Voltage-gated sodium channels (Nav channels) are the molecular target of the largest class of anti-seizure drugs (ASDs). Understanding how these drugs interact with the Nav channel requires understanding the channel's gating states and the pharmacological concept of state-dependent blockade, which explains both their clinical efficacy and the selectivity that makes them tolerable at therapeutic doses.
Voltage-gated sodium channels (Nav channels) are heteromeric transmembrane proteins consisting of a pore-forming alpha subunit and one or more auxiliary beta subunits. The alpha subunit contains four homologous domains (DI–DIV), each with six transmembrane segments (S1–S6). The S4 segment of each domain serves as the voltage sensor, bearing positively charged arginine and lysine residues that move outward in response to membrane depolarization, triggering channel opening. The S5–S6 segments line the ion-conducting pore, and the intracellular loop connecting DIII and DIV contains the inactivation gate, a hydrophobic isoleucine-phenylalanine-methionine (IFM) motif that occludes the pore during fast inactivation.1 The nine Nav channel subtypes (Nav1.1 through Nav1.9) differ in their expression patterns, gating kinetics, and pharmacological sensitivities. Nav1.1, Nav1.2, Nav1.3, and Nav1.6 are the principal subtypes expressed in central nervous system (CNS) neurons and are the primary targets of the ASDs discussed in this module.
Nav channels cycle through three principal functional states during normal neuronal activity. In the resting state, at hyperpolarized membrane potentials, the channel is closed but available for activation. Membrane depolarization drives the channel into the open state, allowing rapid sodium influx that generates the rising phase of the action potential. Within 1–2 milliseconds, the channel spontaneously transitions to the fast-inactivated state, in which the IFM inactivation gate occludes the pore and the channel is refractory to further activation despite continued depolarization. Recovery from fast inactivation, which requires membrane repolarization, returns the channel to the resting state and takes several milliseconds. A fourth state, slow inactivation, develops over hundreds of milliseconds to seconds of sustained depolarization and involves conformational changes in the pore-lining S6 segments rather than the IFM gate; recovery from slow inactivation is much slower, taking seconds to minutes.2
The concept of state-dependent blockade is central to understanding how sodium channel ASDs work. Phenytoin, carbamazepine, oxcarbazepine, lamotrigine, and zonisamide all bind preferentially to the fast-inactivated state of Nav channels, stabilizing them in the inactivated conformation and reducing the pool of channels available for the next action potential. This binding is reversible and voltage-dependent: at normal resting membrane potentials, drug binding is weak, and channels recover rapidly and are available for normal physiological firing. At the depolarized potentials maintained during ictal high-frequency burst firing, a greater proportion of channels occupy the inactivated state, drug binding is stronger and more prolonged, and recovery of channel availability is slowed.3 This voltage- and frequency-dependent profile is the mechanistic basis of the selective depression of high-frequency ictal firing relative to normal neuronal activity, and it explains why these drugs can suppress seizures without causing paralysis or complete anesthesia at therapeutic plasma concentrations.
Use dependence is the related pharmacological phenomenon in which the degree of channel blockade increases with each successive action potential in a train. Each opening exposes binding sites and allows drug molecules to enter the channel, and if the recovery interval between action potentials is shorter than the drug dissociation rate, block accumulates progressively across the train. This means that neurons firing at high ictal frequencies accumulate substantially more block than neurons firing at normal physiological rates, even at the same drug concentration. Use dependence operates on top of state dependence and further sharpens the selectivity of sodium channel ASDs for actively firing epileptic neurons.
All classical sodium channel ASDs (phenytoin, carbamazepine, oxcarbazepine, lamotrigine) enhance fast inactivation. Lacosamide is unique in that it selectively enhances slow inactivation without affecting fast inactivation gating. Slow inactivation is engaged when neurons sustain prolonged depolarization, as occurs in the recruited ictal network during a seizure. The two mechanisms are additive when combined, providing a rational pharmacological basis for using lacosamide as an adjunct to classical sodium channel blockers in patients with drug-resistant focal epilepsy. This mechanistic distinction may also explain some activity in populations where fast inactivation enhancers have failed.
