`
`clinical practice
`
`Initial Management of Epilepsy
`Jacqueline A. French, M.D., and Timothy A. Pedley, M.D.
`
`This Journal feature begins with a case vignette highlighting a common clinical problem.
`Evidence supporting various strategies is then presented, followed by a review of formal guidelines,
`when they exist. The article ends with the authors’ clinical recommendations.
`
`A 29-year-old woman presents for evaluation. The previous evening, her husband,
`who was in the next room, heard unusual sounds and found her lying on the bed look-
`ing dazed. She was confused for a few minutes but quickly returned to normal. On
`questioning, she recalls an unwitnessed event about 1 month previously; at that time,
`she awoke feeling mildly confused, had sore muscles, and discovered she had bitten
`her tongue. How should she be evaluated and treated?
`
`The clinic a l problem
`
`From the Department of Neurology, New
`York University School of Medicine (J.A.F.);
`and the Department of Neurology, Co-
`lumbia University Medical Center (T.P.) —
`both in New York. Address reprint requests
`to Dr. French at the NYU Comprehensive
`Epilepsy Center, 403 E. 34th St., 4th Fl.,
`New York, NY 10016, or at jacqueline.
`french@nyumc.org.
`
`N Engl J Med 2008;359:166-76.
`Copyright © 2008 Massachusetts Medical Society.
`
`Epilepsy, which is defined as two or more seizures that are not provoked by other
`illnesses or circumstances, affects about 45 million people worldwide. In the United
`States, the prevalence of epilepsy is approximately 6 to 8 per 1000 population, and
`the incidence is approximately 26 to 40 per 100,000 person-years, with higher rates
`among infants and persons older than 60 years of age.1-3 Approximately 70% of
`adults with new-onset epilepsy have partial (focal) seizures.3 In the majority of cases
`(62%), the cause is unknown. Stroke (9.0%), head trauma (9.0%), alcohol (6.0%),
`neurodegenerative disease (4.0%), static encephalopathy (3.5%), brain tumors (3.0%),
`and infection (2.0%) account for most remaining cases.4 Although cerebrovascular
`causes are more common in the elderly, the cause is still unknown in 25 to 40% of
`patients who are 65 years of age or older.5
`
`S tr ategies a nd Ev idence
`
`Diagnosis
`The transient occurrence of altered awareness, abnormal behavior, or involuntary
`movements suggests a diagnosis of epilepsy. Because epileptic seizures are rarely ob-
`served by a physician, the diagnosis is typically based on historical information
`supplemented by selected tests. The first step is to answer the question of whether the
`event was a seizure. The second is to determine whether the patient has epilepsy.
`A careful history is the single most important element in diagnosis, with a focus
`on details of the episode and whether there is any history of previous spells that
`may point to a diagnosis of epilepsy. When patients have limited or no recall of
`events, witnesses should be queried about details of the episode. The differential di-
`agnosis varies according to the patient’s age and symptoms (Table 1).
`Seizures are common in metabolic (e.g., uremia, hypoglycemia, hyperglycemia,
`and hepatic failure), toxic (e.g., drug overdose or withdrawal), and infectious (e.g.,
`meningitis and encephalitis) conditions.6 Seizures that occur in patients with these
`conditions do not necessarily confer a diagnosis of epilepsy. Although antiepileptic
`drugs are sometimes necessary to suppress seizures in the short term in these con-
`ditions, medications generally do not need to be continued after the patient has
`recovered.
`
`166
`
`n engl j med 359;2 www.nejm.org
`
`july 10, 2008
`
`The New England Journal of Medicine
`
` Copyright © 2008 Massachusetts Medical Society. All rights reserved.
`
`Neurelis - EX. 2019
`Aquestive Therapeutics, Inv. v. Neurelis, Inc. - IPR2019-00451
`
`
`
`clinical practice
`
`Table 1. Conditions That Can Mimic Epileptic Seizures.
