This page focuses on treatments for the mania of bipolar disorders.  While some of these can also help reduce the more common depressive symptoms, others cannot.  In any case, anti-depressants still play a major role in treating bipolar illness.  They are addressed elsewhere.

Lithium, thanks to its much lower suicide rate, is still the gold standard in treatment for bipolar disorder.  It is quite good at preventing mania and also can have a good anti-depressant effect.  Since lithium is available as a generic and is very inexpensive ($6-$12/month), pharmaceutical advertising has been promoting divalproex (Depakote)($100-$240/month), which has a much higher profit margin.  This advertising has caused a dramatic shift in just the last ten years away from lithium to Depakote without any real science to support the shift.  Both drugs can cause considerable weight gain in some patients, although divalproex is worse.  For patients doing well on divalproex, they should definitely stick with it.  However, lithium is still the best researched and is definitely preferred.  

Non-medication strategies that are worthwhile include fish oil or some other omega-3 fatty acid, exercise, folic acid, bright light therapy and others.

Worthwhile medication alternatives to lithium are numerous.  An anti-depressant is often combined with an anti-manic agent, e.g. an atypical anti-psychotic, or certain seizure medications.  Atypicals may prove better than seizure medications, but this area is currently being researched.  My favorite atypical agents are ziprasidone (Geodon), aripiprazole (Abilitat), amisulpride, and quetiapine (Seroquel) since they cause little or no weight gain.  Ziprasidone is the least expensive in the U.S. and has few side-effects.  Risperidone can be imported at a very low cost for patients who must pay for their own medications. 

Lamotrigine (Lamictal) is an unusual seizure medication in that it treats depression although it does not appear very effective against manic swings.  It has studies showing considerably value, especially for bipolar patients with primarily depressive mood swings, but a very large recent study found quite disappointing results.  It does not cause weight gain although it does cost $100-$170 per month and has other potential side-effects.

Carbamazepine (Tegretol), another seizure medicine, appears better than divalproex, both in terms of cost ($15-$30) and in terms of side-effects.  While it is very likely not as good as lithium in blocking suicidal behavior, it appears better than divalproex.  Carbamazepine also does not cause significant weight gain. Unfortunately, neither carbamazepine nor divalproex work very well against depression.  Lithium does have a significant anti-depressant effect.

Thyroid medication can also be used as an adjunct to lithium to reduce rapid cycling and side-effects.  I expect that lamotrigine will become the big replacement for divalproex instead of a big shift back to lithium.  Lamotrigine is still covered by patent, so there is the huge profit margin.  Lamotrigine does have a rare but serious rash side-effect, but will probably prove a better medication than divalproex and as good as lithium.

Suicide Attempts Not Increased in Placebo Groups in Mania or Bipolar Relapse Prevention Trials: In a review of 11 DB PC studies acute manic episodes, including 1,506 patients (117 person-years) in the combined active compound group and 1,005 patients (71 person-years) in the combined placebo group, no suicides and no suicide attempts occurred. In four placebo-controlled studies of the prevention of manic/depressive episode, including 943 patients (406 person-years) in the combined active compound group and 418 patients (136 person-years) in the combined placebo group, two suicides (493/100,000 person-years of exposure) and eight suicide attempts (1,969/100,000 person-years of exposure) occurred in the combined active compound group, but no suicides and two suicide attempts (1,467/100,000 person-years of exposure) occurred in the combined placebo group. Suicide risk in placebo-controlled trials of treatment for acute manic episode and prevention of manic-depressive episode. Storosum JG, Wohlfarth T, et al. Den Haag, The Netherlands. Am J Psychiatry. 2005 Apr;162(4):799-802.

