Bipolar Disorder Forum Report
The Institute of Psychiatric Services
Philadelphia, Pennsylvania

The Latest Evidence-Based Medicine for the Management of Bipolar Disorder

This report was reviewed for medical and scientific accuracy by David M. Davis, MD , Psychiatry-Addictionology, Private Practice, Atlanta, Georgia.


New Treatment Approaches to the Management of Bipolar Disorder and Chronic Depression

K. N. Roy Chengappa, MD, Associate Professor of Psychiatry, University of Pittsburgh, Western Psychiatric Institute and Clinic, Pennsylvania

Bipolar disorder (BD) is a challenging illness to deal with-both for the sufferer and for those who care for him or her. Inaccurate diagnosis, inappropriate treatments or the lack of response to available agents, very high rates of alcohol and substance dependence, and a near 20 percent rate of completed suicide underscore the tremendous personal, financial, and societal costs associated with BD. Therefore, not unexpectedly, BD is noted to be among the most disabling illnesses worldwide (Murray & Lopez, 1994). Yet, subjects with BD have also been among the most creative and productive citizens over the centuries; for example, Otto Klemperer, Vivien Leigh, Del Shannon, and Virginia Woolf were all suspected to suffer from manic depression.

After nearly three decades of neglect-notably from a research perspective-there has been a resurgence of interest in BD. The key factors that might be spurring this new interest include the availability of alternative treatments to lithium1 (Eskalith®1, Lithobid®1)-the mainstay of mood-stabilizer BD treatments for decades-a greater understanding of the intracellular actions of lithium and that of the antiepileptic drugs, the use of the second generation antipsychotic drugs, and the use of combination and sequential pharmacological treatments. Bipolar depression, a rapid-cycling course, and long-term treatment challenges remain persistent problems in the management of BD, not to mention the extensive prevalence of alcohol and substance abuse among BD sufferers (Chengappa et al., 2000).

The recently concluded 52nd Institute of Psychiatric Services of the American Psychiatric Association focused on some, though not all, of the therapeutic challenges associated with BD. Data from double-blind, placebo-controlled, random assignment studies of the use of the newer antipsychotic agents, which have already been approved for use in schizophrenia, are now available for bipolar illness. Two monotherapy trials of olanzapine1 (Zyprexa®1) versus placebo led to its approval by the USFDA for bipolar type I mania. An adjunctive trial of risperidone1 (Risperdal®1) added to either lithium1 or divalproex sodium1 (Depakote®1) confirmed superiority over placebo for manic episodes that break through with ongoing mood-stabilizer treatment. These and other open data sets, including the impressive Spanish study on risperidone1, confirm the benefits of risperidone1, olanzapine1, and quetiapine1 (Seroquel®1) for hypomanic, mixed, and manic episodes.

Nearly 50 percent of bipolar subjects have psychotic symptoms, in manic or mixed episodes, along with motoric over-activity and severe agitation. It is in these emergent situations that first-generation agents such as haloperidol1 (Haldol®1) or chlorpromazine1 (Thorazine®1) are used successfully but often at high cost. The greater propensity to extrapyramidal symptoms and tardive dyskinesia and depressogenic effects make these first-generation antipsychotic agents particularly problematic in BD patients. Risperidone1 and olanzapine1 have not only proved beneficial in the management of psychotic mania but have proved equally effective in the management of non-psychotic mania. Furthermore, BD subjects with mixed presentations have improved depression scores. Unlike the first-generation agents, these second-generation antipsychotic agents do not appear to have depressogenic effects, and they do have low rates of extrapyramidal side effects and tardive dyskinesia. Their use in bipolar depression and in maintenance treatment, however, remains unclear at this time.

The resurgence of research in and treatment interest for bipolar disorder are reassuring and, hopefully, signs that will likely add to the available treatment options for subjects with BD.


1. Murray CL, Lopez AD. Quantifying disability: data, methods and results. Bull World Health Org. 1994; 72:481-494.

2. Chengappa KNR, Levine J, Gershon S, Kupfer DJ. Lifetime prevalence of substance or alcohol abuse and dependence among subjects with bipolar I and II disorders in a voluntary registry. Bipolar Disorders. 2000; 2:191-195.


Antoni Benabarre, MD, Bipolar Disorders Program, IDIBAPS, Clinical Institute of Psychiatry and Psychology, Hospital Clinic, University of Barcelona, Spain

While mood-stabilizing agents like lithium and antiepileptic agents such as divalproex sodium1 (Depakote®1) and carbamazepine1 (Tegretol®1) are commonly used as pharmacotherapy for bipolar disorder (BD), emergent data continue to garner interest in their off-label use. Atypical antipsychotic agents such as risperidone1 (Risperdal®1), and third-generation antiepileptic drugs (AEDs) such as topiramate1 (Topamax®1), are promising new mood-stabilizing agents in the management of this condition, particularly among refractory BD patients.

