Oxcarbazepine

What it is

Oxcarbazepine is a prescription antiseizure medicine, a keto-analogue of carbamazepine, taken as film-coated tablets, an oral suspension or an extended-release tablet. It is a prodrug: its effect is carried almost entirely by a metabolite called the 10-monohydroxy derivative (MHD). The US label approves it as monotherapy or add-on treatment for partial-onset (focal) seizures in adults, as monotherapy from age 4, and as add-on from age 2.

What the research says

Oxcarbazepine's active metabolite blocks voltage-sensitive sodium channels, which stops over-excitable nerve membranes firing repeatedly. For focal seizures the evidence supports it but is not strong: Cochrane's add-on review found a responder rate ratio of 1.80 (95% CI 1.27 to 2.56) and seizure-freedom ratio of 2.86 (1.19 to 6.87), both rated low certainty, alongside a clearly increased chance of stopping treatment (RR 1.75, 1.44 to 2.13, moderate certainty). As a single first drug it came out behind lamotrigine on high-certainty individual-patient data (hazard ratio for treatment failure 1.30, 1.02 to 1.66). For the common off-label uses the evidence is weaker still: for neuropathic pain Cochrane found little evidence of benefit and clear excess harm (serious adverse effects 8.3% versus 2.5%), and for acute bipolar episodes it found no difference from placebo in the one eligible trial, which was in children. The distinctive harms are low blood sodium, reported at 2.5% below 125 mmol/L in the trials and 8.4% below 128 mmol/L in a real-world cohort; dose-related dizziness, double vision and unsteadiness; and rare but serious skin reactions, with risk linked to the HLA-B*1502 gene variant.

Evidence grade: Limited evidence.

How it works

Drug class: Voltage-gated sodium channel blocking antiseizure medicine (dibenzazepine carboxamide; keto-analogue of carbamazepine)

Oxcarbazepine is converted in the body to a metabolite called MHD, which does the work. MHD blocks the sodium channels that nerve cells use to fire, so an over-excited patch of brain cannot keep firing and cannot pass the excitation on to its neighbours. That is thought to be how it stops a seizure spreading. Extra effects on potassium and calcium channels may contribute, and the precise mechanism is not established. (Source 1)

What it is used for

• Cochrane found oxcarbazepine add-on reduced seizures more than control, with a responder rate ratio of 1.80 and a seizure-freedom ratio of 2.86, but rated both low certainty and recorded 75% more treatment withdrawals. Evidence: limited. (Source 2)
• In Cochrane's network meta-analysis of individual participant data, lamotrigine outperformed oxcarbazepine on time to treatment failure (HR 1.30, 95% CI 1.02 to 1.66, high certainty), and oxcarbazepine was not distinguishable from other newer drugs on time to first seizure. Evidence: limited. (Source 3)
• Cochrane found little evidence of effectiveness, with the only usable 50%-relief estimate resting on a single 146-person trial, and clear excess harm: serious adverse effects 8.3% versus 2.5% on placebo and withdrawal for side effects up to 42.3% versus 14.9%. Evidence: not-supported. (Source 4)
• Cochrane found no difference from placebo in the single eligible placebo-controlled trial, which was in children, no adult placebo-controlled trial at all, and a higher rate of neuropsychiatric side effects than placebo. Evidence: not-supported. (Source 5)
• The US label approves it only for partial-onset seizures. Cochrane's network analysis included oxcarbazepine in the generalised-seizure network but the comparison with sodium valproate was imprecise (HR 1.24, 95% CI 0.72 to 2.14), and the Oxtellar XR label records that primary generalised seizures have been made worse by immediate-release oxcarbazepine. Evidence: unknown. (Source 6)

Interactions

• Combined hormonal contraceptives (ethinylestradiol and levonorgestrel) (pharmacokinetic study): Oxcarbazepine roughly halves the blood levels of both hormones in a combined oral contraceptive, which can make it fail. The label says additional non-hormonal contraception is recommended, and that other oral or implant contraceptives have not been studied. Two of the four intervals the label prints are mis-typeset, with the upper confidence limit equal to the point estimate ("52% [90% CI: 38 to 52]" and "52% [90% CI: 42 to 52]"), so the reductions are reliable but those two ranges are not. (Source 7)
• Alcohol (label): Alcohol and oxcarbazepine can both cause drowsiness, and the label warns the sedative effect may add up. This is the label's own caution rather than a measured interaction study. (Source 8)
• Food (pharmacokinetic study): Food makes no difference to how much oxcarbazepine is absorbed from the tablets, so it can be taken with or without a meal. The suspension was not studied directly but is unlikely to be affected. (Source 9)
• Verapamil and felodipine (calcium channel blockers) (pharmacokinetic study): Verapamil lowers the level of oxcarbazepine's active metabolite by about a fifth, and oxcarbazepine lowers felodipine exposure by about a quarter. (Source 7)
• Drugs that lower blood sodium, including diuretics and drugs causing inappropriate ADH secretion (clinical trial): Oxcarbazepine can drop blood sodium on its own, and the label says sodium measurement should be considered particularly when a person is also taking something else that lowers it. A real-world cohort found diuretics and valproate among the strongest predictors of severe low sodium. (Source 10)
• St John's wort and other supplements (theoretical): Not studied. We found no interaction study, trial or case series of oxcarbazepine with St John's wort, grapefruit, calcium, iron, magnesium, fish oil, turmeric or red yeast rice in the sources we read. The label's interaction sections name only strong CYP3A4 and UGT inducers, phenytoin, hormonal contraceptives, calcium antagonists, cimetidine, erythromycin, dextropropoxyphene and warfarin. St John's wort is itself a CYP3A4 and UGT inducer, so the mechanism the label describes for rifampin and carbamazepine is theoretically available to it, but that is reasoning from the mechanism rather than a measurement anyone has published. (Source 11)

