The Trend of Increasing Placebo Response and Decreasing Treatment Effect in Schizophrenia Trials Continues: An Update From the US Food and Drug Administration

Objective: Concerns of increasing placebo response and declining treatment effect in acute schizophrenia trials have been reported for new drug applications (NDAs) submitted to the US Food and Drug Administration (FDA) during an 18-year period from 1991 through January 2009 (ie, the pre-2009 period). Current exploratory analyses provide an update in the trends observed in placebo response, treatment effect, and dropout rates for NDAs submitted from February 2009 to 2015 (ie, the post-2009 period).

Data Sources: Clinical trial data from all acute schizophrenia trials that were submitted as part of NDAs to the US FDA during a 24-year period from 1991 to 2015.

Study Selection: Aggregate trial-level efficacy data from multicenter, multiregional, randomized, placebo-controlled, 4- to 8-week, fixed- and flexible-dose trials in adult schizophrenia patients were compiled. There were 12 NDAs pre-2009 (32 trials, N = 11,567) and 3 NDAs post-2009 (14 trials, N = 6,434).

Data Extraction: Baseline demographic and disease variables and scores on the Positive and Negative Syndrome Scale (PANSS) were summarized and compared for the two time periods (pre-2009 and post-2009). The primary efficacy measure was mean change from baseline to endpoint in total PANSS score obtained by last-observation-carried-forward analysis. Regional differences in placebo response and treatment effect were explored for the two time periods based on baseline patient characteristics, sample size, and dropout rates.

Results: Trials were predominantly multiregional (10/14; 71%) during the post-2009 period compared to the pre-2009 period (11/32; 34%). The overall trial success rates were 57% (8/14) and 78% (25/32) during the post-2009 and pre-2009 periods, respectively. Comparing the pre-2009 and post-2009 periods, the mean placebo response (change from baseline in PANSS score) increased from −6.4 to −10.5 and the mean treatment effect (drug response – placebo response) declined from −8.6 to −5.8 , with substantial differences observed especially in North American trials. In North American trials, placebo response increased from −4.3 (pre-2009) to −8.5 (post-2009), and treatment effect decreased from −9.0 (pre-2009) to −3.4 (post-2009). The difference in placebo response (pre- and post-2009: −10.0 and −11.3 ) and treatment effect (pre and post-2009: −8.1 and −6.4 ) in multiregional trials for the two time periods remained minimal. Baseline disease severity remained similar in the pre- and post-2009 time periods, with PANSS scores ranging between 85 and 100. Trials with higher mean baseline PANSS scores tended to show higher treatment effect irrespective of the time period and region. Post-2009, dropout rates were higher (55%) in North American trials compared with 33% in multiregional trials, similar to the pre-2009 trend.

Conclusions: The continuing trend of increasing placebo responses and decreasing treatment effects in schizophrenia trials over the 24-year period does remain of great concern, especially with respect to North American trials. However, given the current global nature of drug development, close attention to trial conduct and reexamination of design elements for future trials may be warranted.