Compelling new research findings have added currency to treatment approaches that combine psychotherapy with pharmacotherapy for the optimal management of chronic depression.

Two studies presented by Dr. Antoni Benabarre shed more light on the use of risperidone as adjunctive therapy to mood stabilizers in the treatment of acute BD and schizoaffective patients and on topiramate as an add-on therapy for treatment-resistant BD.

Commenting on results from "Treatment of Bipolar Disorder with Adjunctive Risperidone," a large, open-label study involving 598 patients with either schizoaffective disorder, BD type I or II, Dr. Benabarre said the investigation grew out of previous smaller open-label studies and two recent placebo-controlled, double-blind trials that suggest this atypical antipsychotic may be beneficial and well-tolerated in the manic phase of BD.

He added that risperidone has also shown effectiveness in rapid-cycling BD patients, including those suffering breakthrough episodes-two patient populations that are considered exceedingly difficult to treat. Moreover, the agent has demonstrated effectiveness against affective symptoms in schizophrenia patients, he said, while being relatively free of side effects: notably, extrapyramidal symptoms (EPS), tardive dyskinesia (TD)-often seen with conventional neuroleptics-sedation, and ataxia, which can be induced with some mood stabilizers. However, a need exists for controlled trials to determine the clinical efficacy of these treatment options.

Encouraged by those results, Dr. Benabarre added, the current study scaled up to a much larger patient population (541 were eligible for evaluation at trial's end) in an effort to further substantiate risperidone's1 effectiveness and safety as an add-on therapy in treating affective and psychotic symptoms.

Six months in duration, with evaluations at baseline, weeks one, two, four, and six, and at months three and six, the trial recruited patients 18-65 years of age with signs or symptoms of acute mania, hypomania, or mixed symptoms, with a Young Mania Rating Scale (YMRS) score of > 7. Primary treatment involved lithium1 (Eskalith®1, Lithobid®1), divalproex sodium1 (Depakote®1), or carbamazepine1 (Tegretol®1), with add-on risperidone1 doses ranging from = 3 mg/d to more than 9 mg/d in a small minority of patients, with a mean dose of 3.88 mg/d. Evaluation at six months used various standard treatment efficacy measures, Dr. Benabarre said.

Results showed highly significant improvements (p < 0.0001) in mean YMRS scores (baseline, 26.6 versus 2.4 at month six); the Hamilton Rating Scale for Depression (HAM-D) (baseline, 12.8 versus 4.1 at month six), as well as total scores (baseline, 71.9 versus 40 at month six). Equally significant were Positive and Negative Syndrome Scale (PANSS) sub-scale scores, he said: positive (baseline, 20.8; month six, 8.6), negative (baseline, 13.5; month six, 9.8), and general psychopathology (baseline, 37.8; month six, 21.6), as were the Clinical Global Impression (CGI) scores (2.6 at baseline; 0.9 at month six).

Overall, 76 percent of patients were considered responders according to YMRS and CGI scores by the trial's end, he pointed out, with 40 percent of participants completely asymptomatic at six months, compared with none at the start. Risperidone1 was well-tolerated, he added. Of the 111 (20.5 percent) patients who dropped out of the study, only 3.0 percent left due to side effects, compared to 4.1 percent who were lost to follow-up, 2.6 percent due to lack of response, and 3.0 percent due to hospitalization.

"Based on these findings," Dr. Benabarre said, "we concluded risperidone1 was indeed a safe and effective adjunctive therapy to mood-stabilizing medications in the treatment of manic, hypomanic, or mixed episodes in patients with bipolar and schizoaffective disorders, and showed efficacy against both affective and psychotic symptoms. We believe these results should be further investigated in randomized, placebo-controlled trials. Risperidone1 appears to have mood-stabilizing properties that are useful in conjunction with other mood-stabilizing agents in the treatment of patients with bipolar and schizoaffective disorders."

Dr. Benabarre also reported on preliminary data from the "Use of Topiramate in Treatment-Refractory Bipolar Disorder," one of eight recent open-label studies (n = 224) of this atypical AED, which have collectively suggested the agent possesses mood-stabilizing properties with improvement in mania and depression seen in more than 50 percent of patients who received the drug.

"Our investigation looked at the use of topiramate as a rescue therapy for BD patients who had not responded well to carbamazepine1 and divalproex sodium1-traditional AEDs that are currently widely used as alternatives or adjuncts to lithium1," Dr. Benabarre pointed out.

