THIRD INTERNATIONAL CONFERENCE 
ON BIPOLAR DISORDER

Methodologic Issues and Practical Difficulties in Conducting Clinical Trials in Bipolar Disorder

John Ascher, M.D.
Bibliography
Gary Evoniuk, Ph.D.

Slide Presentation

Abstract
The past decade has seen a resurgence of interest in new treatments for bipolar disorder. Lithium, long regarded as the sole therapeutic mainstay, has been joined by other classes of drugs found to be effective and useful in the treatment of mania and/or depression: the anticonvulsants (e.g. carbamazepine, valproate, lamotrigine), the antidepressants (e.g. SSRIs, bupropion) and the antipsychotics (e.g. haloperidol, clozapine, olanzepine). This growth in interest has led to a substantial increase in the number of clinical trials conducted over the past few years, but has begun to magnify some of the very important methodological, regulatory and ethical difficulties now being faced. This paper will touch on three specific issues which illustrate these areas of difficulty and discuss potential approaches to solutions.

The first issue centers on the difficulty in recruiting appropriate patients for clinical studies. At least two separate issues have converged to create this problem: (1) The prototypical clinical trial subject makes for a very atypical bipolar patient. The best trial patients are compliant and have clear, well-delineated psychopathology (e.g. classic elated mania) uncomplicated by complex features (e.g. rapid, cycling, mixed states) or comorbidity (e.g. alcohol and substance abuse). This profile does not capture the typical bipolar patient. (2) There has been an exponential increase in the number of studies currently underway to test the efficacy of the newest generation of anticonvulsants, antipsychotics and antidepressants in bipolar disorder, many of which are competing for the same patient pool. This pressure to recruit patients doubtless exacerbates the problem of placebo response, which is also driven by other poorly understood factors.

The second issue is a more narrow methodologic issue: to what extent are efficacy data generated in depression generalizable across unipolar vs. bipolar populations? Genetic, epidemiologic and phenomenologic data suggest important differences in bipolar vs. unipolar depression. Yet regulators, especially in the USA, have suggested that they may not be prepared to approve treatments on the basis of studies conducted in bipolar depressed patients alone, in the absence of definitive data which prove or disprove their antidepressant efficacy in the larger unipolar population. Clinical study data will be presented bearing on this issue of "pseudospecifity" and the generalizability of antidepressant effects demonstrated for lamotrigine in bipolar vs. unipolar depressed patients.

The third issue concerns the broad question of how best to design clinical studies which are acceptable to: (1) the bipolar disorder patients who volunteer as research subjects; (2) the clinical investigators who conduct the trials; (3) institutional review boards and ethics committees who approve all clinical research studies; (4) the companies who must determine if a test compound can and should be developed into a medicine; and (5) the regulatory agencies who must ultimately determine if a new medicine can be marketed as an effective and safe treatment for bipolar disorder. The needs and views of each of these groups can and do conflict on many issues (e.g. use of placebo and other controls, length of study treatment). Constructive dialogue between all parties is needed to reach common understanding of these views and to target mutually acceptable study designs.

Combination Treatment in Bipolar Disorder

Atul C. Pande, M.D., F.R.C.P.C.
Bibliography

Slide Presentation

Abstract
Anticonvulsant drugs have become well-established in the treatment of patients with bipolar illness, both as monotherapy and in combination with lithium. Gabapentin is an anticonvulsant that appears to have some beneficial effects on mood and well-being in patients with epilepsy (Dimond et al.). There are also animal behavioral data showing gabapentin may be anxiolytic (Singh et al.). To test whether gabapentin may be useful in the treatment of bipolar illness, a controlled study was carried out. This was a double-blind, placebo-controlled trial of adjunctive gabapentin (dosed flexibly between 900-3600 mg/day). Patients with a lifetime diagnosis of Bipolar Disorder (type I), and who were currently suffering from symptoms of either mania, hypomania or a mixed state, were eligible for inclusion. Patients entered a four week stabilization phase in which the doses of their ongoing treatment (which could be lithium, valproate or lithium+valproate) were optimized. Patients who continued to be symptomatic at the end of this phase were randomized to adjunctive treatment with either gabapentin or placebo for 10 weeks. The clinical ratings included the Young Mania Rating Scale (YMRS), the Hamilton Depression Scale (HAMD), the Internal States Scale (ISS), the NIMH-LCM Life Chart, and the Clinical Global Impression of Change. The primary efficacy analysis was the baseline to endpoint change in the total YMRS score. Both treatment groups had a decrease in total YMRS from baseline to endpoint but this decrease was significantly greater in the placebo group (-9) than the gabapentin group (-6) (p<.05). No difference between treatments was found for the total score on the Hamilton Depression Scale or on the rate of responders on the ISS. Several post-hoc analyses were carried out on the data in a search for explanations for the findings. When the sample was divided into three severity groups (mild, moderate and severe) based on baseline YMRS, the superiority of placebo over drug held for each subgroup but the largest difference in favor of placebo was found in the patients with the highest baseline YMRS scores. Because this was an adjunctive treatment study, we examined whether any changes to "background" treatments may have accounted for the treatment effects. During the stabilization phase 16 patients had changes made to their lithium doses and of these 12 were in the group treated with adjunctive placebo in the double-blind period. Of these 12 placebo patients, 9 had increases in lithium dose, 3 had decreases. When all patients who had a change in lithium doses (n = 16) are removed from the efficacy analysis, the YMRS treatment difference still favors placebo but is no longer statistically significant. This suggests that the patients whose lithium dose was adjusted during the baseline period have a disproportionately large influence on the overall results.