`
`Diagnosis
`
`Hyperventilation
`
`Migraine
`
`Panic attack
`
`Psychogenic seizures
`
`Syncope
`
`Transient global amnesia
`
`Transient ischemic attack
`
`Important Clinical Features
`
`Anxiety and overbreathing evident; often perioral cyanosis, hand paresthesias, and
`carpo pedal spasm are present; environmental trigger may be evident
`
`Slow progression of neurologic symptoms; visual symptoms prominent; basilar migraine
`has unusual features, including confusion, stupor, bilateral blindness; headache
`may be minimal or absent
`
`Abrupt onset with intense feeling of dread or fear; often sense of impending death or
` inability to breathe; prominent autonomic features (e.g., tachycardia, sweating,
` nausea); lasts longer (5–30 min) than typical seizure; no loss of consciousness
`
`Psychiatric history; patient usually motionless with eyes closed at onset; fluttering eye
`movements and forceful eye closure common; out-of-phase, thrashing limb move-
`ments and pelvic thrusting common; urinary incontinence unusual; refractory to
`treatment
`
`Precipitating circumstances usually identifiable; prodrome of wooziness but no aura or
`unilateral symptoms; loss of consciousness brief (<20 sec), with rapid return to nor-
`mal; a few muscle jerks (“convulsive syncope”) can occur at end because of hypoxia
`
`Isolated amnesic syndrome; prolonged duration (several hours); no alteration of con-
`sciousness; no confusion, weakness, or aphasia; persistent memory gap during
` period of attack; recurrence unusual
`
`Sudden onset without progression of symptoms; variable symptoms related to brain
`and vascular anatomy; negative features (e.g., weakness, loss of sensation, aphasia)
`predominate
`
`Evaluation
`The neurologic examination is normal in most pa-
`tients with epilepsy. Findings occasionally point
`to an underlying pathologic condition in the brain
`or a specific disorder such as skin abnormalities
`in neurocutaneous syndromes. According to joint
`recommendations of the American Academy of
`Neurology and the American Epilepsy Society, pa-
`tients with an unprovoked first seizure7 should un-
`dergo electroencephalography (EEG), computed
`tomographic (CT) scanning or magnetic resonance
`imaging (MRI) of the head, and selected blood tests
`according to the clinical circumstances. Epilepti-
`form EEG patterns such as spikes and sharp waves
`can assist in the diagnosis and in classifying sei-
`zures as being either focal or generalized. How-
`ever, neither a normal EEG nor interictal abnor-
`malities alone refute or confirm a diagnosis of
`epilepsy. EEGs are abnormal in about 50% of pa-
`tients presenting with a first seizure, and they show
`epileptiform discharges in only about half of these
`patients.7 The incidence of abnormalities increas-
`es when EEGs are repeated or performed after the
`patient has undergone sleep deprivation.8 Video
`EEG monitoring is necessary if there is concern
`about nonepileptic events (Table 1).
`MRI of the brain is more sensitive than CT in
`identifying structural lesions causally related to
`
`epilepsy.9 CT, however, is appropriate for emer-
`gency situations. Among patients in whom epi-
`lepsy has been newly diagnosed, CT of the head is
`abnormal in 34 to 56%, and cranial CT findings
`affect management in 9 to 17%.10
`Routine blood tests rarely inform the diagno-
`sis in otherwise healthy patients. However, a com-
`plete blood count, liver-function tests, and mea-
`surement of electrolyte levels are useful before
`antiepileptic drug treatment is initiated, since dos-
`age adjustment may be necessary if hepatic or re-
`nal function is abnormal. Albumin levels should
`be measured before administering highly protein-
`bound drugs such as phenytoin and valproate,
`since the fraction of unbound (active) drug is
`higher in patients with hypoalbuminemia. In ad-
`olescents and adults with unexplained generalized
`seizures, screening for substance abuse should
`be considered.
`A diagnosis of epilepsy can have a considerable
`effect on the patient’s mood, interpersonal rela-
`tionships, employability, social functioning, qual-
`ity of life, and ability to drive. Early and repeated
`discussions of these issues are suggested. Patients
`should be discouraged from participating in ac-
`tivities for which a history of seizures increases
`the risk of injury or death; these activities include
`driving, operating high-risk power equipment,
`
`n engl j med 359;2 www.nejm.org
`
`july 10, 2008
`
`167
`
`The New England Journal of Medicine
`
` Copyright © 2008 Massachusetts Medical Society. All rights reserved.
