Background:
Asenapine demonstrated superiority over placebo for mania in bipolar I disorder patients experiencing acute current manic or mixed episodes in 2 randomized, placebo- and olanzapine-controlled trials. We report the results of exploratory pooled post hoc analyses from these trials evaluating asenapine’s effects on depressive symptoms in patients from these trials with significant baseline depressive symptoms.
Methods:
In the original trials (A7501004 [NCT00159744], A7501005 [NCT00159796]), 977 patients were randomized to flexible-dose sublingual asenapine (10 mg twice daily on day 1; 5 or 10 mg twice daily thereafter), placebo, or oral olanzapine 5-20 mg once daily for 3 weeks. Three populations were defined using baseline depressive symptoms: (1) Montgomery-Asberg Depression Rating Scale (MADRS) total score >=20 (n=132); (2) Clinical Global Impression for Bipolar Disorder-Depression (CGI-BP-D) scale severity score >=4 (n=170); (3) diagnosis of mixed episodes (n=302) by investigative site screening. For each population, asenapine and olanzapine were independently compared with placebo using least squares mean change from baseline on depressive symptom measures.
Results:
Decreases in MADRS total score were statistically greater with asenapine versus placebo at days 7 and 21 in all populations; differences between olanzapine and placebo were not significant. Decreases in CGI-BP-D score were significantly greater with asenapine versus placebo at day 7 in all categories and day 21 in population 1; CGI-BP-D score reductions were significantly greater with olanzapine versus placebo at day 21 in population 1 and day 7 in populations 2 and 3.
Conclusions:
These post hoc analyses show that asenapine reduced depressive symptoms in bipolar I disorder patients experiencing acute manic or mixed episodes with clinically relevant depressive symptoms at baseline; olanzapine results appeared to be less consistent. Controlled studies of asenapine in patients with acute bipolar depression are necessary to confirm the generalizability of these findings.