Systematic Review and Network Meta-Analysis: Efficacy and Safety of Antipsychotics vs Antiepileptics or Lithium for Acute Mania in Children and Adolescents

Abstract

To compare second-generation antipsychotics (SGAs) and mood stabilizers (MSs) in youth with a bipolar disorder type I (BD-I) manic/mixed episode. A systematic PubMed/Embase/PsycInfo literature search until December 31, 2023, for randomized trials of SGAs or MSs in patients≤18 years of age with BD-I manic/mixed episode was conducted. The study included a network meta-analysis comparing treatments regarding mania symptoms and mania response (co-primary outcomes), and secondary efficacy and tolerability outcomes. Eighteen studies (n= 2844, mean age= 11.74, female participants= 48.0%, mean study duration= 5.4 weeks) comparing 6 SGAs (aripiprazole, asenapine, olanzapine, quetiapine, risperidone, and ziprasidone) and 4 MSs (lithium, oxcarbazepine, topiramate, and valproate) were meta-analyzed. All 6 SGAs outperformed placebo in reducing manic symptomatology, including risperidone (standardized mean difference [SMD]=-1.18, 95% CI=-0.92,-1.45, Confidence in Network Meta-Analysis [CINeMA]= moderate confidence), olanzapine (SMD=-0.77, 95% CI=-0.36,-1.18, low confidence), aripiprazole (SMD=-0.67, 95% CI=-0.33,-1.01, moderate confidence), quetiapine (SMD=-0.60, 95% CI=-0.32,-0.87, high confidence), asenapine (SMD=-0.54, 95% CI=-0.19,-0.89, moderate confidence), and ziprasidone (SMD=-0.43, 95% CI=-0.17, 0.70, low confidence), whereas no mood stabilizer outperformed placebo. Concerning mania response, risperidone (RR= 2.58, 95% CI= 1.88, 3.54, low confidence), olanzapine (RR= 2.42, 95% CI= 1.33, 3.54, very low confidence), aripiprazole (RR= 2.05, 95% CI= 1.44, 2.92, low confidence), quetiapine (RR= 1.89, 95% CI= 1.45n 2.47, moderate confidence), asenapine (RR= 1.81, 95% CI= 1.28, 2.55, very low confidence) and lithium (RR= 1.35, 95% CI= 1.00, 1.83, p= .049, very low confidence) outperformed placebo, without superiority of other MSs vs placebo. Individually, risperidone was more efficacious in reducing manic symptomatology than all other comparators, except olanzapine and topiramate, yet with low/very low confidence, and was associated with increased prolactin and glucose. Pooled together, SGAs outperformed both placebo and MSs for mania symptom reduction (SMD=-0.68, 95% CI=-0.86,-0.51 and SMD=-0.61, 95% CI=-0.82,-0.40, moderate confidence), and mania response (RR= 1.85, 95% CI= 1.53, 2.24 and RR= 1.65, 95% CI= 1.33, 2.04, moderate confidence) without differences between MSs and placebo. There were no significant treatment-placebo differences for all-cause discontinuation, whereas lithium, ziprasidone, and oxcarbazepine were associated with more adverse event-related drop-outs than placebo. Most SGAs were associated with more sedation, weight gain, and metabolic issues vs placebo and MSs. SGAs were more efficacious than placebo and MSs in treating acute mania symptoms, however, their use must be carefully weighed against important side effects.