Phenytoin was introduced in 1938 and remains in widespread clinical use, particularly for acute seizure management in the emergency setting. Its clinical pharmacology is dominated by a pharmacokinetic peculiarity that distinguishes it from nearly all other drugs in medicine: its metabolism is saturable at therapeutic plasma concentrations, producing zero-order kinetics and making dose adjustments treacherous. No prescriber should use phenytoin without a solid understanding of this kinetic behavior.
Phenytoin is absorbed from the gastrointestinal tract with moderate bioavailability, averaging 70–95%, but absorption is slow and highly variable between formulations. Extended-release capsule formulations produce more stable plasma concentrations than immediate-release tablets and are preferred for chronic oral therapy. Phenytoin is approximately 90% bound to plasma proteins, primarily albumin. This high protein binding has two important clinical implications. First, only the unbound fraction is pharmacologically active and capable of crossing the blood-brain barrier (BBB); therefore, total plasma concentration measurements may be misleading in patients with hypoalbuminemia (nephrotic syndrome, liver disease, malnutrition, pregnancy), in whom the free fraction is higher than predicted from the total concentration. Free phenytoin monitoring or Sheiner-Tozer formula correction is required in these settings.4 Second, drugs that displace phenytoin from albumin binding sites can transiently raise the free fraction and precipitate toxicity at previously well-tolerated total concentrations.
The defining pharmacokinetic feature of phenytoin is its Michaelis-Menten (zero-order) elimination kinetics at therapeutic concentrations. Most drugs follow first-order kinetics, where a constant fraction of the drug is eliminated per unit time, so that doubling the dose doubles the steady-state plasma concentration. Phenytoin is different: it is metabolized primarily by cytochrome P450 2C9 (CYP2C9) and to a lesser extent CYP2C19, and these enzymes become saturated within the therapeutic range, typically at plasma concentrations of 5–10 mg/L. Above saturation, elimination switches from first-order to zero-order: a constant amount of drug (not fraction) is eliminated per unit time, regardless of concentration. The clinical consequence is that small dose increases above the saturation threshold produce disproportionately large and unpredictable increases in steady-state plasma concentration.5 A dose increase of 10% can double or triple the plasma level, driving the patient from a subtherapeutic to a frankly toxic concentration. This is not a pharmacogenomic curiosity; it is a routine clinical hazard that applies to all patients using phenytoin, because the Km of CYP2C9 for phenytoin is within or below the therapeutic concentration range for most individuals.
Therapeutic drug monitoring (TDM) is mandatory for phenytoin. The conventional therapeutic range is 10–20 mg/L (40–79 micromol/L) for total phenytoin, corresponding to a free fraction of approximately 1–2 mg/L. However, these are population averages, and individual patients may achieve seizure control at lower concentrations or tolerate higher concentrations without toxicity. The therapeutic target should always be individualized to the patient's clinical response. Blood samples for TDM should be drawn as trough concentrations, just before the next dose, once steady state has been reached, which takes at least five half-lives. Phenytoin's half-life at therapeutic concentrations is typically 22–36 hours but is concentration-dependent: as concentration rises toward and above the saturation threshold, the apparent half-life lengthens because elimination rate no longer increases proportionately with concentration.5
Given phenytoin's zero-order kinetics above saturation, dose adjustments in chronic therapy must be made in small increments to avoid overshooting into the toxic range. A practical clinical rule: when plasma concentration is below 7 mg/L, increase by up to 100 mg/day; when concentration is 7–12 mg/L, increase by 50 mg/day at most; above 12 mg/L, increases of 25–30 mg/day are sufficient and safer. Allow at least two to three weeks between adjustments to reach a new steady state before reassessing. This incremental approach prevents the sudden jump from subtherapeutic to toxic concentrations that causes a disproportionate number of phenytoin-related adverse events.
