Fycompa MOA: How Perampanel Blocks AMPA Receptors

Fycompa (perampanel) works by blocking AMPA receptors, one of the main types of receptors that respond to glutamate, the brain’s primary excitatory chemical messenger. Unlike older seizure medications that act on sodium channels or enhance inhibitory signaling, perampanel targets excitatory transmission directly and does so through a non-competitive mechanism, meaning it doesn’t compete with glutamate for the same binding spot on the receptor. Instead, it latches onto a separate site and changes the receptor’s shape so that even when glutamate binds, the receptor can’t fully open. That distinction shapes nearly everything about how the drug behaves in practice.

How Perampanel Blocks AMPA Receptors

AMPA receptors sit on the surface of neurons at synapses, and when glutamate binds to them, they open briefly to let charged particles flow into the cell. That influx is the basic spark of excitatory signaling in the brain. Perampanel doesn’t try to outcompete glutamate at the receptor’s glutamate-recognition site. Instead, it binds to a completely different spot on the receptor, an allosteric site, and in doing so prevents the receptor’s ion channel from opening normally even when glutamate is present.¹ Structural studies have confirmed this by imaging the receptor with perampanel physically bound to it, showing the drug nestled into a pocket in the transmembrane region of the receptor complex.²

This non-competitive approach matters for a practical reason. During a seizure, glutamate levels can surge to very high concentrations. A competitive blocker, one that fights glutamate for the same binding pocket, can be overwhelmed when glutamate floods the synapse. Perampanel sidesteps that problem because its effect doesn’t depend on beating glutamate in a binding race. No matter how much glutamate is present, perampanel’s grip on its allosteric site keeps the receptor from conducting normally.

Selectivity for AMPA Over Other Glutamate Receptors

The brain has several families of glutamate receptors, and they don’t all do the same thing. AMPA receptors handle fast excitatory transmission, the quick, moment-to-moment signaling that drives normal brain activity and, when it goes haywire, seizures. NMDA receptors are involved in slower, more complex processes like learning and memory formation. Kainate receptors play more modulatory roles. A drug that indiscriminately blocked all of them would be effective against seizures but would also likely cause severe cognitive and neurological problems.

In lab tests on cultured neurons, perampanel blocked AMPA-driven calcium influx at very low concentrations, with only minimal effects on NMDA-driven responses, and those only at concentrations far higher than what the drug reaches in actual patients.³ When researchers moved from isolated neurons to hippocampal brain slices, a more realistic system, the selectivity held up. Perampanel strongly reduced AMPA receptor-mediated synaptic responses while leaving NMDA and kainate receptor-mediated responses largely intact.⁴ This selectivity is a big part of why perampanel can reduce seizures without wiping out the NMDA-dependent processes that underlie learning and memory.

From Receptor Blockade to Seizure Control

Blocking AMPA receptors on individual neurons is one thing. Stopping a seizure, which involves runaway electrical activity across large networks of neurons, is a more complicated story. Research using a kindling model of epilepsy found something interesting: perampanel doesn’t just quiet excitatory neurons. It also, somewhat counterintuitively, reduces burst firing in inhibitory neurons and diminishes inhibitory inputs onto excitatory pyramidal neurons. The net result is still a powerful suppression of the abnormal synchronized discharges that define a seizure.⁵

This network-level picture helps explain why the drug works in practice. Seizures aren’t just about single overexcited neurons; they involve waves of synchronization spreading through circuits. By dampening excitatory transmission at the synaptic level, perampanel weakens the ability of a seizure focus to recruit neighboring neurons and propagate across the brain. In comparative studies, perampanel’s effect persisted even when seizure thresholds were pushed much higher, suggesting it directly and strongly modulates excitatory transmission both at the seizure’s origin and along its propagation route.⁶ Sodium channel blockers, by contrast, appeared to act mainly by raising the threshold for a neuron to fire in the first place, a fundamentally different strategy.

