Dr. Lembke is clinical instructor, staff physician, and senior research scientist at Stanford University in California.
Disclosure: Dr. Lembke reports no affiliation with or financial interest in any organization that may pose a conflict of interest.
Off-label disclosure: This article includes discussion of unapproved medications for maintenance treatment in adult bipolar disorder (valproate) and for depression in bipolar disorder (lamotrigine).
Please direct all correspondence to: Anna Lembke, MD, Clinical Instructor, Stanford University, 401 Quarry Road, Stanford, CA, 94305; Tel: 650-725-9570; Fax: 650-723-8331; E-mail: email@example.com.
• Lithium, valproate, and lamotrigine are essential medications in the treatment of mood disorders.
• Lithium, valproate, and lamotrigine are all associated with potentially life-threatening side effects.
• Optimizing dosing of lithium, valproate, and lamotrigine in the treatment of mood disorders improves symptom response and compliance as well as decreases the risk of serious adverse events.
Lithium, valproate, and lamotrigine are all effective first-line agents in the treatment of mood disorders, however, they are also associated with serious adverse events. Lithium is known to cause neurotoxicity, renal toxicity, thyroid toxicity, and teratogenic effects. Valproate can result in fulminant hepatic necrosis, acute hemorrhagic pancreatitis, agranulocytosis, thrombocytopenia, and teratogenic effects. Lamotrigine can induce Stevens-Johnson Syndrome, a potentially life-threatening rash. Knowledge of empirically validated dosing strategies for each medication is essential to good care. Optimal dosing of lithium, valproate, and lamotrigine can improve symptom response, decrease time to onset of action, mitigate side effects, limit serious adverse events, and enhance compliance. For example, careful monitoring of lithium serum concentration can reduce the risk of lithium-induced neurotoxicity. Oral loading of valproate can decrease the time to symptomatic response. Slow, gradual titration of lamotrigine can markedly decrease the risk of rash. Knowledge of optimal dosing is essential to finding the safest and most effective treatment for each individual patient.
Tailoring treatment for individual patients is fundamental to good care. The dose that targets the patient’s symptoms without causing intolerable side effects or adverse events is the optimal dose. This article provides evidence-based guidelines for optimal dosing for three mood stabilizing agents: lithium, valproic acid, and lamotrigine. The evidence for optimal dosing derives from numerous double-blind placebo-controlled trials. Clinical case series and peer consensus also inform dosing strategies. The optimal dose is influenced by four major factors: symptom response, side effects/safety, drug-drug interactions, and compliance. “Symptom response” refers to finding the dose of medication that targets the patient’s primary psychiatric symptoms. In the case of mood stabilizers, typical target symptoms are mania, depression, and/or maintaining mood in a euthymic range. “Side effects” and “safety” refer to dosing strategies that offer the fewest number of side effects and minimize the risk of serious adverse events. “Drug-drug interactions” conveys that patients are often taking more than one medication, and these medications can interact in ways that affect serum levels, metabolism, and safety profiles. “Compliance” highlights that medications only work if patients are taking them. Different dosing strategies can enhance compliance by decreasing side effects and increasing convenience. When symptom response, side effects/safety, drug-drug interactions, and compliance are taken into account, optimal dosing can be achieved in most cases.
Lithium was approved by the United States Food and Drug Administration for the treatment of acute mania in 1970. The evidence for lithium as an effective anti-manic agent is incontrovertible. One of the earlier studies to examine the optimal dose of lithium in the treatment of symptoms of acute mania advocated a serum lithium concentration between 0.9 mEq/L and 1.4 mEq/L.1 This serum concentration is higher than what is typically used today. Later studies, like that of Stokes and colleagues,2 illustrated that effective mania response could generally be achieved with doses between 0.5 mEq/kg/day and 0.72 mEq/kg/day, corresponding to serum lithium levels close to 1.0 mEq/L. Numerous controlled, double-blind studies over several decades have now shown a 70% to 80% response rate of lithium monotherapy in acute manic episodes with doses of lithium between 900–1,200 mg a day, corresponding to serum lithium levels of 0.6–1.2 mEq/L.3
The optimal dose of lithium for the treatment of acute mania is different than the optimal dose for maintenance therapy, ie, therapy to prevent future episodes of mania and/or depression. In fact, if “manic doses” are continued once the mania has resolved, the patient is at increased risk to experience side effects, become toxic, or encounter other adverse events associated with too high a dose. The reason for this change is not entirely understood. Mania appears to represent a hyperphysiologic state of arousal, which may in turn cause increased urine blood flow, increased glomerular filtration rate, and/or increased uptake of lithium into excitable tissue.3 Whatever the cause, lithium doses typically need to be lowered as mania resolves.