Fosphenytoin is the water-soluble phosphate ester prodrug of phenytoin, developed specifically for parenteral administration. After intravenous (IV) or intramuscular (IM) injection, fosphenytoin is rapidly hydrolyzed by plasma phosphatases to phenytoin, phosphate, and formaldehyde (in negligible quantities). Fosphenytoin is dosed in phenytoin sodium equivalents (PE) to facilitate direct conversion from phenytoin dosing. The key clinical advantage over intravenous phenytoin is that fosphenytoin can be administered up to three times faster (up to 150 mg PE/min IV vs. 50 mg/min for phenytoin) and does not carry the risk of the severe infusion reactions that occur with the propylene glycol vehicle of IV phenytoin formulations, including purple glove syndrome (progressive limb ischemia and necrosis at the infusion site) and cardiac arrhythmias.6 Fosphenytoin can also be administered intramuscularly when IV access is unavailable, which phenytoin cannot due to tissue necrosis risk from its alkaline pH and poor aqueous solubility.
The adverse effect profile of phenytoin spans acute concentration-dependent toxicities and chronic effects related to long-term exposure. Acute dose-dependent toxicities progress in a predictable sequence with rising plasma concentration: nystagmus on lateral gaze at concentrations above 20 mg/L, ataxia and dysarthria at 30 mg/L, lethargy and mental status changes at 40 mg/L, and seizures paradoxically at very high concentrations above 50 mg/L. These acute toxicities are rapidly reversible with dose reduction. Chronic adverse effects include gingival hyperplasia, which occurs in 20–50% of patients on long-term therapy and is most severe in those with poor oral hygiene; coarsening of facial features; hirsutism; and peripheral neuropathy with long-term use.7 Phenytoin is a potent inducer of hepatic CYP enzymes (CYP2C9, CYP2C19, CYP3A4) and uridine diphosphate glucuronosyltransferase (UGT) enzymes, which has far-reaching consequences for drug interactions discussed in Section 5. Phenytoin is classified as teratogenic (historical FDA category D), associated with fetal hydantoin syndrome (midface hypoplasia, digit and nail hypoplasia, growth restriction, intellectual disability) and an increased risk of congenital malformations. Use in pregnancy requires careful individualized risk-benefit assessment and, when continued, intensive fetal monitoring.
Purple glove syndrome is a rare but potentially limb-threatening complication of peripheral IV phenytoin infusion. It presents as progressive distal edema, discoloration, and ischemia of the infused extremity, beginning at the infusion site and extending proximally. The mechanism involves the alkaline, propylene glycol-containing phenytoin vehicle causing direct endothelial injury and extravasation. Severe cases progress to tissue necrosis requiring fasciotomy or amputation. Prevention is straightforward: use fosphenytoin for all parenteral phenytoin administration whenever available. If phenytoin is the only option, use a large bore vein, infuse at no more than 50 mg/min, and monitor the infusion site continuously. Never infuse phenytoin in a hand or foot vein.
Carbamazepine has been a first-line agent for focal epilepsy since the 1960s and remains widely used globally, particularly in resource-limited settings where it is inexpensive and available. Its pharmacology is complicated by autoinduction of its own metabolism, an active metabolite with its own toxicity, a substantial drug interaction burden, and the HLA-B*1502 cutaneous reaction risk discussed in Module 1. Oxcarbazepine and eslicarbazepine were developed specifically to address carbamazepine's pharmacokinetic and tolerability limitations.
Carbamazepine is a dibenzazepine compound structurally related to the tricyclic antidepressants (TCAs). It acts primarily by enhancing fast inactivation of Nav channels, in the same manner as phenytoin, but with a distinct binding site on the alpha subunit. Carbamazepine is the drug of choice for focal onset seizures and focal to bilateral tonic-clonic seizures in many national guidelines, and it is also a first-line agent for trigeminal neuralgia. It has no role in the management of primary generalized epilepsies and reliably aggravates absence and myoclonic seizures, a risk that must be borne in mind whenever a patient with possible generalized epilepsy features is being considered for treatment.8 Carbamazepine is also used as a mood stabilizer in bipolar disorder, and patients with this dual indication represent a significant portion of long-term carbamazepine users.
