Evaluate the efficacy of aripiprazole combination with lithium/valproate vs. placebo combination in bipolar mania using a titration regimen with a low starting dose (5 mg/day).

Eligible adult patients with bipolar mania receiving lithium/valproate and a Young Mania Rating Scale (YMRS) Total score ≥16 who might benefit from combination treatment with aripiprazole, were randomized to aripiprazole (n=181) or placebo (n=189) with lithium/valproate. Primary endpoint was mean change from baseline to Week 12 in YMRS. Secondary endpoints were Clinical Global Impressions-Bipolar Version (CGI-BP) severity of illness score, response rate (≥50% improvement in YMRS Total score), and remission rate (YMRS ≤12). Safety and tolerability were also assessed. Enrolment yielded a 77% power to detect a 2.6-point change in YMRS Total score at endpoint.

At endpoint, the mean change in YMRS Total Score (last-observation-carried-forward [LOCF]) for aripiprazole vs. placebo was not significant (treatment difference [-2.04] in favour of aripiprazole (95% CI: -4.14, 0.07; p=0.058). Mean change from baseline to endpoint in CGI-BP showed a treatment difference (-0.30) in favour of aripiprazole vs. placebo (95% CI: - 0.59, -0.01; p=0.044). Response rates were 68.8% vs. 61.3% (p=0.128) and remission rates were 69.9% vs. 64.0% (p=0.211) for aripiprazole and placebo, respectively. No unexpected adverse events (AEs) occurred. Treatment-emergent AEs (≥5% and twice the rate of placebo) were akathisia, depression, and nausea.

The target sample size of 388 patients was not achieved in this study and the primary outcome did not reach statistical significance. Now new or unexpected AEs occurred.