Alex Evans, PharmD, MBA. Dr. Evans has no financial relationships with companies related to this material.
REVIEW OF: Terao I et al, J Affect Disord 2024;346:49–56
STUDY TYPE: Network meta-analysis
There are many augmenting agents for treatment-resistant depression (TRD). Which works best? Without many head-to-head studies to draw from, these researchers employed a meta-analysis to compare common medications.
They included 21 placebo-controlled trials of the most-studied augmentation agents in TRD: lithium, intravenous (IV) ketamine, intranasal esketamine, and aripiprazole. One study of lithium vs aripiprazole was also included (in which they had similar efficacy). All trials used standard measures of depression as primary outcomes, but they were heterogeneous in other ways. Study sizes ranged from around 10 to almost 400. They defined TRD as a failure of one antidepressant (AD), except those studies involving IV ketamine, where TRD was defined as the failure of two ADs. IV ketamine studies were also more likely to be shorter—some only one week. The lithium studies tended to be older and involved augmenting tricyclic antidepressants. Patients in the esketamine studies tended to have higher baseline levels of depression.
Unsurprisingly, all medications were more effective than placebo. Response rates were greater for augmentation with IV ketamine than for all other medications except lithium, where ketamine trended toward superiority but did not separate. The odds ratio for response with IV ketamine was 3.20 vs intranasal esketamine (1.43–7.16) and 2.90 vs aripiprazole (1.33–6.27). There was no significant difference in response rate between augmentation with esketamine, aripiprazole, or lithium. Both intranasal ketamine and aripiprazole had significantly more dropouts due to side effects than placebo. Noting the limitations inherent in their analysis, the authors rated the evidence comparing IV ketamine to other medications as low certainty.
CARLAT TAKE
TRD covers a wide variety of patients, and that’s a problem for this study. Network meta-analyses assume that patients and placebo responses are comparable across trials. This study hints at results that we’d like to see elucidated in future research.
Please see our Terms and Conditions, Privacy Policy, Subscription Agreement, Use of Cookies, and Hardware/Software Requirements to view our website.
© 2025 Carlat Publishing, LLC and Affiliates, All Rights Reserved.