Differences between antipsychotic and placebo treatment emerge within 4 weeks of starting therapy, an analysis of randomised, controlled trials shows.

The study also reveals several predictors of treatment response, two of which the researchers say “would come as a surprise to most clinicians.”

Firstly, they found particularly good responses among patients who had both prominent positive and negative symptoms. The difference in Positive and Negative Syndrome Scale (PANSS) scores between active and placebo treatment was 2.32 points (24%) larger for these patients than for those with prominent positive symptoms only and 6.39 points (65%) larger than for those with prominent negative symptoms alone.

“Thus, the current antipsychotics seem more effective in pan-symptomatic patients than in those with predominantly psychotic presentation”, write lead researcher Jonathan Rabinowitz (Bar Ilan University, Ramat Gan, Israel) and colleagues in The Journal of Clinical Psychiatry.

They also found that use of benzodiazepines did not influence treatment response, which they say “would surprise many given the generally widespread use of benzodiazepines as adjunctive medications.”

Women had a larger treatment response than men, while younger age at illness onset and a larger number of hospitalisations were related to illness course, but not to treatment response.

The team analysed data from the NEWMEDS repository, the result of an “unprecedented private–public collaboration”. They used data from 6971 patients given an antipsychotic and 2200 given placebo in 29 randomised trials of second-generation antipsychotic agents.

Twenty-four of these studies lasted at least 6 weeks, and the proportion of eventual difference in PANSS scores between antipsychotic- and placebo-treated patients increased as the weeks passed, from 44.5% of the 6-week response at week 1 to 90.2% and 95.4% at weeks 4 and 5, respectively.

Twenty of these 24 studies showed a significant difference between the active and placebo treatment groups by week 6, but 19 did so by week 4. Of the four studies that did not show a significant difference at week 6, two were significant at week 4 and one at week 5. Furthermore, the retention rate at week 4 was significantly higher than at week 6, at 68.2% versus 61.8% in the antipsychotic group and 60.1% versus 53.2% in the placebo group.

The researchers calculate that proof-of-concept trials of 4 weeks’ duration with 49 patients per treatment group would be sufficient, noting that “recruitment for shorter trials will probably be easier and retention rates should be even higher than shown here, as patients may be willing to stay in a shorter study with the end in sight.”

Such trials would also be cheaper and require less imputation of missing data, they add.

This article first appeared on News Medical on 1 May, 2014.