Nassir Ghaemi: A Harvard psychiatrist very close to the drug companies gave a talk at the APA Convention on 5/16/99, which he entitled "rational polypharmacy." Ghaemi is anything but rational, promoting many totally unproven seizure medications for bipolar disorder and publishing repeated "open trial" in which patients are treated like guinea pigs with no control groups, meaning the studies can prove absolutely nothing. He reported that the manic-depression spectrum concept started with Kraepelin. 1960s bipolar idea. The spectrum goes from single life-time episode to type I mania. Ghaemi claimed that people with depression but family history bipolar respond better to mood stabilizers. He claims that psychiatrists should not start bipolar with stabilizer, anti-psychotic and anti-depressant but start with stabilizer. He said it was not proven that anti-depressants prevent depression, but only that they treated depressed episodes. Only lithium and maybe valproate have been shown to do this. Only a minority of patients respond completely to a mood stabilizer. He claimed 25% get rapid cycling due to anti-depressants. Often try to get off of anti-depressants. Gives case of bipolar cycling pt doing best with d/c of anti-depr and rx carbamazepine alone. Claims anti-depressants should be less used in the future. Lithium alone had 10% relapse of manic or depression but lithium and imipramine found 24% manic relapses. So, he uses anti-depressants only in depressed phases. Valproate helped 66% of depressed patients at 8 weeks in a study of 28 patients with unipolar depressed (Ed: Such a study is way too small to even be worth mentioning). 40% got better with valproate alone or lithium. Two mood stabilitizers lithium and valproate together helped 67% and adding a neuroleptic helped still more. Risperidone was better than adding a traditional anti-psychotics in study. Gabapentin might be useful added mood stabilizer although it did no better than placebo in type I bipolar. Lamotrigine in DB 200 pt type I with 200mg with over 50% response rate. 1/1000 Stevens Johnson syndrome so back up to lithium or valproate. Have to go very slowly in dosing the drug. Topiramate being researched and causes weight loss and lamotrigine no weight gain. Start lithium or valproate, if partial response, add risperidone or olanzepine, then add 2^nd mood stabilizer or substitute. Clozapine and ECT have side-effects problems. Purely depressed pt bipolar start with mood stabiliers. Prefer buproprion and paroxetine for antidepr because lower switch rate. Esp for type I patients. Use broader range for type II bec less fear of a switch. Ed: I have an extremely low opinion of Dr. Ghaemi.  He seems to be a cheerleaders for expensive patented medications and many of his claims have not been supported by the research.  He continues to publish open trial reports which seem highly irresponsible.  

Teach Patients to Recognize Signs of Relapse: Randomized study shows teaching pts recog relapse improved social functioning and employment. Relapse is 50% at one year for bipolar and 70% at five yr. 2-4 wk prodrome idiosyncratic to each pt. BMJ 1/16/99.

Nitric Oxide & Adrenomedullin Increased: Nitric oxide (NO) has been implicated to play a role in the pathogenesis of depressive disorders. Adrenomedullin (AM) induces vasorelaxation by activating adenylate cyclase and also by stimulating the release of NO. AM immune reactivity is present in the brain, consistent with a role as neurotransmitter. Both found markedly higher in plasma of BPAD patients. Turkey. Neuropsychobiology 2002 Mar;45(2):57-61

Transcranial Prefrontal Magnetic Stimulation More Effective on Right: In depression rTMS more effective on left. 16 pt DB study for mania with 10 days of 20 2 second trains per day. Grisaru, Ben Gurion U, Am J Psychiatry 11/98;155:1608

Tiagabine for Mania: Another anti-seizure med affecting GABA used in three cases successfully. Kaufman, NJ, Ann Clin Psychiatry 12/98.  Ed: Open trials are to be distrusted.

Tamoxifen Helps Mania: Lithium, valproate, and verapamil all reduce Protein Kinase C isozymes. Tamoxifen is a selection PKC inhibitor. It reduced mania in blinded study without placebos. Bebchuk, Wayne State, APA 5/30/98 Toronto

Possible Aging Effects in Bipolar: 22 DSM-IV bipolar patients and 22 healthy controls underwent a 1.5-tesla Spoiled Gradient Recalled Acquisition (SPGR) MRI. Evaluators blind to patients' identities measured total brain, gray and white matter volumes using a semi-automated software. No differences were found for total brain volume, gray matter or white matter volumes between bipolar patients and healthy controls (MANCOVA, age as covariate, p > 0.05). Age was inversely correlated with total gray matter volume in patients (r = -0.576, p = 0.005), but not in controls (r = -0.193, p = 0.388). Our findings suggest that any existing gray matter deficits in bipolar disorder are likely to be localized to specific brain regions, rather than generalized. The inverse correlation between age and brain gray matter volumes in bipolar patients, not present in healthy controls, in this sample of mostly middle-aged adults, could possibly indicate more pronounced age-related gray matter decline in bipolar patients. Neuropsychobiology 2001;43(4):242-7; Unipolar and bipolar patients did not differ in measures of general intelligence or global cognitive status. Generally, across tests of memory, young bipolar patients exhibited the best performance and elderly bipolar patients (>60yo) exhibited the poorest performance. Neuropsychiatry Neuropsychol Behav Neurol 2000 Oct;13(4):246-53. Post-mortem studies find fewer and smaller neurons and fewer glial in prefrontal with MDD or bipolar. and hippocampus 13% smaller in MRI of depression. Exercise cases cell division in rodents.

