Article Text

Download PDFPDF

New antiepileptic drugs: focus on ezogabine, clobazam, and perampanel
  1. Leslie A Rudzinski1,
  2. Naymeé J Vélez-Ruiz2,
  3. Evan R Gedzelman3,
  4. Elizabeth A Mauricio4,
  5. Jerry J Shih4,
  6. Ioannis Karakis3
  1. 1Department of Neurology, University of Florida College of Medicine, Gainesville, Florida, USA
  2. 2Department of Neurology, University of Miami, Miami, Florida, USA
  3. 3Department of Neurology, Emory University School of Medicine, Atlanta, Georgia, USA
  4. 4Department of Neurology, Mayo Clinic, Jacksonville, Florida, USA
  1. Correspondence to Dr Ioannis Karakis, Department of Neurology, Emory University School of Medicine, Faculty Office Building at Grady Campus, 49 Jesse Hill Jr. Drive SE, Office 335, Atlanta, GA 30303, USA; ioannis.karakis{at}


Ezogabine, clobazam, and perampanel are among the newest antiseizure drugs approved by the Food and Drug Administration between 2011 and 2012. Ezogabine and perampanel are approved for adjunctive treatment of partial epilepsy. Perampanel is also approved for adjunctive treatment of primary generalized tonic–clonic seizures. Ezogabine and perampanel have novel mechanisms of action. Ezogabine binds to voltage-gated potassium channels and increases the M-current thereby causing membrane hyperpolarization. Perampanel is a selective, non-competitive 2-amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic acid receptor antagonist, which reduces neuronal excitation. Clobazam has been used worldwide since the 1970s and is approved for adjunctive treatment of seizures associated with Lennox-Gastaut syndrome. Clobazam is the only 1,5-benzodiazepine currently in clinical use, which is less sedating than the commonly used 1,4-benzodiazepines. Phase III multicenter, randomized, double-blind, placebo-controlled trials demonstrated efficacy and good tolerability of these 3 new antiepileptic drugs. These drugs represent a welcome addition to the armamentarium of practitioners, but it remains to be seen how they will affect the landscape of pharmacoresistant epilepsy.

  • Epilepsy

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Epilepsy is a common neurological disorder afflicting nearly 4–10 per 1000 people in developed countries.1 Owing to its chronicity and unpredictability, it confers a significant burden to its bearers, their loved ones, and the society as a whole.2 ,3 An examination of recent studies based on claims data from large general populations in the USA identified epilepsy-specific annual cost estimates ranging from $8412 to $11 354 per person.4 It is for such reasons that a recent report from the Institute of Medicine urged for actions to improve the lives of people with epilepsy and their families.5

Antiepileptic drugs (AEDs) are the mainstay for the treatment of epilepsy. The goal of antiseizure drug treatment is seizure freedom with few adverse effects. About half of the patients with newly diagnosed epilepsy become seizure free with the first antiseizure drug, and about two-thirds become seizure free with the first three antiseizure drugs prescribed. The remaining one-third of patients will remain medically refractory or pharmacoresistant, and the percentage of patients achieving seizure freedom is substantially less following subsequent AED trials.6 ,7 Therefore, the development of new antiseizure drugs with novel mechanisms of action has become important in offering practitioners of rational polytherapy new, better-tolerated treatment options for pharmacoresistant patients.

The older first-generation antiseizure drugs (eight were available for use prior to 1982), with mechanisms of action consisting of sodium channel modulation or γ-aminobutyric acid (GABA) potentiation, are potent inducers or inhibitors of hepatic enzymes and therefore have multiple drug–drug interactions. These drugs unfortunately have many adverse effects that can negatively affect the patients’ quality of life.

Beginning with the release of felbamate and gabapentin in 1993, many new antiseizure drugs have emerged over the last two decades (figure 1). Their premise was to tackle the disease through different mechanisms of action, fewer drug–drug interactions, and improved tolerability. The focus has been on drug developments to modulate ionic channels that regulate flow of cations, augment inhibitory and mitigate excitatory neurotransmission, and modify synaptic trafficking (table 1).

Table 1

Mechanism of actions of currently approved AEDs in the US market

Figure 1

Introduction of major AEDs in the USA (information derived from AED, antiepileptic drug.

Such a premise was reinstated over the past 5 years with the US Food and Drug Administration (FDA) approval of three additional antiepileptic drugs, namely ezogabine, clobazam, and perampanel. Ezogabine (Potiga, GlaxoSmithKline, London, UK) was FDA approved in June 2011 for the adjunctive treatment of partial epilepsy in patients 18 years or older (retigabine is the International Nonproprietary Name).8 It opens voltage-gated potassium channels and enhances the M-current, which result in membrane hyperpolarization and reduces neuronal hyperexcitability. Clobazam (Onfi, Lundbeck, Deerfield, Illinois, USA), a 1,5-benzodiazepine, was introduced in the 1970s as an antianxiety drug and AED. First approved in Australia in 1970 and in France in 19749 and also approved for use in over 100 countries, clobazam has indications for the treatment of anxiety and multiple forms of epilepsy outside the USA. Despite worldwide application, clobazam was recently FDA approved in October 2011 for the adjunctive treatment of seizures associated with Lennox-Gastaut syndrome (LGS) in patients 2 years of age or older.8 Perampanel (Fycompa, Eisai, Woodcliff Lake, New Jersey, USA) was FDA approved in October 2012,8 and European Medicines Agency approved in the European Union in July 20128 for adjunctive treatment of partial epilepsy with or without secondary generalization in patients ≥12 years old. In 2015, it also acquired FDA approval as an adjunctive treatment of primary generalized tonic–clonic seizures in patients 12 years old.8 It is a selective, noncompetitive 2-amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic acid (AMPA) receptor antagonist.

This article comprehensively reviews the pharmacology/mode of action, pharmacokinetics, dosage/administration, efficacy, and safety/tolerability of each of these three drugs. Additionally, a succinct summary of the pharmacokinetics of these medications, their dosing and administration, and their efficacy and tolerability is provided in tables 24, respectively.

Table 2

Pharmacokinetics of EZG, CLB, and PER

Table 3

Dosing and administration for EZG, CLB, and PER

Table 4

Efficacy and tolerability of EZG, CLB, and PER


Pharmacology/mode of action

Ezogabine (N-[2-amino-4-(4-flurobenzylamino)-phenyl] carbamic acid ethyl ester) is a first-in-class new anticonvulsant, which demonstrated efficacy in reducing or inhibiting seizure activity in a variety of animal models of epilepsy. Its primary mode of action is activation of the neuronal outward M-current mediated by KCNQ (Kv7) voltage-gated potassium channels, which results in membrane hyperpolarization. The M-current opposes the depolarizing current and reduces the probability of raising membrane potential above the action potential threshold.10 Ezogabine binds within the ion pore of KCNQ2–5 channels, which alters channel gating and enhances the M-current and thereby stabilizing slightly depolarized hyperexcitable cells.11 ,12 This targeted specificity to the KCNQ2–5 (Kv7.2–7.5) channels accounts for its lack of effect on cardiac KCNQ1 (Kv7.1) activity at clinically relevant concentrations in vitro.13 In a study of cultured cortical neurons using the patch-clamp technique, through the augmentation of M-current currents via potassium channels, ezogabine significantly hyperpolarized the resting membrane potentials of the neurons at concentrations as low as 1 μΜ.14 This study also found that higher concentrations of ezogabine (50 μΜ) enhanced inhibition at GABAergic synapses by directly acting on postsynaptic GABAA receptors and enhancing inhibitory postsynaptic currents.14 At higher concentrations, ezogabine also shows weak blocking effects on sodium and calcium channels.11


The mean half-life of ezogabine is 8 hours. It is rapidly absorbed and has a bioavailability of 60% due to first-pass metabolism. Peak concentration is reached within 1.6 hours (range: 1–2 hours). Steady state is achieved by the third day after administration. Plasma concentrations increase linearly with increasing doses up to at least 1200 mg. Serum protein binding does not exceed ∼80%. Ezogabine is metabolized by N-glucuronidation and to a much lesser extent, N-acetylation. The N-acetyl metabolite displays weak pharmacological effects. Ezogabine does not induce or inhibit its own metabolism.10 ,15 ,16