The absorption of carbamazepine from the gastrointestinal tract is slow and erratic, with bioavailability ranging from 75 to 85%. Carbamazepine is lipophilic and distributes widely, with a volume of distribution of approximately 1–2 L/kg. Protein binding is 75–80%, lower than phenytoin and therefore less susceptible to displacement interactions. The critical pharmacokinetic complexity of carbamazepine is its autoinduction of CYP3A4, the enzyme primarily responsible for its own metabolism. When therapy is initiated, the half-life of carbamazepine is approximately 25–65 hours. Over 2–4 weeks of continued dosing, autoinduction progressively reduces the half-life to 12–17 hours at steady state, substantially lowering plasma concentrations for a given dose.9 This means that plasma concentrations measured early in therapy will be significantly higher than those measured at the same dose after autoinduction is complete, and doses that appear therapeutic in the first week may produce subtherapeutic concentrations after a month. Initial dosing must be low to avoid toxicity before autoinduction occurs, and doses must be titrated upward as autoinduction stabilizes.
Carbamazepine is metabolized primarily by CYP3A4 to carbamazepine-10,11-epoxide (CBZ-E), an active metabolite that contributes significantly to both the therapeutic effect and the adverse effect burden of the parent compound. CBZ-E is further hydrolyzed to an inactive trans-diol by epoxide hydrolase. The ratio of CBZ-E to carbamazepine varies considerably between patients and is increased by co-administration of valproate (which inhibits epoxide hydrolase) and decreased by enzyme inducers. Standard carbamazepine TDM measures parent drug concentration but not CBZ-E; in patients with apparent toxicity at therapeutic carbamazepine levels, measuring the epoxide metabolite is clinically informative.9 The therapeutic range for carbamazepine is 4–12 mg/L, though this is a rough guide and seizure control versus toxicity should always guide target setting.
Oxcarbazepine is a keto-analogue of carbamazepine that functions as a prodrug. After oral administration, it is rapidly and completely reduced by hepatic cytosolic enzymes to its active monohydroxy derivative (MHD), also called licarbazepine, which is the pharmacologically active species responsible for Nav channel blockade. Unlike carbamazepine, oxcarbazepine does not undergo epoxidation, eliminating the CBZ-E metabolite and its associated toxicity. MHD has a half-life of 9–11 hours and is eliminated predominantly by glucuronidation via UGT enzymes, with a smaller fraction by CYP3A4. Oxcarbazepine does not autoinduct its own metabolism to any clinically meaningful degree. It is a weaker inducer of CYP3A4 and CYP2C19 than carbamazepine, resulting in a substantially reduced drug interaction burden.10 Its efficacy for focal epilepsy is comparable to carbamazepine, and its tolerability, as measured in head-to-head trials, is somewhat better, with fewer central nervous system (CNS) side effects and no CBZ-E-related toxicity.
Eslicarbazepine acetate is the most recent member of the dibenzazepine family. It is a prodrug that is hydrolyzed almost entirely to S-licarbazepine, the S-enantiomer of MHD, which has higher affinity for the inactivated Nav channel than the R-enantiomer present in the MHD mixture produced from oxcarbazepine. Eslicarbazepine acetate has a longer half-life than oxcarbazepine (approximately 20–24 hours) and is suitable for once-daily dosing, improving adherence. Its enzyme-inducing potential is lower than both carbamazepine and oxcarbazepine. Hyponatremia is a class effect across all three dibenzazepine agents, occurring through a mechanism that includes inappropriate antidiuretic hormone (ADH) secretion and possibly direct tubular effects. The incidence of hyponatremia (sodium at or below 134 mEq/L) is substantially higher with oxcarbazepine than with carbamazepine, and the incidence of severe hyponatremia (sodium at or below 128 mEq/L) was 12.4% in oxcarbazepine-treated patients compared to 2.8% in carbamazepine-treated patients in one comparative study, with advanced age being the principal risk factor in both groups.10 Eslicarbazepine has a comparable hyponatremia risk profile to oxcarbazepine, but all three agents require periodic sodium monitoring, particularly in elderly patients and those receiving concurrent sodium-depleting medications such as diuretics or selective serotonin reuptake inhibitors (SSRIs).
