Treatments for Acute Bipolar Depression: Meta-analyses of Placebo-controlled, Monotherapy Trials of Anticonvulsants, Lithium and Antipsychotics

Abstract

Optimal treatments for bipolar depression, and the relative value of specific drugs for that purpose, remain uncertain, including agents other than antidepressants. We searched for reports of placebo-controlled, monotherapy trials of mood-stabilizing anticonvulsants, second-generation antipsychotics, or lithium for acute major depressive episodes in patients diagnosed with type I or II bipolar disorder and applied random-effects meta-analysis to evaluate their efficacy, comparing outcomes based on standardized mean drug-placebo differences (SMD) in improvement, relative response rates (RR), and number-needed-to-treat (NNT). We identified 24 trials of 10 treatments (lasting 7.5 weeks, with ≥ 50 collaborating sites/trial) that met eligibility criteria: lamotrigine (5 trials), quetiapine (5), valproate (4), 2 each for aripiprazole, olanzapine, ziprasidone, and 1 each for carbamazepine, lithium, lurasidone, and olanzapine-fluoxetine. Overall, pooled drug-over-placebo responder-rate superiority (RR) was moderate (29% [CI: 19-40%]), and NNT was 8.2 (CI: 6.4-11). By SMD, apparent efficacy ranked: olanzapine + fluoxetine ≥ valproate > quetiapine > lurasidone > olanzapine, aripiprazole, and carbamazepine; ziprasidone was ineffective, and lithium remains inadequately studied. Notably, drugs were superior to placebo in only 11/24 trials (5/5 with quetiapine, 2/4 with valproate), and only lamotrigine, quetiapine and valproate had > 2 trials. Treatment-associated mania-like reactions were uncommon (drugs: 3.7%; placebo: 4.7%). Controlled trials of non-antidepressant treatments for bipolar depression remain scarce, but findings with olanzapine-fluoxetine, lurasidone, quetiapine, and perhaps carbamazepine and valproate were encouraging; lithium requires adequate testing.