Stevens-Johnson and Toxic Epidermal Necrolysis Due to Anticonvulsants: 352 cases in 4 European countries in 6 years with 16% fatality rate. 73 (21%) were due to anticonvulsants: phenobarb (36), carbamazepine (21), phenytoin (14), valproic (13), lamotrigine (3) being used. Risk only in first 8 weeks. Lancet ’99;353:2190

Nimodipine Treatment of Bipolar: Authors claim successful treatment of 30 patients with 10 showing moderate to marked improvement. Partial responders added carbamazepine and four more responded. Two regressed when switched to verapamil. Pazzaglia, NIMH, J Clin Psychopharm 10/98;18:404.  Ed: Open trials are to be distrusted.

Droperidol Treatment of Severe Mania: Three cases successfully Rx with one case dystonia. 10-80mg/d. Brown, U Tx Southwestern, Clin Neuropharm 9/98;21:316.  Ed: Open trials are to be distrusted.

92% Psychiatrists Use Adjuvant Antiepileptics: After lithium or divalproex, 92% report they add antiepileptics.  Gabapentin (Neurontin) and topiramate (Topamax) led the list. Psyc Drug Alerts 6/01. Ed: Much of this practice is not supported by the research.  In fact, many psychiatrists never use lithium despite it being the only treatment proven to have a dramatically lower rate of completed suicides.

Cognitive-Behavioral Counseling Helped Kids in Small Study: 16-session interpersonal/cognitive behavioral therapy vs. control in 36 adolescents at risk for BPD: in a MDD and had at least one parent with a lifetime diagnosis of BPD. No psychotropic drugs were used except during an acute depressive or manic episodes. At study end, 39% of the needs-based intervention group had experienced a manic or hypomanic episode vs. 17% of the interpersonal/cognitive behavioral therapy preventive intervention. Kochman FJ, Hantouche EG, Meynard JA, Bayart D. A randomized trial of an interpersonal/cognitive behavior intervention for preventing bipolar disorder in depressed adolescents: preliminary results. Program and abstracts of the Fifth International Conference on Bipolar Disorder; June 12-14, 2003; Pittsburgh, Pennsylvania. Abstract P113.

Genetic

BDNF Gene & G30/72: Two studies have found associations with bipolar disorder. brain-derived neurotrophic factor gene was found related in 2002. A genecomplexat the end of chromosome 13 found found associated with a 25% increased risk. The G30/72 complex also assoc with schiz. 4/24/03 Am J Hum Gen 5/03 U Chic & NIMH.

Gene 5-HT(4) Receptor Linked: Polymorphisms associated with mood disorder were located within the region that encodes the divergent C-terminal tails of the 5-HT(4) receptor. Japan. Mol Psychiatry 2002;7(9):954-61

GRK3 Gene Mutation Linked to Bipolar: DNA samples from more than 400 families. UCSD, Dr Kelsoe. A defect in GRK3 may make one super-sensitive to dopamine. Molecular Psychiatry 6/03

Induced Mania

Ginseng Mania Case Reported: J Clin Psychopharm 01;21:535

Clarithromycin-Induced Mania: One case. Am J Psychiatry 11/98

Clonazepam Mania: One case. Ikeda, Hiroshima U, Int Clin Psychopharm 7/98

Guanfacine Mania in Kids?: Horrigan, J Child Adol Psychopharm ’98;8:149

Atypicals May Induce: 26 cases in literature associated with risperidone or olanzapine. 16 highly suggestive of causal role. Aubry, J Clin Psych 00;61:649, Switzerland.