As covered in Module 1, carbamazepine carries a risk of life-threatening severe cutaneous adverse reactions (SCARs) that is genetically stratified. HLA-B*1502 carriers of Southeast Asian ancestry face an extremely high risk of Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) with carbamazepine, phenytoin, and oxcarbazepine. HLA-A*3101 carriers of European or Japanese ancestry face a moderately elevated risk of a broader spectrum of hypersensitivity reactions. In practice: screen all patients of Southeast Asian ancestry for HLA-B*1502 before initiating any of these three agents. If positive, use alternative ASDs. The cross-reactivity among aromatic ASDs (carbamazepine, phenytoin, phenobarbital) means that a patient who has experienced a SCAR to one of these agents should not receive another without specialist consultation.
Lacosamide was approved for adjunctive therapy of focal onset seizures in 2008 and has become an increasingly used agent in both outpatient and hospital settings, in part because of its IV formulation and low drug interaction burden. Its mechanism of action is mechanistically distinct from all other sodium channel ASDs, enhancing slow rather than fast inactivation of Nav channels, which gives it a complementary pharmacological profile when used in combination with classical agents.
Lacosamide's selective enhancement of slow inactivation is the result of its interaction with a binding site on the Nav channel alpha subunit that is distinct from the local anesthetic binding site used by phenytoin and carbamazepine. Slow inactivation is a conformational state of the Nav channel that develops over hundreds of milliseconds to seconds of sustained membrane depolarization and involves rearrangements in the pore-lining S6 segments. During an ictal discharge, neurons sustain prolonged depolarization, engaging slow inactivation to a greater degree than during normal physiological firing. Lacosamide stabilizes channels in the slow-inactivated state, reducing the availability of channels for the next action potential during sustained high-frequency activity without materially affecting fast inactivation kinetics and therefore having less effect on normal physiological firing.2 This pharmacological profile accounts for lacosamide's activity in some patients who have failed classical sodium channel blockers that act exclusively on fast inactivation.
Lacosamide has highly favorable absorption pharmacokinetics. Oral bioavailability is approximately 100%, with no clinically significant food effect. It is minimally protein bound (less than 15%), which eliminates protein displacement interactions and makes monitoring of free drug concentrations unnecessary. The volume of distribution is approximately 0.6 L/kg. Lacosamide is metabolized primarily by CYP2C19 to an inactive O-desmethyl metabolite, with approximately 40% eliminated unchanged by the kidneys. The half-life is 13 hours, permitting twice-daily oral dosing. Because lacosamide is not a significant inducer or inhibitor of CYP enzymes, its interaction potential with other drugs is substantially lower than carbamazepine or phenytoin.11 Dose reduction is required in severe renal impairment (creatinine clearance below 30 mL/min), and moderate dose reduction is recommended in severe hepatic impairment.
