access is unavailable.^[10]

Drug-resistant epilepsy — refer early, surgery can cure

Drug-resistant epilepsy is failure of two appropriate and tolerated AEDs to achieve seizure freedom, and affects roughly 30 percent of people with epilepsy. Refer early to a specialist epilepsy centre for surgical evaluation: temporal lobectomy for drug-resistant mesial temporal lobe epilepsy, lesionectomy for a resectable structural lesion, corpus callosotomy for drop attacks, hemispherectomy for catastrophic paediatric syndromes, and — for non-resectable cases — vagus nerve stimulation (VNS), deep brain stimulation, responsive neurostimulation (RNS), or the ketogenic diet.^[1]

Wiebe's landmark randomised trial (2001) showed anterior temporal lobectomy far superior to continued medical therapy in drug-resistant temporal lobe epilepsy — about 64 percent seizure-free versus 8 percent. Surgery is the standard of care for the well-selected drug-resistant patient, and delaying the referral is the recurring error.^[13]

Women with epilepsy — the whole reproductive journey

^{[16] [29] [1]}

Withdrawing an AED — slow, supervised, never abrupt

Never stop an AED abruptly — withdrawal precipitates withdrawal seizures and status epilepticus, the very outcome treatment is meant to prevent. If a drug is to be stopped (side-effects, allergy, pregnancy planning, long-term remission), taper it slowly over weeks to months under supervision, one drug at a time. A patient seizure-free for years must not simply stop because they feel well.^[1]

Withdrawal can be considered in a patient seizure-free for at least 2 years, with a normal neurological examination and ideally a normal EEG, in a low-risk syndrome, after counselling about recurrence risk (roughly 40 percent at 2 years after withdrawal) and the implications for driving (the practical issue that most often stops a patient withdrawing). JME, structural epilepsy and symptomatic causes generally argue against withdrawal.^[1]

Driving and lifestyle — the questions patients actually ask

Driving (region-specific): in India the RTO typically requires the patient to be seizure-free at least 1 year for a private licence, with regional variation. In the UK, group 1 driving requires at least 1 year seizure-free or 1 year of sleep-only seizures; group 2 (HGV or PCV) requires 10 years seizure-free off all AEDs. A patient with exclusively nocturnal seizures for at least 1 year may retain a private (group 1) licence in the UK — a specific, examinable exception.^[1]

Lifestyle counselling: bathing or showering (use a shower, not a bath, to avoid drowning; leave the door unlocked); heights and water (no unsupervised swimming, no working at heights); cooking (use the back hobs, avoid carrying hot liquids); alcohol (moderation — heavy intake and withdrawal both lower seizure threshold); and photosensitivity (for JME, avoid flickering lights and screen exposure in those affected).^[1]

How epilepsy patients come to harm (the preventable list)

• Prescribing carbamazepine or phenytoin for absence or myoclonus — these drugs can worsen those seizures^[23]
• Giving valproate to a woman of childbearing potential unless other options are unsuitable and a Pregnancy Prevention Programme is in place^[16]
• Missing long-QT masquerading as epilepsy — misdiagnosis delayed correct diagnosis by over a decade and raised sudden arrest risk in one series^[24]
• Misdiagnosing PNES as refractory epilepsy — diagnostic delay runs to years, and escalating antiseizure drugs adds harm^[25]
• Stopping an AED abruptly — withdrawal roughly doubled two-year seizure risk in the MRC withdrawal study re-analysis (43 vs 21 percent)^[26]
• Failing to refer drug-resistant epilepsy early for surgical evaluation — surgery is superior and referral is often delayed^[13][27]
• Not discussing SUDEP — risk climbs with tonic-clonic seizure frequency, and adherence plus night-time supervision reduce it^[28][15]
• Omitting high-dose folic acid before conception — starting 4 to 5 mg daily in the 12 weeks before pregnancy cut major anomaly risk by about 45 percent^[29]

Prognosis and disposition

About 70 percent of people with epilepsy achieve seizure freedom on the first or second monotherapy AED; the remaining 30 percent are drug-resistant and should be referred for surgical evaluation. Good prognostic features: an identifiable benign syndrome (childhood absence remits), good response to the first AED, no structural lesion, normal imaging and EEG after treatment. Poor prognostic features: a structural cause (mesial temporal sclerosis, tumour, malformation), multiple seizure types, intellectual disability, abnormal neurological examination, psychiatric comorbidity, and non-adherence.^[1]

The named trials and guidelines — one-line takeaways

ILAE 2017 classification of seizure types and of the epilepsies is the international framework guiding AED choice, imaging, genetic work-up and surgical evaluation; it replaced the older 1981 and 1989 classifications.^[2][3] ILAE 2014 practical clinical definition operationalised the diagnosis to guide treatment decisions.^[1]

SANAD (2007) established lamotrigine as preferred first-line for focal epilepsy and valproate as first-line for generalised epilepsy.^[4] SANAD-II (2021) failed to show levetiracetam non-inferior to lamotrigine for 12-month remission, so lamotrigine remains first-line.^[5]