`
`Neurelis - EX. 2019
`Aquestive Therapeutics, Inv. v. Neurelis, Inc. - IPR2019-00451
`
`
`
`T h e ne w e ngl a nd jou r na l o f m e dic i ne
`
`mg/day
`ity at doses >1200
`reducing bioavailabil-
`tion dose-dependent,
`Gastrointestinal absorp-
`
`tabolism
`induces its own me-
`plete blood count;
`function tests, com-
`Monitor sodium, liver-
`
`failure (all rare)
`multiorgan failure, hepatic
`tients); hypersensitivity,
`drome (in 1–3/1000 pa-
`Rash, Stevens–Johnson syn-
`
`None known
`
`patients)
`tremia (in 1.8–40.0% of
`drome (rare),† hypona-
`Stevens–Johnson syn-
`10% of patients),
`rash (in approximately
`hepatic failure (very rare),
`(in 1/500,000 patients)
`tients), aplastic anemia
`mately 1/200,000 pa-
`
`Agranulocytosis (in approxi-
`
`Dizziness, blurred vision,
`
`insomnia, headache
`
`fects (in children)
`ataxia, behavioral ef-
`ness, mild weight gain,
`
`Sedation, fatigue, dizzi-
`
`penia*
`nausea, benign leuko-
`sedation, weight gain,
`blurred vision, ataxia,
`
`Dizziness, diplopia,
`
`3–14
`
`100–200
`
`12–20
`
`900
`
`300–600 mg; increase daily
`
`each wk‡
`dose by 300–600 mg
`
`Gabapentin (Neurontin)
`
`4–12
`
`μg/ml
`
`400–600
`
`mg/day
`
`3 days
`dose by 200 mg every
`
`200 mg; increase daily
`
`Carbatrol, Tegretol XR)
`Carbamazepine (Tegretol,
`
`300–600 mg; increase daily
`
`each wk
`dose by 300–600 mg
`
`250–500 mg; increase daily
`
`each wk‡§
`dose by 250–500 mg
`
`of 25–50 mg each wk
`creases in the daily dose
`2 wk, followed by in-
`then 50 mg/day for
`25 mg/day for 2 wk,
`
`25 mg; initial monotherapy:
`
`Oxcarbazepine (Trileptal)
`
`Levetiracetam (Keppra,
`
`Keppra IV)
`
`Lamotrigine (Lamictal,
`
`Lamictal XR)
`
`risk of osteopenia
`with long-term use;
`sutism may occur
`gum hypertrophy, hir-
`changes in dose;
`fects after small
`tration with toxic ef-
`es in serum concen-
`lead to rapid increas-
`Nonlinear kinetics may
`
`nent
`olite is active compo-
`10-monohydroxy metab-
`
`like syndrome
`patic failure (rare), lupus-
`2–4/10,000 patients), he-
`ic epidermal necrolysis (in
`Johnson syndrome, or tox-
`phoma, rash, Stevens–
`duction block, pseudolym-
`
`Blood dyscrasias (rare), con-
`
`anaphylaxis (rare)
`liter) (in 2.5% of patients),
`um level, <125 mmol per
`ponatremia (serum sodi-
`per million patients), hy-
`necrolysis (0.5–6.0 cases
`drome or toxic epidermal
`Rash; Stevens–Johnson syn-
`
`iting, headache
`diplopia, nausea, vom-
`Fatigue, dizziness, ataxia,
`
`droxy metabolite)
`
`3–40 (10-monohy-
`
`900–1200
`
`Psychosis (rare)
`
`Fatigue, dizziness, irritabil-
`
`ity, anxiety, asthenia
`
`10–40
`
`1000–2000
`
`iting, confusion
`diplopia, nausea, vom-
`Fatigue, dizziness, ataxia,
`
`10–20
`
`200–300
`
`ing dose‡§
`be initiated with a load-
`tion, 3–5 mg/kg; may
`
`If initiated without titra-
`
`Phenytoin (Dilantin,
`
`Phenytek)
`
`Other Considerations
`
`Serious Side Effects
`
`Common Side Effects
`
`Blood Level
`
`Target Dose
`Typical Initial
`
`Starting Daily Dose
`
`and Titration
`
`Drug
`
`Table 2. Initiation of Antiepileptic Drugs as Initial Monotherapy in the Absence of Special Considerations.
`
`168
`
`n engl j med 359;2 www.nejm.org
`
`july 10, 2008
`
`The New England Journal of Medicine
`
` Copyright © 2008 Massachusetts Medical Society. All rights reserved.
`
`Neurelis - EX. 2019
`Aquestive Therapeutics, Inv. v. Neurelis, Inc. - IPR2019-00451
`
`
`
`§ This drug can be administered intravenously.
`‡ This drug can be initiated at a therapeutic dose.
`† HLA-B*1502 testing is now recommended in patients of Asian descent, since this haplotype is associated with a higher risk of Stevens–Johnson syndrome.26
`* There is usually no need for intervention unless the neutrophil count is below 1000 cells per cubic millimeter.