Stopping it

• The label's position is that oxcarbazepine should normally be tapered rather than stopped suddenly, because abrupt withdrawal risks more frequent seizures and status epilepticus. The one exception it allows is a serious adverse event, where rapid discontinuation can be considered. (Source 12)
• Oxcarbazepine is not a drug of dependence in the sense that opioids or benzodiazepines are, but the label's evidence for that is very thin. Its drug abuse and dependence section says the abuse potential has not been evaluated in human studies at all, and the only dependence data it carries is an animal study: intragastric oxcarbazepine given to four cynomolgus monkeys produced no sign of physical dependence, measured by whether they would press a lever to self-administer it. No human dependence or withdrawal syndrome is described anywhere in the label. (Source 13)
• A quarter of adults already taking other antiseizure drugs stopped it because of side effects rather than by plan: approximately 23% of 1537 such adults discontinued for an adverse reaction, most often dizziness, double vision, unsteadiness, vomiting or nausea. Among 295 adults starting oxcarbazepine as their first antiseizure drug the figure was about 9%. The dose matters too: in Cochrane's pooled data the withdrawal risk ratio against control was 2.38 at 2400 mg a day and 1.54 at 1200 mg, while at 600 mg the single available study gave 0.79 with an interval crossing 1. (Source 14)
• When low sodium is the reason for stopping, it reverses quickly. In the clinical trials, patients whose treatment was stopped because of hyponatremia generally had their serum sodium return to normal within a few days with no extra treatment. (Source 10)

What goes wrong

The same Cochrane review found that people given oxcarbazepine as an add-on were considerably more likely to stop treatment than controls, and the excess grew with dose. (Source 15)

• Systematic review, Moderate certainty.
• Size: 1,593 participants across six trials.
• Who: people with drug-resistant focal epilepsy.
• How long: 9 days to 26 weeks.
• Result: treatment withdrawal RR 1.75 (95% CI 1.44 to 2.13, moderate certainty); at 2400 mg/d RR 2.38 (95% CI 1.92 to 2.94); at 1200 mg/d RR 1.54 (95% CI 1.21 to 1.95); at 600 mg/d RR 0.79 (95% CI 0.55 to 1.15)
• Funding: not stated; the record lists Research Support, Non-U.S. Gov't.

Limit of this finding: The 2400 mg and 1200 mg figures are both statistically significant, but the 600 mg ratio of 0.79 has an interval from 0.55 to 1.15 that crosses 1 and comes from a single study, so it is uninformative: it is not evidence that 600 mg a day caused fewer withdrawals than control.

The largest oxcarbazepine dose used, 2400 mg/d, was associated with a higher treatment withdrawal rate (RR 2.38, 95% CI 1.92 to 2.94; fixed-effect model; 2 studies) compared to control, than 1200 mg/d (RR 1.54, 95% CI 1.21 to 1.95; fixed-effect model; 3 studies) or 600 mg/d oxcarbazepine (RR 0.79, 95% CI 0.55 to 1.15; fixed-effect model; 1 study).

Cochrane found sleepiness significantly more common with oxcarbazepine add-on, while ataxia and hyponatraemia were more frequent but not significantly so at the 99% level the review used for adverse effects. (Source 15)

• Systematic review, Moderate certainty.
• Size: six trials, 1,593 participants (ataxia from 5 studies, somnolence and hyponatraemia from 6)
• Who: people with drug-resistant focal epilepsy.
• How long: 9 days to 26 weeks.
• Result: somnolence RR 2.03 (99% CI 1.17 to 3.54; 6 studies; low-certainty evidence), the only one of the three that excludes 1; ataxia RR 2.54 (99% CI 0.86 to 7.54; 5 studies; moderate-certainty evidence), interval crosses 1; hyponatraemia RR 2.53 (99% CI 0.27 to 23.85; 6 studies; moderate-certainty evidence), interval crosses 1.
• Funding: not stated; the record lists Research Support, Non-U.S. Gov't.

Limit of this finding: The review's own sentence describes ataxia as part of "an increased incidence of multiple adverse effects" while reporting a 99% interval of 0.86 to 7.54, which crosses 1. On the review's chosen 99% threshold for adverse effects that is not a significant increase, so ataxia should be read with the same "not significantly so" wording the review applied to hyponatraemia. The point estimates still point upwards; the trials were simply too small to separate them from chance.

Hyponatraemia occurred more frequently with oxcarbazepine treatment but not significantly so (RR 2.53, 99% CI 0.27 to 23.85; fixed-effect model; 6 studies; moderate-certainty evidence).

In the same review, serious adverse effects and withdrawals for side effects were markedly more common on oxcarbazepine than placebo in neuropathic pain. (Source 16)

• Systematic review, Moderate certainty.
• Size: 634 participants across three diabetic neuropathy trials; 145 in the radiculopathy trial; 83 in the mixed trial.
• Who: adults with neuropathic pain.
• How long: trials of at least six weeks.
• Result: serious adverse effects 8.3% versus 2.5% on placebo (RR 3.65, 95% CI 1.45 to 9.20, moderate quality, NNTH 17, 95% CI 11 to 42); withdrawal for adverse effects in diabetic neuropathy 25.6% versus 6.8% (RR 3.83, 95% CI 2.29 to 6.40); in radiculopathy 42.3% versus 14.9% (RR 2.84, 95% CI 1.55 to 5.23)
• Funding: not stated; the record lists Research Support, Non-U.S. Gov't.

Based on moderate-quality evidence from the three DPN trials, serious adverse effects occurred in 8.3% with oxcarbazepine and 2.5% with placebo (RR 3.65, 95% CI 1.45 to 9.20; n = 634; moderate-quality evidence). The number needed to treat for an additional harmful (serious adverse effect) outcome (NNTH) was 17 (95% CI 11 to 42). The RR for serious adverse effects in the radiculopathy trial was 3.13 (95% CI 0.65 to 14.98, n = 145). The fifth trial did not provide data.More people withdrew because of adverse effects with oxcarbazepine than with placebo (DPN: 25.6% with oxcarbazepine versus 6.8% with placebo; RR 3.83, 95% CI 2.29 to 6.40; radiculopathy: 42.3% with oxcarbazepine versus 14.9% with placebo; RR 2.84, 95% CI 1.55 to 5.23

In the bipolar trials, neuropsychiatric side effects were several times more common on oxcarbazepine than on placebo. (Source 17)

• Systematic review, Low certainty.
• Size: one study, 115 participants.
• Who: children and adolescents with acute bipolar episodes.
• How long: not stated.
• Result: 17% to 39% of participants on oxcarbazepine had at least one such event compared with 7% to 10% on placebo.
• Funding: not stated.

There was a higher incidence of adverse effects, particularly neuropsychiatric, in participants randomised to oxcarbazepine compared to those on placebo (1 study, n=115, 17% to 39% of participants on oxcarbazepine had at least one such event compared to 7% to 10% on placebo).

