Double-blind comparison of escitalopram and duloxetine in the acute treatment of major depressive disorder

Escitalopram is the most selective serotonin reuptake inhibitor antidepressant; in contrast, duloxetine inhibits both serotonin and norepinephrine reuptake. Double-blind comparison studies may help guide treatment decisions by revealing the relative benefits of different therapeutic approaches. This study evaluated the efficacy and safety of escitalopram versus duloxetine in the acute treatment of patients with moderate to severe major depressive disorder. A 1-week, single-blind, placebo lead-in period followed by an 8-week, randomised, double-blind, multicentre, parallel-group comparison was conducted from 20 April 2005 to 10 March 2006 in independent psychiatric research facilities with principal investigators who were board certified in psychiatry. A total of 278 outpatients of 382 patients screened with Diagnostic and Statistical Manual of Mental Disorders (4th edition)-diagnosed major depressive disorder (Montgomery-Asberg Depression Rating Scale [MADRS] total score > or =26) were randomised to the two treatment groups. Eight patients received no medication and were excluded from the safety group. Patients were treated with either escitalopram 10-20 mg/day (fixed at 10 mg/day for the first 4 weeks) or duloxetine 60 mg/day. The primary efficacy variable was change from baseline at week 8 in MADRS total score using the last observation carried forward (LOCF) approach. Efficacy, safety and tolerability measures were prospectively defined in the statistical analysis plan prior to study initiation unless otherwise specifically noted as conducted post hoc. A significantly greater proportion of escitalopram-treated patients completed the 8-week study compared with duloxetine-treated patients (87% vs 69%, respectively; p < 0.01). Mean baseline MADRS total scores were 31.0 for the escitalopram group and 31.6 for the duloxetine group. At week 8, escitalopram treatment resulted in significantly greater improvement compared with duloxetine on the prospectively defined primary efficacy endpoint of mean change from baseline in MADRS total score using the LOCF approach (least-squares mean difference [LSMD] -2.42; 95% CI -4.73, -0.11; p < 0.05). There was no difference between treatment groups in the observed cases (OC) analysis (LSMD -0.32; 95% CI -2.71, 2.07; p = 0.79). Significantly fewer escitalopram-treated patients discontinued because of adverse events compared with duloxetine (2% vs 13%, respectively; p < 0.01). These findings suggest that escitalopram is bet