Pharmacology of Emerging Selective Sodium Channel Antagonists for the Treatment of Epilepsy

The development of modern synthetic organic compounds as medications to treat epilepsy began with the serendipitous observation that phenobarbital reduced seizures when it was administered for the purpose of sedation in a psychiatric institution.¹ To minimize the sedation, medicinal chemists undertook successive modifications of the structure of phenobarbital, which resulted in a series of hydantoins (typified by phenytoin) and succinimides (typified by ethosuximide).² The development of these compounds relied on their ability to protect rodents from electrically and chemically induced seizures. Knowledge about the role of ion channels in regulating network excitability in the brain was lacking at that time. We now know that a large number of existing antiseizure medications (ASMs) share the ability to block voltage-gated sodium channels (Na_v), and we understand sodium channel blockade to be an important mechanism by which abnormal excitability leading to epileptic seizures can be regulated.

Sir Alan Hodgkin and Sir Andrew Huxley published a landmark article in 1952³ that described the ionic basis of the action potential on the squid giant axon, demonstrating the role of the sodium current in depolarization and the potassium current in repolarization. The work of Dr. Michael McLean in the laboratory of Prof. Robert Macdonald at the University of Michigan firmly established the sodium channel blocking properties of phenytoin,⁴ and subsequently of carbamazepine⁵ and valproic acid.⁶ In the following years, the pharmaceutical industry continued to rely on developing ASMs based on the ability of compounds to protect rodents from seizures induced by electrical or chemical means. Several newer ASMs such as oxcarbazepine, eslicarbazepine, lamotrigine, and lacosamide also display an ability to block Na_v in a use-dependent manner.⁷

A detailed examination of the molecular construct of the Na_v is provided in a review by Prof. William Catterall,⁸ a prominent investigator on sodium channels from the University of Washington. The mammalian sodium channel consists of a large α subunit protein that consists of the pore-forming components as well as the voltage sensor and 1 or 2 smaller β subunit proteins that serve modulatory functions such as gating, voltage dependence, and kinetics of the α subunits. The genes coding for the α subunits are named SCNnA; those coding for the β subunits are named SCNnB. Both loss of function sequence variations in SCN1A and functional deletion of SCN1B can result in Dravet syndrome, a severe form of childhood developmental epileptic encephalopathy. In this article, we restrict our discussion to the α subunit—the primary conductor of the sodium channel current and the target for ASMs. 

Five of the 9 mammalian sodium channel genes (SCN1A, SCN2A, SCN3A, SCN5A, and SCN8A) are expressed in the brain; of those, the 3 pertinent to this discussion are SCN1A, SCN2A, and SCN8A, which encode channels (currents) Na_v1.1, Na_v1.2, and Na_v1.6, respectively (with v indicating voltage-gated). SCN3A mediating Na_v1.3 is important only during fetal development of the brain. The SCN5A protein in the brain is not the same splice variant as in the heart, and Na_v1.5 is important for cardiac conduction. Na_v1.7 is important in peripheral nerves conducting pain signals. Na_v1.1 is encountered only in GABAergic interneurons; Na_v1.2 and Na_v1.6 are found in the axon initial segment of principal excitatory neurons.^9,10 Therefore, a sodium channel blocker (SCB) that can spare Na_v1.1 and block Na_v1.2 and Na_v1.6 would be of potential interest in epilepsy. The fast-spiking, parvalbumin colocalizing (PV+) GABAergic interneurons provide crucial axo-somatic inhibition. Antagonizing Na_v1.1, which activates those inhibitory neurons, would not contribute favorably to limiting network inhibition and may indeed be counterproductive in disorders where Na_v1.1 is already compromised, such as in Dravet syndrome. Furthermore, epileptic encephalopathies specifically involving gain of function sequence variations of SCN2A (Na_v1.2) and SCN8A (Na_v1.6) may also benefit from isoform-specific SCBs.