In 1974, 4 years after approving the use of lithium in acute mania, the FDA gave the indication for the use of lithium in maintenance treatment in bipolar disorder. FDA maintenance approval was based largely on the work of Prien and colleagues,4 in which lithium proved superior to placebo in prevention of relapse to both mania and depression. Since then, cumulative data from 10 placebo-controlled double-blind studies3 show that 34% of euthymic patients taking lithium relapsed to mania or depression, compared with 81% of placebo patients. The optimal dose of lithium to ward against future episodes of mania or depression is somewhat more controversial. Coppen and colleagues5 showed that effective maintenance doses correspond to serum lithium levels as low as 0.4–0.6 mEq/L, with far fewer side effects. By contrast, Gelenberg and colleagues6 reported relapse rates three times higher (38% vs. 13%) with lithium levels 0.4–0.6 mEq/L versus 0.8–1.0 mEq/L. Most clinicians adhere to the package insert, and maintain lithium serum lithium levels to between 0.6 and 1.2 mEq/L.3
If the need to discontinue lithium arises, for reasons other than acute toxicity or safety concerns, the evidence supports a very slow discontinuation off of the medication. More abrupt discontinuation has been associated with higher risk of relapse. Faedda and colleagues7 demonstrated that relapse rates are higher in patients with bipolar disorder who discontinue lithium versus those who remain on the medication, but relapse rates are mitigated if lithium is discontinued slowly over time.
Lithium carries a “black box warning,” which is short hand for saying that it has potentially lethal side effects. Lithium can cause central nervous system (CNS) toxicity, renal toxicity, thyroid toxicity, and teratogenic effects, all of which can be life threatening. Lithium is also associated with numerous non-life threatening but nonetheless very bothersome side effects, the most common of which include tremor, polyuria (excessive urination), dry mouth, nausea, sedation, exacerbated acne and psoriasis, and cognitive dulling.3
The CNS effects of lithium toxicity carry the highest morbidity. Mild CNS toxicity manifests as restlessness, irritability, and sedation. Severe neurotoxicity can progress to delirium, with ataxia, coarse tremor, seizures, and ultimately coma and death. Severe neurotoxicity is associated with lithium serum concentrations exceeding 1.6mEq/L,8 but can occur even when lithium serum concentrations are in the therapeutic range in susceptible individuals. Oakley and colleagues8 conducted a retrospective chart review of 97 cases of lithium toxicity and found that 26 of 28 cases of severe neurotoxicity were in the context of chronic lithium use, while only the remaining two were due to acute overdose. In other words, most cases of lithium toxicity are likely to occur in the context of patients taking the medication as prescribed. They also found that serum lithium concentrations were significantly higher in patients with severe neurotoxicity (2.3 mEq/L) versus those with milder versions of neurotoxicity (1.6 mEq/L), but that there was significant overlap in serum levels in both groups. In trying to determine correlates of lithium neurotoxicity, the study found that increasing age, diabetes insipidus, impaired renal function, impaired thyroid function, and polypharmacy were independent risk factors for lithium poisoning.
Careful monitoring of serum lithium levels can help prevent lithium toxicity and should be done twice a week until stabilized, and then every 6 months, targeting levels ≤1.2 mEq/L. It is essential to get trough levels, in order to accurately compare to extant safety and efficacy data. Therefore, serum lithium levels should be drawn 8–12 hours after the previous dose, typically in the morning. In addition to the usual serum level monitoring, patients should be carefully screened for symptoms of diabetes insipidus, thyroid dysfunction, renal impairment, and drug-drug interactions. If screened positive, careful consideration should be given to lowering the lithium dose with even more careful blood monitoring. In addition, even people who have been stable on a given dose of lithium for many years should be monitored more carefully as they age, as this appears to be an independent risk factor for neurotoxicity.