`
`clinical practice
`
`stroke (rare)
`necrolysis (rare), heat-
`drome or toxic epidermal
`Stevens–Johnson syn-
`rash (in 1–2% of patients),
`(in 0.2–4.0% of patients),
`Aplastic anemia, renal calculi
`
`in children)
`oligohydrosis (mostly
`vomiting, headache,
`weight loss, nausea,
`rritability, anorexia,
`culty concentrating,
`Drowsiness, ataxia, diffi-
`
` i
`
`bocytopenia
`1/3000 patients), throm-
`(rare), pancreatitis (in
`monemia, aplastic anemia
`polytherapy), hyperam-
`among children and with
`patients, higher rate
`
`Hepatic failure (in 1/20,000
`
`(rare), heatstroke
`tients), acute glaucoma
`calculi (in 1.5% of pa-
`(in 3% of patients), renal
`Significant metabolic acidosis
`
`Spike-wave status epilepticus
`
`Schedule V controlled
`
`substance
`
`None reported
`
`tremor, hair loss
`thrombocytopenia,
`gain, nausea, vomiting,
`Drowsiness, ataxia, weight
`
`(mostly in children)
`acidosis, oligohydrosis
`esthesias, metabolic
`orexia, weight loss, par-
`culty concentrating, an-
`slowed speech, diffi-
`finding difficulty and
`Drowsiness, ataxia, word-
`
`ness, weakness
`somnolence, nervous-
`Fatigue, dizziness, ataxia,
`
`edema
`diplopia, weight gain,
`Fatigue, dizziness, ataxia,
`
`10–40
`
`100–200
`
`50 mg; increase daily dose
`
`by 50 mg each wk
`
`Zonisamide (Zonegran)
`
`40–100
`
`750–2000
`
`250–500 mg each wk§
`crease daily dose by
`kg orally once a day; in-
`250–500 mg, or 10–15 mg/
`
`ER)
`(Depakote, Depakote
`divalproex sodium
`(Depakene, Depacon),
`
`Valproate, valproic acid
`
`5–25
`
`100–200
`
`Not useful
`
`16–36
`
`Not established
`
`150–300
`
`wk
`dose by 25–50 mg each
`
`25–50 mg; increase daily
`
`4 mg; increase daily dose
`
`by 4 mg each wk
`
`75–150 mg; increase daily
`
`each wk‡
`dose by 75–150 mg
`
`Topiramate (Topamax)
`
`Tiagabine (Gabitril)
`
`Pregabalin (Lyrica)
`
`necrolysis, arthritis
`drome or toxic epidermal
`Stevens–Johnson syn-
`sias, hepatic failure, rash,
`Generally rare; blood dyscra-
`
`(in children)
`pression, hyperactivity
`iting, confusion, de-
`diplopia, nausea, vom-
`Fatigue, dizziness, ataxia,
`
`15–45
`
`60–120
`
`Phenobarbital (generic only)30 mg; increase daily dose
`
`by 30 mg every 2 wk
`
`n engl j med 359;2 www.nejm.org
`
`july 10, 2008
`
`169
`
`The New England Journal of Medicine
`
` Copyright © 2008 Massachusetts Medical Society. All rights reserved.
`
`Neurelis - EX. 2019
`Aquestive Therapeutics, Inv. v. Neurelis, Inc. - IPR2019-00451
`
`
`
`T h e ne w e ngl a nd jou r na l o f m e dic i ne
`
`tively inexpensive
`phadenopathy; rela-
`acial features, lym-
`sutism, coarsened
`gival hyperplasia, hir-
`increased risk of gin-
`Reduces bone density;
`
` f
`
`netics‡
`lead to nonlinear ki-
`ed; more likely to
`
`Not always well tolerat-
`
`
`
`seizures
`nus, atypical absence
`
`May aggravate myoclo-
`
`May aggravate general-
`
`ized seizures
`
`Risk of hyponatremia
`
`disorder
`known psychiatric
`tion in patients with
`
`Behavioral issues; cau-
`
`Reduced dose required
`
`May aggravate myoclo-
`
`nus
`
`May aggravate myoclo-
`
`nus
`
`Mild weight gain
`
`May aggravate myoclo-
`
`nus
`
`Reduced dose requiredMay aggravate general-
`
`ized seizures
`
`pensive
`lele; relatively inex-
`with HLA-B*1502 al-
`creased risk of rash
`scent because of in-
`patients of Asian de-
`typing suggested in
`Mild weight gain; geno-
`
`sence seizures
`seizures, atypical ab-
`nus, typical absence
`May aggravate myoclo-
`
`Developmental Delay
`
`Patients with
`
`Generalized Epilepsy
`
`or Symptomatic
`
`May aggravate general-
`
`ized seizures
`
`Increased sedation and
`
`hyponatremia
`
`(Myoclonic or Absence)
`Patients with Generalized
`
`Seizures
`
`Elderly Patients
`
`may cause hypotension
`poglycemia in diabetes;
`mask symptoms of hy-
`or arrhythmia; may
`may worsen heart block
`T4 and free T4 levels;
`liver disease; reduces
`caution in patients with
`with other medications;
`lead to interactions
`Enzyme induction may
`
`and free T4 levels
`natremia; reduces T4
`tients at risk for hypo-
