Comparative Efficacy and Acceptability of 13 Drugs for the Acute Treatment of Generalized Anxiety Disorder in Adults: A Network Meta-Analysis

Background: Drug therapy is the main clinical treatment method for generalized anxiety disorder (GAD), and numerous drugs are used to treat GAD clinically. A mixed-treatment meta-analysis performed in 2011 included randomized controlled trials (RCTs) conducted before February 2009 when evaluating the efficacy of nine drugs. However, the results were inconclusive due to the small number of articles included and the lack of direct comparisons. Many RCTs have been published since 2009, and hence it is necessary to perform a new meta-analysis of the efficacy of all the currently available drugs. Therefore, we aimed to update the evidence for the acute treatment of adults with GAD by comparing and ranking 13 drugs. Methods: The entire MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, PsycINFO, SCI (Science Citation Index), ClinicalTrials.gov, Current Controlled Trials, World Health Organization IRIS (Institutional Repository for Information Sharing), and ISRCTN Registry databases were searched for placebo-controlled and head-to-head trials of 21 drugs used for the acute treatment of adults with GAD from 1980 up to July 3, 2017. Data on demographics, clinical, and treatment information were extracted from each eligible study. We assessed the studies' risk of bias in accordance to the Cochrane Handbook for Systematic Reviews of Interventions. The primary outcomes were efficacy (quantified as the change in the total score on the Hamilton Anxiety Scale from baseline) and acceptability (quantified as treatment discontinuations due to any cause). This study is registered with PROSPERO, number CRD42017059496. Findings: The review identified 6618 citations, and the data on 59 RCTs were sufficient or appropriate for inclusion in the analysis. Analyses compared placebo plus the following 13 drugs: duloxetine, escitalopram, fluoxetine, paroxetine, sertraline, venlafaxine, agomelatine, vilazodone, vortioxetine, lorazepam, pregabalin, tiagabine, and buspirone. In terms of efficacy, all of the drugs except fluoxetine, tiagabine, and tiagabine were more effective than placebo, with the weighted mean difference ranging between -6·6 (95% credible interval [CrI] = -8·3 to -4·9) for agomelatine and -1·8 (95% CrI = -3·2 to -0·43) for vilazodone. For acceptability, only agomelatine (odds ratio [OR] = 0·6, 95% CrI = 0·36-1·0) were associated with fewer dropouts than placebo, whereas lorazepam (OR = 1·6, 95% CrI = 1·0-2·5), tiagabine (OR = 1·4, 95% CrI = 1·0-2·0), and vilazodone (OR = 1·7, 95% CrI = 1·2-2·6) were worse than placebo. In head-to-head comparisons, agomelatine showed better efficacy and acceptability than most of the other drugs. Small sample size effect exerted in the analysis of efficacy, response rate and tolerability. Agomelatine was the highest ranked drug for efficacy, acceptability, and tolerability. The risk of bias was high, moderate, and low in 23 (39%), 14 (24%), and 22 (37%) of the 59 RCTs, respectively. Interpretation: Our data suggest that most drugs for acute GAD are more effective than placebo, and there are few significant differences between the active drugs and placebo on acceptibility. Agomelatine might be the first-choice drug when starting treatment for acute GAD in adults because it provides the most favorable balance between benefits and acceptability. Funding Statement: This work was supported by the National Natural Science Foundation of China (No. 81771471), the National Natural Science Foundation of China (No.81460560) and the Clinical Research Award of the First Affiliated Hospital of Xi’an Jiaotong University, China (No.XJTU1AF-CRF-2016-024). Declaration of Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.