Ezogabine does not cause clinically significant inhibition or alter the pharmacokinetics of drugs metabolized via CYP enzymes.17 It has modest potential to inhibit CYP2A6 but no potential to inhibit other CYP enzymes. Based on the results of a small phase II study, concomitant treatment with carbamazepine and phenytoin led to a respective 31% and 34% decrease in the area under the curve (AUC), a 23% and 18% decrease in peak concentration, and a 28% and 33% increase in clearance of ezogabine. Based on these results, an increased dosage of ezogabine was suggested, when using in conjunction with carbamazepine or phenytoin.18 However, in pooled clinical studies of >800 patients, enzyme-inducing and non-enzyme-inducing AEDs did not have any effect on ezogabine pharmacokinetics, nor did ezogabine alter the pharmacokinetics of phenytoin, carbamazepine, valproic acid, or topiramate. There is a modest pharmacokinetic interaction between ezogabine and lamotrigine. Ezogabine decreases the half-life and AUC of lamotrigine by 15% and 18%, respectively, and increases lamotrigine clearance by 22%. Lamotrigine increases the half-life and AUC of ezogabine by 7.5% and 15%, respectively, and decreases ezogabine clearance by 13%. The decline in ezogabine clearance due to lamotrigine is thought to be due to competition for renal elimination and not competition for glucuronidation.19 There was no interaction between ezogabine and phenobarbital. Oral contraceptives do not alter the pharmacokinetics of ezogabine, nor are oral contraceptives rendered less effective by ezogabine.10 The N-acetyl metabolite of ezogabine (NAMR) may inhibit renal clearance of digoxin.18 NAMR is also a weak inhibitor of P-glycoprotein (P-gp) efflux.18 Ethanol administered with ezogabine increases its AUC and peak concentration by 36% and 23%, respectively.20

Ezogabine is primarily eliminated renally. Seven metabolites of ezogabine and ∼36% of unchanged ezogabine are excreted through the urine.21 The pharmacokinetics of ezogabine is affected by body weight. Young women with a mean of 15 kg less body weight compared to young men had a significantly shorter ezogabine half-life (7.7 vs 8.5 hours), higher peak concentration (56%), and extent of exposure (measured as AUC—20%), while clearance was not different between the two sexes. The pharmacokinetics of ezogabine is also affected by age-related decline of renal function. In elderly patients, ezogabine clearance was 34% lower compared to younger patients resulting in a 42% higher extent of exposure and a 30% longer half-life.22 Geriatric dosing (patients >65 years old) should start at an initial dose of 150 mg daily and can be increased to a maximal dose of 750 mg daily.18 Ezogabine clearance is affected by renal and hepatic dysfunction. Patients with mild renal dysfunction had a 25% reduction in clearance, while those with moderate to severe dysfunction or on dialysis had a 50% reduction in clearance.23 Renal impairment (creatinine clearance (CrCl) <50 mL/min or end-stage renal disease on dialysis) dosing should start at an initial dose of 150 mg daily and can be increased to a maximal dose of 600 mg daily. There is no dosage adjustment needed for patients with mild renal impairment.18 Patients with moderate hepatic impairment had a 30% reduction in clearance and exposure was increased by 50%, while those with severe hepatic impairment had a 50% reduction in clearance and exposure was increased by 100%.10 Hepatic impairment (Child-Pugh score 7–9 or >9) dosing should start at an initial dose of 150 mg daily and can be increased to a maximal dose of 750 mg daily (Child-Pugh score 7–9) or 600 mg daily (Child-Pugh score >9). There is no dosage adjustment needed for patients with mild hepatic impairment.18 Race also affects ezogabine pharmacokinetics. This difference may relate to race differences in the glucuronidation pathway. Additionally, black patients had a 25% lower clearance and 30% lower volume of distribution compared to white patients, which led to higher exposure among black patients. Circadian changes also seem to affect ezogabine pharmacokinetics. The same study found that trough evening concentrations were 30–35% lower than morning concentrations. Race and circadian rhythm differences in ezogabine pharmacokinetics deserve further study.15

The pharmacokinetics of ezogabine have not yet been studied during pregnancy and currently no published data exist in the literature. However, AEDs eliminated by glucuronidation may be expected to significantly decline (similar to lamotrigine) during pregnancy. This mechanism may possibly apply to ezogabine.24 Ezogabine is classified by the FDA as a Pregnancy Category C drug, with no adequate, well-controlled studies in pregnant women. Treatment of pregnant rats and rabbits with ezogabine throughout organogenesis increased the incidences of fetal skeletal variations and resulted in decreased fetal body weights in rabbits. It is not known whether ezogabine is excreted in human breast milk. In preclinical models, ezogabine has not been found to affect fertility, mutagenesis, or carcinogenesis.18

Ezogabine is administered in three divided doses daily, with or without food. A high-fat meal does not affect the extent of absorption but increases peak concentration by ∼38% and delays the time to peak concentration by 45 min. The initial starting dose is 100 mg three times daily for 1 week. Dosage is increased by no more than 150 mg/week until desired maintenance dosage has been achieved.18 In a dose-ranging trial, dose of 600, 900, and 1200 mg/day administered three times daily were evaluated. The lowest effective dose was 600 mg/day and maximal effective dose was 1200 mg/day.25 In a titration study, the safety and tolerability of different titration rates was investigated, achieving the target dose of 1200 mg/day after 13, 25 and 43 days. The slow titration rate appeared to be best tolerated.26 Information for children <18 years of age is unavailable. An intravenous formulation is not available.


Three randomized, double-blind, placebo-controlled, multicenter clinical trials were pivotal to the approval of ezogabine for use in the treatment of partial onset seizures. All three studies evaluated median percent reduction in partial seizure frequency and responder rates (≥50% reduction in partial seizure frequency from baseline). In each study, after an initial 8-week baseline phase, patients were randomized to ezogabine or placebo using different schemas. Patients were required to be receiving stable dosages of one to two or three antiseizure drugs while continuing to experience four or more partial onset seizures during a 28-day period.

The Retigabine Efficacy and Safety Trial for Partial Onset Epilepsy (RESTORE) 1 enrolled 305 patients who received placebo or ezogabine 1200 mg/day in three equally divided doses, titrated over a 6-week period. Forced dose titration was followed by a 12-week maintenance phase. Median percent reduction in monthly seizure frequency over the double-blind (titration) period was ezogabine 44.3% versus placebo 17.5% (p<0.001), and the responder rate was 44.4% versus 17.8% (p<0.001). Among the 256 patients who completed titration and entered the maintenance phase, median percent reduction in monthly seizure frequency was ezogabine 54.5% versus placebo 18.9% (p<0.001), with responder rates of 55.5% versus 22.6% (p<0.001).27

Titration to high-dose ezogabine 1200 mg/day demonstrated efficacy but led to tolerability issues in some patients. The RESTORE 2 evaluated lower doses of ezogabine 600 and 900 mg/day, titrated over 4 weeks in 538 randomized patients. Ezogabine 600 and 900 mg/day significantly increased median percent reduction in 28-day total partial seizure frequency from baseline to double-blind period (27.9% (p=0.007) and 39.9% (p<0.001), respectively) versus placebo (15.9%) and significantly increased the maintenance phase responder rate (38.6% and 47.0%, respectively (both p<0.001)) versus placebo (18.9%).28

