While studies in the past have focused more on treatment of the manic phase of bipolar disorder (BD), recent findings demonstrate the depressive phase to be at least as debilitating. However, in contrast to unipolar depression, depression in bipolar patients exhibits a varying response to antidepressants, raising questions regarding their efficacy and tolerability.
Bipolar disorder (BD) is a devastating illness, carrying an immense burden of both morbidity [
Since their conception, antidepressants have been the mainstay of treatment for depression of any kind; to this day, antidepressants are prescribed to patients suffering from bipolar depression in 50% of cases [
We conducted a MEDLINE and Cochrane Collaboration Library search for any English-language articles published from January 1st 2005 to August 31st 2011, containing the keywords “treatment,” “bipolar,” and “depression” in the title or abstract. The search included clinical trials, meta-analyses, review articles, practice guidelines, conference summaries, editorials, and comments. We further used the “related citation” property for each item to search for other pertinent articles.
Initial screening yielded 1,062 results. Abstracts were then reviewed to exclude articles of low relevance, such as those regarding treatments other than antidepressants or concerning patients whose age is above 65 or below 18 years. A total of 68 items were finally included in the review.
In light of evidence of an increased risk of manic induction resulting from unopposed antidepressant treatment in bipolar I disorder [
In the 2010 EMBOLDEN II study, a total of 740 depressed bipolar patients (478 bipolar I, 262 bipolar II) were treated by monotherapy with either paroxetine (20 mg/d), quetiapine (300 mg/d or 600 mg/d), or placebo. An eight-week followup revealed no statistically significant change in MADRS total score for paroxetine compared with placebo, both in bipolar I and bipolar II patients, in contrast with a marked response observed in the quetiapine arm. Manic/hypomanic switch rates, defined as Young Mania Rating Scale (YMRS) ≥16, did not statistically differ between paroxetine and placebo (10.7% and 8.9%, resp.) [
Concerns about the risk of manic induction have prompted studies of antidepressant monotherapy in recent years to focus on bipolar II patients, where the risk has been estimated to be lower. One large study examining the response to antidepressant monotherapy in BP II patients included a 14-week open-label trial of 148 patients, treated by 10–80 mg fluoxetine daily. Response rate was demonstrated to be 59.5% (95% CI, 51.1%–67.4%,
In a small, randomized, placebo-controlled proof of concept study (
In a randomized open-label clinical trial including 83 BPII patients, 43 were randomized to treatment with venlafaxine and 40 to lithium monotherapy. Following a 12-week observation period, venlafaxine surpassed lithium both in response rates (58.1% versus 20.0%;
In a 2007 randomized controlled trial, 70 BP II patients were treated with the tricyclic antidepressant imipramine (average dose 250 mg/d) or the monoamine oxidase inhibitor phenelzine (average dose 60 mg/d), showing a response rate of 57% and 52%, respectively, compared with 23% in the placebo arm. Data regarding statistical significance was lacking. Although there was no evidence of manic induction [
In a meta-analysis of 12 trials encompassing a total of 1,088 patients, published in 2004 by Gijsman et al. [
The authors concluded that SSRIs may be an effective treatment for acute bipolar depression, with a low risk of manic switch early in the course of treatment.