`sensitivity and in pa-
`carbamazepine hyper-
`Caution in patients with
`
`with renal impairment
`Reduce dose in patients
`
`disease
`necessary in hepatic
`dose reduction may be
`to antiepileptic drugs;
`known hypersensitivity
`
`Caution in patients with
`
`function
`with impaired renal
`Reduce dose in patients
`
`levels
`reduces T4 and free T4
`creased risk of rash);
`in Asian patients (in-
`for hyponatremia, and
`mias, in patients at risk
`dyscrasias, and arrhyth-
`hepatic disease, blood
`caution in patients with
`with other medications;
`lead to interactions
`Enzyme induction may
`
`Patients with Coexisting
`
`Medical Conditions
`
`taken during pregnancy
`tions) in offspring if
`risk of major malforma-
`hyperplasia, and higher
`canthal folds, digital
`ing hypertelorism, epi-
`sant syndrome (includ-
`risk of fetal anticonvul-
`OCP interaction; increased
`
`Phenytoin
`
`OCP interaction
`
`Oxcarbazepine
`
`and palate in babies
`use; reports of cleft lip
`with concomitant OCP
`crease in pregnancy,
`Levels of lamotrigine de-
`
`Levetiracetam
`
`Lamotrigine
`
`Gabapentin
`
`pregnancy
`spring if taken during
`sant syndrome in off-
`risk of fetal anticonvul-
`OCP interaction; increased
`
`Women†
`
`Carbamazepine
`
`Other Considerations
`
`Patient Population
`
`Drug
`
`Table 3. Factors Affecting the Choice of Antiepileptic Drugs in Specific Patient Populations.*
`
`170
`
`n engl j med 359;2 www.nejm.org
`
`july 10, 2008
`
`The New England Journal of Medicine
`
` Copyright © 2008 Massachusetts Medical Society. All rights reserved.
`
`Neurelis - EX. 2019
`Aquestive Therapeutics, Inv. v. Neurelis, Inc. - IPR2019-00451
`
`
`
`* Dose-related and idiosyncratic reactions are listed only if certain populations are at increased risk for these reactions. OCP denotes oral contraceptive pill or patch, PCO polycystic ovary
`
`‡ When the phenytoin metabolic pathway is saturated, metabolism switches to zero-order kinetics and clearance is substantially reduced; this can lead to toxic levels of the drug.
`† Limited data are available on the use of newer agents (e.g., gabapentin, lamotrigine, levetiracetam, oxcarbazepine, pregabalin, tiagabine, topiramate, and zonisamide) during pregnancy.
`
`syndrome, and T4 thyroxine.
`
`clinical practice
`
`Weight loss; rarely oligo-
`
`thermia
`hidrosis or hyper-
`
`May need to reduce
`
`dose
`
`Weight gain, alopecia
`
`Weight loss; rarely oligo-
`
`thermia
`hydrosis or hyper-
`
`Weight gain, peripheral
`
`edema
`
`tic drug
`expensive antiepilep-
`tivity in children; least
`withdrawn, hyperac-
`of seizures when
`
`May cause exacerbation
`
`Reduced dose may be
`
`required
`
`seizures
`nus, atypical absence
`
`May aggravate myoclo-
`
`May aggravate myo-
`
`clonus
`
`Reduced dose requiredMay aggravate myo-
`
`clonus
`
`Increased risk of behav-
`
`ioral issues
`
`Risk of cognitive dys-
`
`function
`
`patic impairment
`to reduce dose in he-
`impairment; may need
`in patients with renal
`Renal calculi; reduce dose
`
`titis
`ing; can cause pancrea-
`tients at risk for bleed-
`dysfunction and in pa-
`in patients with hepatic
`medications; caution
`actions with other
`lead to possible inter-
`
`Enzyme inhibition may
`
`with renal impairment
`duce dose in patients
`metabolic acidosis; re-
`Renal calculi, may cause
`
`with renal impairment
`Reduce dose in patients
`
`renal impairment
`dose in patients with
`liver disease; reduce
`caution in patients with
`with other medications;
`lead to interactions
`Enzyme induction may
`
`Zonisamide
`
`taken during pregnancy
`function in offspring if
`creased intellectual
`teratogenicity, and de-
`Possible PCO syndrome,
`
`sodium
`acid, divalproex
`Valproate, valproic
`
`OCP interaction at doses
`
`>200 mg
`
`Topiramate
`
`Tiagabine
`
`Pregabalin
`
`OCP interaction; possible
`
`teratogen
`
`Phenobarbital
`
`n engl j med 359;2 www.nejm.org
`
`july 10, 2008
`
`171
`
`The New England Journal of Medicine
`
` Copyright © 2008 Massachusetts Medical Society. All rights reserved.
