Remarkable Phenytoin Sensitivity in 4 Children with SCN8A-related Epilepsy: A Molecular Neuropharmacological Approach

Mutations in SCN8A have recently been described in patients with epileptic encephalopathy, intellectual disability, and developmental delay [1‐4]. SCN8A encodes for the sodium channel Nav1.6, one of the major voltage-gated sodium channels (VGSC) in the central nervous system, playing an important role in action potential propagation and spread.

Other VGSC subtypes expressed in the human brain are Nav1.1, Nav1.2, and Nav1.3, encoded by SCN1A, SCN2A, and SCN3A, respectively. Nav1.3 primarily plays a role in the embryonic and early neonatal brain. Nav1.1 is located at inhibitory neurons, while Nav1.2 and Nav1.6 are the main channels in excitatory neurons [5]. Nav1.2 and Nav1.6 are located in the axon initial segment, which is where action potentials are initiated depending on the net result of excitatory and inhibitory signals the neuron receives. Nav1.6 regulates persistent and resurgent current and is voltage dependent. Persistent current is a steady state after firing of the cell, which is important for repetitive firing of the neuron, while resurgent current can occur after initial depolarization and prevents normal inactivation of the channel after depolarization, promoting the firing of another action potential. This allows neurons to fire quickly and repetitively [6]. These properties make Nav1.6 an important channel in neuronal excitability regulation.

In 2012, Veeramah et al. [1] were the first to describe a de novo SCN8A mutation in a patient with infantile epileptic encephalopathy. The patient had seizure onset at 6 months of age, developmental delay, features of autism, ataxia, and intellectual disability. She died at the age of 15 years from sudden unexplained death in epilepsy. Using whole genome sequencing, a p.Asn1768Asp mutation was detected in the patient, which was not found in her healthy parents nor her unaffected sibling. Functional studies showed persistent channel activation and incomplete channel inactivation, hereby contributing to increased firing [1]. After this publication, various other missense SCN8A mutations in patients with epileptic encephalopathy have been described [2, 4, 7, 8]. The only described SNC8A nonsense mutation resulting in haploinsufficiency was found in a patient with cerebellar atrophy, ataxia, and mental retardation, but without epilepsy [9]. Furthermore, functional studies of selected missense mutation detected in an epileptic encephalopathy patient both confirm gain-of-function and loss-of-function effects [3, 7, 8]. Given the fact that the Nav1.6 channel is mainly located in excitatory neurons [5], gain-of-function mutations would lead to increased firing of the neuron, making the brain more sensitive to epileptic discharges.

Dravet syndrome is caused by a loss-of-function mutation in SCN1A, which encodes for the Nav1.1 channel, located in inhibitory neurons. In these patients, a truncating mutation or deletion in SCN1A leads to a lack of well-functioning Nav1.1 channels. Given the fact that Nav1.1 channels are located in inhibitory neurons, this leads to impaired inhibition and therefore to increased excitation and epilepsy. Therefore, mutations in SCN1A and SCN8A both present clinically with epilepsy but probably have an opposite working mechanism (loss-of-function mutation leading to loss of inhibition, and gain-of-function mutation leading to increased excitability of excitatory neurons, respectively), both leading to increased excitability.

Phenytoin is a sodium channel blocker and binds at a receptor site in the pore of sodium channels, decreases the sodium influx, and thereby decreases the excitability of the neuron. Given the fact that phenytoin is one of the few sodium channel blockers targeting only the sodium channel and no other molecular targets in the brain, this might explain the extraordinary response to phenytoin treatment in our patients, in contrast with the response to treatment with other sodium channel blockers. Carbamazepine, another frequently used sodium channel blocker in epilepsy, is thought to have a similar working mechanism to phenytoin. However, the affinity of carbamazepine for inactivated sodium channels is around three times lower than that of phenytoin [10]. Phenytoin works in a frequency-dependent way: it is found primarily to inhibit high-frequency firing but not low-frequency firing, which explains the blocking effect of phenytoin on repetitive discharges as seen in epilepsy, while not interfering with normal neuronal firing. Therefore, phenytoin could be an important treatment option in patients with SCN8A-related epilepsy because it is primarily expected to block the increased sodium current through the mutated Nav1.6, but not to affect the function of other VGSC not affected by the SCN8A mutation. Side effects of phenytoin include ataxia, nystagmus, cerebellar atrophy, and gingival hyperplasia. Cardiotoxicity (hypotension, prolonged QT interval, arrhythmias) is rare after oral administration but is an important side effect after rapid phenytoin administration IV. The therapeutic range of total phenytoin is a serum level between 10 and 20 mg/l (free phenytoin 1–2 mg/l). Total serum levels >20 mg/l may cause nystagmus, ataxia, and tremor, while total serum levels >40 mg/l can lead to lethargy, confusion, seizures, and coma [11].

The reported patients experienced side effects of phenytoin. However, both the physician and the patients’ parents considered the beneficial effects to outweigh the side effects. In patient 1, only very high serum levels of >30 mg/l led to better seizure control. Neurological side effects were acceptable, which could possibly be explained by the alteration in Nav1.6 function and neuronal excitability, making her probably less sensitive to phenytoin toxicity.

A previous functional study of the SCN8A mutation in patient 4 demonstrated a gain-of-function effect in line with the effect observed for the first SCN8A patient that was described [7]. However, both gain- and loss-of-function effects of SCN8A mutations have been described and further functional studies of SCN8A mutations and corresponding phenytoin response must be carefully examined to ensure adverse reactions are avoided [3, 7, 8]. The described loss-of-function effects of 2 different missense mutations may indicate a dominant negative effect leading to a severe phenotype. Alternatively, it remains possible that gain-of-function effects have gone unobserved owing to the nature of the assay and cell system used for functional characterization [3, 7]. Further evaluation of putative loss-of-function mutations is needed, as disease could be aggravated by phenytoin treatment in patients with loss-of-function SCN8A mutations. It therefore remains uncertain if a beneficial effect of phenytoin depends on the functional effects of missense mutations. Future studies are also needed to overcome the low number of observations and lack of control of confounders. However, based on the findings in the abovementioned patients we consider treatment with (high-dose) phenytoin as a possible treatment option in patients with difficult-to-control seizures due to an SCN8A mutation. Dosing should be titrated gradually and plasma levels above the supposed therapeutic window (i.e., “supratherapeutic”) might be required in order to obtain acceptable seizure control. Owing to saturation pharmacokinetics of (phos)phenytoin, it is important to wait until steady state is reached before increasing the dose further, especially when aiming for supratherapeutic plasma levels. Finally, our observations should be interpreted with care given the low number of patients and lack of control of confounders. Further observations and in vitro or in vivo modeling of the response of SCN8A mutations to sodium blockers are needed.