The third pivotal clinical trial, which preceded RESTORE 1 and RESTORE 2, was ‘study 205’. It was conducted to evaluate the 16-week (8-week forced titration and 8-week maintenance) efficacy and safety of adjunctive therapy with ezogabine 600, 900, and 1200 mg/day versus that of placebo in 399 randomized patients with partial onset seizures. The median percent decrease in monthly partial seizure frequency from baseline was 23% for 600 mg/day, 29% for 900 mg/day, and 35% for 1200 mg/day compared with 13% for placebo. The median percent reduction in monthly partial seizure frequency was dose dependent (p<0.001 for overall difference across all treatment groups). A greater reduction in mean monthly partial seizure frequency was noted in patients receiving 900 mg/day (14.15%, p<0.05) and 1200 mg/day (23.55%, p<0.05) but not 600 mg/day versus placebo. However, this was felt to be most likely driven by outliers such as one patient who had >1700% increase in seizure frequency. Responder rates for ezogabine were 23% for 600 mg/day, 32% for 900 mg/day (p<0.05), and 33% for 1200 mg/day (p<0.05), versus 16% for placebo.25

In summary, study 205 demonstrated statistically significant efficacy with ezogabine 900 and 1200 mg/day, with a significant linear dose–response relationship over the dose range of 600–1200 mg/day. RESTORE 1 and RESTORE 2 confirmed the efficacy of ezogabine at 900 and 1200 mg/day. Although ezogabine 600 mg/day did not demonstrate statistically significant efficacy in study 205, this dose was significantly efficacious in RESTORE 2. Altogether, these three controlled trials demonstrated statistically significant efficacy of ezogabine across the 600–1200 mg/day dose in adults with partial onset seizures compared to placebo.13 Integrated data from three pivotal randomized, double-blind, placebo-controlled, parallel-group studies (studies 205, 301, and 302) were used to evaluate the efficacy of ezogabine used in combination with one or more sodium blocking, non-sodium blocking, or sodium and non-sodium blocking AED. Efficacy appeared to be similar across all three groups.29 Similarly, in an open-label uncontrolled study of ezogabine added to 203 patients already on either carbamazepine/oxcarbazepine, lamotrigine, levetiracetam, or valproic acid, ezogabine was an effective adjunct using a flexible dosing regimen.30 Cost utility studies have also recently suggested cost-effectiveness of its use in focal-onset persons with epilepsy with limited response to standard antiepileptic treatment.31


The most common treatment-emergent adverse effects in the RESTORE 1 trial reported by more ezogabine than placebo-treated patients included dizziness, somnolence, fatigue, confusion, ataxia, blurred vision, and tremor. Most adverse effects emerged during titration and declined during the maintenance phase. Rare adverse effects included hallucinations and visual hallucinations, experienced by 2% and 3% of patients, respectively, for ezogabine, with no reports of either adverse effect for placebo. At the end of the double-blind period, mean body weight increased by 2.6 kg (ezogabine) and 0.3 kg (placebo) (an increase of 3.5% and 0.4%, respectively). A small increase in mean post void residual volume was observed with ezogabine early in treatment, which did not increase over time. However, urinary system adverse effects were generally not associated with elevated post void residual volumes collected at specified study visits. Urinary system adverse effects, most commonly urinary tract infections (11.8% vs 8.6%), urinary hesitation (5.9% vs 0.7%), dysuria (5.2% vs 1.3%), and chromaturia, occurred in more patients assigned to ezogabine than placebo.27

Similar to RESTORE 1, in RESTORE 2, the majority of adverse effects were considered mild or moderate in severity and were reported by a larger proportion of patients during titration relative to the maintenance phase. The most common (>10%) adverse effects in the placebo, ezogabine 600 mg/day, and ezogabine 900 mg/day groups were dizziness (7%, 17%, and 26%), somnolence (10%, 14%, and 26%), headache (15%, 11%, and 17%), and fatigue (3%, 17%, and 15%). Similar to RESRORE 1, in RESTORE 2, the most common adverse effects (≥5% in any treatment group) leading to discontinuation were dizziness (1%, 4%, and 7%) and somnolence (1%, 2%, and 6%) in the placebo, ezogabine 600 mg/day, and ezogabine 900 mg/day groups. Minor increases in body weight at Week 16, the end of the double-blind phase (mean change −0.1, +1.1, and +1.4 kg), were observed in the placebo, ezogabine 600 mg/day, and ezogabine 900 mg/day groups. There were no alterations in post void residual measurements in any treatment group. Chromaturia was reported in 1, 1, and 2 patients in the placebo, 600 mg/day, and 900 mg/day groups. Three patients exited the trial because of adverse effects related to the urinary tract (one nephritis and two urinary retention).28

Integrated data from three pivotal randomized, double-blind, placebo-controlled, parallel-group studies (studies 205, 301, and 302) were used to evaluate the safety and tolerability of ezogabine used in combination with one or more sodium blocking, non-sodium blocking, or sodium and non-sodium blocking AEDs, and it was found to be consistent with that previously observed.29 Likewise, in an open-label uncontrolled study of ezogabine added to 203 patients already on either carbamazepine/oxcarbazepine, lamotrigine, levetiracetam, or valproic acid, ezogabine exhibited similar tolerability to what was previously reported, using a flexible dosing regimen.30

In RESTORE 1, with titration over 6 weeks to ezogabine 1200 mg/day, 26.8% of ezogabine-assigned patients discontinued due to adverse effects versus 8.6% for placebo.27 In RESTORE 2, 8%, 17%, and 26% of patients receiving placebo, ezogabine 600 mg/day, and ezogabine 900 mg/day, respectively, discontinued treatment due to adverse effects.28 In study 205, a total of 79 patients withdrew from the study because of adverse effects: 17% from the 600 mg/day arm, 20% in the 900 mg/day arm, and 29.2% in the 1200 mg/day arm. An additional 12.5% withdrew from the placebo arm. Ninety-one percent of the withdrawals due to adverse effects occurred during the forced-titration phase. The most common treatment-emergent adverse effects leading to withdrawal among patients randomized to the ezogabine arms were confusion, dizziness, and somnolence, and confusion for those in the placebo arm. There was a 2–3% increase in body weight in the ezogabine 1200 mg/day arm and 1% in the 900 mg/day arm. There were no clinically relevant findings related to ezogabine treatment on assessments of urinary bladder function.25 Yet, subsequent studies looking retrospectively in the long-term retention of ezogabine in clinical practice suggested a less favorable retention rate compared to data from open-label extension studies of regulatory trials.32 Slow titration rates appear to be best tolerated.26

Similar to other AEDs, the most common adverse effects associated with ezogabine were central nervous system (CNS) related, some of which increased with increased dose. Most reported adverse effects were mild or moderate in severity, with the majority occurring during early titration. Neuropsychiatric symptoms, including psychosis and suicidal ideation, have been reported with ezogabine, and typically occur within the first week of starting therapy.18 Neuropsychiatric symptoms may be dose related.

In an integrated analysis, the rate of sudden unexpected death in epilepsy (SUDEP) was lower in patients treated with ezogabine compared with placebo (4.7 vs 8.0 per 1000 patient-years). Owing to its aforementioned lack of effect on cardiac KCNQ1 (Kv7.1) activity, ezogabine was not associated with cardiac abnormalities in the same studies. Nevertheless, given the NAMR weak inhibition on P-gp efflux and the known drug transporter role of P-gp in the pharmacokinetic susceptibility of QT prolongation,33 the manufacturer recommends that patients receiving concurrent drugs that prolong the QT interval be monitored due to the 7.7 ms QT prolongation found in healthy volunteers, mostly when they received 1200 mg/day.18

Ezogabine may cause dose-related bladder dysfunction in some patients, an effect pharmacologically related to its effect on KCNQ (Kv7) potassium channels, also located in the urinary bladder smooth muscle.34 An integrated analysis of the safety profile and secondary renal effects of ezogabine found that urinary adverse effects were observed more often in patients receiving ezogabine than in those receiving placebo. However, most adverse effects were judged by the investigator as mild and were reported in the first 8 weeks of therapy, with the majority of patients being able to continue treatment. There was an increased risk of urinary retention-related adverse effects/voiding dysfunction, which was most commonly seen at the 1200 mg/day dose in the forced-titration trials aforementioned.35 The cases of urinary retention were not associated with urinary tract infections or nephrolithiasis. A small number of secondary renal adverse effects were reported, but there is no evidence that ezogabine is nephrotoxic. Ezogabine should be used with caution in patients at risk of urinary retention (eg, benign prostate hypertrophy, patients on anticholinergics), and urologic symptoms should be carefully monitored. A risk evaluation and mitigation strategy is in place for ezogabine to inform healthcare providers of this risk of urinary retention.