In the controlled epilepsy trials behind the label, 2.5% of oxcarbazepine-treated patients dropped their blood sodium below 125 mmol/L and none on placebo or active control did. (Source 18)

• Official position, Moderate certainty.
• Size: 38/1524 oxcarbazepine-treated patients across 14 controlled epilepsy studies.
• Who: participants in the 14 controlled adjunctive and monotherapy epilepsy studies; comparators were placebo, carbamazepine, phenobarbital, phenytoin and valproate.
• How long: usually within the first 3 months, but some cases appeared more than a year after starting.
• Result: 2.5% (38/1524) reached a sodium below 125 mmol/L versus none on placebo or active control; most were asymptomatic; sodium normalised within a few days of stopping without further treatment.
• Funding: manufacturer's label (Novartis), a regulatory position dated 2025-12-16.

In the 14 controlled epilepsy studies, 2.5% of TRILEPTAL-treated patients (38/1524) had a sodium of less than 125 mmol/L at some point during treatment, compared to no such patients assigned placebo or active control (carbamazepine and phenobarbital for adjunctive and monotherapy substitution studies, and phenytoin and valproate for the monotherapy initiation studies).

In routine care the rate of severe low sodium on oxcarbazepine is higher than the trials suggested: 8.4% in a two-hospital real-world cohort. (Source 19)

• Cohort study, Low certainty.
• Size: 2,253 adult patients prescribed oxcarbazepine.
• Who: adults with epilepsy at two tertiary hospitals in South Korea, from standardised OMOP-CDM records; see jung-2025-methods for the cohort definition.
• How long: retrospective cohort; follow-up length not stated in the abstract.
• Result: prevalence of severe hyponatremia, defined as serum sodium 128 mmol/L or less, was 8.4%; risk predictors included valproate, diuretics, high oxcarbazepine dose, age and stroke history.
• Funding: not stated. Observational, and the severity threshold (128 mmol/L) differs from the label's (125 mmol/L), so the figures are not directly comparable.

Among 2253 patients, the prevalence of severe hyponatremia was 8.4%.

Serious skin reactions including Stevens-Johnson syndrome and toxic epidermal necrolysis have been reported with oxcarbazepine at a reported rate 3 to 10 times the population background. (Source 20)

• Case series, Low certainty.
• Size: not stated; spontaneous postmarketing reports in children and adults.
• Who: people treated with oxcarbazepine.
• How long: median time of onset 19 days after starting.
• Result: reporting rate exceeds the background incidence by 3- to 10-fold; background incidence is estimated at 0.5 to 6 cases per million person-years. The label notes reporting rates are generally an underestimate.
• Funding: manufacturer's label (Novartis), a regulatory position dated 2025-12-16.

The reporting rate of TEN and SJS associated with TRILEPTAL use, which is generally accepted to be an underestimate due to underreporting, exceeds the background incidence rate estimates by a factor of 3- to 10-fold.

Carrying the HLA-B*1502 gene variant may raise the risk of severe skin reaction with oxcarbazepine, and the variant is common in several Asian populations and negligible in others. (Source 21)

• Official position, Low certainty.
• Size: not stated; the label cites available clinical evidence plus nonclinical binding data.
• Who: people treated with oxcarbazepine, by HLA-B*1502 carrier status.
• How long: not applicable.
• Result: allele frequency 2% to 12% in Han Chinese, about 8% in Thai, above 15% in the Philippines and some Malaysian populations, up to about 2% in Korea and 6% in India; negligible in people of European descent, several African populations, indigenous peoples of the Americas, Hispanic populations and Japan.
• Funding: manufacturer's label (Novartis), a regulatory position dated 2025-12-16. The direct evidence is for carbamazepine; the oxcarbazepine link is inferred from chemical similarity and protein-binding data.

Limit of this finding: The direct risk evidence is for carbamazepine; for oxcarbazepine the label infers the link from chemical similarity and from laboratory binding data, so the size of the risk is not known. The label is also explicit about the limits of the test: genotyping "has important limitations, and must never substitute for appropriate clinical vigilance and patient management", screening "is not generally recommended in patients from populations in which the prevalence of HLA-B*1502 is low, or in current TRILEPTAL users", and the risk of these reactions is largely confined to the first few months of treatment whatever the result. A negative test does not mean a severe skin reaction cannot happen, and most people who carry the allele never get one.

Available clinical evidence, and data from nonclinical studies showing a direct interaction between TRILEPTAL and HLA-B1502 protein, suggest that the HLA-B1502 allele may also increase the risk for SJS/TEN with TRILEPTAL.

A pooled analysis of 199 placebo-controlled trials of 11 antiseizure drugs found roughly double the risk of suicidal thinking or behaviour, an absolute rise of about one case for every 530 people treated. (Source 22)

• Meta-analysis, Moderate certainty.
• Size: 27,863 drug-treated and 16,029 placebo-treated patients across 199 placebo-controlled trials of 11 antiepileptic drugs.
• Who: patients taking antiepileptic drugs for any indication, aged 5 to 100.
• How long: median treatment duration 12 weeks; most trials did not extend beyond 24 weeks.
• Result: adjusted relative risk 1.8 (95% CI 1.2, 2.7); incidence 0.43% among 27,863 drug-treated patients versus 0.24% among 16,029 placebo-treated patients, about one extra case of suicidal thinking or behaviour for every 530 patients treated.
• Funding: regulatory pooled analysis reported in the manufacturer's label, a position dated 2025-12-16. The finding is class-wide across 11 drugs, not specific to oxcarbazepine.

Pooled analyses of 199 placebo-controlled clinical trials (mono- and adjunctive therapy) of 11 different AEDs showed that patients randomized to one of the AEDs had approximately twice the risk (adjusted Relative Risk 1.8, 95% CI:1.2, 2.7) of suicidal thinking or behavior compared to patients randomized to placebo. In these trials, which had a median treatment duration of 12 weeks, the estimated incidence rate of suicidal behavior or ideation among 27,863 AED-treated patients was 0.43%, compared to 0.24% among 16,029 placebo-treated patients, representing an increase of approximately one case of suicidal thinking or behavior for every 530 patients treated.