Topiramate1 was chosen because it combines mechanisms found in both carbamazepine1 and divalproex sodium1, he added. Like both compounds, topiramate1 inhibits voltage-activated NA+ channels, he explained. Further, like divalproex sodium1, topiramate1 enhances GABA neuroinhibition and also modulates glutamate-mediated neuroexcitation, some voltage-activated Ca++ channels, and acts as a weak inhibitor of some carbonic anhydrase isoenzymes.

In design, the trial was a six-week, prospective, multicenter, open-label study involving 22 patients ages 18-65 years with a DSM-IV diagnosis of either BD type I or II, or schizoaffective disorder (bipolar type). All patients had demonstrated poor response to treatment with lithium1, divalproex sodium1, or carbamazepine1 and had YMRS and HAM-D scores of = 12, with a CGI of = 4.

Treatment with topiramate1 commenced at 25 mg/d and was titrated upwards by 25-50 mg/d every three to seven days to a mean dose of 158 mg/d. Other psychotropic medications remained constant during the six-week study phase and patients were evaluated weekly for two weeks and biweekly for up to six weeks using YMRS, HAM-D, CGI scores, and incidence of side effects.

Of 21 evaluable patients, six (29 percent) discontinued treatment due to: lack of efficacy and side effects (1), lack of efficacy (1), poor compliance (1), and loss to follow-up (3). Of the 15 remaining patients, six were rated as responders based on a > 50 percent reduction in either YMRS or HAM-D scores and a = 2 decrease on the CGI.

Overall, said Dr. Benabarre, topiramate was well-tolerated, with all adverse effects related to the central nervous system or gastrointestinal tract. The one patient who discontinued due to side effects did so due to paresthesia, anxiety, and impaired concentration.

"The most important outcomes we saw in this study were the (> 50 percent) reduction in the YMRS or HAM-D scores and the decrease on the CGI (= 2 points)," Dr. Benabarre said. "Overall, the number of seriously ill patients was reduced from 60 percent to about 10 percent, a significant improvement that led us to conclude topiramate1 is both safe and effective in the acute treatment of BD patients with manic, hypomanic, or depressive symptoms. We believe these results suggest this agent may be a useful new therapy for bipolar disorder, with promising results in the most treatment-refractory patients.

"One interesting aspect of this drug is a side effect that results in weight reduction," he added. "This has raised conjecture about the possible use of topiramate in combination with olanzapine1 (Zyprexa®1), which is associated with weight gain; the hypothesis being that combining the two drugs might eliminate that unwanted side effect." However, this has the disadvantage of adding a second drug to what already may be a complicated treatment regimen.

Dr. Benabarre pointed out the current investigation was a small trial designed to establish efficacy and tolerability for topiramate1 in anticipation of a large multicenter, placebo-controlled, Spanish trial.


Charles Bowden, MD, Karren Professor, Chairman, Department of Psychiatry, University of Texas Health Science Center, San Antonio

Taking a macro approach to effective treatment for BD during a luncheon symposium sponsored by the APA and supported through an unrestricted educational grant from Glaxo Wellcome, Dr. Charles Bowden said it is unlikely any one drug or treatment will be available to relieve the symptoms of this debilitating disease anytime soon.

"One treatment simply does not fit all in BD patients," he said, "because it is not a homogeneous entity, but requires a variety of treatment approaches to get the best out of the expanding treatment array now available."

Discussing factors that impact on the course of BD and the potential risks and benefits of agents currently used to treat the disorder, including the newer antipsychotics and AEDs, Dr. Bowden said gathering evidence increasingly suggests the as yet poorly addressed primary cause of treatment refractory illness is not so much the manic side, but the depressive side of this illness.

Drawing on data from clinical trials supported by various pharmaceutical companies, the National Institute of Mental Health (NIMH), the ongoing Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD), and the Stanley Foundation, Dr. Bowden said it has become clear that certain principles should govern treatment of the disease, namely, that clinicians need to:

• treat the illness, not just the episodes;

• help patients learn about destabilizing factors associated with BD;

• be empathetic, but blunt, about illness denial;

• use regimens that yield excellent tolerability and adherence, and

• recognize that acute episode drugs are often different from those used for maintenance therapy.

Focusing on the illness and not on just the episodes is important, he explained, because many patients will have symptoms between episodes. This was clearly demonstrated by a recent trial (Gitlin MJ et al., Am J Psychiatry, 1995) that demonstrated less than 20 percent of BD patients treated prophylactically with lithium had an unequivocal response to treatment (no recurrence and only mild symptoms) over 18 months of follow-up. Further, including patients with no recurrence and sub-threshold or moderate symptoms only increased the responder percentage by another 15 percent-to about 32 percent of the study population-leaving 66 percent of patients to go on to full episode relapses, he added. In addition, the findings suggested sub-threshold symptomatology was most associated with occupational, social, and family function impairment. "This indicates we need to focus more on the presence of sub-threshold symptoms if we want to optimize treatment of BD," he maintained.