There was a recent safety announcement from the FDA warning the public that ezogabine can cause blue skin discoloration (predominantly on or around the lips or in the nail beds of the fingers and toes) and eye abnormalities characterized by pigment changes in the retina that can cause serious eye disease with vision loss.36 The skin discoloration generally occurred after 4 years of treatment with ezogabine but has appeared sooner in some patients. The FDA does not currently know if these changes are reversible. All patients taking ezogabine should have a baseline eye exam, followed by periodic eye exams every 6 months. The FDA is working with the manufacturer to gather and evaluate all available information to better understand these events.

There are minimal data on the safety and efficacy of ezogabine in patients aged ≥65 years. Ezogabine must be used with caution in elderly patients, and a reduction in initial and maintenance dose is recommended. The effects of ezogabine on pregnancy are unknown, and it is recommended that pregnant patients register with the North American Antiepileptic Drug Pregnancy Registry. The safety of ezogabine for use in pediatric patients (<18 years) has not been studied. The Drug Enforcement Agency has determined that ezogabine should be categorized as a schedule V controlled substance under the Controlled Substance Act. Chemically, ezogabine has demonstrated CNS depressant properties and is classified as a sedative-hypnotic. Ezogabine has demonstrated relative potential for abuse in clinical trials, which has been associated with GABA-mediated neurotransmission.37


Pharmacology/mode of action

Clobazam is the only 1,5-benzodiazepine currently in clinical use.38 The unique chemical structure of clobazam renders different pharmacokinetic and pharmacodynamic properties compared to the commonly used 1,4-benzodiazepines such as clonazepam and diazepam. Clobazam was synthesized with the anticipation that its distinct chemical structure would provide better efficacy with fewer benzodiazepine-associated side effects.38 By binding to the GABAA benzodiazepine site, clobazam increases the rate of chloride channel opening, and the resulting chloride channel flux leads to inhibitory hyperpolarization of the postsynaptic membrane. In addition, clobazam serves to upregulate the GABA transporters 1 and 3, thereby further enhancing GABA inhibition.8 ,38


Clobazam is a long-acting benzodiazepine, with a median serum half-life of 36 hours. It is well absorbed and has a rapid onset of action, reaching a peak plasma concentration in 1–4 hours. Absorption is not significantly affected by food, and it is 80–90% protein bound in the serum.8 ,39 Metabolism of the drug occurs in the liver via the cytochrome P450 system. The major metabolic pathway of clobazam involves demethylation. N-Desmethylclobazam is a biologically active metabolite of clobazam with a similar potency, lower efficacy, and longer half-life than its parent compound.40 Plasma concentrations of clobazam and its active metabolite are affected by cytochrome P450 3A4 and 2C19 enzyme activity. Thus, there is potential for drug–drug interactions with inducers and inhibitors of these enzymes, as well as implications for patients who carry defective alleles for cytochrome P450 2C19 which lend them to be poor metabolizers.41 In CYP2C19 poor metabolizers, levels of N-desmethylclobazam were increased, thus indicating that N-desmethylclobazam plays an important role in long-term clobazam therapy. Poor metabolizers have a better response rate compared to intermediate and extensive metabolizers.42 In patients known to be CYP2C19 poor metabolizers, the starting dose should be low and dose titration should proceed slowly, as tolerated.43

In general, plasma concentrations tend to be higher in geriatric patients, and treatment with clobazam should proceed slowly with caution regarding dose escalation. Pharmacological treatment can be complicated by age-related changes in pharmacokinetics and pharmacodynamics and drug–drug and drug–disease interactions.44 Clinical studies of clobazam did not include sufficient numbers of subjects aged 65 and over to differentiate their response from younger subjects. However, elderly subjects appear to eliminate clobazam more slowly than younger subjects based on population pharmacokinetic analysis. No dose adjustment is required for patients with mild and moderate renal impairment. There are few to no trials regarding assessing clobazam in patients with severe renal impairment or end-stage renal disease. It is not known whether clobazam or its active metabolite, N-desmethylclobazam, is dialyzable. Notably, clobazam is hepatically metabolized; however, there are limited data to characterize the effect of hepatic impairment on the pharmacokinetics of clobazam. Therefore, it is recommended to proceed slowly with dosing escalations.45 Specifically, for patients with mild to moderate hepatic impairment (Child-Pugh score 5–9), the starting dose should be 5 mg/day in both weight groups. For patients with severe hepatic impairment, no dosing recommendations can be given, given the lack of available data.45

As with other benzodiazepines, discontinuation of clobazam should not be abrupt, and doses should be tapered 5–10 mg per day each week.45 The capacity of benzodiazepines to produce dependence, addiction, and subsequent withdrawal syndromes is well known and supported in the literature. Abrupt discontinuation of benzodiazepine treatment may show a spectrum of symptoms similar to those observed from withdrawal of alcohol or barbiturates (eg, convulsions, psychosis, hallucinations, behavioral disorder, tremor, and anxiety). Such reactions have been reported with the use of chlordiazepoxide, diazepam, oxazepam, lorazepam, nitrazepam, temazepam, and clobazam. Generally, the greater the dose and duration of use, the greater the risk of developing symptoms. Even so, withdrawal symptoms may occur in patients receiving low doses and/or short-term therapy, and as with all AEDs, clobazam should be withdrawn gradually to minimize the risk of withdrawal symptoms and minimize the risk of precipitating seizures, seizure exacerbation, or status epilepticus.46

Clobazam is a weak CYP3A4 inducer in a concentration-dependent manner. As some hormonal contraceptives are metabolized by CYP3A4, their effectiveness may be diminished when given with clobazam. Additional nonhormonal forms of contraception are recommended when using clobazam. Clobazam inhibits CYP2D6. Dose adjustment of drugs metabolized by CYP2D6 may be needed. Strong and moderate inhibitors of CYP2C19 may result in increased exposure to N-desmethylclobazam, the active metabolite of clobazam. This may increase the risk of dose-related adverse reactions. Concomitant use of clobazam with cytochrome P450 2C19 inhibitors such as valproate, oxcarbazepine, felbamate, fluconazole, fluvoxamine, ticlopidine, and omeprazole may require dose reduction.45 ,47 Following a single oral dose, ∼11% of the dose was excreted in the feces and ∼82% was excreted in the urine.45

There are no adequate and well-controlled studies of clobazam or benzodiazepines in pregnant women regarding either major congenital malformations or cognitive/behavioral deficits with respect to exposure in utero. Additionally there are no adequate developmental toxicity studies of clobazam in animals. The risks of fetal exposure from benzodiazepines to cognition and behavior are unknown secondary to a dearth of adequate data in human studies.48 However, it is known that offspring of women with epilepsy on specific AEDs are at increased risks for major congenital malformation and reduced cognition.50 It should be noted that neither clobazam nor any benzodiazepines were tested in the trial. Despite the paucity of data, the FDA has labeled clobazam a Category C drug. Nevertheless, clobazam should be used during pregnancy only if the potential benefit justifies the potential risk to the fetus. Clobazam is excreted in human breast milk. There is a paucity of data on the effects of this exposure on infants. Data on the use of second-generation and third-generation AEDs in pregnancy and lactation and regarding long-term effects on cognition and behavior are sparse. Weighing the benefits of breast feeding against the potential risk to the nursing infant must be taken into consideration and infant monitoring for potential adverse effects is advisable when the mother is taking AEDs, including clobazam.51 The safety and effectiveness in patients <2 years of age have not been established.45