In the controlled adjunctive trial behind the label, dizziness, double vision, nausea, vomiting, sleepiness, ataxia and nystagmus were all several times more common on oxcarbazepine than placebo, and worse at higher doses. (Source 23)

• Official position, Moderate certainty.
• Size: 163 at 600 mg/day, 171 at 1200 mg/day, 126 at 2400 mg/day, 166 placebo.
• Who: adults with epilepsy receiving oxcarbazepine as adjunctive treatment in a controlled clinical study.
• How long: not stated in the table.
• Result: at 2400 mg/day versus placebo: dizziness 49% versus 13%, diplopia 40% versus 5%, vomiting 36% versus 5%, somnolence 36% versus 12%, ataxia 31% versus 5%, nausea 29% versus 10%, nystagmus 26% versus 5%.
• Funding: manufacturer's label (Novartis), a regulatory position dated 2025-12-16.

Dizziness | 26 | 32 | 49 | 13 | Somnolence | 20 | 28 | 36 | 12 | Ataxia | 9 | 17 | 31 | 5 | Nystagmus | 7 | 20 | 26 | 5 |

About a quarter of adults previously treated with other antiseizure drugs stopped oxcarbazepine because of a side effect in the trials, but only about 9% of adults starting it as their first antiseizure drug did. (Source 14)

• Official position, Moderate certainty.
• Size: 1,537 adult patients.
• Who: adults in the controlled trials behind the label: 1537 previously treated with other antiepileptic drugs (adjunctive therapy and conversion to monotherapy), and 295 not previously treated.
• How long: not stated.
• Result: Approximately 23% of 1537 adults in the adjunctive and conversion-to-monotherapy trials discontinued because of an adverse reaction, most commonly dizziness 6.4%, diplopia 5.9%, ataxia 5.2%, vomiting 5.1%, nausea 4.9%, somnolence 3.8% and hyponatremia 1.0%. In the separate group of 295 adults not previously treated with other antiseizure drugs, approximately 9% discontinued for an adverse reaction, most commonly dizziness 1.7%, nausea 1.7%, rash 1.7% and headache 1.4%.
• Funding: manufacturer's label (Novartis), a regulatory position dated 2025-12-16.

Limit of this finding: The 23% figure belongs only to adults who were already on other antiseizure drugs and had oxcarbazepine added or substituted. For adults starting oxcarbazepine as their first antiseizure drug the label reports about 9%. Quoting the 23% alone roughly doubles the rate a person starting monotherapy would face.

Approximately 23% of these 1537 adult patients discontinued treatment because of an adverse reaction.

Oxcarbazepine substantially lowers blood levels of both hormones in a combined oral contraceptive, which can make the contraceptive fail. (Source 7)

• Blood level study, Moderate certainty.
• Size: two pharmacokinetic studies; participant numbers not given in the label.
• Who: people given an oral contraceptive with and without oxcarbazepine.
• How long: coadministration studies.
• Result: mean ethinylestradiol AUC fell by 48% [90% CI 22 to 65] in one study and 52% [90% CI 38 to 52] in another; mean levonorgestrel AUC fell by 32% [90% CI 20 to 45] in one and 52% [90% CI 42 to 52] in the other.
• Funding: manufacturer's label (Novartis), a regulatory position dated 2025-12-16.

Limit of this finding: Two of the four figures the label prints are mis-typeset: it gives "52% [90% CI: 38 to 52]" and "52% [90% CI: 42 to 52]", where the upper limit of the interval equals the point estimate, which cannot be right. The reductions themselves (roughly a third to a half in both hormones) are the label's finding and are consistent across the studies; the two intervals should not be read as usable ranges.

The mean AUC values of EE were decreased by 48% [90% CI: 22 to 65] in one study and 52% [90% CI: 38 to 52] in another study.

On the extended-release oxcarbazepine label, clinically significant low blood sodium occurred in 1.2% of people in the product's own add-on trial, with one person falling to 117 mEq/L and needing to stop treatment. (Source 24)

• Official position, Moderate certainty.
• Size: 366 adults.
• Who: adults with complex partial seizures in a controlled trial of adjunctive Oxtellar XR.
• How long: not stated in this section of the label.
• Result: overall incidence of clinically significant hyponatremia 1.2%; shifts from normal to low (<135 mEq/L) in 6.5% of the 2400 mg group and 9.8% of the 1200 mg group against 1.7% on placebo; 1 patient on 2400 mg reached 117 mEq/L and 2 on 1200 mg reached 125 and 126 mEq/L, and all three discontinued.
• Funding: manufacturer's label (Supernus Pharmaceuticals), effective 2026-09-15.

Limit of this finding: These are the extended-release product's own trial figures. Most of the rest of this label section describes immediate-release oxcarbazepine, which the label is careful to distinguish. Note also that borderline sodium shifts were commoner in the 1200 mg group (9.8%) than the 2400 mg group (6.5%), so these numbers are not a clean dose-response.

The overall incidence of clinically significant hyponatremia in patients treated with Oxtellar XR was 1.2%, although slight shifts in serum sodium concentrations from Normal to Low (<135 mEq/L) were observed for the 2400 mg (6.5%) and 1200 mg (9.8%) groups compared to placebo (1.7%).

The same label instructs that serum sodium be measured if symptoms of low sodium appear, and considered during treatment generally, particularly alongside other drugs known to lower sodium. (Source 24)

• Official position, Certainty not rated.
• Size: not applicable; this is a labelling instruction rather than a result.
• Who: people treated with extended-release oxcarbazepine.
• How long: throughout treatment.
• Result: no effect size. The symptoms the label says should prompt a sodium measurement are nausea, malaise, headache, lethargy, confusion, obtunded consciousness, or an increase in seizure frequency or severity. The named risk factor for routine measurement is concomitant medication known to decrease serum sodium, for example drugs associated with inappropriate ADH secretion.
• Funding: manufacturer's label (Supernus Pharmaceuticals), effective 2026-09-15.

Measure serum sodium concentrations if patients develop symptoms of hyponatremia (e.g., nausea, malaise, headache, lethargy, confusion, obtunded consciousness, or increase in seizure frequency or severity). Consider measurement of serum sodium concentrations during treatment with Oxtellar XR, particularly if the patient receives concomitant medications known to decrease serum sodium levels (for example, drugs associated with inappropriate ADH secretion).