Besides a need for earlier and more accurate diagnosis of BD by clinicians (a 1994 NDMDA Survey by Lish JD et al., J Affect Disord, for instance, found 73 percent or 363 of 500 respondents received alternative explanations for their symptoms, with 34 percent diagnosed with unipolar depression), sub-optimal patient adherence continues to confound successful management, Dr. Bowden pointed out. This was aptly demonstrated in a study investigating time on lithium1 following the initial prescription (Johnson JG, et al., Am J Psychiatry, 1995), he said, in which investigators found that despite the best efforts of staff to keep patients on treatment, patients stayed with the therapy an average of two months.

Similarly, a 1999 study (Hirschfeld et al.) that compared patients receiving initial therapy with either divalproex sodium1 (Depakote®1) or lithium for one year found modest advantages for divalproex sodium1, but most patients nonetheless stayed on therapy less than three months, dramatically driving up overall costs of care due to new episodes involving return to the emergency room and rehospitalization, he added.

Those findings were further corroborated by a recently published, randomized, blinded comparison study by Dr. Bowden and colleagues (Bowden CL et al., Arch Gen Psychiatry, 2000; 57:481-489), in which the investigators looked at time to relapse among patients receiving maintenance therapy with divalproex sodium1 (n = 187), placebo (n = 94), or lithium1 (n = 91) over one year. Results showed diavalproex sodium1 prevented depressive or manic episodes in 50 percent of patients for 275 days compared to 189 days among those receiving lithium1 and 173 days among placebo (PBO) patients. Similar results were seen in the time to a 25 percent relapse to mania (> 365 days divalproex sodium1, 293 days lithium1, 189 days PBO) but not in time to a 25 percent relapse to depression, where divalproex sodium1 only slightly out performed PBO (126 days versus 101), while the latter actually outperformed lithium1 (101 versus 89 days).

"In the time to 25 percent relapse to depression," Dr. Bowden said, "even divalproex sodium1 might be a hard sell to patients given the small advantage it offers over no treatment at all (126 versus 101), even though it was clearly better in terms of the manic side of the illness."

This issue of what endpoint clinicians should focus on has become increasingly important over the last several years, he added. "The FDA wants the drug companies to identify a single endpoint on which the success or failure of an intervention can be established, but this illness does not lend itself to that," he explained. "We need to pay attention not just to full manic or full depressive episodes, but to the important changes that result from this illness."

As a consequence, researchers are increasingly paying attention to a 'basket' of relevant endpoints, he added, particularly the depressive side of the illness, rather than just one measure in the search for effective prophylactic strategies.

Towards that end, he said, a recent study (Swann et al., Acta Psychiatr Scand, 2000) has shown no difference between divalproex sodium1, lithium1, and PBO in improving the number of depressive episodes in acute mania, while other studies (Dunner and Fieve, Arch Gen Psychiatry, 1976;33:117-20) suggest lithium1 has more benefit in treating the manic side of BD as compared to the depressive side in both patients with rapid-cycling (RC) and non-rapid cycling (NRC) illness. "In the latter study, results showed that compared to PBO, the number of manic episodes decreased with lithium1 in both arms," Dr. Bowden explained, "but the results also showed only two depressive episodes per year occurred among RC patients receiving PBO versus 4.5 with lithium1, with a more modest result that still favored PBO in the NRC group (0.49 versus 0.37, respectively). In short, several lines of evidence suggest using lithium1 as the definitive BD treatment can worsen the depressive side in both RC and NRC patients," he added.

Turning to optimizing BD treatment strategies, Dr. Bowden said large scale studies already underway or being contemplated are proceeding on the principle that the best clinical results for controlling the depressive symptoms of BD will come from combination therapies.

"Rather than an 'either/or' approach to the available agents, we need to think about using combinations of mood stabilizers or drugs with properties that stabilize mood, including antipsychotics, antiepileptics, and antidepressants," he added. This was demonstrated in a one year cross-over study (Denicoff KD et al., J Clin Psychiatry, 1997) in which investigators compared the response of BD patients randomized to lithium1, carbamazepine1, or combination therapy, he said. Results showed carbamazepine1 alone was least effective, while adding divalproex sodium1 to non-responders produced a 39 percent response. Divalproex sodium1 and lithium1 or divalproex sodium1, lithium1, and carbamazepine1 combinations were the most effective, respectively demonstrating a 59.9 percent and 61.9 percent response.

Similarly, a recently completed study that compared the efficacy of risperidone1 as an add-on treatment in acutely manic patients who were receiving either lithium1 or divalproex sodium1 showed that the combination therapy significantly outperformed the mood stabilizer alone based on YMRS scores. This reprised a similar study where olanzapine1 was added to a mood stabilizer and attained similar results, he pointed out. "It supports in evidence-based fashion what is often done in practice," Dr. Bowden said, "using a mood stabilizer with an atypical antipsychotic."

