Placebo-controlled trials do not find association of olanzapine with exacerbation of bipolar mania

Background: Published case reports describe apparent induction or exacerbation of manic-like symptoms during treatment with the atypical antipsychotics olanzapine and risperidone. To date, such reports are from uncontrolled clinical experience and therefore cannot clarify whether the atypical antipsychotics caused such manic-like states or simply failed to prevent them. Presumably, bipolar patients would be at increased risk for this putative adverse event. Therefore, we evaluated the potential of olanzapine to exacerbate symptoms of mania compared to placebo during treatment of bipolar mania. Methods: Two inpatient, double-blind, randomized trials investigating the efficacy of olanzapine 5-20 mg daily versus placebo for the treatment of acute mania were combined. Two hundred and fifty-four subjects participated (placebo n = 129; olanzapine n = 125) in the two studies. Severity of mania was quantified with the 11-item Young-Mania Rating Scale (Y-MRS). In a post-hoc analysis, after double-blind therapy up to 3 weeks, categorical comparison of olanzapine and placebo groups was made for any worsening and worsening by 10 or 20% from baseline Y-MRS scores (LOCF). Results: The percentage of subjects with exacerbation at endpoint were: any worsening, placebo 37.7%, olanzapine 21.8% (P = 0.005); greater than or equal to10% worsening, placebo 24.6%, olanzapine 14.5% (P = 0.039); greater than or equal to20% worsening, placebo 15.6%, olanzapine 8.1% (P = 0.064). Conclusion: Mania rating scores worsened for some patients during olanzapine therapy. However, this was significantly less common with olanzapine than with placebo. These controlled data suggest that clinical case reports of occurrence of 'mania' during treatment with olanzapine, and possibly those with other atypical antipsychotics, reflect exacerbation in the natural history of bipolar illness, rather than an adverse pharmacological effect. Limitations: Post-hoc analysis of pooled data from two different studies. (C) 2002 Elsevier Science B.V. All rights reserved.