Key points are not available for this paper at this time.
OBJECTIVE: To systematically review published and unpublished efficacy studies of agomelatine in people with depression. DESIGN: Systematic review and meta-analysis. DATA SOURCES: Literature search (Pubmed, Embase, Medline), Cochrane Central Register of Controlled Trials, European Medicines Agency (EMA) regulatory file for agomelatine, manufacturers of agomelatine (Servier). ELIGIBILITY CRITERIA: Double blind randomised placebo and comparator controlled trials of agomelatine in depression with standard depression rating scales. DATA SYNTHESIS: Studies were pooled by using a random effects model with DerSimonian and Laird weights for comparisons with placebo and comparator antidepressant. The primary efficacy measure (change in rating scale score) was summarised with standardised mean difference (SMD; a measure of effect size) and secondary outcome measures with relative risks. All results were presented with 95% confidence intervals. Statistical heterogeneity was explored by visual inspection of funnel plots and by the I(2) statistic. Moderators of effect were explored by meta-regression. RESULTS: We identified 20 trials with 7460 participants meeting inclusion criteria (11 in the published literature, four from the European Medicines Agency file, and five from the manufacturer). Almost all studies used the 17 item Hamilton depression rating scale (score 0-50). Agomelatine was significantly more effective than placebo with an effect size (SMD) of 0.24 (95% confidence interval 0.12 to 0.35) and relative risk of response 1.25 (1.11 to 1.4). Compared with other antidepressants, agomelatine showed equal efficacy (SMD 0.00, -0.09 to 0.10). Significant heterogeneity was uncovered in most analyses, though risk of bias was low. Published studies were more likely than unpublished studies to have results that suggested advantages for agomelatine. CONCLUSIONS: Agomelatine is an effective antidepressant with similar efficacy to standard antidepressants. Published trials generally had more favourable results than unpublished studies.
Taylor et al. (Wed,) studied this question.