"In summary, individualizing strategies for BD should begin with maximizing or optimizing our use of standard mood stabilizers, including combinations of these agents. We should also consider using anxiolytics/hypnotics, atypical neuroleptics, and novel anticonvulsants as adjunctive therapy. And if we are to use standard antidepressants, we need to view this intervention as a brief, acute, and intermittent treatment for depression, paying equal attention to the psycho-educational needs of our patients."


Stephen Strakowski, MD, Professor of Psychiatry, Psychology and Neuroscience, University of Cincinnati College of Medicine, Ohio

Expanding on Dr. Bowden's overview, Dr. Strakowski reviewed major risks and benefits inherent in the expanding array of BD therapies, and offered management algorithms that might be considered in a clinical setting.

"In managing BD," he said, "clinicians need to think about treating acute episodes of this disorder as a recurrent event in a chronic condition that can extend over a life time. Data show a typical bipolar patient who receives no treatment following first onset of an affective episode at age 16 can be expected to have up to ten recurrent episodes in the next 14 years of life, which will preclude the person from developing the connections, jobs, and universal relationships that are needed to establish a successful early adult life."

Consequently, he added, it is important to maintain at least some contact with the patient and treat acute episodes as they occur with a goal of minimizing the affective recurrences early. Psycho-education is a crucial part of this ongoing therapy, Dr. Strakowski pointed out, since patients need to understand that recurrences are part of the illness. Clinically, the goal is to stabilize mood to pre-morbid functioning by controlling hypomania without inducing depression, and by treating the depression when it does occur without provoking mania or cycling, he said.

While complete recurrence prevention is not possible at present, he added, the incidence can be minimized by combining safe, easy-to-take treatments with psychotherapy. "Pharmacotherapy is critical in managing initial symptom complexes and syndromes, but over the long haul, recovery of function tends to lag if we simply give medications," he explained. "Most patients will require some form of rehabilitation therapy, including psychotherapy of all kinds, educational therapy, and occupational therapy. In effect, all the interventions currently used in schizophrenia should be considered for patients with BD," he suggested. "Otherwise, we have patients with fewer symptoms, but no life; they sit at home and are unable to do anything."

That said, Dr. Strakowski pointed out that since the 1970s, only one drug-lithium1-has been approved by the FDA for maintenance therapy in the treatment of BD. Still, even though FDA approval has not been achieved, data do indicate a number of agents can be useful in treating BD, he added, notably divalproex sodium1 and olanzapine1, which already have an FDA indication for the treatment of acute mania, as well as carbamazepine1 and clozapine1. Other agents have shown potential, he added, but need more data to demonstrate their value in maintenance therapy. These include atypical antipsychotics [risperidone1, quetiapine1 (Seroquel®1), ziprasidone1 (Zeldox®1)] and new antiepileptic drugs [topiramate1 and lamotrigine1 (Lamictal®1)]. Risperidone1 does not have any maintenance data yet, he added, but has shown evidence of efficacy in acute mania.

"Even though data accumulated since the 1940s show lithium1 is typically twice as effective as placebo in BD maintenance therapy, he added, up to 50 percent of patients will still have a relapse. This has lead clinicians to view the drug as an inadequate maintenance therapy that requires going to other compounds and combinations."

The question is: "Which ones?" Very good data suggest conventional antipsychotics are effective in treating acute mania, Dr. Strakowski pointed out, but no studies suggest these agents are effective for maintenance. Unfortunately, he added, many BD patients stay on these drugs for extended periods of time, typically not alone, but as add-ons given during acute episodes that get extended into the maintenance phase.

Effectively treating the depressive side of the illness is also problematic, he said. "Again, data is scarce," he added, "but at least one open label European study (Ahlfors et al., Acta Psychiat Scand, 1981; 64:226-237) shows conventional antipsychotics are not effective as maintenance therapy in treating the depression phase of BD. Among 93 lithium1-resistant or intolerant patients, the investigators found depression almost doubled over time in the group receiving flupenthixol decanoate1 (Depixol®1, Fluanxol®1). This indicates failure of the conventional antipsychotic to prevent depression, not to cause it."

Given the dearth of data in this regard, Dr. Strakowski suggested clinicians should examine their practices and look for BD patients who might be taken off conventional antipsychotics.

Principle evidence for the use of the new atypical antipsychotics (AAPs) comes out of the olanzapine1 BD trials, he pointed out. "Support for this agent in acute mania treatment is very strong," he added, "but we have no data for maintenance therapy with this drug. The concern, as with conventional antipsychotics, is that AAPs will either cause or fail to treat depression."