ESETT (2019) randomised established convulsive status to levetiracetam, fosphenytoin or valproate and found all three equally effective — the basis of the modern second-line choice.^[6] RAMPART (2012) established IM midazolam for the prehospital setting.^[10]

The NEAD study (Lancet Neurology 2013) prospectively followed children exposed in utero to AEDs and showed valproate lowered offspring IQ in a dose-dependent fashion (roughly 6 to 9 points) — the evidence underpinning the modern restrictions on valproate in pregnancy.^[7]

Wiebe et al. (NEJM 2001) showed anterior temporal lobectomy far superior to continued medical therapy in drug-resistant temporal lobe epilepsy (about 64 percent seizure-free vs 8 percent), establishing surgery as the standard of care for the well-selected patient.^[13]

^[1]

The mantra, and the wrong-drug mnemonic

^[1]

The mantra: classify first, prescribe second — and never carbamazepine for absence.^[2]

Ward-round test — three stems, thirty seconds each

Stem 1 — the girl from the top of the topic (answer)ShowHide

The 16-year-old with dropped mugs, early-morning arm jerks, one generalised tonic-clonic seizure on waking, and photosensitivity. What is the syndrome, the first-line drug, and the trap? Model: Juvenile myoclonic epilepsy (JME) — onset 12 to 18, early-morning myoclonus, GTCS on waking, photosensitivity, polyspike-and-wave on EEG. First-line is valproate (most effective; levetiracetam if she is of childbearing potential and the PPP cannot be met). The traps: JME is lifelong, so do not withdraw the AED even after years of freedom; and carbamazepine or phenytoin can worsen the myoclonus — never prescribe them. Counsel sleep hygiene (sleep deprivation is a trigger) and photosensitivity, and discuss SUDEP.^[1]

Stem 2 — the 8-year-old 'daydreamer' (answer)ShowHide

A teacher reports a child who has dozens of brief blank spells a day, each lasting seconds, with no fall and immediate resumption of activity. EEG shows 3-Hz spike-and-wave provoked by hyperventilation. What is the syndrome and the drug of choice? Model: Childhood absence epilepsy. The drug of choice is ethosuximide, because it blocks the thalamic T-type calcium channels that generate the 3-Hz spike-and-wave. If the child also has generalised tonic-clonic seizures, use valproate instead (ethosuximide treats absence only). The syndrome typically remits by adolescence. The wrong-drug trap applies — carbamazepine would worsen it.^[1]

Stem 3 — the 'refractory' patient on three drugs (answer)ShowHide

A 45-year-old with 'refractory focal epilepsy' is on carbamazepine, lamotrigine and levetiracetam, still having weekly impaired-awareness seizures with an epigastric aura and automatisms. MRI shows left hippocampal sclerosis. What have you missed? Model: This is drug-resistant mesial temporal lobe epilepsy — failure of two appropriate AEDs. Refer to a specialist epilepsy centre for surgical evaluation: concordant EEG and MRI, neuropsychology, intracranial EEG if needed, and anterior temporal lobectomy if the data localise to one hemisphere. Wiebe's trial showed roughly 64 percent seizure-free after surgery versus 8 percent on continued medical therapy. The error was treating a surgically curable syndrome as a pharmacological problem.^[13]