Any substance capable of affecting renal clearance should be used with care with lithium, and can affect dosing strategies. The largest single class of drugs that would produce a clinically significant interaction with lithium is the diuretics. Diuretics decrease renal clearance, which can elevate lithium levels and produce toxicity. A lower dose of lithium is recommended when used concomitantly with diurectics. Non-steroidal anti-inflammatory drugs (NSAIDs), such as ibuprofen and naproxen, can increase plasma lithium levels, thereby potentially increasing side effects and the risk of toxicity. Patients taking lithium should be counseled to use NSAIDs judiciously or not at all. Abrupt cessation of caffeine in lithium-using patients with high chronic levels of caffeine intake can result in a 24% increase in lithium blood levels, thereby theoretically increasing the risk of toxicity.9 Patients on lithium who are daily consumers of caffeine should be advised against making any sudden changes in caffeine consumption, up or down.
Dosing consists not just of the number of milligrams per occasion, but also the number of occasions per day (frequency). Frequency of dosing is well known to affect compliance, which decreases as frequency increases. Claxton and colleagues10 reviewed medication compliance across 76 different studies and found that once daily dosing was associated with 79% compliance, twice daily with 69%, thrice daily with 65%, and 4 times a day with 51% compliance.
Standard lithium carbonate formulations need to be dosed three times a day to achieve steady-state levels in the blood stream. As per the Claxton and colleagues10 data, compliance is significantly lowered with this dosing regimen. In order to enhance compliance, many providers dose lithium carbonate once a day at night, despite its shorter half-life. The result may be better compliance, but the disadvantage is a lower steady-state serum concentration of the medication, with more side effects at peak serum concentration occurring several hours after dosing.
Lithium is also available in an extended release lithium carbonate formulation. The role of extended release formulations is to extend half-lives so that dosing frequency per day can be reduced and compliance theoretically increased. The decreased peak-serum concentration with extended release formulations may reduce dose-dependent side effects that occur at peak concentrations. The increased serum trough concentrations may lead to better management of symptoms.
Valproic acid is FDA approved for the treatment of acute mania, and first became available for use in the US in 1978. Although it has no maintenance indication in bipolar disorder, it is a first-line agent for maintenance treatment as well.11 Numerous placebo-controlled trials demonstrate the efficacy of valproate in the treatment of acute mania, with therapeutic effect occurring several days after achieving serum concentrations of ≥50mg/L. Optimal dosing usually begins at 15 mg/kg/day, which typically corresponds to 500–1,000 mg/day in two to four divided doses. Valproate should be increased for efficacy and tolerability by 250–500 mg/day every 1–3 days, targeting serum concentrations of 50–150 mg/L.3
The evidence suggests that only 30% of individuals will achieve goal serum concentration (50 mg/L) in 3 days using the standard titrations schedule: 250 mg TID times 2 days, followed by standard dose titration of increasing weekly by 5–10 mg/kg/day.12 For a more rapid response in patients with acute mania, valproate can be orally loaded starting at 20–30 mg/kg/day. Eighty percent of individuals will achieve the goal concentration in 3 days by using a loading dose, ie, 30 mg/kg/day on days 1 and 2, followed by 20 mg/kg/day.12
Like lithium, valproic acid and related compounds carry a “black box” warning. The potentially lethal side effects include hepatic failure, teratogenic effects, acute hemorrhagic pancreatitis, and very rarely agranulocytosis and thrombocytopenia. Unlike lithium, these rare and sudden adverse events appear not to be dose related, so routine blood monitoring does not necessarily decrease their risk. It is nonetheless recommended that hepatic and hematologic parameters be monitored every 6–12 months, to check for transaminase elevation, pancreatitis, and the very rare risk of agranulocytosis/thrombocytopenia.
