Lamotrigine: Independent Evidence on Epilepsy, Bipolar & Rash

Table of contents

• Evidence summary
• What lamotrigine is
• How it works
• What it is prescribed for
• What works and what does not
• Benefits by claim
• Risks and all side effects
• All interactions
• Who should avoid lamotrigine
• Dosage and how to take it
• Follow the money: who makes it and who funded the evidence
• Related research
• Frequently asked questions
• Sources and funding notes

Evidence summary

Independent evidence and credibility scorecard

Independence tiers: 1 = regulator, guideline body or academic team with no declared commercial funding for the work; 2 = public funding or regulator-approved text with indirect industry ties (for example, a maker-written label or a trial that took some company money); 3 = independent authors working mainly on maker-funded data or maker-funded registries; 4 = maker-funded or maker-run. Credibility runs from A (highest) to D. Most sources are British; the US and Japan are the other clusters.

What lamotrigine is

Lamotrigine is an anti-seizure medicine (ATC code N03AX09) (UK SmPC 5.1). It is sold as tablets, chewable or dispersible tablets, orally disintegrating tablets, extended-release tablets in the US, and an oral suspension in the UK (DailyMed Lamictal listing; emc listing).

• Brand name: Lamictal, from GlaxoSmithKline (GSK). The US application (NDA 020241) was first approved on 27 December 1994 (openFDA). In the UK the first authorisation dates from 21 October 1991 (UK SmPC).
• Generics: The openFDA database returned 30 generic (ANDA) applications at the time of checking, from sponsors including Torrent, Aurobindo, Alembic, Dr Reddy's, Zydus and Glenmark (openFDA query). UK listings show many generic licence holders (emc).
• Legal status: A prescription-only medicine in the UK (emc) and sold on prescription in the US under an FDA-approved label.
• Name confusion: The US label warns that the names Lamictal and lamotrigine can be confused with other medicines, and that different Lamictal formulations can be mixed up. It advises checking the tablets at every fill (FDA label 5.8).

How it works

The US label says the precise mechanism by which lamotrigine works is unknown. One proposed mechanism, "the relevance of which remains to be established in humans", is blocking voltage-sensitive sodium channels (FDA label 12.1). The UK SmPC describes lamotrigine as a use- and voltage-dependent sodium channel blocker that reduces repetitive neuronal firing and glutamate release, and adds that how it works in bipolar disorder "has not been established" (UK SmPC 5.1).

Proven in humans: the clinical effects in the trials below. Laboratory and animal only: the sodium-channel and glutamate explanation for seizure control, and the heart-rhythm effect that drove the 2021 FDA warning, which came from in vitro tests (FDA 2021). The same sodium channel action is why overdose can widen the QRS complex on an ECG (UK SmPC 4.9).

What it is prescribed for

• US epilepsy: add-on treatment for partial-onset seizures, primary generalised tonic-clonic seizures and Lennox-Gastaut syndrome from age 2. Conversion to monotherapy from certain other drugs from age 16 (FDA label 1.1).
• UK epilepsy: add-on or single-drug treatment of partial and generalised seizures from age 13, add-on treatment from age 2, and single-drug treatment of typical absence seizures in children aged 2 to 12 (UK SmPC 4.1).
• US bipolar disorder: maintenance treatment of type I bipolar disorder to delay mood episodes in adults already treated with standard therapy. Not recommended for acute manic or mixed episodes (FDA label 1.2).
• UK bipolar disorder: prevention of depressive episodes in adults with type I bipolar disorder who mostly have depressive episodes. It is not indicated for acute mania or depression (UK SmPC 4.1).
• Off-label in the UK: NICE notes that in 2014 using lamotrigine for acute bipolar depression was an off-label use, though common in UK practice (NICE CG185, 1.6.3).

What works and what does not

Benefits by claim

Epilepsy: the SANAD trial

SANAD was a UK randomised trial run in hospital clinics, with neither patients nor doctors blinded. Arm A enrolled 1,721 people with focal seizures for whom carbamazepine was the standard choice. Arm B enrolled 716 people with generalised or hard-to-classify seizures, for whom valproate was standard.

