Experimental studies in diverse preparations, including brain slices and whole animals, have led to the view that the generation of focal epileptic activity represents an imbalance between synaptic excitation and inhibition. Either a relative excess of excitation or a relative deficiency of inhibition can predispose individuals to the generation of epileptic activity. γ-aminobutyric acid (GABA) is the major neurotransmitter that mediates fast (millisecond time scale) inhibition in the mammalian central nervous system. The earliest antiepileptic agents, including bromide and phenobarbital, as well as newer drugs, including benzodiazepines, tiagabine, and vigabatrin, are believed to act by enhancing the actions of GABA as an inhibitory neurotransmitter. Glutamate is now well accepted as the neurotransmitter that mediates fast synaptic excitation. Neurotransmitter glutamate excites neurons through an action on ionotropic glutamate receptors, which are tetrameric protein complexes localized at synapses. Glutamate released by the presynaptic neuron diffuses across the synaptic cleft, where it encounters glutamate receptors in the postsynaptic membrane. Binding of glutamate causes a conformational change in the receptors so that they transition from the closed (ion impermeant) state to the open (ion conducting) state. Flow of ions through glutamate-bound open receptors generates the excitatory postsynaptic potential (EPSP) as well as the pathologic neuronal excitation that mediates epileptiform discharges in focal epilepsies. It has been recognized since the early 1980s that drugs that inhibit glutamate-mediated excitation can inhibit epileptiform discharges in in vitro systems and can also protect against seizures in animal models (1).

There are three families of ionotropic glutamate receptors, designated N-Methyl-D-aspartate (NMDA), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), and kainate (2). The different families have different roles. Because of the availability of selective pharmacologic blockers, the earliest experimental studies focused on NMDA receptors as a potential antiseizure target. While NMDA receptor antagonists do protect against seizures in some animal models, NMDA receptor antagonists have not proven to be useful in the treatment of epilepsy in humans (3). NMDA receptors make variable, but generally relatively small, contributions to the EPSP and also to the pathologic depolarization that mediates epileptiform discharges in focal epilepsies. The bulk of these electric events is generated in postsynaptic neurons by AMPA receptors. Consequently, blockade of NMDA receptors fails to eliminate epileptiform activity in many in vitro seizure models. In contrast, blockade of AMPA receptors reliably inhibits epileptiform discharges, providing a basis for the use of AMPA receptors in the treatment of epilepsy. Studies with early AMPA receptor antagonists in animal seizure models (4,5) and also the kindling model of epilepsy (6) demonstrated that selective blockade of AMPA receptors could represent a strategy to confer protection against epileptic seizures. These concepts emerged in the 1990s, but it was only with the introduction in 2012 of perampanel, the first AMPA receptor antagonist approved for human use, that AMPA receptors were validated as a human antiseizure target.

Perampanel was discovered around 2000 at the Tsukuba Science City Research Laboratories of the Eisai Company in Ibaraki Prefecture, Japan (7). A precursor molecule was identified by high throughput screening in cell-based in vitro models assessing activity as an AMPA receptor blocker. The core structure of this initial “hit” was modified to improve the potency for inhibition of AMPA receptors and to enhance metabolic stability. A series of analogues were then assessed for in vivo activity in a mouse seizure model in which the test outcome was prolongation of the latency to clonic seizures during constant intracerebroventricular infusion of AMPA. Among a variety of structures with activity in this model, perampanel had the highest in vitro and in vivo potencies. The IC₅₀ for inhibition of responses to 1 μM AMPA was 0.06 μM and the minimum effective dose in the mouse AMPA infusion model was 2 mg/kg orally. In studies with human liver microsomes, perampanel was predicted to have a low rate of intrinsic clearance and was later found to have a prolonged duration of action following oral dosing. Pharmacokinetic studies in rats indicated good oral bioavailability and excellent blood–brain barrier penetration. In sum, perampanel is a structurally novel, high potency, orally bioavailable, centrally acting AMPA receptor antagonist with favorable pharmacokintetic properties.

CHEMISTRY, ANIMAL PHARMACOLOGY AND TOXICOLOGY, AND MECHANISM OF ACTION

Chemistry

The International Union of Pure and Applied Chemistry (IUPAC) chemical name for perampanel is 2-(2-oxo-1-phenyl-5-pyridin-2-yl-1,2-dihydropyridin-3-yl) benzonitrile hydrate (4:3). The empirical formula is C₂₃H₁₅N₃O • ¾ H₂O (MW = 362.90). Perampanel is a white to yellowish white, nonhygroscopic powder. Perampanel exists in different polymorphic or pseudopolymorphic forms, including five anhydrous forms and one hydrate; the drug substance perampanel is controlled as the 3/4 hydrate (Figure 59.1). Perampanel is practically insoluble in water.