Clobazam is available in oral tablets as well as in oral suspension and a rapid titration to therapeutic dosage can be achieved. A starting dose of 5 mg two times per day can be initiated in patients >30 kg and can be titrated up to 20 mg two times per day if tolerated over the course of 2 weeks. For those patients weighing <30 kg, 2.5 mg two times per day is the recommended starting dose, which can be titrated up to 10 mg two times per day in 2 weeks. Lower initial doses (5 mg daily) and slower titrations (over 3 weeks) are recommended in the elderly, those with hepatic impairment, and in known cytochrome P450 2C19 poor metabolizers. The maximum daily dosage recommended for these patients (>30 kg) is 20 mg (10 mg two times per day).45


Clobazam led to ≥50% seizure reduction for more than half of LGS and refractory epilepsy patients in multiple clinical trials, open-label studies, and retrospective reviews. There is little evidence for the development of tolerance as efficacy is persistent.47 More specifically, in six open-label prospective studies40 testing the efficacy of clobazam as add-on therapy for pediatric patients with refractory epilepsy (including LGS), 54–85% of patients experienced at least a 50% decrease in seizure frequency.51–56

The FDA's approval of clobazam for LGS was based on two randomized controlled trials, OV-100257 and OV-1012 (also known as the CONTAIN trial).58 The OV-1002 study was a phase II trial of patients with LGS 2–30 years of age. Clobazam was well tolerated and showed decreased frequency of drop and non-drop seizures. Two doses were studied (1.0 and 0.25 mg/kg/day). Both doses showed efficacy, with the higher dose found to be more efficacious. OV-1012 (the CONTAIN study) was a phase III placebo-controlled trial of clobazam. It evaluated three doses in adult and pediatric patients 2–60 years of age. Treatment with clobazam 0.25, 0.5, or 1.0 mg/kg/day was well tolerated and efficacious in the treatment of drop seizures for patients with LGS. Clobazam significantly decreased average weekly rates of drop seizures (12.1%, 41.2%, 49.4%, and 68.3% for the placebo, 0.25, 0.5, and 1.0 mg/kg/day dosages, respectively) and total seizures (9.3%, 34.8%, 45.3%, and 65.3% for the placebo, 0.25, 0.5, and 1.0 mg/kg/day dosages, respectively) compared with placebo.58 After participation in study OV-1002 or OV-1012, patients were eligible to enroll in a long-term, open-label extension trial, OV-1004. From the period of 28 December 2005 to 1 July 2010, an interim data analysis of this trial was carried out, and these data were used in the FDA's review of clobazam for approval.

In the OV-1004 trial, patients with LGS who had completed one of two randomized controlled trials (OV-1002 or OV-1012) received clobazam at doses ≤2.0 mg/kg/day (≤80 mg/day). Of 306 eligible patients from OV-1002 or OV-1012, 267 entered the open-label extension. As of the interim date, 1 July 2010, 213 patients (79.8%) had remained in the trial, and 189 had received clobazam for ≥12 months, 128 for ≥18 months, and 94 for ≥24 months.59 The primary efficacy outcome was the percentage decrease in the average weekly rate of drop seizures at various time intervals compared with baseline. The percentage decrease in the average weekly rate of total seizures was also measured. Median percentage decreases in average weekly rates of drop seizures were 71.1% and 91.6% at months 3 and 24. Mean modal and mean maximum daily dosages were 0.94 and 1.22 mg/kg for those who had received clobazam for ≥1 year. Median percentage decreases in total seizures in this population were 64.8% at month 3 and 81.5% at month 24.59 Of note, no diminution of response rates was observed during the first 2 years of the open-label trial. Important highlights of this study were as follows: (1) sustained substantial seizure improvements for patients treated ≥2 years, (2) 80% retention rate consistent with therapy compliance and treatment satisfaction, (3) no new safety concerns versus results of short-term controlled trials, (4) mean modal clobazam dosages did not increase with time, and (5) a trend toward decreases in numbers of concomitant treatments was observed.59 In 2009, the results of a multicenter, double-blind phase II study were reported.57 This study evaluated low-dose (0.25 mg/kg/day) or high-dose (1 mg/kg/day) clobazam in 68 patients with LGS, aged 2–26 years. Compared with baseline, 83% of patients in the high-dose group had a 50% or greater reduction in seizure frequency. Additionally, patients had improved global assessments on high-dose and low-dose clobazam, consistent with prior studies suggesting improved behavioral and cognitive performance on this AED.51 ,52 More recent post hoc analyses of the OV-1012 and OV-1004 data identified no differences in efficacy across age groups,60 and across the spectrum of disease severity in patients with LGS.61 Moreover, adjunctive clobazam used at stable doses in this large cohort of patients with LGS sustained seizure frequency improvement and seizure freedom through 3 years of treatment.62 Given the absence of head-to-head trials in LGS, indirect comparisons of relative efficacies between clobazam and other therapeutic options such as lamotrigine, felbamate, topiramate, and rufinamide suggested that high (1 mg/kg/day) and medium (0.5 mg/kg/day) doses of clobazam may be not only be more efficacious63 but also be more cost-effective.64

Beyond LGS, clobazam was evaluated as an add-on therapy in refractory partial seizures. In a study of 183 patients with intractable complex partial seizures in whom conventional benzodiazepines had been previously discontinued, clobazam was initiated in a range of 5–10 mg and titrated maximally to 20 mg.65 Although complete remission was initially achieved in 61 patients, tolerance developed in almost half (49.2%) within the first 3 months, whereas 23 out of 31 patients (74.2%) who were seizure free for the first 3 months remained so over the next 3 months. Clobazam proved to be significantly more effective when concurrent generalized tonic–clonic seizures occurred at least yearly. The adverse events profile was in agreement with previous studies. These study results support the notion that clobazam is an effective, safe, and inexpensive medication for add-on therapy in patients with intractable complex partial seizures without concurrent administration of conventional benzodiazepine compounds.65 In 1998, the Canadian Study Group for Childhood Epilepsy66 compared the effectiveness of monotherapy clobazam to carbamazepine and phenytoin in children aged 2–16 years old with epilepsy. Children with newly diagnosed epilepsy (partial epilepsies or only generalized tonic–clonic seizures) or previous failure of one drug (for poor efficacy or side effects) were assigned to one of two study arms and then randomized. Fifteen centers entered 235 patients: 159 randomized to clobazam versus carbamazepine and 76 to clobazam versus phenytoin. In total, 119 patients received clobazam, 78 carbamazepine, and 38 phenytoin. Overall, 56% continued to receive the original medication for 1 year. Seizure control and side effects were approximately equivalent for all three medications. Phenytoin and carbamazepine induced more biologic side effects, such as rash, while clobazam induced slightly more behavioral effects. Tolerance developed in 4.2% of patients receiving carbamazepine, 6.7% with phenytoin, and 7.5% with clobazam. Based on these study results, it was suggested that clobazam should be considered as possible first-line monotherapy, along with carbamazepine and phenytoin, for partial epilepsies and selected generalized childhood epilepsies.66 Yet, a recent Cochrane systematic review for clobazam monotherapy concluded that the available evidence is still insufficient to decisively inform clinical practice on that matter.67 A double-blind add-on trial investigated the effect of clobazam on refractory complex partial seizures in 20 patients. At the end of the third month, eight (40%) of the patients had a seizure reduction of >75%, including four patients (20%) who had no seizures during that time. Tolerance to the antiepileptic effect of clobazam was noted in 56% of the patients, and mild transient sedation occurred in 40% of the patients. These findings support that clobazam appears to be an effective add-on drug for patients with refractory localization-related epilepsy.68 Finally, a small study brought up the consideration that clobazam could possibly be utilized as a second-line agent in partial status epilepticus. The study was a case series of four patients with electroencephalogram (EEG)-proven partial status epilepticus refractory to standard AEDs. Seizure cessation occurred in three out of four of the patients within 2 hours of being administered an oral loading dose of clobazam.69 Larger case series have recently corroborated further that potential.70