In the pooled antiseizure-drug analysis the absolute excess of suicidal thinking or behaviour was about 2.4 extra cases per 1000 people treated in epilepsy trials and 2.9 per 1000 in psychiatric trials. (Source 25)

• Meta-analysis, Moderate certainty.
• Size: 27,863 drug-treated and 16,029 placebo-treated patients across 199 placebo-controlled trials of 11 antiseizure drugs.
• Who: participants in placebo-controlled trials for epilepsy, psychiatric and other indications.
• How long: median treatment duration 12 weeks.
• Result: events per 1000 patients, placebo versus drug, with the relative risk and the risk difference: epilepsy 1.0 versus 3.4, relative risk 3.5, 2.4 extra per 1000; psychiatric 5.7 versus 8.5, relative risk 1.5, 2.9 extra per 1000; other 1.0 versus 1.8, relative risk 1.9, 0.9 extra per 1000; total 2.4 versus 4.3, relative risk 1.8, 1.9 extra per 1000.
• Funding: FDA pooled analysis, reproduced in the manufacturer's label.

Limit of this finding: The label's table prints risk differences that do not exactly equal drug minus placebo for two rows: psychiatric 8.5 minus 5.7 is 2.8 but 2.9 is printed, and other 1.8 minus 1.0 is 0.8 but 0.9 is printed. This is the FDA rounding unrounded underlying rates, not a transcription error; do not re-derive the differences from the two rounded rates.

| Epilepsy | 1.0 | 3.4 | 3.5 | 2.4 | Psychiatric | 5.7 | 8.5 | 1.5 | 2.9 | Other | 1.0 | 1.8 | 1.9 | 0.9 | Total | 2.4 | 4.3 | 1.8 | 1.9

What the evidence supports

Cochrane found oxcarbazepine added to existing drugs cut seizure frequency more than control in drug-resistant focal epilepsy, but rated the key efficacy outcomes low certainty. (Source 2)

• Systematic review, Low certainty.
• Size: 1,593 participants across six trials.
• Who: people aged 1 month to 65 years with drug-resistant focal epilepsy.
• How long: treatment periods of 9 days to 26 weeks.
• Result: median percentage seizure reduction per 28 days 26% to 83.3% on oxcarbazepine versus 7.6% to 28.7% on control (3 studies, moderate certainty); 50% or greater responder rate RR 1.80 (95% CI 1.27 to 2.56, 6 studies, low certainty); seizure freedom RR 2.86 (95% CI 1.19 to 6.87, 5 studies, low certainty)
• Funding: not stated; the record lists Research Support, Non-U.S. Gov't. Three of the six trials were judged at high risk of bias, mainly from missing methodological detail and high attrition.

The median percentage seizure reduction per 28 days (3 studies; moderate-certainty evidence) ranged from 26% to 83.3% for participants randomised to experimental oxcarbazepine compared to 7.6% to 28.7% for participants randomised to control treatment. Oxcarbazepine may increase the responder rate for 50% or greater reduction in seizure frequency compared to control treatment (RR 1.80, 95% CI 1.27 to 2.56; random-effects model; 6 studies; low-certainty evidence). For seizure freedom, the RR was 2.86 (95% CI 1.19 to 6.87; random-effects model; 5 studies; low-certainty evidence), suggesting an advantageous effectiveness of oxcarbazepine over control treatment.

What the evidence does not support

In Cochrane's network meta-analysis of individual patient data, oxcarbazepine monotherapy performed worse than lamotrigine on time to treatment failure in focal seizures. (Source 3)

• Systematic review, High certainty.
• Size: individual participant data for 14,789 of 22,049 eligible participants from 39 of 89 eligible trials.
• Who: adults and children with focal onset seizures or generalised tonic-clonic seizures starting a single antiseizure drug.
• How long: not stated per trial; outcomes were time to treatment failure, remission and first seizure.
• Result: hazard ratio for treatment failure for any reason, lamotrigine versus oxcarbazepine 1.30 (95% CI 1.02 to 1.66); lamotrigine also performed better than carbamazepine 1.26 (1.10 to 1.44); high-certainty evidence.
• Funding: not stated. Only 67% of eligible participant data could be obtained, which the review itself records.

Limit of this finding: In this review the reference drug is lamotrigine and a hazard ratio above 1 means the comparator did WORSE than lamotrigine, which is why the same sentence says lamotrigine "performs better than most other treatments". "lamotrigine versus oxcarbazepine 1.30 (1.02 to 1.66)" therefore means oxcarbazepine failed about 30% more often than lamotrigine, not the other way round.

HRs (95% CIs) for treatment failure for any reason for lamotrigine versus: levetiracetam 1.01 (0.88 to 1.20), zonisamide 1.18 (0.96 to 1.44), lacosamide 1.19 (0.90 to 1.58), carbamazepine 1.26 (1.10 to 1.44), oxcarbazepine 1.30 (1.02 to 1.66), sodium valproate 1.35 (1.09 to 1.69), phenytoin 1.44 (1.11 to 1.85), topiramate 1.50 (1.23 to 1.81), gabapentin 1.53 (1.26 to 1.85), phenobarbitone 1.97 (1.45 to 2.67).

For generalised onset seizures the same analysis found no treatment beat sodium valproate, and the comparison with oxcarbazepine was too imprecise to say anything either way. (Source 6)

• Systematic review, Moderate certainty.
• Size: individual participant data for 14,789 participants from 39 trials.
• Who: people with generalised onset tonic-clonic seizures with or without other generalised seizure types.
• How long: not stated.
• Result: hazard ratio for treatment failure for any reason, sodium valproate versus oxcarbazepine 1.24 (95% CI 0.72 to 2.14); moderate certainty.
• Funding: not stated.

Limit of this finding: Here the reference drug is sodium valproate and a hazard ratio above 1 means the comparator did worse than valproate. The oxcarbazepine interval, 0.72 to 2.14, comfortably crosses 1, so this comparison is uninformative rather than unfavourable: it is not evidence that oxcarbazepine is worse for generalised onset seizures, only that the data cannot tell.

For people with generalised onset seizures, evidence was more limited and of moderate certainty; no other treatment performed better than first-line treatment sodium valproate, but there were no differences between sodium valproate, lamotrigine or levetiracetam in terms of treatment failure; HRs (95% CIs) for treatment failure for any reason for sodium valproate versus: lamotrigine 1.06 (0.81 to 1.37), levetiracetam 1.13 (0.89 to 1.42), gabapentin 1.13 (0.61 to 2.11), phenytoin 1.17 (0.80 to 1.73), oxcarbazepine 1.24 (0.72 to 2.14)

Cochrane's overall verdict on oxcarbazepine for neuropathic pain was that there is little evidence it works and that harms are probably more common than on placebo. (Source 26)

• Systematic review, Very low certainty.
• Size: 862 participants across five trials.
• Who: adults with painful diabetic neuropathy, radiculopathy or mixed neuropathies.
• How long: at least six weeks.
• Result: no pooled benefit estimate the authors trust; they report very low confidence in the measures of effect.
• Funding: not stated; the record lists Research Support, Non-U.S. Gov't.

