Delayed-onset hypothesis of antipsychotic drug action. A hypothesis tested and rejected.
Agid O, Kapur S, Arenovitch T and Zipursky R;
Commented by , 30 Dec 2003
Aim of the study
The time course of the antipsychotic effect has for a long time been an underresearched area. Elucidation of this issue is important for two reasons – on the one hand to understand the mechanism of action of antipsychotic drugs, but on the other hand also for recommendations on the duration of a trial with an antipsychotic before it is considered ineffective and switched to another compound.
Models of antipsychotic action in current textbooks are based on the theory that there is a "delayed onset" of action of this class of drugs. In this context Agid and colleagues carried out a meta-analysis of trials providing information on the time course of response to antipsychotic drugs.
Methods
The authors collected all published randomized, controlled, double blind, studies which provided data on the mean change of symptoms over time of patients with schizophrenia treated with antipsychotic medication. An inclusion criterion was that the studies involved one of the following antipsychotics: chlorpromazine, haloperidol, risperidone and olanzapine.
These were chosen because they are the most frequently used agents. In addition the studies had to have evaluations of symptoms during the first four study weeks and they had to use the BPRS or the PANSS as a rating scale.
Results
The authors identified 53 relevant studies that provided data on 8177 patients and allowed the construction of 112 symptom-response curves. Then the weekly percentage reduction of the total scores of either the BPRS or the PANSS for all four drugs combined was calculated and the following values were obtained:
- 14% after the first week of treatment
- 8.4% during the second week
- 4.2% during the third week
- 4.8% during the fourth week
Even when the results of the placebo-treated patients were subtracted to account for the placebo effect, no substantial changes in the results were obtained. In addition the same results were obtained when only the core psychotic symptoms of the rating scales were analysed. Thus an explanation of the early response by purely sedating effects could be ruled out.
Discussion
Over several decades many psychiatric textbooks described a “delayed onset of action” of antipsychotic drugs. The meta-analysis of the authors does not support this view. On the contrary it suggests that antipsychotic drug action starts as early as during the first week of treatment.
Indeed, trials on intramuscular formulations of antipsychotics show effect already after some hours, not only on general symptoms but also on psychotic symptoms. Furthermore the effect obtained in the first two weeks was greater than in the following 2 weeks.
The explanation of this phenomenon by pure sedation has to a large extent been ruled out by a) subtracting the placebo response and by b) showing the same effect on core psychotic symptoms and not only on general symptoms.
It is likely that this manuscript will change textbook recommendations. Speculations on the mechanism of action of antipsychotics need to take this early onset of action into account.
Furthermore, the results may be relevant for treatment decisions. It is possible that patients who do not respond early to an antipsychotic have a bad prognosis, and that it is better to switch treatment. However, this is another hypothesis that needs to be examined by randomised controlled trials.