References29ShowHide

1. [1]Fisher RS, Acevedo C, Arzimanoglou A, et al. ILAE official report: a practical clinical definition of epilepsy Epilepsia, 2014.PMID 24730690
2. [2]Fisher RS, Cross JH, French JA, et al. Operational classification of seizure types by the International League Against Epilepsy: Position Paper of the ILAE Commission for Classification and Terminology Epilepsia, 2017.PMID 28276060
3. [3]Scheffer IE, Berkovic S, Capovilla G, et al. ILAE classification of the epilepsies: Position paper of the ILAE Commission for Classification and Terminology Epilepsia, 2017.PMID 28276062
4. [4]Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of carbamazepine, gabapentin, lamotrigine, oxcarbazepine, or topiramate for treatment of partial epilepsy: an unblinded randomised controlled trial Lancet, 2007.PMID 17382827
5. [5]Marson A, Burnside G, Appleton R, et al. The SANAD II study of the effectiveness and cost-effectiveness of levetiracetam, zonisamide, or lamotrigine for newly diagnosed focal epilepsy: an open-label, non-inferiority, multicentre, phase 4, randomised controlled trial Lancet, 2021.PMID 33838757
6. [6]Kapur J, Elm J, Chamberlain JM, et al. Randomized Trial of Three Anticonvulsant Medications for Status Epilepticus N Engl J Med, 2019.PMID 31774955
7. [7]Meador KJ, Baker GA, Browning N, et al. Fetal antiepileptic drug exposure and cognitive outcomes at age 6 years (NEAD study): a prospective observational study Lancet Neurol, 2013.PMID 23352199
8. [8]Trinka E, Cock H, Hesdorffer D, et al. A definition and classification of status epilepticus--Report of the ILAE Task Force on Classification of Status Epilepticus Epilepsia, 2015.PMID 26336950
9. [9]Glauser T, Shinnar S, Gloss D, et al. Evidence-Based Guideline: Treatment of Convulsive Status Epilepticus in Children and Adults: Report of the Guideline Committee of the American Epilepsy Society Epilepsy Curr, 2016.PMID 26900382
10. [10]Silbergleit R, Durkalski V, Lowenstein D, et al. Intramuscular versus intravenous therapy for prehospital status epilepticus N Engl J Med, 2012.PMID 22335736
11. [11]Steering Committee on Quality Improvement and Management, Subcommittee on Febrile Seizures, American Academy of Pediatrics. Febrile seizures: clinical practice guideline for the long-term management of the child with simple febrile seizures Pediatrics, 2008.PMID 18519501
12. [12]Thurman DJ, Logroscino G, Beghi E, et al. The burden of premature mortality of epilepsy in high-income countries: A systematic review from the Mortality Task Force of the International League Against Epilepsy Epilepsia, 2017.PMID 27888514
13. [13]Wiebe S, Blume WT, Girvin JP, Eliasziw M. A randomized, controlled trial of surgery for temporal-lobe epilepsy N Engl J Med, 2001.PMID 11484687
14. [14]Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of valproate, lamotrigine, or topiramate for generalised and unclassifiable epilepsy: an unblinded randomised controlled trial Lancet, 2007.PMID 17382828
15. [15]Gonzalez-Viana E, Sen A, Bonnon A, Cross JH. Epilepsies in children, young people, and adults: summary of updated NICE guidance BMJ, 2022.PMID 35790233
16. [16]Pack AM, Oskoui M, Williams Roberson S, et al. Teratogenesis, perinatal, and neurodevelopmental outcomes after in utero exposure to antiseizure medication: practice guideline from the AAN, AES, and SMFM Neurology, 2024.PMID 38748979
17. [17]Hernandez-Diaz S, Quinn M, Conant S, et al. Use of antiseizure medications early in pregnancy and the risk of major malformations in the newborn Neurology, 2025.PMID 40669027
18. [18]Pietrafusa N, La Neve A, de Palma L, et al. Juvenile myoclonic epilepsy: long-term prognosis and risk factors Brain Dev, 2021.PMID 33781581
19. [19]Glauser TA, Cnaan A, Shinnar S, et al. Ethosuximide, valproic acid, and lamotrigine in childhood absence epilepsy N Engl J Med, 2010.PMID 20200383
20. [20]Glauser TA, Cnaan A, Shinnar S, et al. Ethosuximide, valproic acid, and lamotrigine in childhood absence epilepsy: initial monotherapy outcomes at 12 months Epilepsia, 2013.PMID 23167925
21. [21]Berg AT, Levy SR, Testa FM, Blumenfeld H. Long-term seizure remission in childhood absence epilepsy: might initial treatment matter? Epilepsia, 2014.PMID 24512528
22. [22]Broicher T, Seidenbecher T, Meuth P, et al. T-current related effects of antiepileptic drugs and a Ca2+ channel antagonist on thalamic relay and local circuit interneurons in a rat model of absence epilepsy Neuropharmacology, 2007.PMID 17675191
23. [23]Guerrini R, Belmonte A, Genton P. Antiepileptic drug-induced worsening of seizures in children Epilepsia, 1998.PMID 9593229
24. [24]Ramos-Maqueda J, Bermúdez-Jiménez F, Ruiz RM, et al. Prognostic impact of misdiagnosis of cardiac channelopathies as epilepsy PLoS One, 2020.PMID 32298319
25. [25]Vilyte G, Ives-Deliperi V, Butler J, Pretorius C. Diagnostic delay and associated clinical factors in a sample of South African patients with functional seizures Epilepsy Behav, 2026.PMID 41380245
26. [26]Terman SW, Wang C, Wang L, et al. Reappraisal of the Medical Research Council Antiepileptic Drug Withdrawal Study: contamination-adjusted and dose-response re-analysis Epilepsia, 2022.PMID 35490396
27. [27]Jehi L, Jette N, Kwon CS, et al. Timing of referral to evaluate for epilepsy surgery: Expert Consensus Recommendations from the Surgical Therapies Commission of the International League Against Epilepsy Epilepsia, 2022.PMID 35842919
28. [28]Sveinsson O, et al. Type, etiology, and duration of epilepsy as risk factors for SUDEP: further analyses of a population-based case-control study Neurology, 2023.PMID 37813583
29. [29]Sun Y, Alvestad S, Cohen JM, et al. Timing of high-dose folic acid supplementation in the periconceptional period among women taking antiseizure medications and risk of major congenital anomalies: a target trial emulation J Neurol Neurosurg Psychiatry, 2026.PMID 41876129