• Focal seizures (arm A): Lamotrigine was significantly better than carbamazepine for time to treatment failure (hazard ratio 0.78, 95% CI 0.63 to 0.97). A hazard ratio below 1 means fewer people stopped or needed extra drugs on lamotrigine. Time to 12-month remission favoured carbamazepine slightly but not significantly (0.91, 0.77 to 1.09). The authors judged lamotrigine non-inferior at 2 and 4 years in the per-protocol analysis (Marson et al. 2007).
• Rash in SANAD A: Rash was the most common side effect leading to treatment failure overall. It affected 7% allocated to carbamazepine and 3% allocated to lamotrigine (full text).
• Generalised and unclassifiable seizures (arm B): Valproate and lamotrigine did not differ significantly for time to treatment failure (1.25, 0.94 to 1.68). Valproate was better for time to 12-month remission (0.76, 0.62 to 0.94), and more clearly in idiopathic generalised epilepsy (0.68, 0.53 to 0.89). The authors said valproate should remain first choice for many people with these epilepsies (Marson et al. 2007).

Since SANAD, valproate has been restricted for women and girls who could become pregnant, and the UK regulator has added precautions for men. NICE guideline NG217 (recommendation dated 2022) offers lamotrigine, levetiracetam or sodium valproate as first-line options for generalised tonic-clonic seizures, and notes that in January 2025 lamotrigine use under age 13 was off-label (NG217).

Bipolar maintenance: preventing relapse

The most recent dedicated review is a 2021 Cochrane review of 11 randomised trials with 2,314 people (1,146 on lamotrigine, 869 on placebo, 299 on lithium). The authors rated all studies as having unclear risk of bias in at least one area, and five as high risk for outcome assessment (Hashimoto et al. 2021).

• Versus placebo: Recurrence of manic symptoms at one year was lower on lamotrigine (risk ratio 0.67, 95% CI 0.51 to 0.87; 3 studies, 663 people; low certainty). Clinical worsening needing extra psychiatric treatment was also lower (0.82, 0.70 to 0.98; moderate certainty). Stopping treatment for any reason was lower (0.88, 0.78 to 0.99).
• Versus lithium: Overall efficacy was similar, but manic recurrence at one year was higher on lamotrigine (2.13, 1.32 to 3.44; 3 studies, 602 people; moderate certainty). Fewer people on lamotrigine reported at least one adverse effect (0.70, 0.51 to 0.96).
• Network comparison: A 2014 network meta-analysis of 33 trials (6,846 people, 17 treatments) found most drugs reduced relapse versus placebo, and that stopping because of side effects was about 3.8 times as likely on lithium as on lamotrigine (risk ratio 3.76, 95% credible interval 1.13 to 12.66). Even so, the authors said lithium should remain first choice, citing its effect on both mania and depression relapse and stronger supporting evidence (Miura et al. 2014).

In plain English: lamotrigine seems to help prevent relapse, especially into depression, and people tolerate it well. It does less than lithium to prevent mania. The licensing wording matches this: the UK licence is for preventing depressive episodes (SmPC 4.1). Most of the underlying trials were funded by the maker; see the money section. This page does not cover lithium's own evidence; see our lithium review.

Bipolar depression: treating an episode

An independent patient-level analysis pooled five GSK-run trials (1,072 people). More people on lamotrigine responded than on placebo (Hamilton scale risk ratio 1.27, 95% CI 1.09 to 1.47; Montgomery-Åsberg 1.22, 1.06 to 1.41). The benefit appeared only in people with baseline Hamilton scores above 24 (RR 1.47, 1.16 to 1.87), not at or below 24 (1.07, 0.90 to 1.27). The authors called the overall effect modest (Geddes, Calabrese and Goodwin 2009). One author sits on a GSK advisory board, and the paper states GSK ran and funded the trials.

The MRC-funded CEQUEL trial randomised 202 people with bipolar depression, all starting quetiapine, to add lamotrigine or placebo. At 12 weeks the self-rated QIDS score difference was −1.73 points (95% CI −3.57 to 0.11; p=0.066), which did not reach statistical significance. At 52 weeks it was −2.69 (−4.89 to −0.49). Folic acid appeared to cancel lamotrigine's effect (interaction p=0.028). The authors concluded lamotrigine improved outcomes (Geddes et al. 2016). We would read the primary 12-week result as uncertain and the folic acid finding as unconfirmed.

NICE says to consider lamotrigine alone or added to lithium or valproate for bipolar depression in specific circumstances, usually after other options have not worked or if the person prefers it (CG185, 1.6.3 to 1.6.8). See our bipolar depression guide for the wider picture.