Clobazam's safety profile is known to be similar to other benzodiazepines, but with better psychomotor performance and substantially decreased sedation. Studies suggest that clobazam acts mechanistically in a manner similar to other benzodiazepines through potentiation of GABAA receptors. During the open-label study OV-1004,59 a total of 219 (82.0%) patients experienced ≥1 treatment-emergent adverse event and 140 (52.4%) patients experienced ≥1 adverse event that was treatment related. The most common treatment-emergent adverse events (≥10% of patients) listed by descending incidence were upper respiratory tract infection (18.4%), fall (14.2%), pneumonia (13.9%), somnolence (12.7%), otitis media (12.0%), fever (10.5%), and constipation (10.1%). The upper respiratory tract infection and pneumonia events occurred predominantly in pediatric patients.59

In OV-1004, many of the adverse events with an incidence ≥10% generally decreased in their prevalence and incidence after the first 6 months of treatment. These included upper respiratory tract infection, somnolence, otitis media, and fever. Mild or moderate adverse events were reported for 160 patients (59.9%), and severe adverse events were reported for 59 patients (22.1%). Reported severe treatment-emergent adverse events for ≥1.0% of patients were pneumonia and convulsion (4.1% each), status epilepticus and aspiration pneumonia (1.5% each), and lobar pneumonia, sepsis, septic shock, urinary tract infection, dehydration, sedation, somnolence, and aggression (1.1% each). Serious adverse events were reported for 85 patients (31.8%) during clobazam exposure in the same study. Those that occurred in >2% of patients were pneumonia (9.4%), convulsion (8.2%), pneumonia aspiration (3.7%), lobar pneumonia (2.2%), and urinary tract infection (2.2%). Fifteen patients (5.6%) reported treatment-related serious adverse events, of which pneumonia (four patients), convulsion (three patients), and status epilepticus (two patients) were the only serious adverse events reported for more than one patient. As of the interim data cutoff (1 July 2010), six patients (2.2%) died. The presumed etiologies for these deaths were pneumonia (two patients); epilepsy (one patient); pneumonia, sepsis, and acute respiratory distress syndrome (one patient); and death of unknown origin or etiology (two patients). None was considered related to clobazam treatment. The adverse events most commonly reported with clobazam use, such as falls and respiratory infections, were to be expected in a patient with LGS population in a long-term trial. Additionally, upper respiratory tract infection and pneumonia events occurred mostly in pediatric patients, which is a common finding in trials in pediatric populations with epilepsy. When compared with the results of short-term controlled trials, no new safety concerns become apparent with the interim OV-1004 open-label extension trial data. Over the first 2 years of OV-1004, the clobazam safety profile and potential for clinical benefit taken in conjunction with the decades of global use in LGS and other seizure disorders indicate that clobazam is a safe long-term treatment option.59 More recent post hoc analyses of the OV-1012 and OV-1004 data identified no differences in adverse events across LGS age groups.60 A randomized, double-blind, dose-ranging study57 evaluated safety and efficacy of clobazam in patients with LGS. Five serious adverse events were reported in four patients, but in no case was clobazam discontinued. Adverse events experienced by ≥5% patients in descending frequency in the high-dose group include somnolence, lethargy, sedation, salivary hypersecretion, constipation, aggression, hypomania, and insomnia. Another study was performed in five European countries using a double-blind crossover design of 7 months duration. Clobazam was compared to placebo as AED adjunct treatment in 129 refractory epilepsy patients mainly suffering from complex partial seizures. The most frequent adverse reactions to clobazam were drowsiness and dizziness. Of note, the sedating effects of clobazam seemed less pronounced in comparison with other benzodiazepines.71 In general, somnolence and sedation begin within the first month of treatment with clobazam and may diminish with continued treatment. Patients on clobazam should be monitored for somnolence and sedation, particularly with concomitant use of other CNS depressants (eg, alcohol). Additionally, patients should be cautioned against engaging in hazardous activities requiring mental alertness, such as operating dangerous machinery or motorized vehicles.45 Albeit rare, Stevens-Johnson syndrome and toxic epidermal necrolysis have been reported with clobazam use, in children and adults. Close monitoring, particularly in the first 2 months of initial or recurrent introduction to the medication, is recommended.45

AEDs including clobazam may increase the risk of suicidal thoughts or behavior in patients taking these drugs for any indication. In 2008, the FDA issued an alert to healthcare professionals about an increased risk of suicide ideation and suicide behavior in people treated with antiseizure drugs. Since then, a number of retrospective cohort and case–control studies have been published that endeavored to address this issue, but gathered results are contradictory. An expert consensus statement developed by an ad hoc task force of the Commission on Neuropsychobiology of the International League Against Epilepsy determined that some (but not all) AEDs can be associated with treatment-emergent psychiatric problems that can lead to suicidal ideation and behavior. The actual suicidal risk is yet to be established, but it seems to be very low.72 However, this does not preclude the fact that patients treated with any AED for any indication should be monitored for the emergence or worsening of depression, suicidal thoughts or behavior, and/or any unusual changes in mood or behavior. Clobazam is a Schedule IV controlled substance and can be abused in a similar manner as other benzodiazepines for its rewarding psychological or physiological effects. This may lead to overdosage that could lead to signs and symptoms of CNS depression, drowsiness, confusion and lethargy, possibly progressing to ataxia, respiratory depression, hypotension, and rarely coma or death.45

Withdrawal-related adverse event rates were investigated following abrupt clobazam discontinuation in phase I trials and gradual (over 2–3 weeks) tapering from phase II (OV-1002), phase III (OV1012) and open extension (OV-1004) trials. When clobazam was abruptly discontinued in phase I trials, only mild withdrawal-related adverse events were generally observed. When clobazam was gradually tapered, regardless of short-term or long-term (up to 5 years) use, no withdrawal-related adverse events occurred.73


Pharmacology/mode of action

Perampanel (2-(2-oxo-1-phenyl-5-pyridin-2-yl-1,2-dihydropyridin-3-yl) benzonitrile hydrate) inhibits 2-amino-3-(3-hydroxy-5-methyl-isoxazol-4-yl)propanoic acid-induced increases in intracellular Ca2+ and selectively blocks AMPA receptor-mediated synaptic transmission.74 Perampanel demonstrates a broad-spectrum of activity in rodent and other preclinical seizure models.75 The precise mechanism of action for seizure reduction has not been fully determined. Perampanel reduces calcium influx mediated by AMPA receptors in cultured cortical neurons. In vitro studies revealed its mechanism of action as a selective noncompetitive AMPA receptor antagonist that affects neurotransmission by high potency reduction of neuronal excitability in the brain.76 Perampanel also acts at the N-methyl-d-aspartate and the kainate receptors in the excitatory postsynaptic membranes of the neurons, though the principal ionotropic glutaminergic receptor activity involves AMPA.77 In mouse models, perampanel showed a protective effect against audiogenic-induced, phenylenetetrazol-induced, and maximal electroshock-induced seizures studied in monotherapy and in combination with other AEDs. In amygdala-kindled rats, perampanel significantly increased afterdischarge threshold and significantly reduced motor seizure duration, afterdischarge duration, and seizure severity recorded at 50% higher intensity than afterdischarge threshold current.


Perampanel has a long half-life of ∼105 hours (range: 52–129 hours in the single-dose study, and 66–90 hours in the multiple-dose study).78 Perampanel is rapidly and completely absorbed after oral administration with minimal first-pass metabolism. Perampanel is absorbed from the gastrointestinal tract following oral administration and rapidly reaches its maximal plasma concentration about 60–90 min after ingestion. Steady-state plasma concentrations are reached after 2 weeks. Perampanel is 95% bound to plasma proteins and is extensively metabolized via oxidation and sequential glucuronidation. It is primarily metabolized by CYP3A4 of the P450 enzyme system, although other CYP enzymes may also be involved. Perampanel does not act as an enzyme inducer or inhibitor. However, enzyme-inducing drugs such as carbamazepine, oxcarbazepine, and phenytoin increase the clearance of perampanel and decrease its plasma concentration by at least half. Topiramate reduces the pharmacokinetic AUC of perampanel by 20%, while lamotrigine, levetiracetam, valproate, zonisamide and benzodiazepines had no significant effect on perampanel clearance. Perampanel reduces the clearance of oxcarbazepine by 26%, but has minimal effect on carbamazepine, valproate, and lamotrigine clearance. Concomitant use of perampanel with other potent CYP3A4 inducers such as rifampin and St. John's wort should be avoided to minimize drug–drug interactions. At doses of 12 mg/day, perampanel reduces levonorgestrel serum levels by ∼40%, and therefore, hormonal contraceptives containing levonorgestrel may be compromised with perampanel at this dose.

