Antipsychotics are a cornerstone of treatment for patients with schizophrenia. However, in recent years there have been several failed trials: antipsychotics commonly used in clinical practice did not show statistically significant improvements compared with placebo. A systematic review suggested an increasing placebo response rate as a potential explanation for this finding.1 However, this review focused on predictors of placebo response, but not antipsychotic response.
Meta-analysis of placebo-controlled antipsychotic trials
To identify predictors of drug-placebo differences, Leucht and colleagues2 performed a systematic review of all placebo-controlled antipsychotic trials in adults with acute exacerbations of schizophrenia since 1953. They excluded relapse prevention studies in stable patients taking maintenance antipsychotics, studies of patients with predominant negative symptoms, studies of patients with major psychiatric or medical comorbidities, and Chinese studies. They included all non-clozapine antipsychotics, and both fixed- and flexible-dose studies, but excluded studies of both short- and long-acting intramuscular agents.
Studies were identified by searching the Cochrane Schizophrenia Group Controlled Trials Register, the FDA website, ClinicalTrials.gov, MEDLINE, EMBASE, PsycINFO, and Cochrane CENTRAL. The primary outcome was the mean overall change in symptoms, measured by the change in rating scale scores (typically Positive and Negative Syndrome Scale [PANSS], Brief Psychiatric Rating Scale [BPRS], and/or Clinical Global Impressions [CGI] scale). Other outcomes included response rates, discontinuation, symptom subscale scores, quality of life, social functioning, and side effects. Data were analyzed using Bayesian hierarchical random-effects model meta-analysis. Meta-regression analyses were also performed to investigate potential moderating patient-, drug-, and design-related factors.
The authors identified 167 studies (28,102 participants) published between 1955 and 2016. Mean subject age was 39, mean illness duration was 13 years, and median study duration was 6 weeks (range, 3 to 28 weeks). Antipsychotics were significantly more efficacious than placebo with a small to moderate effect size (standardized mean difference = 0.38, after accounting for small trial effects and publication bias). Regarding the primary outcome: 51% in the antipsychotic group versus 30% in the placebo group had a “minimal” response (defined as a ≥ 20% decrease in PANSS/BPRS or at least minimal improvement on CGI). By contrast, 23% in the antipsychotic group versus 14% in the placebo group had a “good” response (defined as a ≥ 50% decrease in PANSS/BPRS or at least much improvement on CGI).
Considering potential moderating factors in a multivariate regression, higher placebo response and industry sponsorship were each associated with significantly lower effect sizes for antipsychotic-placebo differences. Antipsychotic response rates remained stable over time. Positive symptoms improved more than negative symptoms and depression with antipsychotic treatment. There were also short-term improvements in quality of life and functioning in the antipsychotic group.
The bottom line
The authors concluded that approximately twice as many acutely ill patients with schizophrenia and related disorders improved with antipsychotics versus placebo, but less than one-quarter of patients experienced a good response. They also found that increasing placebo response and industry sponsorship, but not decreasing drug response, reduced effect sizes for antipsychotic-placebo differences. The authors noted that a major limitation of the study is that all antipsychotics were analyzed as a single class. They suggest that smaller sample sizes with better patient selection may facilitate drug development.
Dr. Miller is Associate Professor in the Department of Psychiatry and Health Behavior at Augusta University in Augusta, GA, and Schizophrenia Section Editor for Psychiatric Times. He reports no conflicts of interest concerning the subject matter of this article.
1. Agid O, Siu CO, Potkin SG, et al. Meta-regression analysis of placebo response in antipsychotic trials, 1970-2010. Am J Psychiatry. 2013;170:1335-1344.
2. Leucht S, Leucht C, Huhn C, et al. Sixty years of placebo-controlled antipsychotic drug trials in acute schizophrenia: systematic review, Bayesian meta-analysis, and meta-regression of efficacy predictors. Am J Psychiatry. 2017 May 25. doi: 10.1176/appi.ajp.2017.16121358.