Augmentation agents to serotonin reuptake inhibitors for treatment-resistant obsessive-compulsive disorder: A network meta-analysis.

BACKGROUND:Various agents for augmentation of serotonin reuptake inhibitors have been investigated for treatment-resistant obsessive-compulsive disorder (OCD). We aimed to comprehensively compare different augmentation agents for treatment-resistant OCD in adults. METHODS:PubMed, Embase, Web of Science, CENTRAL, the WHO's ICTRP, and ClinicalTrials.gov were searched on February 20, 2018. Pairwise meta-analysis and Bayesian network meta-analysis were performed. The primary outcome was efficacy measured by the Yale-Brown Obsessive Compulsive Scale. The secondary outcomes were tolerability (side-effect discontinuation) and acceptability (all cause discontinuation). Mean differences (MDs) and odds ratios (ORs) were reported with 95% confidence intervals (CIs). RESULTS:Thirty-three articles with 34 trials (1216 patients) were included. Memantine (MD, -8.94; 95% CI, -14.42 to -3.42), risperidone (-4.47, -8.75 to -0.17), topiramate (-6.05, -10.89 to -1.20), lamotrigine (-6.07, -11.61 to -0.50), and aripiprazole (-5.14, -9.95 to -0.28) were significantly superior to placebo. Antipsychotic (-4.09, -6.22 to -1.93) and glutamatergic (-5.22, -7.53 to -2.84) agents were significantly superior to placebo. Considerable heterogeneity was found across studies, and baseline symptom severity was identified as a significant moderator. After baseline severity adjustment, quetiapine (-5.00, -8.59 to -1.29) and olanzapine (-8.28, -15.34 to -1.13) became significantly superior to placebo. CONCLUSIONS:Our study supports the use of antipsychotic or glutamatergic agents as augmentation agents for treatment-resistant OCD. Topiramate, lamotrigine, aripiprazole, olanzapine, risperidone, memantine, and quetiapine are alternative augmentation drugs; however, a definitive conclusion of the best drug remains undetermined because of the considerable heterogeneity and limited numbers of studies and patients for each agent.