Still, an olanzapine1 trial by Tohen et al. (European Neuropsychopharmacol, 1999;9[suppl 5]:S247) showed the risk of relapse within the 49-week trial was very low (16 percent), which may well indicate olanzapine1 is prophylactic, he added. "The trial was open-label and did not include comparators," Dr. Strakowski pointed out, "but it suggests this agent and other AAPs have enough difference in their pharmacologic properties that they may have a role in maintenance therapy for BD."

In contrast to the dearth of maintenance data on depression in BD, very good data exist to guide clinicians in the instance of acute mania, he added. For instance, a summary study by Zornberg and Pope (J Clin Psychopharmacol, 1993) reviewed the world literature on the treatment of BD depression and confirmed that lithium outperformed PBO in eight out of nine controlled studies 36 percent of the time. "But again," added Dr. Strakowski, "Sixty percent of the patients had depression for which we have no clear indication of what to do next."

One approach that may not be indicated, he added, is the use of antidepressants during the depressive phase of BD. The concern is that they may perpetrate a switch to mania. "This is not yet clear," Dr. Strakowski said, "but depression should be seen as a high risk time for developing mania." Two studies, for instance (Lewis JL and Winokur G, Arch Gen Psychiatry, 1982; Prien RF et al., Arch Gen Psychiatry, 1984) suggest patients are more likely to become manic on antidepressants. Compared to a natural switch rate to depression of 41 percent without these medications, switch rates on imipramine1 (Tofranil®1) alone were 53 percent-almost twice the rate seen on lithium1 and imipramine1 combined (28 percent) and on lithium1 alone (26 percent).

"Although lithium1 seemed to defeat the risk to some extent," Dr. Strakowski added, "it seems fair to say we should no longer give bipolar patients antidepressants in the absence of an agent that will stabilize mood-primarily mania."

In fact, studies suggest tricyclic antidepressants (TCAs) have increased the rate at which rapid cycling between mania and depression occur (see Angst J. Psychopathology, 1985), he said, while evidence from Europe (Prien RF et al., Arch Gen Psychiatry, 1973) indicate lithium1 may treat rapid-cycling in the absence of antidepressant exposure, but may fail in the agent's presence. "Potentially then, antidepressants may change the neurophysiology of this illness by worsening its course. Indeed, antidepressants in placebo-controlled, double blind studies (Wehr TA, Goodwin FK, Psychopharmacol Bull, 1987) have been shown to induce reversible rapid-cycling.

"The take-home message from this," Dr. Strakowski added, "is the time-limited nature of rapid-cycling in BD patients will be typically related to the agent causing the cycling-get that under control and the cycling will likely not recur throughout the entire life-course of the illness."

What then, are the options for treating bipolar depression? According to Dr. Strakowski, current evidence suggests the most reasonable initial intervention should involve mood stabilizers (lithium1, carbamazepine1, divalproex sodium1, and electroconvulsive therapy [ECT]) maximized to a dose adequate to resolve the depression. If that does not occur, clinicians should at least be sure the mood stabilizer will effectively prevent the risk of a manic switch before they add an antidepressant.

He noted that ECT can cause mania in depressed patients, but is also an effective anti-manic treatment in contrast to all the other antidepressants. "Often, even if the switch occurs," he pointed out, "you can continue with the treatment, and the patient will stabilize."

"Antidepressants offer many choices," he added, "but none have been shown to not induce rapid cycling or mania. However, large summaries of different kinds of treatment indicate TCAs carry the highest propensity for causing patients to either switch into mania or rapid cycling. Conversely, my suspicions are that the newer antidepressants carry a lower risk of inducing cycling and mania. Although the switching rates seem almost the same as that seen with TCAs, the switches appear to be less severe."

Further, the atypical antipsychotics all seem to possess antidepressant properties, he said. "We remain hopeful they will show the ability to relieve depression in BD patients, but this has yet to be demonstrated in controlled clinical trials. In like vein, Omega-3-fatty acids have shown promise in treating the depressive phase of BD, but this also needs to be pursued in clinical trials."

Although gabapentin1 has failed a recent double-blind efficacy study in patients with acute mania, he pointed out, most clinicians still consider the agent as effective as an adjunctive anxiolytic. "This may be [gabapentin's1] ultimate role in treating BD," he added, "since anxiety is often a part of depression."

Dr. Strakowski concluded by suggesting a possible treatment algorithm for depressive symptoms in rapid cycling/mixed BD states. In the event of a poor response to lithium1 monotherapy, consider divalproex sodium1 or carbamazepine1, possibly in combination with a newer AED (lamotrigine1 or topiramate1) and/or an atypical antipsychotic (olanzapine1, risperidone1, quetiapine1 or ziprasidone1). Avoid TCAs and stabilize patients before adding any antidepressant.

