Recent placebo-controlled acute trials in bipolar depression: focus on methodology

The completion of three recent large-scale, double-blind controlled acute trials in bipolar I depression has improved our understanding of the management of major depressive episodes associated with bipolar disorder. In contrast to the cross-over designs used in the early studies of lithium in bipolar depression, the designs utilized in these recent studies have employed random assignment to parallel arms including the use of placebo as a monotherapy in one study. The analyses of recent studies have all been conducted on intent-to-treat data, and included two types, change from baseline analyses and responder analyses. Lamotrigine monotherapy was shown to be superior to placebo with both types of analyses on the Montgomery-Asberg Depression Rating Scale (MADRS) and the Clinical Global Impressions (CGI) scales, but not the 17-item Hamilton Depression Rating Scale (HAMD) (n=195). The percentage of patients responding to placebo as a monotherapy were 29, 26 and 37%, respectively; there were no differences in switch rates (5% vs. 5%). Paroxetine augmentation was no better than placebo augmentation overall with both types analyses on the CGI and HAMD (n=117); the MADRS was not used. In patients with lithium levels < or =0.8 mequiv./l, the change from baseline analysis showed paroxetine to be superior to placebo, but responder analyses were negative; switch rates with paroxetine, imipramine, and placebo were 0, 8 and 2%. Moclobemide monotherapy was similar in efficacy to imipramine (n=156), but had a lower rate of switching (4% vs. 11%).