Why This Matters During Status Epilepticus

One of the more striking implications of perampanel’s mechanism relates to prolonged seizures, or status epilepticus. When seizures continue for an extended period, the brain’s receptor landscape rapidly changes. Inhibitory GABA receptors get pulled off the cell surface, which is why benzodiazepines, the standard first-line treatment, lose their effectiveness the longer a seizure goes on. At the same time, AMPA and NMDA receptors are increasingly trafficked to the surface, making excitatory signaling even stronger.⁷

This receptor trafficking essentially makes the brain more resistant to drugs that rely on inhibition while potentially making it more responsive to drugs that target excitation. AMPA receptor antagonists remain effective treatments even after benzodiazepines have failed, precisely because the receptors they target are still on the surface and, if anything, more abundant. That pharmacological logic has driven growing clinical interest in using perampanel as a later-line option in refractory status epilepticus, although clinical trial data for this specific use are still accumulating.

Getting Into the Brain and Staying There

A seizure medication is useless if it can’t cross the blood-brain barrier. Perampanel does cross it, with studies in rats showing a brain-to-plasma ratio of about 0.6, meaning the drug reaches the brain at concentrations roughly proportional to what’s circulating in the blood.⁸ Once there, it is cleared from brain tissue relatively quickly, but plasma levels sustain the drug’s presence over time.

In the bloodstream, perampanel is heavily protein-bound, with more than 95% attached to plasma proteins at any given moment.⁹ The drug is broken down almost entirely by the liver enzyme CYP3A, and its elimination half-life averages around 105 hours, roughly four to five days.¹⁰ That long half-life is why perampanel is dosed once daily, typically at bedtime. It also means the drug takes a couple of weeks to reach a stable level in the body after a dose change, and roughly the same amount of time to wash out if it’s stopped.

Drug Interactions and Dose Adjustments

Because perampanel depends almost entirely on CYP3A for metabolism, any other drug that revs up or slows down that enzyme will change perampanel’s blood levels. This is not a theoretical concern. Many commonly used seizure medications are potent enzyme inducers. In therapeutic drug monitoring studies, patients taking enzyme-inducing seizure medications had significantly lower perampanel concentration-to-dose ratios than patients not on such drugs, meaning their bodies were clearing perampanel much faster.¹¹ In practice, patients on enzyme-inducing co-medications often need higher doses of perampanel to reach the same therapeutic blood levels.

Interestingly, the enzyme inhibitor valproate did not significantly change perampanel levels in the same monitoring data, suggesting that CYP3A inhibition in vivo may not boost perampanel concentrations as much as one might predict from test-tube experiments. When both enzyme inducers and inhibitors were present together, the inducing effect appeared to dominate. These interactions matter because inadequate perampanel levels mean breakthrough seizures, while excessive levels increase side effects. Clinicians managing patients on multiple seizure medications often rely on blood level monitoring to fine-tune the dose.

What Perampanel Does to Brain Waves

Blocking a major class of excitatory receptors inevitably changes the brain’s background electrical activity, and quantitative EEG studies have mapped out those changes. In adults with epilepsy, perampanel increased power in the theta frequency band across all brain regions while decreasing alpha power, and the alpha rhythm’s peak frequency slowed. Despite this electrophysiological slowing, standardized cognitive testing did not show a corresponding decline in thinking ability.¹² The degree of alpha power change correlated with the drug’s blood level, suggesting a dose-dependent effect on background brain rhythms.

A separate study looking at both children and adults found a different pattern: an increase in faster beta frequencies along with a decrease in slow delta activity.¹³ The increase in theta power has also been confirmed in patients with focal epilepsy.¹⁴ These EEG signatures are useful for researchers and could eventually help clinicians gauge whether the drug is reaching effective concentrations in the brain, but for patients, the reassuring finding is that the EEG slowing doesn’t seem to translate into noticeable cognitive problems at standard doses.

Aggression and the Glutamate Connection

One side effect that has drawn particular attention is aggression and irritability, reported in a subset of patients taking Fycompa. This is not just an anecdotal concern; it appears in clinical trial data and has been serious enough to warrant a label warning. Animal research has shed some light on why this might happen. In mice given perampanel chronically, aggressive behavior increased measurably. When researchers examined the hippocampus of these animals, they found significant changes in roughly 93 proteins involved in synaptic function, including an upregulation of GluA1, one of the AMPA receptor subunits, and changes in signaling proteins downstream of the receptor.¹⁵

The working hypothesis is that chronic AMPA receptor blockade prompts the brain to compensate by ramping up excitatory receptor components and altering synaptic scaffolding. That compensatory remodeling could shift the balance of excitation and inhibition in circuits involved in emotional regulation and impulse control. This is one area where the mechanism of action has a direct clinical consequence that both prescribers and patients need to be aware of, particularly at higher doses.