The intravenous formulation of lacosamide is bioequivalent to the oral formulation and can be given at the same doses. This makes lacosamide practical in hospitalized patients who cannot take oral medications, for acute seizure clusters, and as a second-line IV agent for the treatment of status epilepticus (SE) in settings where IV fosphenytoin or IV valproate have been used as first-line second-stage agents. In the Emergency Treatment with Levetiracetam, Fosphenytoin, or Valproate (ESETT) trial, these three agents showed equivalent efficacy for benzodiazepine-refractory convulsive SE, which is the current evidence basis for second-line SE management. Lacosamide has not yet been evaluated in a similarly powered randomized trial for SE, but retrospective and observational data are accumulating, and it is used in clinical practice when the above three agents have failed or are contraindicated.12
Lacosamide causes dose-dependent prolongation of the cardiac PR interval, reflecting its slowing of slow inactivation in cardiac Nav channels in addition to neuronal channels. In clinical trials, PR prolongation of approximately 3–5 milliseconds was seen at therapeutic doses. Clinically significant atrioventricular (AV) block has been reported, particularly in patients with pre-existing conduction abnormalities or those receiving other PR-prolonging drugs. Before initiating lacosamide, obtain a baseline electrocardiogram (ECG) in patients with known cardiac conduction disease, those on antiarrhythmic drugs, or those on other ASDs that affect cardiac conduction. Symptoms of palpitations, bradycardia, or syncope should prompt ECG evaluation. This monitoring requirement distinguishes lacosamide from other ASDs in the outpatient chronic management setting.
Lacosamide is currently approved only for focal onset seizures, including focal to bilateral tonic-clonic seizures. Evidence for efficacy in primary generalized epilepsies is limited and inconsistent. Unlike carbamazepine, there is no strong evidence that lacosamide aggravates absence or myoclonic seizures, but it should not be assumed to be broadly effective across generalized seizure types until adequate evidence accumulates. Its clinical positioning is as a well-tolerated agent for focal epilepsy — either as an adjunct when first-line focal epilepsy ASDs are partially effective, as monotherapy in patients who cannot tolerate carbamazepine or oxcarbazepine, or as a parenteral option in the acute hospital setting. Its cost is substantially higher than carbamazepine or phenytoin, which is a consideration in resource-limited settings.
Phenytoin and carbamazepine are among the most potent enzyme inducers in clinical pharmacology. Their induction of CYP3A4, CYP2C9, CYP2C19, and UGT enzymes accelerates the metabolism of a large number of co-administered drugs, reducing their plasma concentrations and potentially compromising efficacy. Understanding these interactions is not optional for any prescriber who uses these agents; it is a core clinical pharmacology competency.
Enzyme induction by phenytoin and carbamazepine occurs through activation of the pregnane X receptor (PXR) and constitutive androstane receptor (CAR), nuclear receptors that upregulate the transcription of CYP1A2, CYP2C9, CYP2C19, CYP3A4, and multiple UGT isoforms in the liver and intestinal wall. The induction is not immediate: it develops over 2–4 weeks as hepatocytes synthesize new enzyme protein and reaches a new steady state when induction is balanced by normal enzyme turnover. When an inducing ASD is discontinued, enzyme activity returns to baseline over a similar timeframe. This temporal profile of induction and de-induction means that drug levels of co-administered agents will fall over weeks after starting an inducer and will rise over weeks after stopping one, creating two distinct windows of pharmacokinetic instability.13
Oral contraceptive failure is one of the most clinically consequential and underappreciated interactions involving enzyme-inducing ASDs. Both phenytoin and carbamazepine substantially accelerate the metabolism of ethinyl estradiol and progestins, reducing plasma concentrations of these hormones by 40–60% and rendering standard-dose combined oral contraceptive pills and progestin-only pills unreliable for contraception. Women of reproductive age who require both an enzyme-inducing ASD and contraception should use an intrauterine device (copper or levonorgestrel), injectable depot medroxyprogesterone at standard dosing intervals, or a subdermal implant — recognizing that etonogestrel implant efficacy may also be reduced. Barrier methods should supplement hormonal contraception as a precaution. Prescribers must counsel patients explicitly on this interaction; it is a common source of unintended pregnancies in women with epilepsy.13
Anticoagulant interactions with enzyme-inducing ASDs are clinically significant and require careful management. Carbamazepine and phenytoin both induce CYP2C9, the enzyme primarily responsible for the metabolism of the more potent S-enantiomer of warfarin, reducing warfarin plasma concentrations and anticoagulant effect. Patients stabilized on warfarin who are started on carbamazepine or phenytoin will require higher warfarin doses to maintain their target international normalized ratio (INR). Conversely, if an inducing ASD is discontinued, warfarin doses must be reduced as enzyme induction wanes, or the INR will rise dramatically, increasing hemorrhage risk. Direct oral anticoagulants (DOACs) including rivaroxaban, apixaban, and dabigatran are all substrates of CYP3A4 and/or P-glycoprotein; their plasma concentrations are substantially reduced by enzyme-inducing ASDs, and these combinations are generally contraindicated or require specialist management with alternative anticoagulation.
