A middle-of-the-road dose of an antipsychotic appears to be optimal for relapse prevention in stable schizophrenia, new research suggests.
Results of a meta-analysis show a 5-mg/day equivalent risperidone dose worked best. Higher doses were associated with more adverse events without showing substantial gains in relapse prevention, and lower doses were associated with greater relapse risk.
“The safest approach is to just to carry on with 5 mg,” which in many cases represents a full dose, lead author Stefan Leucht, MD, professor, department of psychiatry and psychotherapy, Technical University of Munich School of Medicine, Germany, told this news organization.
However, he added, patient preferences and other factors should be considered in dosage decision-making.
The findings were published online August 18 in JAMA Psychiatry.
Unique meta-analysis
Antipsychotic drugs are effective for short-term treatment of schizophrenia and prevention of relapse but are associated with movement disorders, weight gain, and other metabolic changes. They are also associated with even more severe adverse events, including tardive dyskinesia and increased cardiovascular risk.
For years, researchers have tried to find the optimal dose of antipsychotic drugs to prevent relapse in patients with stable schizophrenia while mitigating adverse event risk.
For the meta-analysis, researchers searched for fixed-dose, randomized, blinded, or open trials that lasted longer than 3 months and compared two first-generation antipsychotics – haloperidol or fluphenazine – or a second-generation antipsychotic with placebo or a different dose of the same drug.
The analysis included 26 studies with 72 individual dose arms and 4,776 participants with stable schizophrenia.
Researchers used a dose-response meta-analysis. Unlike a simple meta-analysis that provides an “arbitrary” cut-off of superiority of one drug over placebo or another drug, a dose-response meta-analysis gives a plot or curve “that shows how this evolves with different doses,” Dr. Leucht noted.
The investigators estimated dose-response curves for each antipsychotic drug compared with placebo separately and as a group.
They did not have enough data for most of the single antipsychotics, so they converted doses to risperidone equivalents for a pooled analysis across drugs. They chose risperidone because its equivalents “are pretty well-defined,” said Dr. Leucht.
Go slow to go low
For the primary outcome of relapse, the dose-response curve showed a hyperbolic shape with a clear plateau. Initially, the plot decreased sharply but then flattened at about 5-mg/day risperidone equivalent (odds ratio, 0.20; 95% confidence interval, 0.13-0.31; relative risk, 0.43; 95% CI, 0.31-0.57).
“We were a little disappointed because we hoped that a dose lower than 5 mg would be most efficacious in terms of relapse rate because this would have reduced the side-effect burden,” Dr. Leucht said.
Nevertheless, he emphasized that doses lower than 5 mg/day risperidone equivalent are not completely ineffective. For example, the 2.5-mg dose reduced risk to relapse in relative terms by about 40% (RR, 0.63).
Dr. Leucht also pointed out there is “huge interindividual variability.” Therefore, “It just means for the average patient it’s safest, let’s say, to keep her or him on 5 mg,” he said.
When lowering the dose, Dr. Leucht noted clinicians should “be very careful and to do it very slowly. It should be very small reductions every 3 to 6 months.”
For the secondary endpoint of rehospitalizations, the shape of the curve was similar to the one for relapse but with lower rates.
“If patients need to be rehospitalized, it usually means that the relapse was major and not only a minor increase in symptoms,” said Dr. Leucht.
The curves for all-cause discontinuation and reduction in overall symptoms were also similar to that of relapse.
However, the curve for dropouts because of adverse events showed that higher doses led to more adverse events. For example, with 5-mg/day dose, the OR was 1.4 (95% CI, 0.87-2.25) and the RR was 1.38 (95% CI, 0.87-2.15), but for the 15-mg/day dose, the OR was 2.88 (95% CI, 1.52-5.45) and the RR was 2.68 (95% CI, 1.49-4.62).