May 2011  << Back  

  Can't open the PDF? Click here for help.



Role of atypical antipsychotics in rapid cycling bipolar disorder: A review of the literature

Melanie L. Zupancic, MD

Department of Internal Medicine, Division of Medicine/Psychiatry, Southern Illinois University School of Medicine, Springfield, IL, USA

BACKGROUND: The role of atypical antipsychotics in the management of bipolar disorder continues to expand. This review summarizes the literature on use of atypicals in rapid cycling bipolar disorder in clinical practice and highlights areas for future study.

METHODS: A PubMed search was done using keywords rapid cycling, atypical antipsychotics, refractory bipolar, aripiprazole, clozapine, olanzapine, risperidone, quetiapine, and ziprasidone. Reference lists from original peer-reviewed articles, review articles, and book chapters were reviewed and articles were extracted.

RESULTS: Data on the use of atypical antipsychotics in rapid cycling bipolar disorder are sparse. Atypical antipsychotics may be effective as anti-manic agents during acute mania and may reduce depressive symptoms when used for short and intermediate durations. Their efficacy as mood stabilizers in maintenance therapy has not been demonstrated.

CONCLUSIONS: The study of atypical antipsychotics in rapid cycling bipolar disorder is in its infancy. Although atypical antipsychotics are useful in acute mania, current data do not support their use as maintenance agents. Future double-blind, randomized studies are needed to establish their efficacy relative to traditional mood stabilizers and their utility as adjuvant agents in this subset of patients.

KEYWORDS: rapid cycling, atypical antipsychotics, refractory bipolar disorder



Rapid cycling (RC) bipolar disorder is characterized by 4 or more distinct mood episodes during a 12-month span. It is estimated to affect 15% to 20% of bipolar disorder patients.1-3 This subset of patients includes a disproportionate number of women and patients with bipolar II disorder.3 Relative to patients with non-rapid cycling (NRC), patients with RC have an earlier onset of disease and a higher incidence of substance abuse, anxiety, and suicidal behavior.1,4,5 Similar to NRC patients, patients with RC tend to spend more time in depressive episodes than manic or hypomanic episodes; however their response to medication is poor relative to NRC patients.3,5,6 An estimated 70% to 80% of RC patients are lithium nonresponders.3,5 Historically, anticonvulsants were thought to be more effective than lithium in RC patients, however, more recent studies do not support this.5,7 The elevated risk of morbidity and mortality and the poor response of RC patients to traditional therapies have increased interest in the role of atypical antipsychotics during the acute and maintenance phases. Despite several studies of individual atypical antipsychotics in RC, there currently is no literature review that summarizes these findings.


Articles were identified through a PubMed search using the keywords rapid cycling, atypical antipsychotics, refractory bipolar, aripiprazole, clozapine, olanzapine, risperidone, quetiapine, and ziprasidone. Reference lists from original, peer-reviewed articles, review articles, and book chapters also were reviewed and the relevant articles were extracted. Given the relative novelty of the atypical agents, no date restrictions were used. Inclusion criteria included original research, case reports, and case series of patients with RC bipolar disorder. Exclusion criteria included articles not in the English language and review articles. A total of 16 articles were identified. Four articles contained data from randomized, double-blind, placebo-controlled trials. Data regarding the specific atypical antipsychotics used, doses, duration of therapy, primary and secondary outcomes, and limitations were extracted by the author.



One study examining the role of aripiprazole in RC was identified.8 Twenty-eight RC patients, who had been stabilized with aripiprazole for at least 6 weeks after an acute manic or mixed episode, were randomized to aripiprazole (n=14) or placebo (n=14) for an initial 26 weeks. Patients who completed the initial 26 weeks were eligible for an additional 74 weeks of double-blind treatment under the same regimen. The primary efficacy end point was the time to relapse during the initial 26-week treatment block, defined as a mood episode requiring hospitalization or addition of psychotropic medication other than aripiprazole. Additional measures included time to relapse during the additional 74 weeks and mean change in Young Mania Rating Scale (YMRS) and Montgomery-Åsberg Depression Rating Scale (MADRS) scores. Seven out of 14 aripiprazole-treated patients and 5 of 14 placebo-treated patients completed the initial 26 weeks of treatment and entered the 74-week phase.