Although the recommendation to use antidepressants as adjuncts to mood stabilizers in the treatment of acute bipolar depression did not conflict with common practices at the time, as reflected in the 2003 British Association for Psychopharmacology guideline [
Over the course of the next few years, numerous trials and meta-analyses sought to clarify the question of the role of antidepressants as adjuncts to mood stabilizers in the acute treatment of bipolar depression. One small double-blind randomized trial (
Perhaps one of the most quoted studies on the topic is the 2007 trial conducted by the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) collaborators, published in the New England Journal of Medicine by Sachs et al. [
In a recent meta-analysis by Sidor and MacQueen [
In addition to the general question regarding the efficacy and tolerability of antidepressants in the treatment of bipolar depression, several studies in recent years have attempted to compare the individual properties of various antidepressants. Among them is the 2006 publication by Post et al. [
Another recent comparison of antidepressant classes as adjuncts to mood stabilizers is the 2010 randomized clinical trial published by Pilhatsch et al. [
A more intricate look into the potential use of antidepressants as adjuncts to treatment is a comparative trial employing lithium, lamotrigine, and paroxetine in two different treatment algorithms [
Since the introduction of Symbyax, an olanzapine/fluoxetine combination (OFC), it has gained widespread use for the treatment of bipolar depression, becoming the first treatment to be approved by the U.S. Federal Drug Administration for this indication in 2003 [
Another study worth mentioning in this context is a Lilly-sponsored open-label continuation trial of 114 bipolar patients in Puerto Rico. The first phase of the trial included a 7-week treatment course with OFC, demonstrating a response rate of 69% and a remission rate of 59%, in accordance with earlier findings. Responders were then randomized to either OFC continuation or olanzapine alone for 12 weeks, showing maintenance of response to be significantly higher for the OFC group than the olanzapine group (31.3% versus 12.5%). Metabolic adverse effects were highly prevalent, with 33% of the OFC-treated patients gaining over 7% of their body weight over the 4-month course [
It is astounding how, despite numerous trials and meta-analyses conducted on the subject in recent years, the role of antidepressants in the treatment of bipolar depression still remains unclear. Since the 2004 meta-analysis by Gijsman et al. [
A marked disparity exists between the results of studies examining the efficacy of antidepressants in acute bipolar depression, whether as monotherapy or as an adjunct to mood stabilizers (Table
Summary of recent studies examining the efficacy of antidepressants in the treatment of acute bipolar depression.
Positive studies | Negative studies | ||||||
Study | No. of participants | Drug | Comments | Study | No. of participants | Drug | Comments |
Amsterdam and Shults 2005 [ | 34 ( | Fluoxetine, olanzapine/fluoxetine combination | No placebo control | McElroy et al. 2010 (EMBOLDEN II) [ | 740 ( | Paroxetine | |
Parker et al. 2006 [ | 10 BPII | Escitalopram | Small sample size | ||||
Agosti and Stewart 2007 [ | 70 BP II | Imipramine, phenelzine | No significance demonstrated | ||||
Amsterdam and Shults 2008 [ | 83 BPII | Venlafaxine | No placebo control | ||||
Amsterdam and Shults 2010 [ | 148 BPII | Fluoxetine | No placebo control | ||||
Schaffer et al. 2006 [ | 20 | Citalopram | Small sample size, no placebo control | Sachs et al. 2007 [ | 366 ( | Paroxetine, bupropion | |
Fonseca et al. 2006 [ | 20 | Escitalopram | Small sample size, no placebo control | Sidor et al. 2011 [ | 1,034 (Meta- Analysis) | Fluoxetine, paroxetine, bupropion, imipramine | |
Tamayo et al. 2009 [ | 114 | Olanzapine/fluoxetine combination | Conducted by Lilly Research Laboratories | ||||
Pilhatsch et al. 2010 [ | 40 | Paroxetine, amitriptyline | No placebo control | ||||
Perlis et al. 2010 [ | 410 | Olanzapine/fluoxetine combination | Conducted by Lilly Research Laboratories |
Finally, the heterogeneous inclusion of patients treated with various “mood stabilizers,” some of them possessing antidepressant activity, such as quetiapine [
The majority of recent studies do not demonstrate a significant risk of manic/hypomanic switch as a result of antidepressant treatment in acute bipolar depression, both as monotherapy and in conjunction with a mood stabilizer (Table
Summary of recent studies examining the risk of manic/hypomanic switch following the use of antidepressants in the treatment of acute bipolar depression.