For minimizing mania risk in BD, he suggested a three-month course of mood stabilizers, in combination as necessary, may obviate the need for antidepressants. When the latter are required, duration should be brief (probably = 3 months) in classic mania and briefer yet (probably = 3 weeks) in rapid-cyclers. Antidepressants should be stopped if patients are hypomanic or mixed, and TCAs or any destabilizing antidepressant should be avoided.

"In summary, the backbone of treatment still requires one of the classic mood stabilizers," Dr. Strakowski said, "but a sizable proportion of patients will fail this initial intervention, so be prepared. Almost all patients with BD will require multiple medications and psychotherapy for successful treatment, but beware of antidepressants and consider adding them as your last intervention therapy. Conventional antipsychotics are not a good maintenance therapy and should be minimized. At present, systematic, rational, trial-and-error therapy remains the mainstay of maintenance therapy for BD."


Martin Keller, MD, Professor, Chairman, Department of Psychiatry and Human Behavior, Brown University Medical School, Providence, Rhode Island

At a luncheon symposium sponsored by the APA and supported by an unrestricted educational grant from Bristol-Myers Squibb, Dr. Martin Keller presented results from the ongoing nefazodone1 (Serzone®1) Chronic Depression Study (NCDS), the world's largest psychotherapy/pharmacotherapy combination-treatment trial to date for the management of chronic depression. In design, the NCDS is a 12-site, 681-patient investigation that was undertaken to resolve which of the following treatment modalities is most effective for the management of chronic depression: psychotherapy monotherapy, pharmacologic monotherapy, or combination therapy.

Patients randomly received 12 weeks of acute-phase treatment with either:

• antidepressant nefazodone1 alone;

• monotherapy with the Cognitive Behavioral Analysis System of Psychotherapy (CBASP), a new psychotherapy model developed specifically for chronic depression (See McCullough JP, Treatment for Chronic Depression. New York, Guilford Press, 2000); or

• a combination of nefazodone1 and CBASP.

Responders to the acute phase then received 16 weeks of continuation-phase treatment, while non-responders to either CBASP or nefazodone1 monotherapy were crossed over to 12 weeks of alternative treatments.

In the second (continuation) phase, responders entered a one-year maintenance phase in which nefazodone1 alone or combination-treatment patients were randomized to double-blind nefazodone1 or placebo, while CBASP patients were randomized to either maintenance CBASP or assessment only.

Nefazodone1 dosing schedules began at 100 mg BID in week one, 150 mg BID in week two, and were titrated during weeks three to 12 by 100 mg/day at weekly intervals to a maximum of 600 mg/day in two doses, based on response and tolerance. Average final total daily dose was 520 mg/d, nefazodone1 alone and 479 mg/d in patients receiving combination therapy (nefazodone1 plus CBASP).

Unlike prior studies involving psychotherapy, Dr. Keller pointed out, the investigators opted for the maximum tolerated CBASP dose. For the first six weeks, all participants received two 50-minute sessions per week. "We opted for that dose, even though it is unusual, to ensure the effects of psychotherapy would be apparent," Dr. Keller added.

Baseline demographics indicated a lower marriage rate (38 percent) among the 681 participants compared to that seen in the general population, which was interpreted as a proxy of the consequences of suffering from an illness for over 20 years with dysthymia and about ten years with major depression, he said. Similarly, the unemployment rate among participants was about three times higher than that seen in the general population.

Baseline clinical characteristics also showed a comorbidity reflective of previous chronic depression studies: about 33 percent had a lifetime anxiety disorder, about 60 percent had a lifetime personality disorder, and about 30 percent had a lifetime history of alcohol and substance abuse in addition to being chronically and severely ill, Dr. Keller pointed out.

"This was reflected in the HAM-D and CGI scores," he added, "which were in the moderate to moderate-marked ranges."

In exploring the advantages seen to date with each modality, Dr. Keller focused on data from the acute phase of the trial and preliminary results from the study's secondary phases, concentrating on findings related to chronic major depression (CMD), defined as illness lasting at least two years, double depression (DD), described as major depression superimposed on dysthymia for at least two years, and major double depression (MDD) without complete inter-episode recovery (that is, recurrent episodes of major depression without full recovery between episodes over two or more years).

Data from the acute phase of the study provided a number of surprises, he added, one of the first being that no difference in premature discontinuation was apparent between the three treatment groups (CBASP, 24 percent; nefazodone1, 26 percent; combination, 21 percent).

"When we tried to explain this," he said, "we saw adverse events were about 14 percent on [the] drug alone-typical of what occurs in any 12-week study involving a first-line antidepressant. But when psychotherapy was added to the drug, the adverse event rate was cut in half. This raises questions about the genesis of these symptoms, which are physiologic in nature."