Only 3 of 14 aripiprazole patients completed the full 74 weeks, compared with 0 of 14 placebo-treated patients. After 26 weeks, only 2 of 7 aripiprazole patients had discontinued the study due to lack of efficacy, compared with 7 of 9 placebo-treated patients. The number of patients continuing to the 74-week phase was too small to draw any meaningful conclusions. The mean dose during the initial phase was 25.3±7.2 mg/d. Authors report that the time to relapse was significantly longer with the aripiprazole group than placebo at 26 weeks (26-week hazard ratio [HR], 0.21; 95% confidence interval [CI], 0.04 to 1.03), although it should be noted that the confidence interval is wide and crosses 1. At day 101, 81% of the aripiprazole-treated patients had not yet relapsed. Median survival period (time to relapse) for placebo-treated patients was 118 days. There were no statistically significant differences in mean changes in YMRS and MADRS scores at the end of 26 weeks between the 2 groups. The increases in YMRS and MADRS were consistent with those observed in the NRC population as well. These findings suggest that patients who are initially stabilized on aripiprazole after a manic or mixed phase may elect to continue monotherapy with the expectation of reasonable mood stabilization during the first 4 months after the index episode.


The largest study of clozapine in RC was done by Suppes et al.9 Thirteen RC and 7 NRC patients were randomized to either add-on clozapine or treatment as usual (TAU), which included any intervention with the exception of clozapine. Groups were nonblinded. Patients were followed for 12 months. During the 12 months, median clozapine doses were 286 mg/d in the NRC group and 227 mg/d in the RC group and were not statistically significantly different at any point in the study. Both groups received a median of 4 additional medications. Although both groups improved at least somewhat during the follow up, the NRC patients showed significantly greater improvements in Brief Psychiatric Rating Scale (BPRS) scores. Improvement seen in the RC group was restricted to the first month of study entry. The RC group only experienced significant improvement in BPRS scores during the first month, after which the scores remained the same or increased only slightly. These findings suggest that relative to NRC patients, RC patients have a quicker initial response to clozapine. However, this effect appears to attenuate over time, which may reflect the refractory nature of the disorder in this subset of patients. These findings are limited by the small sample size, the lack of a placebo control group, and the lower median dose of clozapine in the RC group.

Calabrese et al10 reported on an open label study of 7 RC patients and 3 NRC patients during a manic cycle. None of the patients had responded to lithium, carbamazepine, or valproate, or 2 typical antipsychotics in the past. The median clozapine dose was 525 mg/d among RC patients and 500 mg/d among NRC patients. After 13 weeks, BPRS scores showed more improvement in the NRC group than in the RC group (median improvements were 42 and 15, respectively). Significant improvements were seen in both groups on the YMRS (median improvement of 23 for RC and 34 for NRC). These results suggest that clozapine may be an effective anti-manic agent for acute mania in RC patients.

There have been several additional case reports and series demonstrating the efficacy of clozapine in this population.1,9,11-13 The findings of these studies are reviewed in TABLE 1. Many of these studies are limited by small sample sizes, open design, brief follow-up, and the publication bias inherent in all case reports. There were no reports of agranulocytosis. Frequently reported side effects included sedation, weight gain, constipation, and hypersialorrhea.10 Limited conclusions can be drawn from these studies. It appears that clozapine can be an effective anti-manic agent during acute episodes of mania in RC patients. Its role as a mood stabilizer in maintenance therapy for RC bipolar disorder is supported by a few case studies. Given the metabolic side effects and risk of agranulocytosis, this therapy should be reserved for extreme cases.