FIGURE 59.1 Structural formula of perampanel.

Preclinical Efficacy Studies

Perampanel has broad spectrum antiseizure activity in animal seizure models, including audiogenic seizures in DBA/2J mice, the maximal electroshock test, the pentylenetetrazol test, and the 6 Hz test (Table 59.1). On the basis of dose, perampanel is among the most potent antiseizure agents known. However, perampanel, as is the case for other AMPA receptor blockers evaluated in animal seizure models (5), caused motor impairment at doses in the range of those conferring seizure protection. The results in the 6 Hz test should therefore be approached with caution as the endpoint is difficult to reliably assess in the presence of neurologic side effects. As has been observed previously for other AMPA receptor antagonists in models of absence seizures (11), perampanel did not have activity in the genetic absence epilepsy rat from Strasbourg (GAERS), an absence seizure model. Whether perampanel can protect against absence seizure in humans remains to be determined.

Preclinical Toxicology Studies

Except where otherwise referenced, data summarized are from the compiled U.S. Food and Drug Administration (FDA) pharmacology reviewer reports of the perampanel new drug application (10).

Acute and Chronic General Toxicology in Adult Animals. Single and repeat-dose toxicology studies were performed in mouse, rat, rabbits, dog, and monkey. In male and female mice, decreased activity and abnormal gait was observed with single oral doses of 100 mg/kg and greater. In male and female rats, single oral doses of 10 mg/kg and higher resulted in abnormal gait, prostration, and decreased activity. In female rabbits, abnormal gait and decreased activity were observed with oral doses greater than 10 mg/kg. In male and female beagle dogs, vomiting, abnormal gait, and decreased activity occurred with doses of 1 mg/kg orally. In a male and female cynomolgus monkey, single oral doses greater than 0.3 mg/kg were associated with ataxia, prostration, and drowsiness.

TABLE 59.1

Source: From Ref. (8) and Ref. (9) except for GAERS, AMPA-induced clonic convulsions, and corneal kindling, which are from Ref. (10).

Daily repeat dosing in mice resulted in ataxia and decreased activity, decreased body weight, and excoriation believed to be due to excessive grooming at doses of 10 mg/kg and greater. Mice receiving a daily dose of 30 mg/kg orally lost weight and some animals died. Chronic toxicity studies in rats demonstrated abnormal gait and decreased activity at daily doses of 6 mg/kg orally and greater. The incidence of abnormal gait and decreased activity tended to diminish progressively over 1 to 4 weeks, indicating the development of tolerance. Chronic daily oral dosing in rats for longer than 99 days led to an increase in the incidence of convulsions in relation to controls (although controls were scored as having convulsions on a small percentage of days). The increase in incidence of convulsions occurred at 10 and 30 mg/kg in males and 3 and 10 mg/kg in females; a lower incidence occurred at higher doses, presumably due to the antiseizure effect of the drug. This raises the possibility that long-term exposure to subtherapeutic doses of perampanel may have a risk of convulsions. No consistent organ or histopathologic changes were found. Beagle dogs exhibited vomiting with oral daily doses of 1 mg/kg or more. At a dose of 1 to 3 mg/kg, the dogs exhibited abnormal gait, decreased activity, and prostration. Some dogs exhibited a transitory abnormality in gait at a dose of 0.3 mg/kg. Cynomolgus monkeys receiving perampanel orally once daily for 4 weeks at doses of 0.3 mg/kg exhibited ataxia and decreased activity. Mean C_max values at this dose were 83 ng/mL in males and 98 ng/mL in females.

Genotoxicity and Carcinogenicity. No evidence of genotoxicity or carcinogenicity was found with perampanel in studies with Sprague-Dawley (SD) rats and CD-1 mice. However, perampanel was found to be a clastogen but not a mutagen in V79 Chinese hamster cells in the presence of UV irradiation, indicating that it is a photo-clastogen but not a photo-mutagen.