`
`Neurelis - EX. 2019
`Aquestive Therapeutics, Inv. v. Neurelis, Inc. - IPR2019-00451
`
`
`
`T h e ne w e ngl a nd jou r na l o f m e dic i ne
`
`working at heights, and swimming or bathing
`alone. In most states, patients who are not seizure-
`free are prohibited from driving; the required dura-
`tion of time without seizures varies across states
`and ranges from 3 months to 1 year.11
`As many as 55% of patients with uncontrolled
`seizures are depressed.12,13 Even patients with
`well-controlled seizures have rates of depression
`that are higher than rates among the general
`population, and suicide rates are tripled, with the
`highest rates in the 6 months after diagnosis.14-16
`Patients should be observed for signs of depres-
`sion and queried specifically about their mood,
`with attention to the potential need for psychiat-
`ric referral and treatment. A simple screening tool
`developed specifically for use in people with epi-
`lepsy may facilitate prompt recognition of major
`depression.17,18
`A recent Food and Drug Administration ad-
`visory indicated an increased risk of suicidal
`thoughts among patients who were in the treat-
`ment group in add-on studies of new antiepilep-
`tic drugs.19 Over the 2 to 6 months of treatment
`in the various studies, the absolute risk was 0.43%
`among patients receiving active treatment as com-
`pared with 0.22% among patients in the placebo
`group. These findings provide support for the
`assessment of mood in patients who are starting
`antiepileptic drug therapy.
`
`Pharmacologic Therapy
`Opinion remains divided about treating patients
`who have had only a single seizure, since only
`about 25% of patients will have a recurrence with-
`in 2 years in the absence of factors that predict a
`high probability of recurrence (e.g., epileptiform
`activity detected on an EEG or a known cause such
`as remote major head trauma).20 Even with one or
`more risk factors, the recurrence rate at 2 years is
`no more than 40%. Furthermore, although ran-
`domized trials have shown that treatment reduces
`the risk of seizure recurrence by 30 to 60%, the
`likelihood of being seizure-free at 3 to 5 years
`after a first or second seizure was similar whether
`treatment was started after the first or second sei-
`zure or deferred initially and started only if a sei-
`zure recurred.21 Treatment is almost always justi-
`fied when a diagnosis of epilepsy has been made.
`In the past two decades, nine new antiepilep-
`tic drugs have been marketed, making the choice
`of initial therapy complex. Antiepileptic drugs are
`classified as being either broad-spectrum or nar-
`
`row-spectrum drugs with regard to efficacy against
`different seizure types and epilepsy syndromes.
`Broad-spectrum antiepileptic drugs are particu-
`larly useful because they are reasonable initial
`choices in most adult patients, regardless of the
`type of seizure or syndrome. These drugs include
`valproate, lamotrigine, topiramate, and levetirac-
`etam (the efficacy of which [in generalized sei-
`zures] is supported by randomized, controlled
`trials), and zonisamide (the efficacy of which [in
`generalized seizures] is based on open studies and
`clinical experience). In contrast, narrow-spectrum
`drugs, which include carbamazepine, phenytoin,
`gabapentin, tiagabine, oxcarbazepine, and pregab-
`alin, should be restricted to patients who have lo-
`calization-related (focal) epilepsy with partial and
`secondarily generalized seizures.22 These drugs are
`less effective than broad-spectrum agents in the
`idiopathic generalized epilepsy syndromes (e.g.,
`juvenile myoclonic epilepsy and childhood ab-
`sence epilepsy), and they may even exacerbate
`some seizure types in these patients.23 About half
`of patients in whom epilepsy is new ly diagnosed
`become seizure-free while receiving the first anti-
`epileptic drug. Failure of the first antiepileptic
`drug for reasons other than tolerability increases
`the likelihood of nonresponse to other drugs, but
`nearly two thirds of patients become seizure-free
`after receiving the second or third drug.24
`Head-to-head trials suggest similar efficacy
`among the various antiepileptic drugs against par-
`tial seizures.22 However, a recent large, pragmatic,
`randomized, controlled trial involving patients
`with generalized epilepsy showed valproic acid to
`be more effective than lamotrigine and topira-
`mate. Lamotrigine had almost twice the failure
`rate because of inadequate seizure control, where-
`as topiramate was similarly effective in control-
`ling seizures but had a higher failure rate caused
`by discontinuation because of side effects.25
`The selection of medication should be informed
`by patient characteristics, including sex, age, and
`coexisting conditions that affect the likelihood of
`adverse events. Table 2 provides useful informa-
`tion regarding the initiation of antiepileptic drugs.