Use of clozapine in rapid cycling bipolar disorder

Reference No. patients Duration Adjunctive meds (Y/N) Median dose (mg/d) Findings Limitations
Chen et al, 20051 1 156 weeks Y = topiramate 350 mg/d 50 psychiatric hospitalizations prior to topiramate/clozapine, none in 3 years since Limited number, generalizability
Suppes et al, 20049 20 (7 NRC, 13 RC) 12 months Y = unspecified NRC = 329 ± 220 mg/d, RC 238 ± 134 mg/d (not significant) NRC showed statistically significant decrease in BPRS while RC did not Small number, lack of control, open design allowing for adjuvant medication adjustments
Calabrese et al, 199610 10 manic (7 RC, 3 NRC) 13 weeks N NRC 483 mg/d, RC 560 mg/d NRC/RC showed mean decrease in YMRS of 31.67/20.86, mean decrease in BPRS of 34.67/15.41 Small sample size, limited duration, no control
Frye et al, 199611 4 52 to 104 weeks Y = lithium (1) 383 mg/d (dose of one patient not given) Decreased cycle frequency in 2 cases, decreased severity in 2 cases, decreased hospitalization in 3 patients, decreased polypharmacy in all patients Limited size, limited follow-up, lack of control group
Suppes et al, 199412 3 12 to 20 months Y = lithium (2), thyroid (2) 275 mg/d Subjective improvements in functioning, decreased cycle numbers and intensity, more sustained euthymia Lack of control group, limited size
Calabrese et al, 199113 2 3.5 to 5.5 months N 350 mg/d Remission of manic symptoms, improvement in depressive symptoms Limited size, limited follow-up, lack of control group
BPRS: Brief Psychiatric Rating Scale; NRC: non-rapid cycling; RC: rapid cycling; YMRS: Young Mania Rating Scale.

Several studies have evaluated the role of olanzapine in RC bipolar disorder14-16 and are summarized in TABLE 2. Sanger et al14 conducted a double-blind, placebo-controlled trial of patients with RC bipolar I disorderwho presented with a baseline YMRS ≥20. Nineteen patients were randomized to the olanzapine arm and 26 received placebo. Doses ranged from 5 to 20 mg/d. There were no significant differences in age, race, presence of psychotic features, or number of previous episodes. Thirty-six patients presented in a manic phase and 9 were mixed. After 3 weeks of monotherapy, 54% of patients in the placebo group discontinued the medication because of lack of efficacy compared with 21% in the olanzapine arm. The primary end point was mean change in YMRS. The mean change in the olanzapine arm was 13.89 compared with 4.12 for the placebo arm (P=.011). A clinical response, defined as a 50% reduction in YMRS, was achieved in 58% of olanzapine-treated patients vs 28% of placebo-treated patients (P=.066). Changes in secondary outcomes, such as the Hamilton Rating Scale for Depression (HRSD), Positive and Negative Syndrome Scale (PANSS) total and subscores, and the Clinical Global Impression-Severity of Bipolar Disorder (CGI-BP) scale, did not reach statistical significance. There was statistically significant mean weight gain of 1.78±1.85 kg in the olanzapine arm compared with 0.89±2.38 kg in the placebo arm. Changes in fasting blood glucose were not significant. These results support a role for olanzapine during acute mania in RC patients. However, the short duration of the study is not sufficient to provide information about the efficacy of olanzapine in reducing affective episodes during the maintenance phase of treatment.

Gonzalez-Pinto et al15 conducted an open-label study of olanzapine in 13 RC patients during a mixed phase for 4 weeks. The primary outcomes measures were changes in HRSD, YMRS, and CGI. Patients were allowed a concurrent mood stabilizer. Olanzapine doses ranged from 10 to 30 mg/d, with a mean dose of 16 mg/d. Baseline YMRS and HRSD scores were 29.61 and 21.15, respectively. At the study’s conclusion, mean YMRS and HRSD scores were 3.15 and 4.78. After 4 weeks, 10 of 13 patients were deemed responders, defined as a 50% reduction in HRSD and YMRS, and a CGI of 1 or 2. These findings suggest that olanzapine may reduce symptoms of acute mania and depression during a mixed phase in RC patients. However, similar to Sanger et al,14 this study was too short to address the role of olanzapine in the maintenance phase of treatment.