Several published clinical reports (totaling nearly 200 patients) suggest that gabapentin is useful in the treatment of patients with bipolar illness. That observation is in direct contrast to the findings of the present study. The present study included a heterogeneous patient population who, despite being ‘refractory’ to mood-stabilizing treatments, showed a robust response to placebo, making it difficult to conclude definitively that gabapentin is not effective in treating bipolar disorder. Alternative possible explanations may be that (1) gabapentin treats some other symptoms dimension (such as anxiety) not captured by the YMRS or HAMD, but which clinicians and patients perceive to be valuable, or (2) the mood-stabilizing effect of gabapentin is not additive to that of lithium or valproate. Further studies are required to clarify these questions before gabapentin can be recommended as an effective mood-stabilizing drug.

Novel Endpoints for Clinical Trials in Bipolar Disorder

Ellen Frank, Ph.D.
David J. Kupfer, M.D.
Alexandre Gerebtzoff, M.D.
Uwe Meya, M.D.
Fouzia Laghrissi-Thode, M.D.
Bibliography
Victoria J. Grochocinski, Ph.D.
Patricia R. Houck,M.S.H.
Alan G. Mallinger, M.D.
Robert D. Gibbons, Ph.D.

Slide Presentation

Abstract
Over the last decade, the ratio of published clinical trials in manic patients versus schizophrenic patients is approximately one to ten reflecting the difficulties in conducting studies in the bipolar population. Recently, there is an increasing interest on the part of academic and pharmaceutical researchers to evaluate new antimanic compounds. However the classical approach, i.e., evaluating mean rating scale score changes over time between groups, requires lengthy trials and large number of patients. The exposure of acute manic patients to placebo, possible ineffective drugs or dosages have raised ethical, methodological as well as clinical concerns. Accordingly, we set out to define "exit" criteria or novel clinical endpoints to assess the efficacy of antimanic compounds in more safe, economical, sensitive and specific approach.

Among bipolar patients enrolled in two ongoing research protocols, 76 hospital admissions for a manic or mixed episode experienced by 49 subjects were identified. Data used for the development of the "exit" criteria included twice weekly Bech-Rafaelsen Mania Scale (BRMS) and Hamilton Rating Scale for Depression (17 and 25-items score) (HRSD) ratings for the first 14 days of admission and weekly thereafter. We fit a mixed-effects regression model to all available data on the BRMS and the HRSD from admission to day 28 of treatment. Using the estimated model coefficients we obtained Empirical Bayes (EB) estimates of each subject’s trend coefficients based on : 1) all available data; 2) data available through day 11 of treatment for mania outcome; and 3) data available through day 28 of treatment for depression outcome.

We found a high correlation (r = .67) between EB estimates of final response at day 28 and actual day 28 scores on the BRMS. When subjects were categorized as full, partial or non responder, we were able on day 11 to predict 46 of 53 responders based on the BRMS score. In a second step, we examined the HRSD scores on the assumption that a true remission has been achieved only in those patients who do not switch into depression. Using the EB estimate we were able to correctly classify 100% of the responders in remission.

On the basis of this chart review study, our results suggest that "exit" criteria for trials assessing the efficacy of antimanic compounds could be defined. We found a strong relationship between failure to respond early (Day 11) and subsequent failure to remit (Day 28). Thus, patients whose early treatment is not characterized by a rapid decline in BRMS scores could confidently be removed early from the trials. The "exit" criteria need of course to be further developed and tested in prospective clinical trials.

Use of Antipsychotics in Bipolar Disorder

Mauricio Tohen, MD
Bibliography

Slide Presentation

Abstract
Antipsychotic agents are used in the treatment of bipolar disorder. Prior to the introduction of lithium, typical antipsychotic agents comprised the primary treatment for bipolar disorder. Their role later became secondary to lithium. However, clinical evidence has accumulated indicating that a significant number of patients are either unresponsive to or intolerant of lithium. In spite of the adverse effect profile (akathisia, tardive dyskinesia, and neuroleptic malignant syndrome) of the typical antipsychotics, a substantial number of patients are exposed to these agents. In part, this may be explained by the fact that a significant number of patients fail to respond to conventional therapies (e.g. lithium or anticonvulsant agents). The atypical antipsychotic agents (e.g., clozapine, risperidone and olanzapine) appear to show promise as alternative therapies for patients with bipolar disorder. Clozapine may exert beneficial effects on patients with bipolar disorder who respond inadequately to or are unable to tolerate conventional antipsychotic drugs. Risperidone has been reported to have a more favorable profile compared with that of clozapine and conventional antipsychotics, and has apparent efficacy in the management of acute mania. Another novel antipsychotic, olanzapine, appears to be safe and effective in the acute (placebo-controlled) and long-term (open-label) treatment of bipolar I patients.

In this presentation, the use of antipsychotics in bipolar disorder will be discussed. An emphasis will be placed on the newer atypical antipsychotics in terms of their efficacy and safety in the treatment of mania.

Top of Page