Also interesting were discontinuations based on withdrawn consent, he said. While this occurred among five percent of patients receiving nefazodone1 alone, eight percent withdrew on combination therapy, while 14 percent withdrew among patients receiving CBASP alone. "The best explanation we could find for this was the inability of some participants to tolerate the psychotherapy," he added.

Response rates to therapy were also surprising, Dr. Keller said. Top-line findings among patients completing all 12 weeks of the trial indicated a 30 percent advantage in response for people receiving combination therapy versus drug alone or psychotherapy alone (85 percent, 52 percent and 55 percent, respectively). "We had anticipated no more than a 15 percent difference between any two cells in the study," Dr. Keller explained. "In fact, no other chronic depressive disorder study has shown a 12-week response rate much greater than 50 percent."

This held true in the intention-to-treat (ITT) analysis, he added, where patients receiving combination therapy had a much higher response than those in the other treatment arms (73 percent versus 48 percent for drug or psychotherapy alone). "Just as the 85 percent response rate is greater than any published study with major depression disorder, even if it is not chronic, the 25 percent difference seen in the ITT results is the highest response yet seen in any study of people with primary depression."

Time course of the response was also interesting, he said. "We were surprised that the nefazodone1 (alone) and the combination therapy groups both showed advantages after the first week compared to psychotherapy alone, which was maintained throughout the 12-week trial."

Similarly, he added, a significantly greater reduction in anxiety was seen in patients on drug alone versus psychotherapy alone, beginning in the second, third, and fourth weeks. "After that time frame, the differences became less significant, except with the combination therapy, where the differential improvement continued to accelerate over the other treatment arms as the trial moved through to week 12."

Continuation Phase

Turning to early results from the continuation phase, Dr. Keller pointed out that patients qualified only if they had a remission or satisfactory response to treatment after 12 weeks. Eligible participants were maintained on the same treatment they were given during the acute phase.

During this 16-week period, he added, the rate of discontinuation was lower than that seen in the acute phase, with no significant difference between the three cells. What was particularly interesting was the adverse event drop-out rate, he added, which declined to very low levels on both drug alone and combination therapy. "A clinical point to take from this is the need to help patients understand they are likely to have adverse events during the early treatment phases, but the data suggest these events will decrease or disappear over time if they stay the course," he pointed out.

As for weight gain related to therapy, no statistically significant overall difference between the three groups was evident, Dr. Keller said, although the researchers were surprised by the increased weight loss observed in the combination cell. "At present, we have no good explanation to account for this," he added.

Similarly, no major differences in sexual dysfunction were seen in the three treatment groups, Dr. Keller said. "Based on spontaneous reports and questionnaires, all three reported a significant improvement, with only minor differences apparent, in any of the treatment cells."

For those who attained remission in the three cells, he added, the rates remained the same-about 17 percent across the board, while a significant response rate was seen in about 15-20 percent of participants.

It should be noted, he said, that with the best treatments currently used, people with chronic major depression have at least a ten percent likelihood after 12 weeks of going from remission to developing what is close to full depression syndrome.

"So the good news is that many people maintained remission in this study," he added. "The bad news is how careful you need to be in keeping patients engaged once they have an initial response. We know that a very high proportion of these patients get better and are well for three months, but then leave treatment, regardless of their ability to afford it."

While 52-week maintenance phase results are not yet available, Dr. Keller concluded, the early NCDS data already suggest people who have been chronically depressed with major depression for up to 20 years can be successfully treated with nefazodone1, a structured form of psychotherapy (CBASP) and, particularly, by a combination of the two.

"The most compelling point we have made is that combination therapy is dramatically more effective than either treatment used alone," he added. "How this will ultimately translate into clinical practice remains to be seen, given factors such as cost and the fact that many practitioners are not sufficiently trained in major forms of psychotherapy to be comfortable administering it. Still, it is hard to look away from data that indicate there is a 30 percent increase in the probability of response after 12 weeks of combination therapy in this patient population. But regardless of what form of therapy a clinician is most comfortable using, we now have strong evidence to suggest they should use it both vigorously and aggressively to help patients with chronic depression."

The opinions expressed in this publication are those of the participating faculty and do not necessarily reflect the opinions or the recommendations of their affiliated institutions: University of Medicine & Dentistry of New Jersey; MMC, Inc.; or any other persons. Any procedures, medications, or other courses of diagnosis or treatment discussed or suggested in this publication should not be used by clinicians without evaluation of their patients' conditions, assessment of possible contraindications or dangers in use, review of any applicable manufacturer's product information, and comparison with the recommendation of other authorities. This educational monograph includes discussion of treatment and indications outside of current approved labeling. This CME monograph was made possible through an unrestricted educational grant from Ortho-McNeil Pharmaceutical, Inc.

(c) 2000 Millennium Medical Communications, Inc. and UMDNJ-Center for Continuing Education in the Health Professions