Mania

Lamotrigine is not a treatment for mania. The FDA label recommends against it for acute manic or mixed episodes, and NICE says not to offer it for mania (FDA label; NICE 1.5.12).

Risks and all side effects

Who is at higher risk: Both regulators list valproate, exceeding the starting dose and exceeding the step-up rate. A history of allergy or rash to other anti-seizure drugs makes non-serious rash about three times more likely. The HLA-B*1502 gene variant, found mainly in people of Han Chinese and Thai origin, is linked to Stevens-Johnson syndrome and toxic epidermal necrolysis with lamotrigine (UK SmPC 4.4; FDA label). Cases also occur without any of these factors. If a person stopped because of a rash, the label says not to restart unless the benefits clearly outweigh the risks.

Other serious warnings on the US label: hemophagocytic lymphohistiocytosis (a life-threatening immune overreaction); multiorgan hypersensitivity reactions (also called DRESS); blood disorders; suicidal thoughts and behaviour, which are a class warning for anti-seizure drugs; aseptic meningitis; withdrawal seizures; and status epilepticus (FDA label, section 5).

Common side effects. In adults with epilepsy (10% or more): dizziness, headache, double vision, unsteadiness, nausea, blurred vision, sleepiness, runny nose, sore throat and rash. In adults with bipolar disorder (over 5%): nausea, insomnia, sleepiness, back pain, tiredness, rash, runny nose and abdominal pain (FDA label, section 6). The UK SmPC notes that dizziness and double vision can affect driving, and advises people to see how it affects them first (SmPC 4.7).

Overdose: doses 10 to 20 times the maximum have caused unsteadiness, reduced consciousness, convulsions and coma, with QRS widening and QT prolongation on ECG. Fatal cases are reported (SmPC 4.9).

All interactions

Lamotrigine is cleared mainly by a liver process called glucuronidation. Drugs that speed or slow that process change lamotrigine levels. The UK SmPC finds no clinically significant effect of lamotrigine on liver cytochrome P450 enzymes (SmPC 4.5).

Does lamotrigine weaken the contraceptive pill? In a study of 16 volunteers taking 300 mg of lamotrigine, the oestrogen component was unaffected. The levonorgestrel component fell modestly (AUC down 19%). Some blood hormone markers showed a loss of ovarian suppression in some women, but progesterone measurements showed no hormonal evidence of ovulation in any of the 16 (UK SmPC 4.5). The small sample means this is not proof of safety in everyone. The SmPC also suggests considering contraception without a pill-free week as first choice (SmPC 4.2).

Who should avoid lamotrigine

• Anyone who had a serious rash on lamotrigine before: the label advises against restarting unless benefits clearly outweigh risks (FDA label 2.1).
• People with significant structural or functional heart disease: the FDA says to avoid it in certain underlying cardiac disorders or arrhythmias unless the benefit outweighs the risk (FDA 2021).
• People taking valproate: not forbidden, but they need a different, slower schedule (SmPC 4.2).
• Pregnancy: the MHRA review calls lamotrigine and levetiracetam the safer of the medicines it reviewed, but advises monitoring lamotrigine blood levels before, during and after pregnancy (MHRA). EURAP found major malformations in 2.9% of lamotrigine monotherapy pregnancies (74 of 2,514), rising with dose, and with lamotrigine at 325 mg a day or less showing lower rates than all doses of carbamazepine and valproate (Tomson et al. 2018). A 2023 Cochrane review pooled cohort studies and found a risk ratio of 1.99 (95% CI 1.16 to 3.39) versus women without epilepsy, but 1.04 (0.66 to 1.63) versus women with untreated epilepsy (Bromley et al.). The US label cites registry rates of 2.0% to 2.2% (FDA label 8.1). Neurodevelopment data are more limited than malformation data (MHRA).
• Why levels fall in pregnancy: a small study of 8 women on lamotrigine alone found the dose-to-concentration ratio rose 295% by mid-pregnancy, meaning the same dose gave much lower blood levels. With valproate it rose only 60% (Tomson et al. 2006). After birth levels may rise quickly. The MHRA says to monitor for dose-related effects after delivery (MHRA). The funding of this small study could not be verified.
• Breastfeeding: lamotrigine passes into milk. The US label says infants can have high levels and reports rash, apnoea, drowsiness and poor feeding in some breastfed infants, without proof of cause (FDA label 8.2). NICE says breastfeeding is generally safe for the baby for most women on anti-seizure medicines and should be encouraged (NG217 4.3.12). The two bodies take different tones; infant monitoring is advised.
• Children: higher serious rash rates than adults. It is licensed for epilepsy from age 2, and not for bipolar disorder under 18 in the UK (SmPC 4.1).
• Liver and kidney disease: the label advises reducing starting, step-up and maintenance doses by about 25% in moderate to severe liver impairment, and says lower maintenance doses may work in significant kidney impairment (FDA label 2.1).