Perampanel is metabolized by first-order elimination pharmacokinetics. Elimination occurs primarily through the feces for 70% of the dose. The remainder is excreted in the urine with <2% of the dose eliminated unchanged. Drug clearance decreases by 27% and AUC increases by 37% in mild to moderate renal dysfunction. No adjustments in mild renal dysfunction (CrCl 50–80 mL/min) are required. However, perampanel has not been adequately studied in patients with severe renal dysfunction or on dialysis and is not recommended in these patients. In hepatic disease, the AUC increases almost twofold in mild dysfunction and 3.3-fold in moderate dysfunction, and perampanel half-life is prolonged. Patients with mild hepatic impairment should have a maximum of 6 mg daily, while those with moderate hepatic impairment should have a maximum recommended daily dose of 4 mg once daily. Perampanel is not recommended in patients with severe hepatic impairment.

There are no well-controlled studies on the effects of perampanel on the developing human fetus. The FDA currently classifies perampanel as a pregnancy Category C drug. Oral administration of perampanel to pregnant rats at any dose throughout organogenesis resulted in an increase in visceral abnormalities. In a dose-ranging study in rabbits, embryo lethality was observed at extreme doses. However, no evidence of adverse effects to embryo–fetal developmental toxicity in rabbits was seen with a dose comparable to 16 mg/day of perampanel in humans. It is unknown whether perampanel is excreted in breast milk. Perampanel has not been found to affect fertility, mutagenesis, or carcinogenesis.

Perampanel is administered in a single dose that is typically taken at night. Food slows the rate of absorption but does not affect the extent. Perampanel serum concentrations are not affected by sex, race, or age between 12 and 74 years old. Patients older than 12 years may be dosed as adults. In dose-ranging studies, 2–12 mg of active drug per day were evaluated. Efficacy studies suggest a minimum effective dose of 4 mg daily.79 The FDA recommended maximum dose of perampanel is 12 mg/day. Information is unavailable for children below 12 years and for the geriatric population. An intravenous formulation is not available.


The primary goals of two consecutive, multicenter, phase II, dose-escalation and dose-finding, randomized, double-blind, placebo-controlled studies (studies 206 and 208) were safety and tolerability.80 Patients with refractory partial epilepsy (n=153 in study 206 and n=48 in study 208) were randomized for 12–16 weeks between adjunctive perampanel 1–12 mg versus placebo.80 While not powered to provide efficacy data, phase II studies provided preliminary evidence of it, but the small numbers of participants precluded statistical significance.80 Subsequently, efficacy was the primary goal of three concurrent, multicenter, multinational, phase III, randomized, double-blind, placebo-controlled studies (studies 304, 305, and 306).81–83 All three were conducted in patients with refractory partial epilepsy (n=388 for study 304, n=386 for study 305, and n=706 for study 306), already treated with a stable regimen of 1–3 antiseizure drugs. They all had a similar design of a 6-week baseline period prior to randomized dose escalation over 6 weeks starting at 2 mg/day and a 13-week maintenance period. In study 30481 that was conducted in the Americas and 30582 that took place in four additional continents, patients were randomized to receive once daily perampanel 8 mg, perampanel 12 mg, or placebo. Study 306 was also universal and targeted lower doses with randomization between perampanel 2 mg, perampanel 4 mg, perampanel 8 mg, and placebo. Analysis was performed in the intention-to-treat population. A pooled analysis of the phase III data also ensued.84 The primary end points in all three phase III studies were median reduction of seizure frequency per 28 days from baseline to the end of the 19-week double-blind period and responder rate in the 13-week maintenance period relative to baseline, defined as proportion of patients achieving at least 50% reduction in seizure frequency. Median reduction of seizure frequency was statistically greater with perampanel 4 mg/day (23.3%, study 306), 8 mg/day (26.3–30.8%, studies 304–306), and 12 mg/day (17.6–34.5%, studies 304 and 305) than with placebo (9.7–21%, studies 304–306), but not with 2 mg/day (13.6%, study 306). The responder rate was significantly greater with perampanel 4 mg/day (28.5%, study 306), 8 mg/day (33.3–34.9%, studies 305 and 306 but not study 304), and 12 mg/day (33.9%, study 305 but not study 304) than with placebo (14.7–17.9%, studies 305 and 306). The lack of statistically significant responder rate effect in study 304 was attributed to regional variations.81 The pooled analysis of these trials demonstrated clinical efficacy of perampanel at 4, 8, and 12 mg/day.84 Perampanel efficacy appeared to be unaffected by the use of concomitant non-enzyme-inducing AEDs, though it may be lower in the presence of numerous concomitant AEDs, likely indicative of more refractory disease.85 Interestingly, the dose-dependent effect seen in the individual trials was not corroborated in that combined analysis, likely reflecting the wide variability in placebo response,84 ,86 though some patients who tolerated higher doses could benefit from increasing the dose from 8 to 12 mg.87 Further post hoc subanalysis by gender of the three phase III, double-blind, randomized trials revealed that seizure frequency was reduced with perampanel treatment regardless of gender, with female participants demonstrating greater reduction in seizure frequency and higher responder rates compared to male participants.88 Subanalysis in the elderly group demonstrated similar efficacy, in terms of seizure frequency and the 50% responder rate, to the overall population.89 At the other end of the age spectrum, subanalysis in the adolescent patients with refractory focal epilepsy from the three double-blind, placebo-controlled, phase III studies as well as the combined extension study demonstrated better seizure control than placebo and sustained improvements in seizure frequency, comparable to the overall study population.90

The placebo-controlled trials were followed by long-term, open-label, extension studies. Specifically, patients (n=138) from both phase II studies were offered participation in an extension study (study 207)91 to allow detection of longer term safety signals for a planned maximum duration of ∼8 years. Moreover, patients (n=1218) from all phase III studies were offered participation in study 307,79 with long-term safety/tolerability and efficacy maintenance being the primary and secondary end points of this study, respectively. In the 207 extension study, the responder rates varied from 43.8% to 51.5% within 4 years of follow-up with a mean (SD) perampanel dose of 7.3 (3.3) mg/day. Seizure freedom was <3%.91 In the interim analysis of the 307 extension study, the responder rates varied from 47.6% to 63.2% within the 2 years of follow-up with a mean (SD) perampanel dose of 10.1 (2.3) mg/day. Seizure freedom for a whole year was seen in 7.1% of patients.79 In this open-label extension study, patients transitioned from placebo in the core studies achieved similar seizure control at the end of the conversion period to that previously seen with patients on perampanel in the core studies. Also, patients who continued on long-term exposure of perampanel during the open-label phase manifested sustained improvements in seizure control.92 ,93

In 2015, the results of a multicenter, double-blind, randomized, placebo-control trial of perampanel in patients 12 years or older with primary generalized tonic–clonic seizures were announced. One hundred sixty-two patients (81 placebo and 81 perampanel) comprised the full analysis set. During a 4-week period, they were uptitrated from 2 to 8 mg/day, or highest tolerated dose with a 13-week maintenance period. The primary outcome was percent change in seizure frequency per 28 days from baseline. The secondary end point (primary for European Union registration) was the 50% responder rate compared to baseline. Compared to placebo, perampanel resulted in greater median percent change of tonic–clonic seizure frequency per 28 days (−38.4% vs −76.5%; p<0.0001) and greater 50% tonic–clonic seizure responder rate (39.5% vs 64.2%; p=0.0019). During maintenance, 30.9% of perampanel-treated patients compared to 12.3% of placebo-treated patients achieved seizure freedom.94 This study provided class I evidence for the adjunctive use of perampanel for primary generalized tonic–clonic seizures, leading to its subsequent FDA approval in patients 12 years of age or older with drug-resistant idiopathic generalized epilepsy.8 Recently, preliminary efficacy data for the use of perampanel in the pediatric population with refractory epilepsies (focal, unclassified generalized, Lennox-Gastaut, West and Dravet) have emerged, reporting 31% responder rate (18/58 patients) after the first 3 months of therapy.95 The first clinical experience of its use in 12 patients with refractory and super-refractory status epilepticus has not been particularly rewarding.96