`A common practice in patients who present with
`a first seizure has been to provide a loading dose
`of phenytoin in the emergency department. How-
`ever, clinical trials in newly diagnosed epilepsy
`have not shown any advantage associated with
`phenytoin,27 and it is generally preferable to initi-
`ate whichever antiepileptic drug is considered to
`
`172
`
`n engl j med 359;2 www.nejm.org
`
`july 10, 2008
`
`The New England Journal of Medicine
`
` Copyright © 2008 Massachusetts Medical Society. All rights reserved.
`
`Neurelis - EX. 2019
`Aquestive Therapeutics, Inv. v. Neurelis, Inc. - IPR2019-00451
`
`
`
`clinical practice
`
`be most appropriate with regard to other patient
`characteristics. Table 3 lists considerations rele-
`vant to selecting initial therapy in particular pa-
`tient populations.
`
`Adverse Effects
`Table 2 lists dose-related, idiosyncratic, and long-
`term adverse effects of antiepileptic drugs. Loss of
`bone density may occur during treatment with phe-
`nytoin and possibly with other hepatic enzyme–
`inducing antiepileptic drugs such as carbamazepine
`and phenobarbital.28,29 Both men and women who
`take enzyme-inducing drugs should receive sup-
`plemental vitamin D (up to 2000 IU daily) and
`calcium (up to 1200 mg per day), and they should
`have periodic bone-density measurement.29
`
`Choice of Antiepileptic Drugs in Women
`Antiepileptic drugs, and especially valproate, have
`been associated with reproductive endocrine dis-
`orders, most notably features of polycystic ovary
`syndrome (e.g., irregular menstrual cycles, weight
`gain, and hirsutism).30,31 This association appears
`to be related at least in part to the epilepsy itself,32,33
`but in the majority of women, medication seems
`to play the major role.34,35 Observational studies
`have shown clinically important associations be-
`tween the use of valproate, alone or in combination
`with other drugs, and the development of poly-
`cystic ovaries, anovulatory cycles, and hyperan-
`drogenism.34,36,37
`Hepatic enzyme–inducing antiepileptic drugs
`such as phenytoin, carbamazepine, and phenobar-
`bital, as well as topiramate and oxcarbazepine,
`increase the clearance of oral contraceptive pills.
`Thus, women taking these drugs who use oral
`contraceptive pills are advised to use preparations
`containing at least 50 µg of ethinyl estradiol in
`order to reduce the chance of pregnancy.38 How-
`ever, the contraceptive efficacy of higher-dose oral
`contraceptive pills has not been well studied, and
`alternative methods (e.g., barrier contraception)
`should be discussed. The dosage of lamotrigine
`requires adjustment when oral contraceptive pills
`are started or discontinued, because oral contra-
`ceptives enhance the clearance of lamotrigine.38
`Serum concentrations of lamotrigine should be
`followed in this setting and in pregnancy,38 which
`increases the clearance of many antiepileptic
`drugs, but particularly that of lamotrigine.
`Babies born to women with epilepsy have an
`increased rate of malformations; this is believed
`
`to be attributable mostly to antiepileptic drugs.39
`Studies of the effect of specific drugs during preg-
`nancy are hampered by confounding factors such
`as the type and severity of epilepsy and the use
`of more than one agent in many patients. No anti-
`epileptic drug can be considered to be absolutely
`safe. Newer drugs are less well studied, but the
`evidence linking valproate to an increased risk of
`birth defects is most convincing and sufficient
`to advise against its use in women of childbear-
`ing age unless there is no alternative.40 The risk
`of birth defects is likely to be minimized further
`by treating with monotherapy and drug dosages
`as low as possible during pregnancy, although the
`evidence to support these recommendations is
`limited. Retrospective analyses of school-age chil-
`dren have suggested associations between in-
`trauterine exposure to valproate (but not other
`antiepileptic drugs) and lower IQ scores and de-
`velopmental delay41,42; this finding warrants con-
`firmation in prospective studies.
