Review of: Chia-Ling Y et al, BMJ Ment Health 2023;26(1):e300546

Study Type: Meta-analysis of ­placebo-controlled trials

This meta-analysis sought to answer two questions: 1) Which antipsychotics are the most effective in acute mania? and 2) How does their optimal dosing differ between mania and schizophrenia?

To answer the first question, the authors searched for blinded, randomized controlled trials using flexibly dosed antipsychotics in acute bipolar mania, and they selected 42 studies totaling 11,396 patients with a mean age of 39 years. They compared antipsychotics by their antimanic efficacy at week three of treatment and created the following rank list: risperidone (effect size: 0.82), haloperidol (0.74), paliperidone (0.71), cariprazine (0.57), aripiprazole (0.53), olanzapine (0.49), quetiapine (0.42), asenapine (0.37), ziprasidone (0.33), brexpiprazole (0.12), and placebo (0.05). Notably, risperidone was significantly more effective than olanzapine (standardized mean difference = -0.22, 95% confidence interval = -0.41 to -0.02).

Because these were flexibly dosed studies, the authors used the mean dose to arrive at the optimal dose for each antipsychotic. Those were: quetiapine 618 mg, ziprasidone 119 mg, aripiprazole 24.6 mg, asenapine 18.3 mg, haloperidol 10.1 mg, paliperidone 9.1 mg, cariprazine 7.9 mg, risperidone 4.8 mg, and brexpiprazole 3.3 mg.

To answer the second question, the authors converted the optimal doses from mania studies to olanzapine equivalents and compared them to published data on optimal olanzapine dose equivalents in schizophrenia. Optimal doses for acute mania were higher for quetiapine (28.5%, p<0.001) and aripiprazole (17.0%, p<0.001) but lower for haloperidol (-8.1%, p<0.001) and risperidone (-15.8%, p<0.001). In other words, treatment of acute mania relative to schizophrenia required a higher mean dose of quetiapine (618 mg vs 482 mg) and aripiprazole (24.6 mg vs 21.0 mg) but a lower mean dose of haloperidol (10.1 mg vs 11.0 mg) and risperidone (4.8 mg vs 5.7 mg). Dosing of ziprasidone was not different between acute mania and schizophrenia (1%, p=0.32), and there were not enough published data for the other antipsychotics in schizophrenia to compare their dosing between mania and schizophrenia.

The main limitation is built into the “network” design, which assumes that outcomes are comparable across studies. In reality, many aspects of a study’s design influence a drug’s effect. While other analyses support the high ranks of risperidone and haloperidol in mania and the low rank for brexpiprazole, we should note that paliperidone (ranked third here) failed to obtain FDA approval in mania due to inconsistent results in its trials. The authors also did not look at long-term trials, which would have identified a higher risk of precipitating depression when haloperidol is used in mania (Mohammad O and Osser DN, Harv Rev Psychiatry 2014;22(5):274–294).

Carlat Take

Risperidone and haloperidol are reliable options if antimanic efficacy is the goal, with the caution that haloperidol may cause depression. This study also offers guidance on antipsychotic dosing. Most antipsychotics have similar target levels for acute mania and schizophrenia, with the exceptions of quetiapine and aripiprazole (higher dosing in mania) and haloperidol and risperidone (slightly lower dosing in mania).