Acute valproate toxicity is characterized primarily by sedation and cognitive dulling. Unlike lithium, there is no specific valproate serum concentration that is associated with toxicity, but clinical consensus is to target levels <150 mg/L. Vaproate is a strongly protein-bound anticonvulsant, and in patients with conditions which may impact the level of valproate-albumin binding, it is strongly recommended that free serum valproate levels, as oppose to total serum valproate, be measured. For example, patients with chronic liver disease and patients with hypoalbuminemia (burn patients, elderly, pregnancy, AIDS, etc.) should have free drug monitoring.13
Common dose-related side effects include gastrointestinal (GI) symptoms, sedation, hair loss, and weight gain.3 A pooled analysis by Smith and colleagues14 showed improved tolerability with enteric-coated divalproex sodium. There were significant reductions of weight gain, tremor, hair loss, and GI symptoms.14 GI side effects can further be targeted with divalproex sprinkle capsules on food. In general, a lower dose is associated with fewer side effects and a better safety profile. Valproic acid, like all anticonvulsants, should be discontinued with a slow taper, as abrupt discontinuation can result in a seizure, even in the absence of an underlying seizure disorder.
As above, valproic acid is highly protein bound and extensively metabolized by the liver, and, therefore, toxicity can be precipitated by administration with other highly protein-bound and/or hepatically metabolized drugs. For example, aspirin, which is also highly protein bound, can displace valproate from its protein-binding sites and precipitate toxicity. Serum levels of valproate can be decreased drastically in the presence of hepatic-inducing agents. Fluoxetine can increase valproic acid concentrations by inhibiting liver metabolism. Valproate can drastically increase serum levels of lamotrigine and increase the risk of life-threatening rash when administered concomitantly with lamotrigine. Risk of hepatic failure may be increased when valproic acid is administered with other anti-epileptic drugs.3
Valproic acid is available in the US in five oral preparations: valproic acid, sodium valproate, divalproex sodium, extended-release (ER) divalproex sodium, and divalproex sodium sprinkles. It is also available as a suppository and an intravenous preparation. Valproic acid and sodium valproate are rapidly absorbed after oral ingestion and obtain peak plasma concentrations within 2 hours. Divalproex sodium, an enteric-coated compound containing equal amounts of valproic acid and sodium valproate, reaches peak serum concentrations within 3–8 hours, which can be slower with the sprinkles, particularly if taken with food. All oral preparations except for ER should be dosed twice a day once the dose reaches >250 mg/day.3
Divalproex ER is approved for once daily dosing, with sustained release over more than 18 hours. The once daily dosing and improved side-effect profile with divalproex ER is likely to enhance compliance. When changing from delayed-release divalproex to divalproex ER, increase the dose by 8% to 20% to compensate for the overall decrease in drug exposure.15
Lamotrigine was initially approved in 1994 as an add-on therapy in partial epilepsy in adults, with optimal dose amounts between 200 and 400 mg/day.16 In 2003, the FDA approved lamotrigine for the maintenance treatment of adults with bipolar I disorder to delay the time to occurrence of mood episodes (depression, mania, hypomania, mixed episodes). Double-blind, placebo-controlled trials support the use of lamotrigine as prophylaxis against future episodes.17 However, lamotrigine is more effective in preventing relapse to depression, and may indeed be no better than placebo in preventing relapse to mania or hypomania in recently manic or hypomanic patients.18 Double-blind controlled trials have also demonstrated efficacy in the treatment of depression, but not mania.19
Starting dose and titration schedule with lamotrigine are dependent on the presence and type of concomitant medications (see lamotrigine’s drug-drug interactions below). In the absence of drug-drug interactions, lamotrigine should be started at 25 mg/day for 2 weeks, and then increased by 25–50 mg/day every 1–2 weeks until target dose is achieved. The literature on efficacy supports doses between 100 and 200 mg/day, with limited efficacy beyond doses of 400 mg/day.20 There is no specific therapeutic plasma concentration for lamotrigine, and dosing should generally be based on therapeutic response. Serum concentrations of lamotrigine can be obtained, but they are of limited clinical utility in most cases. Clinical trials in the treatment of patients with epilepsy have shown mixed results in the correlation of lamotrigine serum levels and efficacy, with two studies21,22 showing significant correlation between serum concentration and efficacy, and four studies23-26 not showing significant correlation. In the neurology literature, clinical consensus supports a therapeutic serum concentration between 1.5 and 10 micrograms/mL in patients with epilepsy.27 How this information translates to patients with mood disorders is unknown.