This review found little evidence to support the effectiveness of oxcarbazepine in painful diabetic neuropathy, neuropathic pain from radiculopathy and a mixture of neuropathies. Some very-low-quality evidence suggests efficacy but small trials, low event rates, heterogeneity in some measures and a high risk of publication bias means that we have very low confidence in the measures of effect.

For acute bipolar episodes, another off-label use, Cochrane found no difference between oxcarbazepine and placebo in the one eligible trial, which was in children. (Source 5)

• Systematic review, Low certainty.
• Size: 368 participants across seven studies; the primary placebo comparison rests on one trial of 110 children.
• Who: people with mania, hypomania, mixed episodes or rapid-cycling bipolar disorder; no adult placebo-controlled trial existed.
• How long: not stated.
• Result: 50% or greater fall in Young Mania Rating Scale, oxcarbazepine versus placebo OR 2.10 (95% CI 0.94 to 4.73), one study, n=110; versus valproate OR 0.44 (95% CI 0.10 to 1.97, n=60, P=0.273)
• Funding: not stated. The review describes the methodological quality of the included studies as relatively low.

There was no difference in the primary outcome analysis - a fall of 50% or more on the Young Mania Rating Scale (YMRS) - between oxcarbazepine and placebo (N=1, n=110, OR =2.10, 95% CI 0.94 to 4.73) in one study, conducted in children; no studies were available in adult participants.

Cochrane concluded there is not enough good-quality evidence to say whether oxcarbazepine works for acute bipolar episodes, and that it may be worse tolerated than placebo. (Source 27)

• Systematic review, Low certainty.
• Size: seven studies, 368 participants.
• Who: people with acute bipolar episodes.
• How long: not stated.
• Result: no usable pooled estimate; the review reports insufficient trials of adequate methodological quality.
• Funding: not stated.

Currently, there are insufficient trials of adequate methodological quality on oxcarbazepine in the acute treatment of bipolar disorder to inform us on its efficacy and acceptability. Studies predominantly examine the treatment of mania: there are data from subgroup analysis on mixed affective, hypomania and rapid-cycling states.From the few studies included in this review, oxcarbazepine did not differ in efficacy compared to placebo in children and adolescents. It did not differ from other active agents in adults. It may have a poorer tolerability profile compared to placebo.

Where the evidence is mixed

Cochrane's own conclusion was that the true efficacy of oxcarbazepine as an add-on is uncertain and its tolerability is a concern. (Source 28)

• Systematic review, Low certainty.
• Size: six trials, 1,593 participants.
• Who: people with drug-resistant focal epilepsy.
• How long: 9 days to 26 weeks.
• Result: no pooled figure; the review states the 50%-reduction and seizure-freedom estimates rest on low-certainty evidence.
• Funding: not stated; the record lists Research Support, Non-U.S. Gov't.

The efficacy outcomes - 50% or greater seizure reduction and seizure freedom - were derived from low-certainty evidence. We are, therefore, uncertain whether the estimated effect size is representative of the true effect. In contrast, the evidence for median percentage seizure reduction and treatment withdrawal were of moderate certainty: thus, we are fairly certain of the effect estimates' reliability. Overall, we are unsure of the true efficacy of oxcarbazepine, but have concerns about its tolerability.

For painful diabetic neuropathy, an off-label use, the only trial with usable data showed about a third of people reaching 50% pain relief against under a fifth on placebo, with a number needed to treat of 6. (Source 4)

• Systematic review, Very low certainty.
• Size: 862 participants across five trials; the 50% outcome rests on a single trial of 146 people.
• Who: adults with painful diabetic peripheral neuropathy (634 across three trials), radiculopathy (145) or mixed peripheral neuropathic pain (83)
• How long: 16 weeks of treatment for the pain outcomes.
• Result: at least 50% pain reduction 34.8% on oxcarbazepine versus 18.2% on placebo (RR 1.91, 95% CI 1.08 to 3.39, NNTB 6, 95% CI 3 to 41); at least 30% reduction 44.9% versus 28.6% (RR 1.57, 95% CI 1.01 to 2.44, NNTB 6, 95% CI 3 to 114)
• Funding: not stated; the record lists Research Support, Non-U.S. Gov't. Two trials that found little or no benefit did not provide usable data, and the review flags serious imprecision and a high risk of publication bias.

Limit of this finding: The review attaches three qualifiers to this estimate: incomplete outcome data and possible unblinding from obvious side effects put it at high risk of bias, and there was "also serious imprecision and a high risk of publication bias". Both figures come from one 146-person trial, because two other trials that found little or no benefit did not supply usable data. The radiculopathy trial reported no benefit at all for 50% pain relief.

For painful DPN, compared to the baseline, the proportion of participants who reported at least a 50% or 30% reduction of pain scores after 16 weeks of treatment in the oxcarbazepine group versus the placebo group were: at least 50% reduction: 34.8% with oxcarbazepine versus 18.2% with placebo (risk ratio (RR) 1.91, 95% confidence interval (CI) 1.08 to 3.39, number of people needed to treat for an additional beneficial outcome (NNTB) 6, 95% CI 3 to 41); and at least 30% reduction: 44.9% with oxcarbazepine versus 28.6% with placebo (RR 1.57, 95% CI 1.01 to 2.44; NNTB 6, 95% CI 3 to 114; n = 146).