Dosage and how to take it

Everything below is what the label says or what trials used. It is not personal dosing advice. A prescriber sets the schedule using the maker's titration kits or tables.

• Why slow: The label says the recommended initial dose and step-up rate should not be exceeded, to lower rash risk. Starter and titration kits are provided to help (FDA label 2.1). NICE says to carry out a full blood count and kidney and liver tests before starting, to be aware of the valproate interaction and to follow the SmPC/BNF titration (CG185, 1.10.36).
• Label schedule for adults with bipolar disorder (not on valproate or enzyme inducers): 25 mg daily for weeks 1 and 2, 50 mg daily for weeks 3 and 4, 100 mg daily in week 5 and 200 mg daily in week 6. With valproate: 25 mg every other day for weeks 1 and 2, then 25 mg daily, 50 mg, then 100 mg. The target dose is 200 mg a day (100 mg with valproate, 400 mg with some enzyme-inducing drugs) (FDA label 2.4).
• Epilepsy: epilepsy schedules depend on age and other medicines. Maintenance doses in the label tables reach as high as 700 mg a day with enzyme-inducing drugs without valproate (FDA label 2.2).
• Combined pill users: the label says maintenance doses will in most cases need to be increased by as much as 2-fold when starting an oestrogen pill, and reduced when it is stopped. Dose changes limited to the pill-free week are not recommended (FDA label 2.1).
• Stopping: the label advises a step-wise reduction over at least 2 weeks (about 50% a week) unless safety requires faster. NICE advises at least 4 weeks to reduce relapse risk in bipolar disorder (CG185, 1.10.39).
• Blood levels: NICE says not to routinely measure lamotrigine levels in bipolar disorder unless there is ineffectiveness, poor adherence or toxicity (CG185, 1.10.38). In pregnancy, both NICE and the MHRA advise monitoring (NG217 4.5).
• Generics: the US label and the UK SmPC both describe the brand. Generic products must meet regulator standards, but we did not verify any brand-switch advice here.

Follow the money: who makes it and who funded the evidence

Who made it, and who sells it now

• Originator and brand owner: GSK. The UK licence holder is the Wellcome Foundation Ltd, trading as GlaxoSmithKline UK, London (UK SmPC). The US sponsor is GlaxoSmithKline LLC (openFDA). The label notes Lamictal is a trademark of the GSK group of companies (FDA label).
• Timeline: A US federal appeals court recorded that GSK holds the patent, began selling Lamictal in 1994 and that its patent was set to expire in early 2009 (Third Circuit 2020).
• Today: Lamictal is still on the market, alongside many generics (see above). We did not find a citable market-size figure and give none.

Who funded the evidence

• Bipolar maintenance trials: In the Cochrane characteristics tables, GSK or Glaxo Wellcome is named as a funder for 10 of the 11 included studies (the review's own discussion says eight were funded by the same sponsor), sometimes alongside other drug makers and foundations; one lists no funder. One included trial's lead investigator reported research support, consultancy and speaker fees from GSK and many other makers (Cochrane full text). The Cochrane review itself had no external funding.
• Bipolar depression: the five trials in the 2009 re-analysis "were conducted and funded by GlaxoSmithKline", which supplied the data. The authors state GSK gave no other support, and that one author is on a GSK advisory board (BJPsych 2009). The public MRC-funded CEQUEL trial is the main non-maker test (CEQUEL).
• Epilepsy: SANAD was funded mainly by the NHS Health Technology Assessment programme. About 20% of resources came from companies with drugs under test, including GSK. The statement says the funders had no role in design, analysis or writing (SANAD A; SANAD B).
• Pregnancy: EURAP was funded by eight drug makers (including GSK) plus a Dutch epilepsy charity and Stockholm County Council (EURAP). The Cochrane pregnancy review is NIHR-supported, though several authors declare fees from UCB and GSK (Bromley 2023 full text).