Neuroprotection Beyond Seizure Control

AMPA receptors don’t just matter for seizures. When the brain is deprived of blood flow during a stroke, neurons release massive amounts of glutamate, and the resulting overstimulation of glutamate receptors, particularly AMPA receptors, can kill neurons through calcium overload. This process, called excitotoxicity, has made AMPA receptor blockade an appealing strategy for protecting the brain after injury. In rat models of stroke, perampanel significantly reduced neuronal death in the hippocampus and improved long-term potentiation, the cellular process underlying memory formation.¹⁶ These effects appear to work partly by preventing AMPA receptor subunits from being pulled off the cell surface during ischemia, a process that normally makes neurons more vulnerable to damage.

What’s particularly interesting is the dose range involved. In brain slice experiments simulating stroke conditions by depriving tissue of oxygen and glucose, perampanel provided neuroprotection at doses low enough that they didn’t significantly reduce normal excitatory transmission or interfere with learning-related synaptic strengthening.¹⁷ The drug also reduced neuroinflammation by dampening the activation of microglia and astrocytes, the brain’s immune cells, after ischemic injury. These findings remain preclinical, and translating neuroprotection from animal models to human stroke patients has historically proven difficult for the entire glutamate-targeting drug class. But perampanel’s advantage is that it’s already an approved, orally available drug with a known safety profile, lowering the barrier to clinical testing.

The ALS Story and Its Cautionary Lesson

Perhaps the most ambitious extension of perampanel’s mechanism has been in amyotrophic lateral sclerosis, where motor neurons progressively degenerate and die. In the majority of ALS patients, a molecular error means that the GluA2 subunit of AMPA receptors is improperly edited, making the receptors permeable to calcium. The resulting chronic calcium influx is believed to contribute to motor neuron death. In genetically engineered mice that model this process, oral perampanel significantly slowed the ALS-like disease progression and prevented the death of motor neurons.¹⁸

Those animal results generated real excitement, and a randomized phase 2 trial was conducted in people with sporadic ALS. The outcome, however, was sobering. Perampanel was actually associated with a decline in functional scores, and the higher dose group showed worsening of bulbar symptoms, the muscles controlling speech and swallowing.¹⁹ The trial’s failure illustrates a broader challenge: a drug mechanism that looks right on paper can still miss the mark in a complex human disease. It’s possible that blocking all AMPA receptors broadly, rather than selectively targeting the abnormal calcium-permeable ones, removes too much excitatory drive from motor circuits that are already struggling. The ALS work remains a valuable example of how understanding a drug’s mechanism in detail can generate plausible but ultimately incorrect therapeutic hypotheses.

Combination Therapy and Synergy

Because perampanel’s mechanism is fundamentally different from most other seizure medications, it’s a natural candidate for combination therapy. In a mouse kindling model, pairing perampanel with lacosamide, a sodium channel blocker, proved more effective than either drug alone at reducing seizure progression. The combination also outperformed monotherapy on behavioral measures, reducing anxiety and depression associated with repeated seizures, and produced less oxidative stress in brain tissue.²⁰ The rationale is straightforward: one drug reduces a neuron’s ability to fire rapidly, while the other weakens the excitatory synaptic transmission that triggers firing in the first place. Hitting the seizure network at two different points should, in theory, produce better suppression at lower individual doses, and that’s what the preclinical data suggest.

Targeting Specific AMPA Receptor Mutations

One of the more precise applications of perampanel’s mechanism involves rare genetic conditions where a specific AMPA receptor subunit is mutated. In a case involving a gain-of-function mutation in the GluA2 subunit, the mutated receptors produced abnormally large currents. Because perampanel works at an allosteric site that doesn’t depend on the normal glutamate binding process, it was able to fully block the overactive mutant receptors. When perampanel was added to the patient’s treatment, seizure burden dropped markedly, failure to thrive resolved, and developmental progress followed.²¹ This kind of precision use, matching a drug’s mechanism to a specific molecular defect, represents the most targeted version of what perampanel can do and hints at a future where genetic diagnosis guides seizure medication choices more routinely.