Antidepressant Induced Mania: According to Psyc Drug Alerts 5/97, antidepressants induce hypomanic/manic reactions in about 50% of bipolars and 9% of unipolars. PDR ’95 nefazodone associated with 0.3% mania, similar to placebo.

Bupropion Less Switch to Mania than Desipramine: DB bupropion up to 450mg/d vs. desipramine up to 250mg/d for patients on lithium who became depressed. Both meds did equally well but 42% of patients on desipramine switched to mania in following year vs. 14% on bupropion. Dufault, MGH, APA 5/30/98 Toronto.

High Dose Bupropion Increases Switch to Mania: Case report on 600mg/d similar to other cases. Goren, Annals Pharmacoth 00;34:619

Switch to Mania More Common After D/C SSRI or MAOI: Chart review found 11% of patients on SSRIs and 33% (2/6) on MAOIs switched to mania within 14 days of tapering off meds but only 4% (1/25) on tricyclics and 0%(0/23) on atypical antidepressants did. Amy Shriver, MGH, APA 5/30/98 Toronto.

Venlafaxine More Mania Than Paroxetine: Random assignment, single-blind. 55 pt bipolar on mood stabilizer 6 months before depression. Paroxetine aver. 32mg/d, venlafaxine 179mg/d. Response >50% decrease in depression in 43% and 48%. 3% vs. 13% became manic. Vieta, U Barcelona, J Clin Psychiatry 02;63:508

Anti-Depressant May Reduce, Not Increase Bipolar Relapse: Current guidelines recommend d/c anti-depressant 3-6 months after remission. A chart review of 44 outpatients successfully rx with anti-depressant added to mood stabilizer found those remaining on meds for year had 32% relapse vs. 68% relapse for those who d/c meds an average of 6 weeks after recovery. J Clin Psych 01;62:612

TCA Switches High, Mood-Stabilizers Decrease: Maniform switches during inpatient treatment were observed in 39 (25%) patients out of the total of 158 patients. Results indicate that especially patients receiving tricyclic antidepressants are at risk of switching to maniform states. This risk was shown to be significantly less when patients also received a mood stabilising medication (lithium, carbamazepine or valproic acid). Bottlender, Munich, J Affect Disord 2001 Mar;63(1-3):79-83

Comorbid OCD-Bipolar: clomipramine and, to a lesser extent, with selective serotonin reuptake inhibitors was associated with hypomanic switches in OCD-bipolar patients, especially in those not concomitantly treated with mood stabilizers. A combination of multiple mood stabilizers was necessary in 16 OCD-bipolar patients (42.1%) and a combination of mood stabilizers with atypical antipsychotics was required in 4 cases (10.5%). U Pisa. Obsessive-compulsive-bipolar comorbidity: a systematic exploration of clinical features and treatment outcome. Perugi G, Toni C, Frare F, Travierso MC, Hantouche E, Akiskal HS. J Clin Psychiatry 2002 Dec;63(12):1129-34

Rapid Cycling

Rapid Cycling 15%: Meta-analysis of lithium studies found similar response both males 61% and females 66% to lithium.

Rapid Cycling From Anti-Depressants in Women Only: A chart review of 129 bipolars at Massachusetts General Hospital (Harvard) found that 45% had experienced rapid cycling.  They found a gender-specific relationship between antidepressant use prior to first manic/hypomanic episode and rapid-cycling bipolar illness. When antidepressants are prescribed to depressed women who have a risk of bipolar disorder, the risk of inducing rapid cycling is a possibility that should influence therapy. Do antidepressants induce rapid cycling? A gender-specific association. Yildiz A, Sachs GS. J Clin Psychiatry. 2003 Jul;64(7):814-8

Rapid Cycling Not Helped by Lithium or Lamotrigine: More than 4 episodes per year in 3 years before maintenance lithium vs lamotrigine trials didn’t benefit. Joseph Calabrese, Case Western Reserve Univ, 638 patients with bipolar I disorder DB PC 18 months lamotrigine (n = 280; 50-400 mg/day fixed and flexible dose), lithium (n = 167; 0.8-1.1 mEq) or placebo (n = 191). GSK funded researcher. European College of Neuropsychopharmacology Congress held in Prague 9/25/03

High Dose L-thyroxine: Initial study by Bauer & Whybrow, AGP 90;47:435. Case report of woman on clozapine, valproic and 0.15mg/d levothyroxine having marked improvement with increase to 0.25mg/d in 1 week. Did well for 7 months, but then needed increase to 0.30mg/d. Am J Psychiatry 00;157:1704.  Ed: Open trials are to be distrusted.