The inward sodium current that passes through Na_v may be transient (I_NaT)—the large, fast current responsible for the upstroke of the action potential—or a smaller, noninactivating current called the persistent sodium current (I_NaP). The I_NaP is a sustained current that raises the resting membrane potential and brings it closer to the threshold for firing an action potential, thus contributing to what has been termed a paroxysmal depolarizing shift.^11,12 The paroxysmal depolarizing shift was initially thought to be a calcium-mediated current, but evidence has been accumulating that show that the I_NaP may be a significant contributor to this phenomenon, resulting in hyperexcitability.¹³ The I_NaP amplifies a neuron’s response to synaptic input and enhances its capacity for sustained repetitive firing.

The role of I_NaP in epileptogenesis has been noted in genetic epilepsies associated with gain-of-function sequence variations in SCN2A¹⁴ and SCN8A.¹⁵ An increase in I_NaP has also been demonstrated in both experimental and human-acquired epilepsy. In a rat model of temporal lobe epilepsy acquired after a bout of pilocarpine-induced status epilepticus, Chen et al¹⁶ demonstrated an increase in I_NaP in CA1 pyramidal cells that displayed high-threshold bursting behavior. Such bursting behavior prevailed in 86.6% of status epilepticus–experienced neurons sampled several weeks after status epilepticus, but in only 33% of control neurons. Other channel changes observed in epileptogenesis after status epilepticus include an increase in low-threshold calcium channels (Ca_v3.2)¹⁷ and a loss of dendritic HCN channels.¹⁸

I_NaP is a relevant target for the treatment of seizures, and selective blockade of I_NaP over I_NaT may preserve the intrinsic network inhibition provided by fast-spiking, PV+ GABAergic interneurons while addressing hyperexcitable principal neurons. In another model of experimental epileptogenesis, kindling and increased expression of Na_v1.6 protein and increased I_NaP was demonstrated by Blumenfeld et al.¹⁹ Increased I_NaP has also been identified in resected human epileptic tissue.²⁰ The authors hypothesized that the abundant presence of persistent sodium current in half of the subicular neurons could lead to a larger number of neurons with intrinsic burst firing. The extraordinarily large amplitude of the persistent sodium current in this subset of subicular neurons might explain why these individuals are susceptible to seizures and resistant to medication and thus had been selected as candidates for surgical therapy.

Figure 1. Transient sodium current (A). Persistent current (B). Resurgent current (C).
Panels A and B reproduced with permission from Patel RR. Nav1.1 and Nav1.6: electrophysiological properties, epilepsy-associated mutations and therapeutic targets. Dissertation. Indiana University; 2016.
Panel C reproduced with permission from Afshari FS, Ptak K, Khaliq ZM, et al. Resurgent Na currents in four classes of neurons of the cerebellum. J Neurophysiol. 2004;92(5):2831-2843. doi:10.1152/jn.00261.2004. 

Another form of sodium current that can lead to neuronal hyperexcitability is the resurgent current I_NaR resulting from an unstable initial closure of the channel.²¹ This form of instability has been shown to be associated with gain-of-function sequence variations in SCN2A and SCN8A, which cause genetic epilepsies. Pictorial descriptions of I_NaT, I_NaP, and I_NaR are provided in Figures 1 and 2.

Figure 2. Transient, persistent, and resurgent currents in a single illustration.
Reproduced with permission from Baeza Loya S. Voltage-gated ion channels influence firing patterns of vestibular afferents. Signaling in the Inner Ear. Accessed October 15, 2025. https://voices.uchicago.edu/eatocklab/current-projects/selina-baeza-loyas-projec/

With respect to established SCB ASMs, data are sparse with respect to specific blockade of I_NaP vs I_NaT. A review by Rogawski and Löscher²² provides a limited amount of tabular data that show the effect of a number of ASMs on I_NaP and I_NaF, the latter presumably referring to the transient current as the fast current. The table has limited data on the effect of those ASMs on I_NaP, but all those are registered as blockers of I_NaF (ie, I_NaT). A review by Wengert and Patel²³ provides a large list of compounds that attenuate I_NaP, but the effect on I_NaT is not included. A more recent article by Goodchild et al²⁴ shows that phenytoin and carbamazepine have remarkably similar dose–response relationships with all 7 isoforms of the sodium channel (Na_v1.1 through Na_v1.7). These data are consistent with the clinical observation that traditional SCBs worsen conditions like Dravet syndrome in which there is a preexisting deficit in Na_v1.1 due to a loss-of-function sequence variation in SCN1A in 80% to 85% of individuals. Such a sequence variation would severely compromise the firing of the aforementioned GABAergic interneurons furnishing important network inhibition.