Documented integrity events

• Antitrust litigation: In the Third Circuit's 2020 opinion, direct purchasers challenged GSK and Teva's settlement of a patent dispute over Lamictal and generic lamotrigine. The court said it did not reach the antitrust questions, and sent the class-certification decision back for a redo (Third Circuit 2020). That is an allegation under litigation, not a finding of wrongdoing, and we could not verify the final outcome.
• Safety-label history: the FDA required cardiac testing after reports of abnormal ECGs and serious events, and added warnings in 2020 and 2021 (FDA 2021).

Related research

For wider, evidence-graded context, see these Pure City Research guides:

• Bipolar depression: assessment, treatment evidence and safety
• Lithium: evidence on bipolar disorder, suicide and kidney risk
• Cyclothymia: diagnosis, treatment and risk
• Aripiprazole and antipsychotic add-ons for depression
• Self-harm: assessment, DBT, CBT and safety
• L-methylfolate for depression
• Clonazepam: panic, epilepsy and withdrawal
• Ketamine for depression
• Stress, anxiety and depression: evidence and care options
• Supplements for stress, anxiety and depression

Frequently asked questions

Sources and funding notes

What we could not verify. The US Department of Justice press release on the 2012 GSK settlement blocked automated access, so we make no statement about it. The web-search allowance ran out before we could look for a separate UK regulator communication on the 2021 heart-rhythm finding, or for ILAE, Health Canada or TGA documents. We did not open the original maker-funded bipolar trial papers; their funding is taken from Cochrane's tables. We did not check author conflict statements for Miura 2014, how the FDA, MHRA and NICE are funded, or the funding of the small 2006 pregnancy pharmacokinetics paper. The "Geddes meta-analysis" cited in the request concerns bipolar depression (2009); for maintenance we rely on Hashimoto 2021 and Miura 2014. We found no head-to-head maintenance data against quetiapine in the sources we opened. We make no claim about market size.

1. Lamictal prescribing information (GlaxoSmithKline LLC; DailyMed; text revised October 2025) — US regulator-approved label; written by the maker.
2. FDA Drug Safety Communication, 31 March 2021 — US government regulator.
3. Lamictal tablets SmPC (GSK UK; revised 5 November 2025) — MHRA-approved; licence holder GSK.
4. MHRA Drug Safety Update, 7 January 2021: antiepileptic drugs in pregnancy — UK government regulator.
5. NICE CG185: Bipolar disorder: assessment and management — UK public body.
6. NICE NG217, chapter 5 and chapter 4 — UK public body.
7. Marson et al., SANAD arm A, Lancet 2007 — NHS HTA programme plus about 20% from companies; see full text.
8. Marson et al., SANAD arm B, Lancet 2007 — same funding; see full text.
9. Hashimoto et al., Lamotrigine in the maintenance treatment of bipolar disorder, Cochrane 2021 — Kobe Gakuin University; no declared interests; see full text.
10. Miura et al., Lancet Psychiatry 2014 — funding line reads "None".
11. Geddes, Calabrese and Goodwin, Br J Psychiatry 2009 — trials GSK-funded; analysis by the authors; one GSK advisory-board member. See journal page.
12. Geddes et al., CEQUEL, Lancet Psychiatry 2016 — Medical Research Council.
13. Tomson et al., EURAP, Lancet Neurology 2018 — eight drug makers plus Netherlands Epilepsy Foundation and Stockholm County Council.
14. Bromley et al., Cochrane 2023: monotherapy treatment of epilepsy in pregnancy — NIHR; some author fees from UCB and GSK; see full text.
15. Tomson et al., Neurology 2006: valproate effects on lamotrigine kinetics in pregnancy — small study; funding not verified.
16. In re Lamictal Direct Purchaser Antitrust Litigation, Third Circuit, 22 April 2020 — US federal court opinion.
17. openFDA: Lamictal applications and openFDA: generic lamotrigine applications — US government data.
18. emc lamotrigine listing — UK product information database.

Educational research reviewed 6 October 2026. This article is not medical advice; talk to a qualified clinician or pharmacist before starting, stopping or combining treatments.