Levetiracetam (Keppra) Used for Rapid Cycling: A German Univ. of Dresden report of two cases of rapid cycling resolving in 3-4 weeks after the addition of levetiracetam 2000 mg/d to valproate after long histories of bipolar rapid cycling not stopped by lithium, carbamazepine, lamotrigine, valproate, and atypicals. Braunig P, Kruger S: Levetiracetam in the treatment of rapid cycling bipolar disorder. J Psychopharm 2003;17:239-41. Keppra is a anti-epileptic medication.  Ed: Open trials are to be distrusted.

Meds General

Polypharmacy Way Up at NIMH: Chart review of 187 refractory bipolar or unipolar patients. The percentages of patients discharged on treatment with 3 or more medications were 3.3% (1974-1979), 9.3% (1980-1984), 34.9% (1985-1989), and 43.8% (1990-1995). J Clin Psychiatry 2000 Jan;61(1):9-15

Metformin for Valproate, Risperidone, Quetiapine, Olanzapine Obesity: Metformin was assessed as a treatment for weight gain in children taking olanzapine, risperidone, quetiapine, or valproate. 19 10-18yos; 12 boys. 12-week open-label metformin, 500 mg t.i.d. Of the 19 patients, 15 lost weight, three gained 1.6 kg or less, and one had no change. The mean changes in weight and body mass index at 12 weeks were highly significant. Metformin for weight loss in pediatric patients taking psychotropic drugs. Morrison JA, Cottingham EM, Barton BA. Am J Psychiatry 2002 Apr;159(4):655-7

Donepezil: Helpful as Add-On: 11 outpatients bipolar I partially responsive to 2 or more agents started 5mg/d for 4 weeks then increased to 10mg/d if not improved. Six markedly improved. Open study. Burt, Biol Psych ’98;45:959, Harvard. Ed: Open trials are to be distrusted.

Donepezil Helps: A open chart review of 8 pts with cholinesterase inhibitor donepezil for mania poorly controlled by lithium or valproate found 6 showed marked improvement with 5-10mg. Tal Burt, MGH, APA 5/30/98 Toronto.

Donepezil Helps Anti-depressant S-E but May trigger Mania: 5mg/d for 3 weeks then 10mg if nec. 13 pt. 12 had had memory loss rated dramatic improvement with donepezil. Also reduced comstiption and dry mouth. S-E insonia, nausea, vomiting, diarrhea and two bipolars became manic within 12 hr of donepezil. Jacobsen, APA 5/30/98 Toronto

Choline Bitartrate Helps Rapid Cycling: Open study of 6 patients on lithium with rapid cycling. Five had marked decr mania and four and marked reduction of all mood symptoms. Stoll, Harvd, BiolPsychiatry ’96;40:328. Also worked for several patients in a prelim report of a DB Harvd study. Domopulos, MGH, APA 5/30/98. Phosphotidylinositol second messenger system disrupted in bipolar and choline works there hypothetically. Erythrocyte choline elevated in some manic patients who tend to do much worse. Also, those low in erythrocyte choline tend to have 4 manic episodes to each depressive one. Stoll, Harvd, Biol Psychiatry ’91;29:1171. Lithium causes a dramatic rise in erythrocyte choline. A subgroup of psychotic, depressed, and manic patients have unusually high erythrocyte cholines and can be distinguished from those without high levels. Stoll, Havrd, Biol Psych ’91;29:309; High RBC choline has been found to be a marker for doing poorly on lithium for patients measured when already on lithium. Haag, Prog Neuropsychopharm Biol Psych ’87;11:209. Lithium interferes with phosphotidylinositol cycle. Choline helped one patient and physostigmine infusion helped 6 of 9 but no schiz patients. Davis, Biol Psychiatry ’78;13:23

Fish Oil: DB EPA for Rapid Cycling: Keck Jr PE, Freeman MP, McElroy SL, Altshuler LL, Denicoff KD, Nolen WA, Suppes T, Frye M, Kupka R, Leverich GS, Grunze H, Walden J, Post RM. A double-blind, placebo-controlled trial of eicosapentanoic acid in rapid cycling bipolar disorder. Bipolar Disord 2002: 4(Suppl. 1): 26-27. A Medscape article states that Keck failed to replicate Stoll’s findings in treating 59 patients. Keck PE. Omega 3 fatty acids in bipolar disorder. Program and abstracts of the 3rd International Stanley Foundation Bipolar Conference; September 12-14, 2002; Freiberg, Germany.