Phototoxicity. In in vitro studies, perampanel was phototoxic to BALB/3T3 cells in the presence of UVA irradiation (IC₅₀, 0.39 μg/mL) but did not kill the cells in the absence of irradiation. However, in in vivo studies there was no evidence of phototoxicity, contact hypersensitivity, or photoallergy in male hairless guinea pigs.

Acute Effects on Behavior. In male SD rats, a 0.3 mg/kg single oral dose of perampanel was not associated with signs of abnormal behavior, whereas a 1 mg/kg dose caused a minimal change in abdominal tone at 1 and 2 hours after dosing and a 5 mg/kg dose was associated with altered alertness, spontaneous activity, touch response, body position, staggering gait, limb tone, grip strength, body tone, and palpebral opening from 0.5 to 4 hours after dosing.

Reproductive Toxicology. Oral administration of radioactive perampanel (1 mg/kg) as a single dose to pregnant rats is associated with distribution of radioactivity throughout the fetus, indicating that the drug is readily transferred across the placenta. Radioactivity is recovered in the fetal blood, brain, heart, lung, liver, kidney, and digestive track at levels that are a substantial fraction of the levels in the corresponding maternal tissues (typically, 10%–50%). Administration of a single oral dose of radioactive perampanel to lactating rats 4 days postparturition result in levels in milk that are 2.5 to 3.9 times those in plasma during the period 24 hours after dosing. Up to 4 hours following dosing, most of the radioactivity in milk is associated with the parent compound, indicating that unbiotransformed perampanel is excreted in breast milk for at least 4 hours after dosing.

Perampanel at doses up to 30 mg/kg daily in male and female rats for 14 days prior to mating and during mating did not affect fertility, although prolonged and irregular estrous cycles were observed in female rats at 30 mg/kg. Oral daily dosing for 12 days in pregnant rats at doses of 10, 30, and 60 mg/kg was associated with fetal loss and reduced fetal body weight at the 30 and 60 mg/kg doses. Perampanel was found to be teratogenic in SD rats in doses as low as 1 mg/kg. A small number of offspring of pregnant animals receiving oral perampanel on gestation days 6 to 17 exhibited intestinal diverticuli. Pregnant New Zealand white rabbits dosed orally with perampanel at doses of 3 mg/kg and greater on gestational days 6 to 18 had fetal loss.

Developmental Toxicology. SD rat pups born to dams treated orally with perampanel at doses up to 10 mg/kg per day on gestation day 6 to postnatal day 6 were delivered normally and there was no effect on gestation, the number of live offspring at birth, sex ratio, the number of stillbirths, birth or viability measures, or number of external abnormalities, despite the dams exhibiting abnormal gait, decreased activity, and prostration.

Toxicity in Juvenile Animals. Several nonclinical studies in rats and dogs have been conducted to assess the safety of perampanel in pediatric populations. SD rat pups were dosed orally with perampanel for 12 weeks from postnatal day 7 to 90 followed by a 4-week recovery period. The juvenile rats exhibited reduced activity, incoordination, excessive grooming, and excessive scratching even at a dose of 1 mg/kg/day, the lowest dose tested, and the incidence of these clinical signs increased in a dose-dependent fashion with higher doses. At high doses (titrated up to 30 mg/day), there was reduced growth and body weight, and delayed sexual maturation (preputial separation, vaginal opening). Hindlimb grip strength was reduced during dosing, but this did not persist in the recovery period. Reproductive performance was not affected. Adverse effects of chronic treatment on learning and memory as assessed by the Cincinnati water maze were present but were not dramatic. During dosing, there appeared to be a dose-dependent reduction in learning, particularly at doses of 3 mg/kg and greater; the impact was present but much reduced following discontinuation of drug treatment. C_max and AUC_0-24h values associated with the effects on learning were comparable to those expected to be achieved with clinically relevant doses in humans, indicating that persistent effects on learning could occur with chronic treatment. However, it is noteworthy that other AMPA receptor antagonists have generally not been found to impact memory formation or retrieval in animal models even at doses that impair motor function (13). Whether perampanel affects memory function in humans remains to be determined. Nevertheless, it is apparent that juvenile rats are more susceptible to the sedative-like clinical signs and sterotypies of perampanel than are adult rats.

Juvenile dogs dosed daily with oral perampanel for 33 weeks from postnatal day 42 to postnatal week 39 showed incoordination, abnormal gait, altered activity, excessive scratching, and tremors even at doses as low as 1 mg/kg/day, which are associated with C_max values of 24 to 99 ng/mL (area under the curve, AUC_0-24h: 300–681 ng•h/mL). However, the animals gained weight normally, there was no organ pathology, and their brain measurements were normal with no apparent histopathologic changes (although there was slight ventricular dilation in animals titrated to a dose of 10 mg/kg/day).