Lamotrigine also has a black-box warning with respect to life-threatening rash, but the risk can be mitigated with a conservative starting/titrating dosing schedule.28 Most rashes are benign drug rashes, but rash leading to hospitalization occurs in 0.3%, with 0.1% Stevens-Johnson Syndrome.15 Rash almost always occurs in the first 8 weeks of therapy and is commonly related to concomitant anti-epileptic medication, specifically valproic acid. The evidence supports a dose-related concentration-dependent effect in the development of rash; the higher the initial dose, the higher the risk of rash. A review of initial lamotrigine dose and rash rates revealed the following: 25 mg (1% rash), 50 mg (9% rash), 100 mg (12% rash), and 200 mg (38% rash).15 In addition, the faster the lamotrigine is titrated, the higher the risk of rash. These findings have led to the recommendation to start lamotrigine low and titrate upward slowly to decrease risk of rash, particularly in the present of other anti-epileptic drugs. Since these recommendations were made, studies15 show that risk of rash, including life-threatening rash, with lamotrigine is now comparable to phenytoin, carbamazepine, and phenobarbital.
Beyond rash, side effects associated with lamotrigine include nausea, fatigue, abdominal pain, dry mouth, constipation, vomiting, ataxia, and dizziness. There is also a risk of seizure with abrupt discontinuation of lamotrigine, as with all antiepileptic medication. Hirsch and colleagues27 studied lamotrigine serum levels and tolerability, and found in a chart review of 811 patients who took lamotrigine for epilepsy that toxicity (defined as causing side effects significant enough for the prescribing provider to reduce the dose or change to an alternative medication) increased with increasing lamotrigine serum concentrations. However, other clinical trials23,29 did not find a relationship between serum levels and toxicity.
As alluded to above, the risk of rash, including life-threatening rash, is increased when lamotrigine is given in combination with valproic acid. This phenomenon may be related to acute increases in dose concentration. The half-life of lamotrigine, which is 24 hours, is increased to 60 hours when given with valproic acid. In the presence of valproate-containing concomitant medications, lamotrigine should be started at 25 mg every other day for 2 weeks, and increased to 25 mg a day for weeks 3 and 4. Beyond the first 4 weeks, lamotrigine can be increased by 25–50 mg/day every 1–2 weeks.
Conversely, lamotrigine’s half-life is shortened from 24 hours to 12 hours when hepatic inducers are on board, such as phenytoin, carbamazepine, phenobarbital, synthetic estrogen and progestins, HIV protease inhibitors, rifampin, sertraline, escitalopram, risperidone, and gingko. In the presence of inducing drugs, lamotrigine doses should be increased and in some cases dosed more frequently in a given day, to achieve a more stable serum concentration time curve.15 For example, when administered concurrently with estrogen-containing oral contraceptives, lamotrigine should be increased by as much as two-fold above the target maintenance dose, as oral contraceptives have been shown to decrease lamotrigine by ~50%.30 Drug-drug interactions with lamotrigine may be the one instance in which monitoring serum concentrations of lamotrigine may have real clinical utility. Serum concentrations before and after starting valproic acid or estrogen-containing contraceptives, for example, may serve as a useful guide in regaining therapeutic levels in the presence of polypharmacy.27
Optimal dose frequency for immediate release formulation in the absence of valproate is twice per day, although for convenience it is frequently dosed all at once at night. In the presence of valproate, optimal dose frequency is once per day, due to the longer half-life of lamotrigine in the presence of valproate. Patients on inducing drugs should take lamotrigine more than once per day. An extended release formulation of lamotrigine is currently being developed. Although not yet available in the US, it allows for once daily dosing, although again on inducing drugs needs to be dosed more than once per day.
Optimal dosing of lithium, valproic acid, and lamotrigine depends upon careful attention to many factors, including but not limited to starting dose, rates of titration, serum concentration for efficacy (lithium and valproate), serum concentrations for toxicity (lithium), drug-drug interactions, dosing frequency, and rates of discontinuation. Lithium, which is FDA approved for the treatment of mania and maintenance therapy, requires careful monitoring to avoid neural, renal, and thyroid toxicity. Most cases of lithium neurotoxicity occur in the context of chronic therapeutic use. Caution should be used when lithium is prescribed concurrently with any medication that affects renal clearance.