Magnesium: IV Mg Adjunct in Open Trial: Vienna, 10 pt rx-refractory severe mania on lithium and clonazepam and 5 also on haldol given magnesium sulfate to Mg 200/hr in continuous IV infusion for 7-23 days. 7 of 10 much improved. Five had bradycardia resolve with reduced Mg. Psychiatry Res 1999 Dec 27;89(3):239-46

Magnesium: Magnesium + Verapamil was Better than Verapamil Alone in DB: DB PC p<.015. Giannini. Psychiatry Res 2000 Feb 14;93(1):83-7

Nimodipine Helps Some in Open Trial: Of 30 patients with treatment-refractory affective illness, 10 showed a moderate to marked response to blind nimodipine monotherapy compared with placebo on the Clinical Global Impressions Scale. Fourteen inadequately responsive patients (3 unipolar [UP], 11 bipolar [BP]) were treated with the blind addition of carbamazepine. Carbamazepine augmentation of nimodipine converted four (29%) of the partial responders to more robust responders. Patients who showed an excellent response to the nimodipine-carbamazepine combination included individual patients with patterns of rapid cycling, ultradian cycling, UP recurrent brief depression, and one with BP type II depression. When verapamil was blindly substituted for nimodipine, two BP patients failed to maintain improvement but responded again to nimodipine. NIMH. J Clin Psychopharmacol 1998 Oct;18(5):404-13

Stimulants Make Bipolar Kids Worse:Charts of 195 children and adolescents (mean age, 10.9 years; 2.4-19 yo) diagnosed as having BPD to examine the effects of psychostimulants on bipolar symptoms. 63% had been on a stimulant. 70% of trials resulted in an adverse response: increased cycling, increased aggressive behavior, increased oppositionality, hyperactivity, and sleep difficulties. Greenberg R, Papolos DF, Tresker S. Psychostimulant-induced adverse effects in juvenile-onset bipolar disorder. Program and abstracts of the Fifth International Conference on Bipolar Disorder; June 12-14, 2003; Pittsburgh, Pennsylvania. Abstract P150

Verapamil No Benefit in DB: 3-week double-blind, random-assignment, parallel-group, placebo-controlled inpatient trial of verapamil for patients with acute mania. Of the 32 study patients, 15 were given placebo and 17 were given verapamil. RESULTS: Mean absolute change scores on the Mania Rating Scale at endpoint, with baseline scores as the covariates, did not differ.Janicak, U Ill, Chic. Am J Psychiatry 1998 Jul;155(7):972-3

Verapamil: Lithium Did Better in DB: 40 pt DSM-IV mania 28-day randomized, controlled, single-blind trial of either lithium or verapamil. Patients receiving lithium showed a significant improvement on all rating scales (Brief Psychiatric Rating Scale [BPRS], Mania Rating Scale [MRS], Global Assessment of Functioning [GAF], and Clinical Global Impression [CGI]) compared with those receiving verapamil. The mean MRS score at Day 28 for lithium lower than verapamil (17.47 vs. 24.43). A similar pattern was seen with the BPRS (12.68 vs. 20.57). U Witwatersrand, J Clin Psychiatry 1996 Nov;57(11):543-6

Verapamil: Did as Well as Lithium in Very Small DB: Am J Psychiatry 1992 Jan;149(1):121-2, 20 pt, 4 weeks; Similar in two other early studies.

Verapamil No Benefit in DB: DB 3 wk 32 pt acute mania vs placebo. Verapamil no better than placebo. Other studies DB 15 pt vs lithium found better with lithium, worse with verapamil (Arkonac). 4 wk single-blind 40 pt lithium better than verapamil on all assessments (Walton). Janicak, U Ill, Chic, Am J Psychiatry 7/98;155:972-3