Where the research disagrees

How often oxcarbazepine causes severe low blood sodium

• The FDA-approved TRILEPTAL label, from 14 controlled epilepsy studies, position: In the 14 controlled epilepsy studies, 2.5% of TRILEPTAL-treated patients (38/1524) had a sodium of less than 125 mmol/L at some point during treatment (Source 18)
• Jung and colleagues, from a 2,253-patient real-world cohort at two South Korean hospitals, cohort: Among 2253 patients, the prevalence of severe hyponatremia was 8.4%. (Source 19)

Whether oxcarbazepine is an acceptable first-choice single drug for focal seizures

• The FDA-approved TRILEPTAL label, position: TRILEPTAL is indicated for use as monotherapy or adjunctive therapy in the treatment of partial-onset seizures in adults (Source 29)
• Nevitt and colleagues, Cochrane network meta-analysis of individual participant data, in which lamotrigine is the reference drug and a hazard ratio above 1 means the comparator did worse than lamotrigine, systematic-review: lamotrigine versus: levetiracetam 1.01 (0.88 to 1.20), zonisamide 1.18 (0.96 to 1.44), lacosamide 1.19 (0.90 to 1.58), carbamazepine 1.26 (1.10 to 1.44), oxcarbazepine 1.30 (1.02 to 1.66) (Source 3)

How much

• Reference intake: Dosing is set by the prescriber, not by a reference intake. The FDA-approved label starts adults at 600 mg a day given twice daily, with a recommended add-on maintenance dose of 1200 mg a day and up to 2400 mg a day when converting from other antiseizure drugs; initiation as monotherapy rises in 300 mg/day steps every third day to 1200 mg a day. Recorded here as the manufacturer's position dated 2025-12-16. (Source 30)
• Upper limit: There is no tolerable upper intake level for a drug. The highest adult dose in the label is 2400 mg a day, reached during conversion to monotherapy; in children aged 2 to under 4 the label says the maintenance dose should not exceed 60 mg/kg/day. In the trials 2400 mg a day was the dose at which withdrawals more than doubled against control, so the ceiling is set by tolerability as much as by the label. (Source 30)
• Studied: Cochrane's add-on review pooled trials using 600 mg/day, 1200 mg/day and 2400 mg/day of oxcarbazepine, and reported withdrawal risk separately at each. (Source 15)
• Studied: The label's controlled adjunctive study randomised adults to 600, 1200 or 2400 mg a day of oxcarbazepine or to placebo, and reports side-effect rates for each arm. (Source 23)
• Studied: Cochrane's neuropathic pain review included five placebo-controlled trials totalling 862 participants, with pain outcomes measured at 16 weeks. (Source 4)
• Studied: The real-world hyponatremia cohort defined severe hyponatremia as a serum sodium of 128 mmol/L or less in adults prescribed oxcarbazepine, and identified high oxcarbazepine dosage as one of the key predictors. (Source 31)

A common belief, and what the research shows

The belief: Oxcarbazepine is a safer, cleaner carbamazepine, so it can be used freely for nerve pain and mood as well as seizures.

What the research shows: It avoids some of carbamazepine problems but it is not free of them, and the evidence for the extra uses is poor. Cochrane verdict on nerve pain was that “This review found little evidence to support the effectiveness of oxcarbazepine in painful diabetic neuropathy, neuropathic pain from radiculopathy and a mixture of neuropathies.” In those same trials, “serious adverse effects occurred in 8.3% with oxcarbazepine and 2.5% with placebo (RR 3.65, 95% CI 1.45 to 9.20; n = 634; moderate-quality evidence)”. For bipolar disorder, “There was no difference in the primary outcome analysis - a fall of 50% or more on the Young Mania Rating Scale (YMRS) - between oxcarbazepine and placebo (N=1, n=110, OR =2.10, 95% CI 0.94 to 4.73) in one study, conducted in children; no studies were available in adult participants.” The skin-reaction risk it shares with carbamazepine also persists: the label states “Available clinical evidence, and data from nonclinical studies showing a direct interaction between TRILEPTAL and HLA-B1502 protein, suggest that the HLA-B1502 allele may also increase the risk for SJS/TEN with TRILEPTAL.”

Questions and answers

What is it?

Oxcarbazepine is a prescription antiseizure medicine, a close chemical relative of carbamazepine, taken as tablets or a liquid. It is a prodrug: almost all of its effect comes from a metabolite called the 10-monohydroxy derivative, or MHD. An extended-release version is sold separately. (Source 1)

What does it do in the body?

Its active metabolite blocks voltage-sensitive sodium channels in nerve cells, which steadies over-excitable membranes and stops repeated firing from spreading. Increased potassium flow and effects on calcium channels may add to this. The precise mechanism is not actually known, and it does not act on brain neurotransmitter receptors. (Source 1)

Is it good or bad for you?

For focal (partial-onset) seizures it has a real effect: Cochrane found add-on oxcarbazepine roughly doubled the chance of halving seizures, though it rated that estimate low certainty. Against that, people stop it much more often than controls (RR 1.75), and the higher the dose the worse this gets. As a first-choice single drug it came out behind lamotrigine in Cochrane's individual-patient network analysis. For nerve pain and bipolar disorder the evidence does not support it while the harms persist, which makes those uses a poor trade. (Source 28)

How do you get more of it?

Oxcarbazepine comes only as a prescribed medicine and no food or supplement contains it. The label's adult starting dose is 600 mg a day split into two, going up to a recommended 1200 mg a day as add-on, or to 2400 mg a day when converting from other drugs; children start at 8 to 10 mg per kg a day. Food does not change how much is absorbed, so it can be taken with or without a meal. (Source 30)

If it is harmful, what reduces it?

Reducing or stopping it is a prescriber's decision, and the label is explicit that it should normally be tapered rather than stopped abruptly, because stopping suddenly risks more seizures and status epilepticus. The exception the label allows is a serious adverse event, where rapid discontinuation can be considered. Drugs that induce liver enzymes, such as rifampin, carbamazepine, phenytoin and phenobarbital, lower the active metabolite's level by 25% to 49%. (Source 12)

Why might someone be low in it or missing it?

This does not apply in the nutrient sense: oxcarbazepine is a drug, not something the body stores. What does make levels fall is co-treatment with strong enzyme inducers, which cut the active metabolite by a quarter to a half, and the label recommends monitoring the metabolite's plasma level when such a drug is started, changed or stopped. Many people also end up off it because they could not tolerate it: about 23% of the 1537 adults who had oxcarbazepine added to or substituted for other antiseizure drugs discontinued for a side effect, against about 9% of the 295 adults who started it as their first antiseizure drug. (Source 14)

Which whole foods contain it or feed it?

No whole food contains oxcarbazepine or feeds it. Food is relevant in only one way the evidence addresses: it makes no difference to absorption, so the tablets and liquid can be taken with or without a meal. Alcohol is the one ingested substance the label warns about, because the sedative effects can add up. (Source 9)

What happens if you do not have it?

