A Review of Sensitivity and Tolerability of Antipsychotics in Patients With Bipolar Disorder or Schizophrenia: Focus on Somnolence

Keming Gao, M.D., Ph.D.; Stephen J. Ganocy, Ph.D.; Prashant Gajwani, M.D.; David J. Muzina, M.D.; David E. Kemp, M.D.; and Joseph R. Calabrese, M.D.

Objective: This study compared the sensitivity and tolerability of antipsychotics in patients with bipolar disorder or schizophrenia.

Data Sources: English-language literature from January 1966 to December 2006 cited in MEDLINE was searched for the terms antipsychotics, typical antipsychotics, atypical antipsychotic, generic and brand names of antipsychotics, safety, tolerability, discontinuation due to adverse events, somnolence, and bipolar mania, bipolar depression, bipolar disorder, manic-depressive illness, or schizophrenia, randomized, double blind, and controlled clinical trial.

Study Selection: Randomized, double-blind, placebo-controlled, monotherapy studies of antipsychotics in both bipolar disorder and schizophrenia were prioritized.

Data Extraction: Absolute risk increase (ARI) or reduction (ARR) and the numbers needed to treat to harm (NNTH) or benefit (NNTB) for the discontinuation due to adverse events and somnolence relative to placebo were estimated.

Data Synthesis: Ten acute trials in mania, 3 in bipolar depression, and 8 in schizophrenia were identified, along with 2 maintenance studies in bipolar disorder and 2 in schizophrenia. In schizophrenia, ziprasidone caused significantly more discontinuations due to adverse events than placebo, with an NNTH of 19, while aripiprazole caused significantly fewer discontinuations due to adverse events than placebo, with an NNTB of 12. In mania, there was no statistically significant difference in discontinuation due to adverse events between antipsychotics and placebo. However, in bipolar depression, both quetiapine and olanzapine caused more discontinuations due to adverse events than placebo, with NNTHs of 7 and 24, respectively. All atypical antipsychotics caused a significantly greater incidence of somnolence than placebo in mania and depression, with NNTHs from 5 to 8 for mania and 2 to 6 for depression. In schizophrenia, only olanzapine, ziprasidone, and aripiprazole (NNTHs from 5 to 14) caused a significantly higher incidence of somnolence. There was no significant difference between schizophrenia and mania in the discontinuation due to adverse events or somnolence of all studied antipsychotics. However, there was a significantly higher incidence of discontinuation due to adverse events and somnolence caused by quetiapine in bipolar depression than that in schizophrenia or mania.

Conclusion: Patients with bipolar disorder appear more sensitive to antipsychotics, and depressed patients are less tolerant to somnolence than those with either mania or schizophrenia.

(J Clin Psychiatry 2008;69:302-309; online ahead of print January 16, 2008)

Received Jan. 16, 2007; accepted May 7, 2007. From the Department of Psychiatry, Bipolar Disorder Research Center at the Mood Disorders Program, University Hospitals Case Medical Center/Case Western Reserve University, School of Medicine (Drs. Gao, Ganocy, Gajwani, Kemp, and Calabrese); and the Department of Psychiatry and Psychology, Cleveland Clinic Foundation, Cleveland Clinic Lerner College of Medicine at Case Western Reserve University (Dr. Muzina), Cleveland, Ohio.

Support for this manuscript includes grant # P20 MH-66054 (Dr. Calabrese) from the National Institute of Mental Health, Bethesda, Md.

Financial disclosure appears at the end of the article.

Corresponding author and reprints: Keming Gao, M.D., 11400 Euclid Ave., Suite #200, Cleveland, OH 44106 (e-mail: keming.gao@uhhospitals.org).