Valproate is FDA-approved for the treatment of mania, and a non-approved first-line agent for maintenance therapy. Like lithium, valproic serum concentrations are helpful in determining doses related to optimal symptom response. The evidence supports a minimum valproate serum concentration of 50 mg/L in the treatment of mania. Unlike lithium, serious adverse events are not clearly linked to serum concentration levels and cannot be predicted by blood monitoring. Nonetheless, clinical consensus suggest valproate serum concentrations ≤150 mg/L to avoid toxicity. Patients with compromised albumin for any reason need free valproate drug monitoring. Any concomitant medication that is protein bound (eg, aspirin) or hepatically metabolized has a potential drug-drug interaction with valproate and may increase the risk of toxicity.
Lamotrigine is FDA approved for maintenance therapy in bipolar disorder. Lamotrigine is also a first-line agent in the treatment of bipolar depression, although it is not FDA approved for this use. The most important adverse event with lamotrigine is rash, including life-threatening rash. The risk of rash appears to be dose related and can be mitigated by starting low and going slowly. Although lamotrigine does not affect the metabolism of other drugs, its own metabolism and side-effect profile is significantly affected by other medications. Concomitant valproate can increase the risk of serious rash, and hepatic inducers such as estrogen can reduce effective lamotrigine levels by as much as half. Although serum lamotrigine levels are not useful in most cases, in the context of polypharmacy, monitoring serum concentrations can help readjust dosing.
Knowledge of the evidence on dosing in the use of lithium, valproate, and lamotrigine allows providers to optimize care by tailoring treatment to each individual patient. Individualized care is essential to good care, particularly in psychiatry where the inter-individual variability in response to psychotropic medications poses an additional challenge. Keeping the evidence in mind, it is essential to acknowledge that the optimal dose is the dose that allows the patient to function at his or her highest level while preserving safety. In some cases, the optimal dose may be higher or lower than the recommended therapeutic window or FDA-approved range (Table). Therein lies the art in the science. PP
1. Prien RF, Caffey EM Jr, Klett CJ. Relationship between serum lithium level and clinical response in acute mania treated with lithium. Br J Psychiatry. 1971;120(557):409-414.
2. Stokes PE, Kocsis JH, Arcuni OJ. Relationship of lithium chloride dose to treatment response in mania. Arch Gen Psychiatry. 1976;33(9):1080-1084.
3. Schatzberg AF, Nemeroff CB, eds. Textbook of Psychopharmacology. 2nd ed. Washington, D.C: American Psychiatric Press; 1998.
4. Prien RF, Caffey EM Jr. The current status of lithium prophylaxis. Dis Nerv Syst. 1974;35(10):470-471.
5. Coppen A, Abou-Saleh M, Milln P, Bailey J, Wood K. Decreasing lithium dosage reduces morbidity and side effects during prophylaxis. J Affect Disord. 1983;5(4):353-362.
6. Gelenberg AJ, Kane JM, Keller MB, et al. Comparison of standard and low serum levels of lithium for maintenance treatment of bipolar disorder. N Engl J Med. 1989;321(22):1489-1493.
7. Faedda GL, Tondo L, Baldessarini RJ, Suppes T, Tohen M. Outcome after rapid vs gradual discontinuation of lithium treatment in bipolar disorders. Arch Gen Psychiatry. 1993;50(6):448-455.
8. Oakley PW, Whyte IM, Carter GL. Lithium toxicity: an iatrogenic problem in susceptible individuals. Aust N Z J Psychiatry. 2001;35(6):833-840.
9. Mester R, Toren P, Mizrachi I, Wolmer L, Karni N, Weizman A. Caffeine withdrawal increases lithium blood levels. Biol Psychiatry. 1995;37(5):348-350.
10. Claxton AJ, Cramer J, Pierce C. A systematic review of the associations between dose regimens and medication compliance. Clin Ther. 2001;23(8):1296-1310.
11. Gyulai L, Bowden CL, McElroy SL, et al. Maintenance efficacy of divalproex in the prevention of bipolar depression. Neuropsychopharmacology, 2003;28(7):1374-1382.