Vieta et al16 pooled data from 2 double-blind, placebo-controlled, 3- to 4-week trials of olanzapine in manic NRC and RC patients with an open label extension up to 1 year. Ninety RC patients and 164 NRC patients were included. RC patients were younger at intake, less likely to have psychotic features, more likely to comorbid substance abuse, and had more prior hospitalizations. Symptom remission was defined as a YMRS and HRSD score ≤7. Sustained recovery was defined as remission for at least 8 weeks. RC patients had slightly lower YMRS scores and higher HRSD scores at baseline than NRC patients. After the initial 3 to 4 weeks of therapy, the YMRS score was reduced by at least 50% in 76.7% of olanzapine-treated RC patients, compared with 60.5% of NRC patients (P=.069) and 50% of placebo-treated RC patients (P=.01). These findings suggest that olanzapine may have a role in treating acute mania in RC patients; however, the effect of olanzapine was not evaluated during the maintenance phase.

Thirty-nine RC and 74 NRC patients elected to continue open-label olanzapine for 1 year.16 There were no significant differences in sex, baseline rating scales, initial treatment with olanzapine or placebo, and degree of improvement from baseline between those who continued open treatment vs those who did not. The mean dose of olanzapine was 12.4 mg/d; adjuvant medications were allowed during the open label phase. YMRS improvement >50% was observed in 37 of 39 RC and 62 of 74 NRC (P=.05). Overall, across all end-of-study assessments, 36.6% of RC patients achieved remission compared with 52.5% of NRC patients (P=.006). More RC patients experienced mood episodes during follow-up than NRC patients. Among 39 RC patients, 27 experienced a total of 56 mood episodes, compared with a total of 42 episodes seen in 23 of 74 NRC patients (P=.002); most of these episodes were the depressed type. Thirty-eight percent of RC patients required rehospitalization compared with 16% NRC patients (P=.01) and 22% RC patients attempted suicide compared with 7% of NRC patients (P=.03). These findings are consistent with the refractory nature of the disorder in this subset of patients, and they do not support the use of olanzapine in the maintenance phase.

Taken together, the data on olanzapine suggest that this agent has a role in RC bipolar disorder during acute mania. Evidence supporting its role as a mood stabilizer during maintenance therapy is lacking.


Use of olanzapine in rapid cycling bipolar disorder

Reference No. patients Duration Adjuvant medications (Y/N) Mean dose (mg/d) Outcomes Limitations
Sanger et al, 200314 45 RC (olanzapine = 19, placebo = 26) (manic = 36, mixed = 9) 3 weeks N Range: 5 to 20 mg/d (mean not given) 21.1% olanzapine group discontinued vs 53.8% placebo (P = .035) Mean YMRS change in olanzapine–13.89 vs–4.12 placebo (P = .011) ≥50% reduction in YMRS achieved in 58% olanzapine vs 28% placebo (P = .066) Limited number, limited follow-up
Gonzalez-Pinto et al, 200215 13 RC (mixed) 4 weeks Y = Valproate (9), valproate + lithium (3), lithium + carbamazepine (1) 16.15 mg/d Response as measured by 50% decrease in YMRS and HRSD and CGI of 1 or 2 seen in 76.9% patients Limited size, limited number, lack of placebo, short duration follow-up
Vieta et al, 200416 90 RC 164 NRC (manic = 181, mixed = 73) 3 to 4 weeks N   ≥50% improvement in YMRS seen in 76.7% olanzapine RC vs 50% placebo RC (P = .01) vs 60.5% olanzapine NRC (P = .069) Limited follow-up
Vieta et al, 200416 39 RC, 74 NRC (manic) 52 weeks Y = lithium (16 NRC, 0 RC), fluoxetine (4 NRC, 10 RC) both fluoxetine and lithium (20 NRC, 10 RC) 12.4 mg/d 27/39 RC experienced new episode vs 23/74 NRC (P = .002) 15/39 RC required rehospitalization vs 12/74 NRC (P = .01) Lack of placebo, limited numbers
CGI: Clinical Global Impression scale; HRSD: Hamilton Rating Scale for Depression; NRC: non-rapid cycling; RC: rapid cycling; YMRS: Young Mania Rating Scale.