Both juvenile rats and juvenile dogs were more susceptible to neurologic toxicity than adult animals. For example, in adult dogs, a daily oral dose of 10 mg/kg, which is associated with C_max values of about 130 ng/mL (AUC_0-24h: 819 ng•h/mL), caused only abnormal gait, whereas, as noted, this and other toxicities were observed in juvenile dogs at a 10-fold lower dose associated with lower exposures. The results in animals indicate that caution is warranted in the use of perampanel in children and young adults. As discussed later in this chapter, there is evidence from clinical studies that juveniles may be at greater risk for adverse behavioral effects of perampanel than are adults.

Cardiac Safety. Drug-induced prolongation of the QT interval (delay of cardiac repolarization) presents a risk for cardiac arrhythmias, most commonly torsade de pointes, which can be fatal. The potential for QT interval prolongation was assessed with cell-based assays, animal safety studies, a study in healthy volunteers, and in a pooled analysis of data from the Phase III clinical trial program (14). Studies with HEK-293 cells transfected with hERG potassium channels indicate that perampanel inhibits hERG tail current with an IC₅₀ of 15.8 μM. In healthy volunteers, the maximum (12 mg) dose of perampanel administered under fasting conditions is associated with perampanel concentrations of less than 1,400 ng/mL (3.9 μM) [mean C_max levels <400 ng/mL (1.1 μM)], whereas the maximum concentration observed in the Phase III clinical trials was ~2,500 ng/mL (6.9 μM) (14), providing a margin of safety, particularly in relation to expected free concentrations (4%–5% of total plasma levels). In male and female beagle dogs, a 1 mg/kg single oral dose of perampanel failed to affect mean arterial blood pressure or the electrocardiogram. A dose of 10 mg/kg caused a transitory (4 hours after dosing) increase in heart rate but had no other cardiac effects, including no effect on the QT interval. In healthy male and female human volunteers, oral daily perampanel at doses of 6 and 12 mg did increase the QT interval corrected for heart rate. In the Phase III clinical trials, there was no difference in corrected QT interval between patients with partial seizures treated with placebo and those treated perampanel. Moreover, there was no correlation between plasma perampanel concentrations and QT interval duration with plasma concentrations up to 2,500 ng/mL. Overall, there is no evidence that perampanel presents a clinically relevant risk of QT prolongation.

Preclinical Absorption, Distribution, Metabolism, and Excretion

Perampanel has moderate oral bioavailability in rats (~47%) and dogs (~49%) and higher bioavailability in monkeys (~74%). Plasma protein binding is 94% in mice, 87% in rats, 90% in dogs, and 90% in monkeys. Perampanel mainly binds to albumin and α1-acid glycoprotein and, to a much lesser extent, γ-globulin. It distributes widely throughout the body, including to the fetus of the pregnant rat, and also appears in breast milk (see Toxicology, discussed earlier). Perampanel is rapidly eliminated from the brain, but there is extended residence in various other tissues (up to 106 weeks), including in the eyes of pigmented rats and cynomolgus monkeys, but there is no evidence of ocular toxicity or retinal damage. In rats, but not monkeys, perampanel also appears to exhibit prolonged binding to arteries, but not veins. This unusual observation was not associated with any histopathologic abnormalities assessed by light microscopy or any evidence of cardiovascular toxicity.

In SD rats, following a single oral dose of radioactive perampanel (1 mg/kg), the radioactivity is completely recovered by 7 days in the feces (88%) with the remainder in the urine (12%). In rats with bile duct cannulae, the radioactivity is completely recovered by 2 days in the bile (92%) and in the urine (7%). In cynomolgus monkeys receiving a single oral dose of radioactive perampanel (0.3 mg/kg), 94% of the radioactivity is recovered by 7 days in the feces (57%) and urine (37%). At this time, no radioactivity is detected in the brain or spinal cord, although there is some radioactivity in the eye as well as in the liver and gallbladder. The metabolic profile of perampanel in rats and monkeys is generally similar to that in humans as discussed subsequently.