Increased risk of switch | No increased risk of switch | ||||||
Study | No. of participants | Drug | Comments | Study | No. of participants | Drug | Comments |
Amsterdam and Shults 2005 [ | 34 ( | Fluoxetine, olanzapine/fluoxetine combination | No Placebo Control | ||||
Parker et al. 2006 [ | 10 BPII | Escitalopram | Small sample size | ||||
Agosti and Stewart 2007 [ | 70 BP II | Imipramine, phenelzine | No valid tool used to assess switch | ||||
Amsterdam and Shults 2008 [ | 83 BPII | Venlafaxine | Compared to Lithium, no placebo control | ||||
McElroy et al. 2010 (EMBOLDEN II) [ | 740 ( | paroxetine | |||||
Amsterdam and Shults 2010 [ | 148 BPII | Fluoxetine | No placebo control, subsyndromal hypomania in 19.6% | ||||
Post et al. 2006 [ | 174 | Venlafaxine | 15%; no placebo control | Schaffer et al. 2006 [ | 20 | Citalopram | Small sample size, no placebo control |
Truman et al. 2007 [ | 366 ( | Paroxetine, bupropion | By self-report only | Fonseca et al. 2006 [ | 20 | Escitalopram | Small sample size, no placebo control |
Amsterdam and Shults 2010 [ | 148 BPII | Fluoxetine | No placebo control, subsyndromal hypomania in 19.6% | Sachs et al. 2007 [ | 366 ( | Paroxetine, bupropion | |
Perlis et al. 2010 [ | 410 | Olanzapine/fluoxetine combination, lamotrigine | Conducted by Lilly Research Laboratories | ||||
Sidor et al. 2011 [ | 1,034 (Meta- Analysis) | Fluoxetine, paroxetine, bupropion, imipramine |
If antidepressants are indeed redundant in the treatment of bipolar depression, it should have a direct impact on several aspects of the care of depressed bipolar patients, as well as depressed patients in general. First is the heightened importance of reaching the correct initial diagnosis of bipolar depression, often misdiagnosed as unipolar depression upon first presentation [
Aside from the importance of making a correct diagnosis, several other factors have been suggested as relevant in treating patients suffering from bipolar depression. One is the identification of clinical factors potentially associated with antidepressant resistance, including the severity of the current episode, presence of melancholic features, current suicidal risk, and psychiatric comorbidity, including social phobia [
Following initiation of treatment, assessment of response poses another challenge. In one large trial, early improvement of depressive symptomatology did not appear to be a reliable predictor of eventual response or remission due to an unacceptably high false-positive rate. However, the absence of early improvement appeared to be a highly reliable predictor of eventual nonresponse, demonstrating the need for close monitoring of patient status during the initial phase of treatment [
Another issue in the treatment of acute bipolar depression is the duration of treatment. In a trial conducted as part of the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) study, patients responding to treatment with antidepressants plus mood stabilizers, and euthymic for 2 months, were randomly assigned to antidepressant continuation versus discontinuation. After a follow-up period of 1–3 years, antidepressant continuation showed a mildly delayed depressive episode relapse (HR = 2.13 [1.00–4.56]) and trended toward less severe depressive symptoms (mean difference = −1.84 [95% CI, −0.08 to 3.77]), without increased manic symptoms [
In conclusion, it is worth noting that bipolar disorder is a complex condition, requiring a multimodal approach. The tools used in the treatment of bipolar disorder include various pharmacological treatments, including antidepressants, mood stabilizers, and antipsychotics, nonpharmacological treatment modalities, such as electroconvulsive therapy and transcranial magnetic stimulation, as well as psychotherapeutic approaches. Proper integration of all available modalities is necessary for optimal treatment response.
Probably in light of the inconclusive nature of the evidence, a review of guidelines published in recent years has not revealed major changes in the recommendations regarding antidepressant treatment. While the American Psychiatric Association (APA) guideline for the treatment of bipolar disorder was not updated since 2005 [
In a 2010 update of the World Federation of Societies of Biological Psychiatry (WFSBP) guideline on the treatment of acute bipolar depression, only the olanzapine/fluoxetine combination has been approved as a first-line therapy for this indication. There were no additional recommendations regarding antidepressant therapy, aside for mentioning possible efficacy in bipolar II patients [
In summary, while the major trend in recent years has been the adoption of quetiapine monotherapy as a first-line agent, most guidelines still advocate the use of antidepressants as potential first-line agents in the acute treatment of bipolar depression, in adjunction to mood stabilizers. Specific references were made for the SSRIs paroxetine [
Studies conducted in recent years have failed to demonstrate significant beneficial effects of antidepressants in the treatment of acute bipolar depression. The rate of manic/hypomanic switch is probably low, especially with concurrent use of mood stabilizers. However, the considerable disparity between studies should prompt further large-scale, long-term, double-blind, randomized clinical trials, including comparison between various classes of antidepressants.
The authors would like to thank the staff of the European College of Neuropsychopharmacology (ECNP) 2011 School of Neuropsychopharmacology, in Oxford, UK, where many of the topics presented herein were thoroughly discussed.