Interactions between sodium channel ASDs and other ASDs are also clinically important. Carbamazepine induces CYP3A4 and UGT enzymes, reducing the plasma concentrations of lamotrigine (which is eliminated by UGT1A4 glucuronidation) by approximately 40–50%, requiring higher lamotrigine doses when these agents are combined. Valproate inhibits epoxide hydrolase, increasing carbamazepine-10,11-epoxide (CBZ-E) concentrations without substantially changing parent carbamazepine levels; patients on carbamazepine-valproate combination therapy may experience CBZ-E toxicity (diplopia, dizziness, nausea) at apparently therapeutic carbamazepine concentrations. Phenytoin and carbamazepine also reduce plasma concentrations of topiramate and zonisamide.14 Oxcarbazepine and eslicarbazepine have fewer inductive interactions but still reduce ethinyl estradiol concentrations and can reduce lamotrigine concentrations modestly.
Phenytoin or carbamazepine + oral contraceptive: contraceptive failure; use non-hormonal IUD or injectable depot. Carbamazepine + valproate: elevated CBZ-E with toxicity symptoms at normal carbamazepine levels; monitor CBZ-E if toxicity suspected. Enzyme-inducing ASD + warfarin: increased warfarin requirement; monitor INR closely with any ASD change. Enzyme-inducing ASD + DOAC (rivaroxaban, apixaban, dabigatran): major reduction in anticoagulant exposure; generally contraindicated. Enzyme-inducing ASD + lamotrigine: lamotrigine doses must be substantially higher; abrupt ASD withdrawal will cause lamotrigine toxicity. Carbamazepine + voriconazole (azole antifungal): major reduction in voriconazole exposure; combination contraindicated. Phenytoin + fluconazole (CYP2C9 inhibitor): substantial rise in phenytoin levels; toxicity risk.
Clinical drug selection among sodium channel ASDs in focal epilepsy requires weighing efficacy, pharmacokinetic complexity, interaction burden, tolerability, and cost. For new-onset focal epilepsy in adults, carbamazepine and oxcarbazepine remain first-line options supported by the most extensive clinical trial data, though lamotrigine is now preferred in many guidelines for women of reproductive age due to a better teratogenicity profile and is preferred in the elderly due to its linear kinetics and lower interaction burden. Lacosamide is a reasonable first-line option in patients where the interaction burden of carbamazepine is clinically problematic. Phenytoin, while effective, is generally falling from favor as a chronic oral agent due to its zero-order kinetics, cosmetic adverse effects, and high interaction burden, though it retains an important role for acute parenteral management of seizures and status epilepticus.15 In patients already on enzyme-inducing ASDs who require additional agents, pharmacokinetic interactions must be anticipated proactively and doses of the newly added agent adjusted from the outset rather than reactively after concentrations are measured.
New focal epilepsy, no interaction concerns, cost-sensitive setting: carbamazepine or oxcarbazepine. Oxcarbazepine preferred for better tolerability and absence of CBZ-E toxicity.
New focal epilepsy, woman of reproductive age: lamotrigine preferred (better teratogenicity profile) or levetiracetam; avoid carbamazepine and phenytoin if possible.
New focal epilepsy, elderly patient or polypharmacy: lamotrigine or lacosamide preferred; avoid phenytoin (zero-order kinetics hazardous in elderly) and carbamazepine (high interaction burden).