Not taking oxcarbazepine is normal for almost everyone. For someone whose focal seizures it controls, stopping abruptly is the danger: the label says it should generally be withdrawn gradually because of the risk of more frequent seizures and of status epilepticus. In the add-on trials, people on oxcarbazepine were about 2.9 times more likely to become seizure-free than controls, so losing it means losing that. (Source 2)

How can you test for it?

There is no routine test for whether oxcarbazepine is 'working' other than seizure counts, but two measurements matter. Serum sodium should be considered during maintenance treatment, especially alongside other drugs that lower sodium or if symptoms appear, because 2.5% of treated patients in the trials fell below 125 mmol/L. The plasma level of the active metabolite MHD can be measured, and the label recommends doing so when a strong enzyme inducer is being started or changed. HLA-B*1502 genotyping exists and is relevant to skin-reaction risk in people of the ancestries where the allele is common. The label adds that this genotyping has important limitations and must never substitute for clinical vigilance. (Source 10)

References

1. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 12.1 Mechanism of Action (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
2. Cochrane Database of Systematic Reviews. Oxcarbazepine add-on for drug-resistant focal epilepsy.. 2020. PMID 32129501, DOI 10.1002/14651858.CD012433.pub2. Read the source
3. Cochrane Database of Systematic Reviews. Antiepileptic drug monotherapy for epilepsy: a network meta-analysis of individual participant data.. 2022. PMID 35363878, DOI 10.1002/14651858.CD011412.pub4. Read the source
4. Cochrane Database of Systematic Reviews. Oxcarbazepine for neuropathic pain.. 2017. PMID 29199767, DOI 10.1002/14651858.CD007963.pub3. Read the source
5. Cochrane Database of Systematic Reviews. Oxcarbazepine for acute affective episodes in bipolar disorder.. 2011. PMID 22161387, DOI 10.1002/14651858.CD004857.pub2. Read the source
6. Cochrane Database of Systematic Reviews. Antiepileptic drug monotherapy for epilepsy: a network meta-analysis of individual participant data (generalised onset seizures).. 2022. PMID 35363878, DOI 10.1002/14651858.CD011412.pub4. Read the source
7. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 12.3 Pharmacokinetics, Hormonal Contraceptives and Other Drug Interactions (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
8. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 17 PATIENT COUNSELING INFORMATION (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
9. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 12.3 Pharmacokinetics, Absorption and Distribution (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
10. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 5.1 Hyponatremia, postmarketing cases and monitoring (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
11. DailyMed (US National Library of Medicine), label of record from Novartis Pharmaceuticals. TRILEPTAL prescribing information, Highlights DRUG INTERACTIONS and sections 7.1 Effect of TRILEPTAL on Other Drugs and 7.2 Effect of Other Drugs on TRILEPTAL (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
12. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 5.6 Withdrawal of Antiepileptic Drugs (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
13. DailyMed (US National Library of Medicine), label of record from Novartis Pharmaceuticals. TRILEPTAL prescribing information, section 9 DRUG ABUSE AND DEPENDENCE (9.2 Abuse and 9.3 Dependence) (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
14. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 6.1 Clinical Trials Experience, discontinuations (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
15. Cochrane Database of Systematic Reviews. Oxcarbazepine add-on for drug-resistant focal epilepsy (MAIN RESULTS, adverse effects) .. 2020. PMID 32129501, DOI 10.1002/14651858.CD012433.pub2. Read the source
16. Cochrane Database of Systematic Reviews. Oxcarbazepine for neuropathic pain (MAIN RESULTS, adverse effects).. 2017. PMID 29199767, DOI 10.1002/14651858.CD007963.pub3. Read the source
17. Cochrane Database of Systematic Reviews. Oxcarbazepine for acute affective episodes in bipolar disorder (MAIN RESULTS, adverse effects).. 2011. PMID 22161387, DOI 10.1002/14651858.CD004857.pub2. Read the source
18. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 5.1 Hyponatremia (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
19. Seizure. Predicting oxcarbazepine-induced hyponatremia in adult epilepsy patients: A multicenter machine learning analysis using real-world CDM data.. 2025. PMID 41166858, DOI 10.1016/j.seizure.2025.10.004. Read the source
20. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 5.4 Serious Dermatological Reactions (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
21. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 5.4 Serious Dermatological Reactions, Association With HLA-B*1502 (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
22. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 5.5 Suicidal Behavior and Ideation (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
23. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 6.1 Clinical Trials Experience, Table 3 (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
24. Supernus Pharmaceuticals, Inc., via DailyMed (US National Library of Medicine). OXTELLAR XR (oxcarbazepine) extended-release tablets, section 5.1 Hyponatremia (SPL set id aa610e56-1d1d-11e1-8bc2-0800200c9a66, effective 2026-09-15). 2026. Read the source
25. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, section 5.5 Suicidal Behavior and Ideation, Table 2 (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
26. Cochrane Database of Systematic Reviews. Oxcarbazepine for neuropathic pain (AUTHORS CONCLUSIONS).. 2017. PMID 29199767, DOI 10.1002/14651858.CD007963.pub3. Read the source
27. Cochrane Database of Systematic Reviews. Oxcarbazepine for acute affective episodes in bipolar disorder (AUTHORS CONCLUSIONS).. 2011. PMID 22161387, DOI 10.1002/14651858.CD004857.pub2. Read the source
28. Cochrane Database of Systematic Reviews. Oxcarbazepine add-on for drug-resistant focal epilepsy (AUTHORS CONCLUSIONS).. 2020. PMID 32129501, DOI 10.1002/14651858.CD012433.pub2. Read the source
29. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL (oxcarbazepine) tablets and oral suspension, section 1 INDICATIONS AND USAGE (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
30. Novartis Pharmaceuticals Corporation, via DailyMed (US National Library of Medicine). TRILEPTAL prescribing information, Highlights section 2 DOSAGE AND ADMINISTRATION (SPL set id 4c5c86c8-ab7f-4fcf-bc1b-5a0b1fd0691b, effective 2025-12-16). 2025. Read the source
31. Seizure. Predicting oxcarbazepine-induced hyponatremia in adult epilepsy patients: A multicenter machine learning analysis using real-world CDM data (METHODS section).. 2025. PMID 41166858, DOI 10.1016/j.seizure.2025.10.004. Read the source