Background: While clinical trial evidence has firmly established the efficacy of several atypical antipsychotics (AAPs) for treating bipolar depression, no randomized controlled trials (RCT's) comparing AAPs have been conducted. This Bayesian network meta-analysis (NMA) compared the relative efficacy and tolerability of AAP monotherapy in adults with bipolar depression.
Methods: Efficacy measures included change in Montgomery Åsberg Depression Rating Scale (MADRS), Clinical Global Improvement - Bipolar Disorder (CGI-BP), response, and remission. Multiple tolerability outcomes were examined. Results from random effects models were reported as difference in change from baseline for continuous variables or odds ratios for dichotomous variables. Treatments were ranked using the surface under the curve cumulative ranking probabilities. Number needed to treat (NNT) and harm (NNH) were calculated.
Results: Eighteen RCT's met inclusion criteria of the systematic literature review. On change in MADRS, lurasidone (- 4.71 [95% Crl - 6.98, - 2.41]), quetiapine (- 4.80 [- 5.93, - 3.72]), olanzapine (- 4.57 [- 5.92, - 3.20]), and cariprazine (- 2.29 [- 3.47, - 1.09]) were more efficacious than placebo. Lurasidone was associated with a significantly greater odds of response (≥50% improvement in MADRS) compared to cariprazine (1.78 [95% Crl 1.08, 2.77]), aripiprazole (2.38 [1.38, 3.85]), and ziprasidone (2.47 [1.41, 3.98]), but was similar to olanzapine (1.68 [0.99,2.65]) and quetiapine (1.25 [0.78, 1.90]). For change in CGI-BP-S-overall score, lurasidone was significantly better than cariprazine (- 0.38 [95% Crl - 0.66,-0.10]) and ziprasidone (- 0.58 [- 0.91,-0.26]), but similar to quetiapine (- 0.08 [- 0.36, 0.19])and olanzapine (- 0.04 [- 1.41, 1.46]). Lurasidone (0.34 kg [95% Crl - 0.22, 0.89]) and aripiprazole (0.20 kg [- 0.59, 1.00]) had a similar weight change compared to placebo, but olanzapine (2.88 kg [2.40, 3.36]), quetiapine (1.17 kg [0.84, 1.49]), and cariprazine (0.65 kg [0.34, 0.96]) were associated with greater weight gain. The NNT for response was the lowest for lurasidone (NNT = 5) followed by quetiapine (NNT = 6), olanzapine (NNT = 10) and cariprazine (NNT = 12).
Conclusions: In this NMA in adults with bipolar depression, which evaluated change in depressive symptoms (assessed by MADRS) across short-term trials, the largest improvement versus placebo was observed for lurasidone, olanzapine and quetiapine with cariprazine, showing a smaller treatment effect. Aripiprazole and ziprasidone were ineffective for the treatment of bipolar depression. Improvement in CGI-BP-S score for lurasidone was larger than cariprazine and ziprasidone but similar to quetiapine and olanzapine. Based on short term studies lurasidone and aripiprazole had similar weight gain compared to placebo.