Acute parenteral seizure management / status epilepticus: fosphenytoin IV (second-line after benzodiazepines), lacosamide IV as an alternative; never use IM phenytoin.
Focal epilepsy, partially controlled on fast inactivation blocker: add lacosamide for complementary slow inactivation mechanism; anticipate no pharmacokinetic interaction with levetiracetam, modest interaction with enzyme-inducing ASDs.
Catterall WA. From ionic currents to molecular mechanisms: the structure and function of voltage-gated sodium channels. Neuron. 2000;26(1):13–25.
doi:10.1016/S0896-6273(00)81133-2Sheets PL, Heers C, Stoehr T, Cummins TR. Differential block of sensory neuronal voltage-gated sodium channels by lacosamide [(R)-2-(acetylamino)-N-benzyl-3-methoxypropanamide], lidocaine, and carbamazepine. J Pharmacol Exp Ther. 2008;326(1):89–99.
doi:10.1124/jpet.107.133413Ragsdale DS, Avoli M. Sodium channels as molecular targets for antiepileptic drugs. Brain Res Brain Res Rev. 1998;26(1):16–28.
doi:10.1016/S0165-0173(97)00054-4Anderson GD. Phenytoin: mechanisms of action. In: Levy RH, Mattson RH, Meldrum BS, Perucca E, eds. Antiepileptic Drugs. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2002:554–563.
Ludden TM. Nonlinear pharmacokinetics: clinical implications. Clin Pharmacokinet. 1991;20(6):429–446.
doi:10.2165/00003088-199120060-00001Fischer JH, Patel TV, Fischer PA. Fosphenytoin: clinical pharmacokinetics and comparative advantages in the acute treatment of seizures. Clin Pharmacokinet. 2003;42(1):33–58.
doi:10.2165/00003088-200342010-00002Asadi-Pooya AA, Rostaminejad M, Zeraatpisheh Z, Mirzaei Damabi N. Cosmetic adverse effects of antiseizure medications: a systematic review. Seizure. 2021;91:9–21.
doi:10.1016/j.seizure.2021.05.010Marson AG, Al-Kharusi AM, Alwaidh M, et al. The SANAD study of effectiveness of carbamazepine, gabapentin, lamotrigine, oxcarbazepine, or topiramate for treatment of partial epilepsy: an unblinded randomised controlled trial. Lancet. 2007;369(9566):1000–1015.
doi:10.1016/S0140-6736(07)60460-7Patsalos PN, Berry DJ, Bourgeois BF, et al. Antiepileptic drugs: best practice guidelines for therapeutic drug monitoring. Epilepsia. 2008;49(7):1239–1276.
doi:10.1111/j.1528-1167.2008.01561.xDong X, Leppik IE, White J, Rarick J. Hyponatremia from oxcarbazepine and carbamazepine. Neurology. 2005;65(12):1976–1978.
doi:10.1212/01.wnl.0000188819.45330.90Halasz P, Kalviainen R, Mazurkiewicz-Beldzinska M, et al. Adjunctive lacosamide for partial-onset seizures: efficacy and safety results from a randomized controlled trial. Epilepsia. 2009;50(3):443–453.
doi:10.1111/j.1528-1167.2008.01951.xKapur J, Elm J, Chamberlain JM, et al. Randomized trial of three anticonvulsant medications for status epilepticus. N Engl J Med. 2019;381(22):2103–2113.
doi:10.1056/NEJMoa1905795Hachad H, Ragueneau-Majlessi I, Levy RH. New antiepileptic drugs: review on drug interactions. Ther Drug Monit. 2002;24(1):91–103.
doi:10.1097/00007691-200202000-00016Johannessen SI, Landmark CJ. Antiepileptic drug interactions: principles and clinical implications. Curr Neuropharmacol. 2010;8(3):254–267.
doi:10.2174/157015910792246254Perucca P, Gilliam FG. Adverse effects of antiepileptic drugs. Lancet Neurol. 2012;11(9):792–802.
doi:10.1016/S1474-4422(12)70153-9