Both phase II studies indicated no major safety issues and feasibility of flexible titration schedules.80 Akin to other antiseizure drugs, CNS side effects, predominantly in the form of dizziness, somnolence, headache, and fatigue, prevailed.80

Similar to phase II studies, all phase III studies demonstrated that the majority of treatment-emergent adverse events were mild to moderate and CNS mediated.81–83 Pooled analysis of all phase III data indicated that serious adverse events at any dose were reported by 5.5% of perampanel-treated patients versus 5% of placebo-treated patients.84 Overall, a dose-dependent relationship in the frequency of adverse events was seen.84 Perampanel 12 mg/day was the only dose that was significantly linked with higher withdrawal rates than placebo.84 Falls were infrequently reported. An increased rate of severe neuropsychiatric adverse effects of ∼15% was noted. These include agitation, aggression, affective and psychotic disorders, independent of prior risk factors. Suicidality was observed in five patients total (three on perampanel and two on placebo),97 leading to a warning from regulatory agencies.9 The risk of dependency from perampanel use is low.9 Idiosyncratic drug reactions were not seen, despite the breakdown of perampanel to active metabolites.9 Rash occurrence was rare and not dose dependent.84 More perampanel-treated patients reported >7% weight gain than placebo patients.84 Perampanel at 12 mg/day may decrease the effectiveness of progesterone-containing hormonal contraceptives.9 Laboratory values, electrocardiographic parameters, and vital signs were not adversely affected.84 Subanalysis of the phase III studies by gender suggested that common side effects such as dizziness and headaches occur more frequently in female subjects, possibly accounting for 17% lower perampanel clearance in female than male patients not receiving enzyme-inducing AEDs.88 Similarly, subanalysis of those studies in the elderly participants suggested more pronounced dizziness and fatigue despite the largely similar adverse events rates, advocating for careful titration in individuals above the age of 65 years.89 When pooled data of those phase III studies plus the open-label extension study were reviewed particularly for the adolescent participants, a generally favorable safety profile was concluded.90

The extension studies overall reinforced the side effect profile discerned by the phase II and III trials.92 Additionally, four deaths (one in study 207, possibly linked to SUDEP and three in study 307, one due to SUDEP, one from an intracerebral hemorrhage, and one due to motor vehicle accident) were reported.79 ,91 There was no evidence of prolonged QT interval duration during perampanel treatment.98 The most common treatment-emergent adverse events were again dizziness, somnolence, fatigue, irritability, and weight increase.99 ,100 For most subjects, these were observed during the 6-week titration and were mild or moderate in severity. For severe adverse events, a dose–response relationship was not observed. In general, past 6–12 months of perampanel treatment, low to no incidence of those side effects was observed.99 Long-term extension studies showed a 62.5–69.6% adherence after 1 year of treatment, favorably comparing with other second-generation AEDs.101 No new safety signals emerged during up to 3 years of perampanel exposure in 39 countries.93

In the randomized trial of perampanel for tonic–clonic seizures in idiopathic generalized epilepsy, a similar safety profile was reported, with dizziness (32.1%) and fatigue (14.8%) being the most common adverse events.94 Likewise, during its off-label use for refractory pediatric epilepsies95 or for refractory status epilepticus,96 no additional adverse events were reported.

Despite the favorable efficacy and safety profile demonstrated by the aforementioned well-designed studies, there are limitations to acknowledge. From an interpretational standpoint, seizure freedom was restricted to single digit percentages, casting doubts on the potential of this compound as a game changer. From a methodological standpoint, the pivotal phase III trials were vulnerable to regional reporting variations.102 The extension trial results do not reflect the progressive dropout of nonresponders and can thus overestimate true response rates.86 Also, these long-term studies did not include a placebo comparison and they permitted concurrent antiseizure drug changes.103 On the other hand, in all studies, many patients were concurrently on a multitude of other antiseizure drugs and that may have exaggerated their adverse events.84 Moreover, randomized trials are commonly not adequate to detect idiosyncratic drug reactions104 and enduring surveillance is therefore necessary. Head-to-head comparison of perampanel with other antiseizure drugs has not been yet performed in terms of efficacy and safety. Its role as monotherapy for newly diagnosed epilepsy patients merits evaluation.105 It is unclear if it will ultimately gain a role in emergent situations such as status epilepticus.106 The plausible synergistic effect of perampanel with other antiseizure drugs warrants clinical confirmation.102 Finally, its use for special populations such as children <12 years of age is currently being evaluated,95 while pregnancy registry data collection is anticipated.

The so far accumulated evidence suggests that (1) perampanel is an efficacious adjunct in reducing frequency and improving responder rate of partial onset seizures with or without secondary generalization and in primary generalized tonic–clonic seizures in patients ≥12 years of age, although seizure freedom is infrequently achieved; (2) the safety and tolerability profile appears favorable; (3) treatment and adverse effects seem to be overall dose dependent; and (4) safety and efficacy sustain over time. For those reasons plus its innovative mechanism of action as an inhibitor of glutamatergic excitation, perampanel emerges as a promising add-on drug for partial onset and primary generalized tonic–clonic seizures in adults with refractory epilepsy.


Three new antiseizure drugs have been recently FDA approved. Ezogabine and perampanel are licensed for adjunctive treatment of partial epilepsy. Additionally, perampanel recently gained approval as adjunctive treatment for primary generalized tonic–clonic seizures. Clobazam is licensed for adjunctive treatment of seizures associated with LGS in children 2 years of age or older. Efficacy measured as percent reduction of seizure frequency and responder rate is similar for all three drugs, but appears greatest for clobazam when used to treat seizures associated with LGS. Dizziness and somnolence are the most common adverse effects of ezogabine; upper respiratory tract infections, falls, pneumonia, somnolence, dizziness, otitis media, fever, and constipation for clobazam; and dizziness, somnolence, headache, and fatigue for perampanel. All three AEDs have a low to moderate propensity for drug–drug interactions. Albeit not approved for monotherapy, these drugs represent a welcome addition to the armamentarium of practitioners. It remains to be seen how these new drugs will affect the landscape of pharmacoresistant epilepsy. Cumulative clinical experience in the next years will clarify whether these promising drugs can truly play an important role in rational polytherapy of refractory epilepsy, and whether they could ever become first-line treatment options.

Future directions

Pharmacotherapy of epilepsy has substantially evolved over the past two decades yielding medications with unique mechanisms of action for synergistic use, fewer pharmacokinetic interactions, and better tolerability. In addition to the currently marketed drugs, there are at least 20 compounds under clinical development. Some of them target existing mechanisms of action, others novel or undeterminated modes of action, and plenty are repurposed for the treatment of seizures (table 5).107

Table 5

Potential antiseizure drugs currently in clinical development107

Nevertheless, despite the advent of newer antiepileptic agents in the past decades, one-third of patients still remain intractable.108 Regretfully, there has not been significant progress in preventing epileptogenesis or reversing its underlying mechanism.109

These unfulfilled needs in the treatment of epilepsy require a paradigm shift for drug discovery that is based on our expanding knowledge of epileptogenesis, seizure generation, propagation and emergence of pharmacoresistance, and comorbid conditions. It is expected that the use of proteomics and molecular genetics will lead to new medication targets, focused drug delivery, and gene and cell therapies.110 In parallel, a review of the methodology of clinical AED development is required to address ethical and scientific challenges.111


View Abstract


  • Competing interests None declared.

  • Ethics approval This is a review paper of medication trials performed under institutional review board approval.

  • Provenance and peer review Not commissioned; externally peer reviewed.