Two studies2,17 were identified that addressed the use of risperidone in RC bipolar disorder. Vieta et al2 conducted an open label trial of risperidone in 10 RC patients. The mean age was 36 and 70% of the patients were female. Adjuvant medications were allowed. Seven patients were on concurrent mood stabilizers, 2 patients were on concurrent benzodiazepines, 2 patients were on a combination of benzodiazepines and mood stabilizers, and 1 patient was on risperidone monotherapy. The mean dose of risperidone was 3.2 mg/d (range, 2 to 6 mg/d). Patients were followed for 6 months before and after initiating risperidone. The mean number of affective episodes in the 6 months prior to treatment was 5.5, compared with 2.0 episodes in the 6 months after treatment initiation. Limitations of this study included small size, open label design, and short-term follow-up. However, the decrease in number of affective episodes in the post-treatment phase warrants consideration.

Gianfrancesco et al17 conducted a retrospective analysis of hospitalization among bipolar disorder patients on risperidone, olanzapine, quetiapine, and ziprasidone. Among RC patients, those on risperidone had a higher rate of hospitalization than those taking olanzapine (HR, 3.31; P < .05). The retrospective nature of this study is a limitation, but the findings do suggest decreased efficacy of risperidone in RC patients. These findings seem to contradict those of Vieta et al,2 although differing methodologies limit comparison.

In summary, the data on risperidone in RC patients are sparse and contradictory. Randomized prospective studies are needed to clarify the efficacy of risperidone in RC patients. To date, its use is not supported by existing evidence.


Four studies3,18-20 have examined the role of quetiapine in RC patients (TABLE 3). Vieta et al18 randomized 542 depressed bipolar disorder patients to quetiapine, 300 mg/d or 600 mg/d, or placebo for 8 weeks. Of these, 119 patients were classified as RC. Concomitant medications were prohibited. The primary end point was the mean change in the MADRS. Of these 119 patients, 108 had at least one post-baseline assessment and were analyzed in an intent-to-treat basis. All 3 groups were matched at baseline for demographic and clinical characteristics including baseline MADRS, YMRS, and HRSD. At the end of 8 weeks, the mean MADRS change was significantly greater in the 600 mg/d (–21.1) and 300 mg/d (–20.7) groups than in the placebo group (–11.6, P < .001). The percentage of patients meeting MADRS response criteria (≥50% improvement) at week 8 was greater in both treatment arms than in the placebo group: 75% of patients receiving 300 mg/d responded, compared with 60% of patients taking 600 mg/d and 30% receiving placebo. However, the improvement was statistically significant only in the 300 mg/d group (P < .001). Patients with NRC bipolar I disorder who received 300 or 600 mg/d also showed statistically significant improvements in MADRS compared with placebo.21 Interestingly, although RC patients often are considered treatment resistant, their response to quetiapine was similar to that of NRC bipolar I patients. Limitations of this study included the small number of patients and exclusion of patients with comorbid Axis I diagnosis, which limits the generalizability of the results to real world practice. Despite these limitations, these findings support the role of quetiapine for depressive symptoms in RC patients at least in the acute phase. With only 8 weeks’ follow-up, this study does not address the role of quetiapine in the maintenance phase of treatment.

Goldberg et al19 conducted an open label study of 41 RC patients who received quetiapine monotherapy with flexible dosing (n=19) or add-on therapy with quetiapine (n=22) for up to 1 year. Mean age was 41.1 and 83% of patients were female. Patients already on antidepressants or mood stabilizers (lithium, valproate, carbamazepine, and lamotrigine) were allowed to continue with fixed doses of these agents; no other medication “add ons” were permitted during the study. Patients who did not tolerate or refused standard mood stabilizers, and those whose disorder was refractory to standard mood stabilizers were assigned quetiapine monotherapy. Primary outcome measures were changes in YMRS and HRSD scores. The mean maximal dose of quetiapine for all patients was 196.6±195.1 mg/d (range 25 to 900). There were no significant differences in baseline HRSD scores between the add-on therapy and monotherapy groups. Baseline YMRS scores were lower in patients taking concomitant antidepressants. The study dropout rate was 68%. The mean duration of study participation was 18±16.9 weeks and high dropout after 28 weeks limits the reliability of findings. For the total sample, 11 of 27 patients were responders for manic symptoms and 4 of 14 were responders, as measured by YMRS and HRSD scores, for depressive symptoms. Changes in YMRS scores from baseline to end were statistically significant, however, changes in HRSD scores did not reach statistical significance. Mean changes in YMRS scores were similar between patients on monotherapy and add-on therapy. Primary reasons for dropout included inefficacy (defined as the need for changes to the current treatment or insufficient progress, as determined by a clinician’s judgment), nonadherence, and side effects. Common side effects included dry mouth (10/41) and weight gain (7/41). Although these findings are encouraging at face value, the high attrition rate limits optimism. Other study limitations include the small sample size, non-random assignment, and open label design.