Mechanism of Action

Perampanel was first demonstrated to inhibit AMPA receptors using a cell culture assay where changes in intracellular free Ca²⁺ concentration were measured using the fluorescent Ca²⁺ indicator dye fura-2 (8). Exposure of rat cortical neurons to AMPA (2 μM) caused a Ca²⁺ response that was blocked in a concentration-dependent fashion by perampanel with IC₅₀ of 0.093 μM. The source of the free Ca²⁺ in these studies is not defined but could relate to entry through various pathways activated by the depolarizing action of AMPA receptor activation (such as voltage-activated Ca²⁺ channels) or to release from intracellular stores. Perampanel was shown more directly to block AMPA receptors in whole-cell voltage-clamp studies in cultured rat hippocampal neurons (15). Perampanel block (IC₅₀, 0.56 μM) was unaffected by the agonist concentration when kainate, a nondesensitizing agonist, was used to activate AMPA receptors, demonstrating a noncompetitive blocking action. AMPA receptor currents exhibit rapid desensitization when activated by the natural agonist glutamate or with AMPA, but perampanel did not alter the time course of the currents. In contrast to its effect on AMPA receptor currents, perampanel had no effect on NMDA receptor currents. Experiments in brain slices have confirmed that perampanel blocks AMPA receptor-mediated synaptic transmission (16). Field EPSPs in hippocampal slices were inhibited with an IC₅₀ of 0.23 μM, whereas perampanel did not affect synaptic responses mediated by NMDA or kainate receptors. With chronic daily administration in human patients with epilepsy who are not taking enzyme-inducing concomitant medications, mean perampanel serum concentrations are in the range of 500 ng/mL (1.43 μM) (17). As protein binding is 95%, the unbound concentration is estimated to be about 0.07 μM. If this is taken as an estimate of extracellular concentrations in the brain representing the concentration available at AMPA receptors, it is apparent that in clinical use perampanel only blocks a fraction of AMPA receptors (perhaps <25%). Complete block of AMPA receptors would be incompatible with brain function. However, a degree of partial block seems to be tolerated, although there is clinically relevant protection against seizures.

Studies on Drug-Metabolizing Enzymes

In vitro studies with human liver microsomes have indicated that perampanel (30 μM) does not inhibit cytochrome P450 (CYP) 1A2, CYP2A6, CYP2C9, CYP2C19, CYP2D6, CYP2E1, uridine 5’-diphospho-glucuronosyltransferase (UGT) 1A1, UGT1A4, or UGT1A6 (15). While it is a weak inhibitor of CYP2C8, UGT1A9, and UGT2B7, and possibly CYP3A4, the IC₅₀ values are greater than 30 μM, indicating that the inhibitory actions on these enzymes are not clinically relevant. However, studies with recombinant human CYP isoforms indicated that in primary cultured human hepatocytes, perampanel (3 μM but not 0.3 μM or lower concentrations) weakly induces CYP2B6, CYP3A4/5, CYP2B5, UGT1A1, and UGT1A4, but not CYP1A2. The inducing actions are not expected to cause significant drug–drug interactions because the magnitude of the effects are modest, but effects on the pharmacokinetics of drugs that are metabolized by those enzymes that are induced cannot be ruled out.

Studies on Drug Transporter Interactions

Except where otherwise referenced, data summarized in the remainder of this chapter are from the compiled FDA clinical pharmacology reviewer reports of the perampanel new drug application (12).

Various transporters may be involved in the absorption, excretion, distribution, and intracellular concentration of drugs. Actions of a drug at these transporters may influence its pharmacokinetic properties and may play a role in drug–drug interactions. There is no evidence that perampanel interacts with such transporters in a way that is of clinical relevance. Studies with cell lines overexpressing human P-glycoprotein and breast cancer resistance protein indicated that perampanel is not a substrate of either multidrug transporter (18). Perampanel also is not a substrate or inhibitor of OATP1B1 or OATP1B3 transporters; or of organic anion transporters OAT1, OAT2, OAT3; or of organic cation transporters OCT1, OCT2, and OCT3. It does inhibit OAT1, OAT3, OCT1, and OCT3 in a concentration-dependent manner, with OAT3 being the most sensitive (K_i, 8.5 μM). Perampanel stimulated OAT2-mediated transport at concentrations of 1 μM and above. Because the unbound concentration of perampanel is estimated to be much lower than the affinity values, these transporter interactions are not believed to be clinically relevant.

CLINICAL EFFICACY AND TOLERABILITY

Stay updated, free articles. Join our Telegram channel

Join