`
`Concurrent Medical Conditions
`Some patients — particularly many elderly pa-
`tients — may be poor candidates for some anti-
`epileptic drugs because of coexisting conditions
`or the use of medications with which a given an-
`tiepileptic drug may interact.
`In patients with hepatic dysfunction, dosing
`adjustments of drugs metabolized by the liver may
`be necessary, although their use is not necessar-
`ily contraindicated; valproate, however, should be
`avoided, since it can increase ammonia levels.
`Many antiepileptic drugs (particularly valproate,
`phenytoin, phenobarbital, and carbamazepine) can
`cause elevations in hepatic enzyme levels, particu-
`larly alanine aminotransferase and γ-glutamyl-
`transferase.43 Stable mild elevations (even up to
`two times the normal range) are not a substan-
`tive concern, but they may complicate monitor-
`ing of patients with underlying hepatic disease.
`A history of renal calculi is a relative contraindi-
`cation to the use of topiramate and zonisamide,
`which can predispose to stone formation.44 Both
`carbamazepine and oxcarbazepine can cause hy-
`ponatremia and should generally be avoided in
`patients with preexisting hyponatremia or risk
`factors for hyponatremia (e.g., older age, history
`of excess water intake, renal failure, or concur-
`rent use of other medications associated with
`hyponatremia).45
`Levels of drugs metabolized by hepatic mi-
`
`n engl j med 359;2 www.nejm.org
`
`july 10, 2008
`
`173
`
`The New England Journal of Medicine
`
` Copyright © 2008 Massachusetts Medical Society. All rights reserved.
`
`Neurelis - EX. 2019
`Aquestive Therapeutics, Inv. v. Neurelis, Inc. - IPR2019-00451
`
`
`
`T h e ne w e ngl a nd jou r na l o f m e dic i ne
`
`Table 4. Common Drugs with Either Increased or Reduced Clearance in the Presence of Antiepileptic Drugs.*
`
`Type of Medication
`
`Cardiac
`
`Psychiatric
`
`Antineoplastic
`
`Antiinfective
`
`Other
`
`Increased Clearance
`(and Need for Higher Doses) with Phenytoin,
`Phenobarbital, Carbamazepine
`
`Decreased Clearance
`(and Need for Lower Doses)
`with Valproic Acid
`
`Mexiletine, quinidine, amiodarone, propranolol, meto-
`prolol, nifedipine, felodipine, nimodipine, digoxin,
`lovastatin, simvastatin, dicoumarol, warfarin
`
`Amitriptyline, nortriptyline, imipramine, desipramine,
`clomipramine, citalopram, paroxetine, buproprion,
`haloperidol, chlorpromazine, clozapine, olanzapine,
`risperidone, quetiapine
`
`Cyclophosphamide, busulfan, etoposide, methotrexate,
`teniposide, some vinca alkaloids
`
`Praziquantel, albendazole, doxycycline, nevirapine,
`efavirenz, delavirdine, indinavir, ritonavir, saquinavir
`
`Cyclosporine, tacrolimus, diazepam, alprazolam,
`prednisone, oral contraceptive pills, theophylline,
`methadone
`
`Nimodipine
`
`Amitriptyline, nortriptyline, clomip-
`ramine, paroxetine
`
`Zidovudine, possibly others
`
`Lorazepam, diazepam
`
`* Data are from Patsalos and Perucca.46 This list is not comprehensive.
`
`crosomal enzymes (e.g., cytochrome P-450 and
`glucuronyl transferases) can be altered consider-
`ably by concurrent antiepileptic drug use (Table 4).
`Enzyme-inducing antiepileptic drugs should be
`avoided, whenever possible, in patients receiving
`antiretroviral therapy for human immunodefi-
`ciency virus infection, in organ-transplant recipi-
`ents, and in patients with cancer being treated
`with chemotherapy.
`Other medical conditions may also influence
`the choice of antiepileptic drugs. Carbamazepine
`can cause partial or complete heart block and
`aggravate sick sinus syndrome.47 Carbamazepine
`may reduce the white-cell count and should prob-
`ably be avoided in patients with blood dyscrasias,
`because a change in the white-cell count may be
`difficult to interpret. Valproate causes a dose-
`related thrombocytopenia in up to 17% of patients
`and should be avoided in patients who are at risk
`for bleeding.48,49
`Carbamazepine and gabapentin are associated
`with modest weight gain (5 to 10 lb [2.3 to 4.5 kg]),
`and valproate and pregabalin are associated with
`more substantial weight gain (10 to 50 lb [4.5 to
`23.0 kg]) in about one third of patients. These
`drugs should be avoided, if possible, in patients
`with diabetes or eating disorders. Felbamat