The only study to compare an atypical antipsychotic with a “gold standard” mood stabilizer was done by Langosch et al,20 who conducted an open label, randomized pilot study of 38 remitted or partially remitted RC patients recovering from an acute phase on quetiapine (n=22) or valproate (n=16) for 12 months. Patients with comorbid substance abuse, schizophrenia, borderline personality disorder, or antisocial personality disorder were excluded. Mean quetiapine dose was 465±167 mg/d and mean valproate dose was 1340±393 mg/d. Haloperidol was allowed for acute manic episodes and paroxetine for acute depressive episodes. Efficacy measures included CGI, Life Chart Method (LCM), HRSD, YMRS, and MADRS; analysis was done in an intent-to-treat manner. Compared with the valproate-treated individuals, patients treated with quetiapine had a longer duration of illness (19.5±12.2 years vs 10.6±8.4 years, P ≤ .05) and were more often inpatients (52% vs 13%, P ≤ .05). There were no significant differences at baseline between the 2 groups in history of suicide attempt or number of depressive, hypomanic, or manic episodes. According to LCM analysis, the treatments produced no statistical difference in the number of euthymic days, subclinically depressed days, hypomanic, or moderate-to-severe manic days. Quetiapine-treated patients had significantly fewer days of moderate-to-severe depression (P=.04). Mean changes in YMRS, HRSD, and MADRS did not differ between the groups. Responder rates on the depressive dimension of CGI-BP-II were higher in the quetiapine arm (43%) than the valproate arm (25%), but this did not reach statistical significance.

The significance of these findings is limited by the small sample size, which was probably not sufficient to detect true differences, as well as the open label design and absence of a placebo group. Additionally, the broad exclusion criteria limits the generalizability of these findings. However, this study suggests that in RC patients, quetiapine may be an acceptable alternative to valproate, particularly for depressive symptoms. Patients in the quetiapine arm appeared to have a heavier burden of disease at entry, as evidenced by a longer duration of illness and greater amount of time spent as an inpatient; this observation supports a role for quetiapine. However, before quetiapine can be recommended as equivalent therapy to valproate, future studies with sound methodology need to corroborate these findings.

Vieta et al3 conducted an open label study assessing the impact of add-on quetiapine in 14 patients for an average of 112 days. Patients were manic (n=5), mixed (n=3), depressed (n=3), hypomanic (n=2), and euthymic (n=1). After quetiapine was added, concomitant medication doses remained the same. Efficacy was measured by CGI-BP, YMRS, and HRSD. The addition of quetiapine produced a statistically significant reduction in 2 of the 3 subscales of the CGI-BP, including general and manic subscales. No significant difference was observed in the depressive subscale or the HRSD, either in the sample as a whole or in patients with a depressive index episode. The YMRS showed a statistically significant improvement in manic symptoms for the sample as a whole. The average maximum dose was 443±235 mg/d and the final mean dose was 268±190 mg/d. The lack of efficacy in depressive domains is surprising in light of quetiapine’s approval for bipolar depression. However, these findings are limited by the small sample size.


