Remission, dropouts, and adverse drug reaction rates in major depressive disorder: a meta-analysis of head-to-head trials

Objective: To summarize remission rates and dropouts due to adverse drug reactions (ADRs) or lack of efficacy (LoE) of serotonin-norepinephrine reuptake inhibitors (SNRIs), selective serotonin-reuptake inhibitors (SSRIs), and tricyclic antidepressants (TCAs) in treating major depressive disorder. Methods: We searched MEDLINE, EMBASE, IPA, and the Cochrane International Library from 1980-2005. Meta-analysis summarized outcomes from head-to-head randomized clinical trials comparing ! 2 drugs from three antidepressants classes (SNRIs, and/or SSRIs, and/or TCAs) followed by ! 6 weeks of treatment. Remission was a final Hamilton Depression Rating Scale (HAMD) score <= 7 or Montgomery-Asberg Depression Rating Scale (MADRS) <= 12. Intent-to-treat data were combined across study arms using random effects models, producing point estimates with 95% confidence intervals. Results: We obtained data from 30 arms of 15 head-to-head trials with 2458 patients. SNRIs had the highest ITT remission rate (49.0%), then TCAs (44.11%), and SSRIs (37.7%) (p > 0.05 for SNRIs versus TCAs; p < 0.001 for TCAs versus SSRIs and SNRIs versus SSRIs). When categorized as inpatients (n = 582) and outpatients (n = 1613), SNRIs had the highest remission rates (52.0% for 144 inpatients and 49.3% for 559 outpatients). SNRIs had lowest overall dropouts (26.1%), followed by SSRIs (28.4%), and TCAs (35.7%). Dropouts due to ADRs and LoE were 10.3% and 6.2% for SNRIs, 8.3% and 7.2% for SSRIs, and 19.8% and 9.9% for TCAs, respectively (p > 0.05 for ADR dropouts only). One limitation was the inclusion of only venlafaxine-XR; results may not be the same for immediate release forms. In addition, few studies reported remission rates. Conclusions: SNRIs had the highest efficacy remission rates (statistically significant for inpatients and outpatients), and the lowest overall dropout rates, suggesting clinical superiority in treating major depression.