12. Hirschfeld RM, Allen MH, McEvoy JP, Keck PE Jr, Russell JM. Safety and tolerability of oral loading divalproex sodium in acutely manic bipolar patients. J Clin Psychiatry. 1999;60(12):815-818.
13. Dasgupta A, Usefulness of monitoring free (unbound) concentrations of therapeutic drugs in patient management. Clin Chim Acta. 2007;377(1-2):1-13.
14. Smith MC, Centorrino F, Welge JA, Collins MA. Clinical comparison of extended-release divalproex versus delayed-release divalproex: pooled data analyses from nine trials. Epilepsy Behav. 2004;5(5):746-751.
15. Werz MA. Pharmacotherapeutics of epilepsy: use of lamotrigine and expectations for lamotrigine extended release. Ther Clin Risk Manag. 2008;4(5):1035-1046.
16. Fraser AD. New drugs for the treatment of epilepsy. Clin Biochem. 1996;29(2):97-110.
17. Goodwin GM, Bowden CL, Calabrese JR, et al. A pooled analysis of 2 placebo-controlled 18-month trials of lamotrigine and lithium maintenance in bipolar I disorder. J Clin Psychiatry. 2004;65(3):432-441.
18. Bowden CL, Calabrese JR, Sachs G, et al. A placebo-controlled 18-month trial of lamotrigine and lithium maintenance treatment in recently manic or hypomanic patients with bipolar I disorder. Arch Gen Psychiatry. 2003;60(4):392-400.
19. Geddes JR, Calabrese JR, Goodwin GM. Lamotrigine for treatment of bipolar depression: independent meta-analysis and meta-regression of individual patient data from five randomised trials. Br J Psychiatry. 2009;194(1):4-9.
20. Marangell LB, Martinez JM, Ketter TA, et al. Lamotrigine treatment of bipolar disorder: data from the first 500 patients in STEP-BD. Bipolar Disord. 2004;6(2):139-143.
21. Loiseau P, Yuen AW, Duché B, Ménager T, Arné-Bès MC. A randomised double-blind placebo-controlled crossover add-on trial of lamotrigine in patients with treatment-resistant partial seizures. Epilepsy Res. 1990;7(2):136-145.
22. Schapel GJ, Beran RG, Vajda FJ, et al. Double-blind, placebo controlled, crossover study of lamotrigine in treatment resistant partial seizures. J Neurol Neurosurg Psychiatry. 1993;56(5):448-453.
23. Jawad S, Richens A, Goodwin G, Yuen WC. Controlled trial of lamotrigine (Lamictal) for refractory partial seizures. Epilepsia. 1989;30(3):356-363.
24. Messenheimer J, Ramsay RE, Willmore LJ, et al. Lamotrigine therapy for partial seizures: a multicenter, placebo-controlled, double-blind, cross-over trial. Epilepsia. 1994;35(1):113-121.
25. Reunanen M, Dam M, Yuen AW. A randomised open multicentre comparative trial of lamotrigine and carbamazepine as monotherapy in patients with newly diagnosed or recurrent epilepsy. Epilepsy Res. 1996;23(2):149-155.
26. Smith D, Baker G, Davies G, Dewey M, Chadwick DW. Outcomes of add-on treatment with lamotrigine in partial epilepsy. Epilepsia. 1993;34(2):312-22.
27. Hirsch LJ, Weintraub D, Du Y, et al. Correlating lamotrigine serum concentrations with tolerability in patients with epilepsy. Neurology. 2004;63(6):1022-1026.
28. Joe SH, Chang JS, Won S, Rim HD, Ha TH, Ha K. Feasibility of a slower lamotrigine titration schedule for bipolar depression: a naturalistic study. Int Clin Psychopharmacol. 2009;24(2):105-110.
29. Binnie CD, Debets RM, Engelsman M, et al. Double-blind crossover trial of lamotrigine (Lamictal) as add-on therapy in intractable epilepsy. Epilepsy Res. 1989;4(3):222-229.
30. Sabers A, Ohman I, Christensen J, Tomson T. Oral contraceptives reduce lamotrigine plasma levels. Neurology. 2003;61(4):570-571.