Use of quetiapine in rapid cycling bipolar disorder

Reference No. patients Duration Adjuvant medications (Y/N) Mean dose(mg/d) Outcomes Limitations
Vieta et al, 20023 14 RC (manic = 5, mixed = 3, depressed = 3, hypomanic = 2, euthymic = 1) 112 ± 33 days Y = lithium (n = 8), valproate (n = 7), carbamazepine (n = 5) alone or in combination Mean maximum = 443 ± 235 mg/d Final mean = 268 ± 190 mg/d 6/14 patients met criteria for remission via CGI-BP Small sample size, open design, lack of placebo
Vieta et al, 200718 108 (placebo = 35, quetiapine = 73) 8 weeks N 600 mg/d (n = 31) 300 mg/d (n = 42) Mean decrease in MADRS significantly greater in 600 mg (–21.1) and 300 mg (–20.7) vs placebo (–11.6) P < .001
Percentage of patients meeting response (≥50% reduction in MADRS) was significantly greater in 300-mg group vs placebo
Restricted to patients in depressed episode, short length follow up, exclusion of comorbid substance abuse and Axis I diagnosis limits ability to generalize results
Goldberg et al, 200819 41 (monotherapy = 19, add-on = 22) (manic = 11, hypomanic = 5, mixed = 11, depressed = 14) 52 weeks Y = lithium, valproate, carbamazepine, lamotrigine, antidepressant 196.6 ± 195.1 mg/d 68% patients terminated the study prematurely, most frequently for lack of efficacy Open label design, lack of placebo, high attrition rate
Langosch et al, 200820 38 remitted or partially remitted RC (quetiapine = 22, valproate = 16) 52 weeks N Quetiapine = 465 ± 167 mg/d Valproate = 1340 ± 393 mg/d Quetiapine treated patients had fewer days with moderate-to-severe depressive symptoms (11.7±16.9 vs 27.7 ± 24.9, P = .04) No significant differences in days with subclinical depression, hypomania, or moderate to severe mania. Responder rates on depressive dimension of CGI-BP and manic dimension did not reach statistical significance Small sample size, exclusion of comorbid substance use, open label, lack of placebo
CBI-BP: Clinical Global Impression Severity of Bipolar Disorder Scale; HRSD: Hamilton Depression Rating Scale; MADRS: Montgomery-Åsberg Depression Rating Scale; NRC: non-rapid cycling; RC: rapid cycling; YMRS: Young Mania Rating Scale.

There were no studies identified.


The use of atypical antipsychotics in the treatment of bipolar disorder continues to evolve. Their role in the management of RC patients is far from clear. The studies to date are limited by small sample sizes, brief follow-up periods, and high attrition rates. Furthermore, most studies are nonrandomized, open label design, and many lack a placebo control. Data were insufficient to draw conclusions about different management approaches for patients with bipolar I and II disorder. The role of atypical antipsychotics after depressive, mixed, hypomanic, and manic episodes is also not clear based on existing data. With the exception of 1 study involving quetiapine, no studies compared atypical antipsychotic monotherapy with traditional mood stabilizers. Furthermore, there are no prospective studies comparing one atypical antipsychotic to another. However, based on the existing data, clozapine and olanzapine may be effective anti-manic agents during acute mania in RC patients; their role as mood stabilizers in maintenance phase is not clear. Quetiapine may have a role in the treatment of RC patients during acute depressive phases and may limit the number of depressed days during maintenance therapy. However, supporting evidence is inconsistent.

Prospective, randomized, double-blind studies are needed to assess the role of atypical antipsychotics in the acute and maintenance phase of RC bipolar disorder. These studies must compare the atypical antipsychotics with traditional mood stabilizers such as lithium, valproate, and carbamazepine. They also must examine the use of the atypical antipsychotics as auxiliary agents in different phases, and compare the efficacy of one atypical vs another.

At best, based on existing information, atypical antipsychotics can be considered second-line therapy for individuals who intolerant of multiple traditional mood stabilizers and those with refractory bipolar disorder. There is limited evidence that they may be effective during acute affective episodes, however, their use during the maintenance phase is not evidence based.

ACKNOWLEDGEMENT: Stephen Soltys, MD, Southern Illinois University School of Medicine, Department of Psychiatry for his careful review of this manuscript.

DISCLOSURE: The author reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.


