To date, several papers about treatment guidelines for bipolar disorder have been published 8910111213141516171819202122232425262728293031323334353637383940. There are also a number of guideline documents developed by national bodies that have been published. The CANMAT 37 and the NICE 34 guidelines are the most recent, but even they fail to incorporate all recent findings and approvals 41.

The gradual acceptance of the use of atypical antipsychotics such as monotherapy and of antidepressants for a limited period of time, and in combination with antimanic agents, seems to be the trend 42. A summary of guidelines is shown in Table 1.

It is generally accepted and supported by the literature that lithium is moderately useful against all phases of BD. It is also believed to exert a specific action on suicide prevention 364344454647484950 and its use is strongly endorsed by all published treatment guidelines 42. It seems to be a somewhat more effective against classic mania (the response rate being around 40%) than against depression 36395152. It has a relatively slow onset of action; clinical improvement generally occurs within 1 to 3 weeks of treatment.

A potential problem may be that after several years of successful use, a number of patients seem to develop a tolerance to lithium, while up to 15% of patients report a lithium discontinuation-induced refractoriness 53.

Resistance to lithium treatment could be predicted by the presence of mixed or dysphoric mania, rapid cycling, many prior episodes, poor interepisode functioning, an episode pattern of depression-mania-euthymia, comorbid substance abuse, and comorbid personality disorder 554. By contrast, patients with an episodic course with euthymic intervals and the absence of rapid cycling may be better responders.

The recommended therapeutic Li blood levels for the treatment of acute mania range from 0.6–1.2 mEq/L, whereas maintenance levels could be lower, ranging from 0.6 to 0.9 mEq/L. Levels higher than 1.2 mEq/L are potentially toxic. When treating a patient with lithium, creatinine clearance is regarded to be the most reliable marker of kidney function to take into consideration.

Adverse events are to be expected during treatment with lithium 55, the most frequent being neurological, endocrinological (usually concerning the thyroid), cardiovascular, renal, gastrointestinal, hematological and dermatological manifestations and lithium intoxication. However, only about 30% of patients have more than minor complaints, whereas less than 20% of have no adverse effects at all.

While lithium seems to be more specific to euphoric mania, specific anticonvulsants (but not all) seem to have a broad spectrum of effectiveness, including mixed, dysphoric and rapid-cycling forms.

Valproic acid is FDA approved for the treatment of acute manic episodes. Its response rate in acute mania is around 50%, compared to a placebo effect of 20–30% 48545657585960616263. Patients respond relatively rapidly (within 1–2 weeks and often a few days). Valproate appears to have a more robust antimanic effect than lithium in rapid cycling and mixed episodes 6364. Concerning bipolar depression, there is only one controlled study supporting the effectiveness of valproate 57, whereas uncontrolled data suggest that it may be less effective than against mania (response rate close to 30%) 5765. Although valproate seems to have significant prophylactic antimanic properties, its prophylactic antidepressant ones are low-to-moderate 656667. Therapeutic serum levels range between 50 and 150 mg/mL. Gastrointestinal symptoms, sedation, tremor, weight gain, hair loss, ataxia, dysarthria and persistent elevation of hepatic transaminases are among its common adverse effects.

Carbamazepine is approved by the FDA only for the treatment of bipolar mania. It is widely used, especially in continental Europe. The response rate against acute mania is close to 50% (similar to that of valproic) 68697071. However, the response rate against bipolar depression appears to be lower (roughly 30% or less) 7273. Carbamazepine seems to be less effective in the prophylaxis against depressive than against manic/mixed episodes 69 and less effective than lithium 7475767778798081. The MAP study in 1997 8182 and a replication in 2003 74 are the most important among studies comparing carbamazepine and lithium. Both studies showed a superiority of lithium over carbamazepine for the treatment of classic mania. A secondary analysis of the MAP data demonstrated that patients that don't respond to lithium may have a favourable response to carbamazepine 77, although its actual long-term efficacy is under question. The recommended dosage against acute mania is 600–1800 mg daily (blood concentration 4–12 mg/mL). Hepatic enzymes (CYP 3A4) induction occurs after several weeks, resulting to a lowering of drug levels. This may require additional upward dose titration 83. Adverse effects are dose-related and include double or blurred vision, dizziness, sedation, ataxia, and diplopia, vertigo, gastrointestinal disturbances, cognitive impairment and hematological effects 58485. The induction of the metabolism of antidepressants, antipsychotics and other anticonvulsants is yet another major problem which makes the use of carbamazepine during combination treatment problematic.

Lamotrigine, at a daily dosage of 50–200 mg may be effective in the treatment of acute bipolar depression but not mania 458687888990919293. Moreover, it may be equally effective to lithium in the prophylaxis of any mood episode 2245. In depression, response rates are double than those observed under placebo (close to 50%). Lamotrigine may also be effective against rapid cycling 54. Treatment should be initiated slowly; 25 mg daily for the first 2 weeks and then 50 mg for another 2 weeks, followed by slow increases, in order to avoid a moderately high incidence of rash.

Topiramate and gabapentin can only be used as supplementary therapy for the treatment of weight gain (topiramate) and anxiety (gabapentin), as data on them is negative 94959697. Data regarding other anticonvulsants is not reliable. It must be pointed out that unlike antipsychotics, that seem to have a possibly antidopaminergic 'class effect' limited to the treatment of acute mania, anticonvulsants have no such effect in any phase of bipolar disorder. Each agent has a very distinct pharmacologic profile, thus should be considered separately.

First generation (typical) antipsychotics (FGAs) are considered to be the traditional first-line treatment for acute mania, especially in Europe. TGAs, mostly haloperidol, have been used for long and are generally regarded to act faster than mood stabilizers. Nevertheless, many psychiatrists share the anecdotal clinical impression that FGAs induce depression.

Unlike FGAs, second generation (atypical) antipsychotics (SGAs) do not induce depression. Moreover, several recent studies support their usefulness in all phases of bipolar illness, either as monotherapy or as an adjunct to conventional mood stabilizers. They have a lower incidence of extrapyramidal symptoms and signs, thus considered to have a more favourable adverse effects profile. Improvement is reported to be similar among different antipsychotic agents, irrespective of whether the antipsychotic was utilized as monotherapy or adjunctive therapy 98. Olanzapine, risperidone, quetiapine, ziprasidone and aripiprazole have already been approved by the FDA for the treatment of acute mania. These drugs are also approved for the treatment of mania in most European countries. Although available data is still limited, SGAs are considered a rather promising option for treating bipolar depression.

The use of adjunct SGAs on anticonvulsants produces a response rate increase of about 20%, while, when used as monotherapy, SGAs produce a roughly 20% difference from placebo.

Risperidone effectiveness in acute mania is supported in several studies 99 with remission rates of 42% vs 13% for placebo 85100101102103104105106107. Dose-related extrapyramidal symptoms, weight gain, sedation and hyperprolactinemia seem to be its main disadvantages 99. There are also a number of studies that included patients with mixed states 101102106.

Olanzapine has the highest number of published randomized control trials (RCTs) 1608587108109110111112113114115116117118119120121122123124125126127128 and a solid basis supporting its use in bipolar disorder 129, hence it is the most well-studied atypical antipsychotic. It is approved by the FDA, but not the EMEA, for the treatment of bipolar depression (only in combination with fluoxetine), and for the maintenance phase for those patients that responded well to olanzapine during an acute manic episode 118122130. Regarding mixed episodes, there are some available data, however its use is not well established. The most common adverse effects reported include dry mouth, weight gain, increased appetite and somnolence 60.

Quetiapine effectiveness in both mania and depression as monotherapy is supported by RCTs 5969131132133134135136137138139. It is currently the only SGA approved by the FDA as a monotherapy (300–600 mg/daily) for both acute mania and bipolar depression. In depression trials, 600 mg/day were found not to be more effective than 300 mg/day. Concerning mixed episodes and rapid cycling, only some uncontrolled data is available 21135. The most common adverse effects include somnolence and hypotension.

The use of aripiprazole and ziprasidone as monotherapy in manic or mixed episodes is supported by existing data 42140141142143144. The most common adverse events are akathisia (Aripiprazole), somnolence and extrapyramidal symptoms (Ziprasidone).

Currently, fluoxetine, as part of the fluoxetine plus olanzapine combination, is the only antidepressant medication officially approved by the FDA for the treatment of bipolar depression 87112126.

In spite of the fact that there are some double-blind studies supporting their effectiveness against bipolar depression 145146147, this is still an open issue. Thus, their use and usefulness in bipolar disorder is still controversial 102. Guidelines suggest their cautious use, always in combination with an antimanic agent 139, as antidepressants may induce switching to mania or hypomania, mixed episodes and rapid cycling 148149150151. In patients receiving a mood stabilizer, the outcome of depression could be improved by the addition of an antidepressant without significantly altering the risk of switch 152. According to earlier studies, switching to mania or hypomania was a considerable risk, especially with tricyclics 46153. However, this may not apply to newer agents. Switching to mania or hypomania may occur in 7–30% of patients. This depends on the antidepressant agent and dose used and the personal (prepubertal onset) and family history 154155. Nevertheless, it is supported by some authors that the true rate of switching is rather low, if any 154156157158. The general concept however, is that dual action agents (TCAs or Serotonin and Noradrenaline Reuptake Inhibitors – SNRIs) may be more potent in increasing the risk for switching to mania or hypomania 148159 and to development of suicidal ideation 38160161. An adjunctive antimanic agent (atypical antipsychotic or anticonvulsant) may protect against switching or mixed symptoms, but this is not always the case 148162.

A warning regarding the possible induction of suicidality (ideas and behavior but not completed suicide) by antidepressants in children and adolescents and possibly in all age groups, has been recently issued by the FDA 163, however data from the STEP-BD program does not support the idea of increased suicidality in bipolar patients treated with antidepressants 164. Thus, this issue remains controversial.

Hard data concerning the effectiveness of psychosocial interventions in BD are emerging. Psychoeducation is what appears to be the first line of psychosocial intervention. In bipolar patients under medication, psychoeduation, family-focused psychoeducation and cognitive-behavioral therapy seem to be the most efficacious interventions for relapse prevention. Moreover, they can help both the patient and family members to learn to recognize early warning signs of oncoming episodes, thus obtain earlier treatment interventions, and to identify possible triggering factors 165.

Although there are no definite data, the efficacy of electroconvulsive therapy (ECT) in acute mania is supported by several older clinical observations and some more recent clinical trials 166167168. Transcranial magnetic stimulation (rTMS) of the brain at 20 Hz over the right but not left frontal cortex or 1 Hz bi-frontally is reported to be effective, however data are still insufficient and no conclusions can be drawn 169170171.