Among the available SCBs, cenobamate (Xcopri; SK Life Science, Paramus, NJ) is the only drug for which data show a clear preference for blocking I_NaP over I_NaT.²⁵ At a 100-µM concentration of cenobamate, the I_NaT in hippocampal CA3 neurons was unaffected, whereas I_NaP was decreased to ~25% of control (Figure 3). The IC₅₀ value for I_NaP was 53.1 µM; that for I_NaT was >500 µM. Consistent with these in vitro findings, Makridis et al²⁶ reported successful treatment of adults with Dravet syndrome. The applicability of these findings to children has been challenged in a report by Cagigal et al,²⁷ who found no benefit in their 6 pediatric participants, and status epilepticus in one. However, individuals with Dravet syndrome are prone to status epilepticus, and there have been other unpublished reports of children with Dravet syndrome benefiting from treatment with cenobamate. Furthermore, impressive seizure control as well as nonseizure improvements (in alertness, sleep, and muscle tone) were reported in a series of 12 individuals with SCN8A gain-of-function–related developmental epileptic encephalopathy.²⁸ The clinical efficacy of cenobamate may stem from both selective sodium channel antagonism as well as its effect augmenting extrasynaptic GABAA-mediated tonic inhibition.²⁹ A detailed discussion of the implications of augmented tonic inhibition is not within the scope of this article.

Figure 3. Cenobamate has no discernible effect on transient sodium current (I_NaT) at a 100-μM concentration. The same concentration measurably attenuates persistent sodium current (I_NaP) (A). Concentration–response relationships of cenobamate against the peak (open circles) and steady-state (closed circles) components of I_NaT are shown (B).
Reproduced with permission from Nakamura M, Cho JH, Shin H, Jang IS. Effects of cenobamate (YKP3089), a newly developed anti-epileptic drug, on voltage-gated sodium channels in rat hippocampal CA3 neurons. Eur J Pharmacol. 2019;855:175-182.25 

There has been intense focus on creating isoform-selective (Na_v1.2 as well as Na_v1.6) compounds that spare Na_v1.1 (which is crucial for GABAergic inhibition) by the medicinal chemistry group at Xenon Pharmaceuticals Inc.^24,30,31 One of the Xenon products (XEN901; Vancouver, British Columbia, Canada), licensed to Neurocrine Biosciences (San Diego, CA) as NBI-921352 for clinical development, failed to provide a meaningful reduction in seizure frequency in humans despite strong efficacy in a mouse model of SCN8A gain-of-function epilepsy.^31,32 However, whether issues with pharmacokinetics and the selected dosing regimen rather than the intrinsic pharmacology led to this negative result are unclear. 

Relutrigine (PRAX-562; Praxis Medicines, Boston, MA), which preferentially blocks I_NaP over the peak current (I_NaT), has shown potent preclinical anticonvulsant activity and good tolerability compared with standard Na_v blockers like carbamazepine and lamotrigine.³³ According to recent topline data presented at scientific meetings (the 36th International Epilepsy Congress; Lisbon, Portugal; 2025) and the company website (https://praxismedicines.com), in an ongoing trial of children with SCN2A or SCN8A developmental epileptic encephalopathy, ~1 in 3 became seizure-free while taking the medication.

The ability to generate isoform-selective sodium channel inhibitors that spare GABAergic interneuron function is in its early stages but has exciting possibilities. Selectivity for Na_v1.2 or Na_v1.6 over other isoforms may be likely to lower the cardiac toxicity of these ASMs. This has renewed interest in the development of SCBs, which have been used for nearly a century in the treatment of epilepsy.