  1. Chen CK, Shiah I, Yeh CB, et al Combination treatment of clozapine and topiramate in resistant rapid-cycling bipolar disorder. Clin Neuropharmacol. 2005;28:136–138.
  2. Vieta E, Gasto C, Colom F, et al Treatment of refractory rapid cycling dipolar disorder with risperidone. J Clin Psychopharmacol. 1998;18:172–174.
  3. Vieta E, Parramon G, Padrell E, et al Quetiapine in the treatment of rapid cycling bipolar disorder. Bipolar Disord. 2002;4:335–340.
  4. Hajek T, Hahn M, Slaney C, et al Rapid cycling bipolar disorders in primary and tertiary care treated patients. Bipolar Disord. 2008;10:495–502.
  5. Tondo L, Hennen J, Baldessarin RJ. Rapid-cycling bipolar disorder: effects of long-term treatments. Acta Psychiatr Scand. 2003;108:4–14.
  6. Calabrese JR, Shelton MD, Bowden CL, et al Bipolar rapid cycling: focus on depression as its hallmark. J Clin Psychiatry. 2001;62(suppl 14):34–41.
  7. Calabrese JR, Shelton MD, Rapport DJ, et al A 20-month, double-blind, maintenance trial of lithium versus divalproex in rapid-cycling bipolar disorder. Am J Psychiatry. 2005;162:2152–2161.
  8. Muzina DJ, Momah C, Eudicone JM, et al Aripiprazole monotherapy in patients with rapid-cycling bipolar I disorder: an analysis from a long-term, double-blind, placebo-controlled study. Int J Clin Pract. 2008;62:679–687.
  9. Suppes T, Ozcan ME, Carmody T. Response to clozapine of rapid cycling versus non-cycling patients with a history of mania. Bipolar Disord. 2004;6:329–332.
  10. Calabrese JR, Kimmel SE, Woyshville ML, et al Clozapine for treatment-refractory mania. Am J Psychiatry. 1996;153:759–764.
  11. Frye MA, Altshuler LL, Bitran JA. Clozapine in rapid cycling bipolar disorder. J Clin Psychopharmacol. 1996;16:87–90.
  12. Suppes T, Phillips KA, Judd CR. Clozapine treatment of nonpsychotic rapid cycling bipolar disorder: a report of three cases. Biol Psychiatry. 1994;36:338–340.
  13. Calabrese JR, Meltzer HY, Markovitz PJ. Clozapine prophylaxis in rapid cycling bipolar disorder. J Clin Psychiatry. 1991;11:396–397.
  14. Sanger T, Tohen M, Vieta E, et al Olanzapine in the acute treatment of bipolar I disorder with a history of rapid cycling. J Affect Disord. 2003;73:155–161.
  15. Gonzalez-Pinto A, Tohen M, Lalaguna B, et al Treatment of bipolar I rapid cycling patients during dysphoric mania with olanzapine. J Clin Psychopharmacol. 2002;22:450–454.
  16. Vieta E, Calabrese JR, Hennen J, et al Comparison of rapid-cycling and non-rapid-cycling bipolar I manic patients during treatment with olanzapine: analysis of pooled data. J Clin Psychiatry. 2004; 65:1420–1428.
  17. Gianfrancesco F, Rajagopalan K, Goldberg JF, et al Hospitalization risks in the treatment of bipolar disorder: comparison of antipsychotic medications. Bipolar Disord. 2007;9:252–261.
  18. Vieta E, Calabrese JR, Goikolea JM, et al Quetiapine monotherapy in the treatment of patients with bipolar I or II depression and a rapid-cycling disease course: a randomized double-blind, placebo-controlled study. Bipolar Disord. 2007;9:413–425.
  19. Goldberg JF, Kelley ME, Rosenquist KJ, et al Effectiveness of quetiapine in rapid cycling bipolar disorder: a preliminary study. J Affect Disord. 2008;105:305–310.
  20. Langosch J, Drieling T, Biedermann NC, et al Efficacy of quetiapine monotherapy in rapid-cycling bipolar disorder in comparison with sodium valproate. J Clin Psychopharmacol. 2008;28:555–560.
  21. Calabrese JR, Keck PE Jr, Macfadden W, et al A randomized, double-blind, placebo-controlled trial of quetiapine in the treatment of bipolar I or II depression. Am J Psychiatry. 2005;162:1351–1360.

CORRESPONDENCE: Melanie L. Zupancic, MD Department of Internal Medicine Division of Medicine/Psychiatry Southern Illinois University School of Medicine 901 W. Jefferson Street Springfield, IL 62794 USA E-MAIL: