Comparative efficacy of racemic ketamine and esketamine for depression: A systematic review and meta-analysis

Ketamine's rapid antidepressant effects at subanaesthetic doses have attracted attention as an option for treatment-resistant depression, but its dissociative effects, abuse potential and the logistical challenges of intravenous administration have prompted investigation of alternative formulations and routes. One such alternative is the S-enantiomer esketamine, which is more potent at the NMDA receptor and is available as an intranasal spray; intranasal esketamine received regulatory approval for treatment-resistant depression in 2019. Despite these developments, uncertainty remains about comparative efficacy, tolerability and acceptability between intravenous racemic (R,S)-ketamine and intranasal esketamine, and no robust head-to-head clinical trials had been completed at the time of this review. Bahji and colleagues set out to synthesise the randomised controlled trial evidence to compare racemic ketamine and esketamine for unipolar and bipolar depression. The study aimed to estimate comparative effects on remission and symptom change, to examine safety and acceptability (including all-cause and adverse-event withdrawals), and to explore moderators such as treatment resistance, trial design and mode of administration using systematic review and meta-analysis methods registered on the Open Science Framework and performed in accordance with PRISMA guidance.

This is a systematic review and meta-analysis of randomised controlled trials assessing ketamine for adults (aged 18+) with primary unipolar or bipolar depression. The investigators searched MEDLINE, Embase, PsycINFO, CENTRAL and the Cochrane Database via Ovid up to 13 December 2019, and supplemented this with searches of ClinicalTrials.gov, the EU Clinical Trials Register, the Australian and New Zealand Clinical Trials Registry and reference lists. The review was limited to randomised trials; observational studies, mechanism reviews, commentaries and trials combining ketamine with neurostimulation were excluded. Both intravenous racemic ketamine and intranasal esketamine trials were eligible, whether administered as monotherapy or adjunctive to other psychotropic treatments. Two reviewers independently screened records and extracted data using a piloted Excel tool; disagreements were resolved by consensus. Extracted items included participant and study characteristics, dosing and route, outcomes, and withdrawals. Risk of bias within trials was assessed using the Cochrane tool for randomised trials and risk-of-bias plots were generated with RevMan. The authors registered the protocol and followed PRISMA. Seven prespecified outcomes were analysed: change in depression severity (continuous), response (≥50% reduction in score), remission (MADRS ≤12 or HAM-D ≤7), change in suicidality score (continuous), study completion, all-cause drop-outs, and drop-outs due to adverse events. Continuous outcomes were pooled as standardised mean differences (SMDs) and dichotomous outcomes as rate ratios (RRs). Heterogeneity was assessed with the I2 statistic and anticipated to be high, so random-effects models were used. Pairwise meta-analyses applied Mantel–Haenszel methods and a DerSimonian–Laird estimator for heterogeneity, implemented in R (meta package); a continuity correction of 0.5 was used for zero-event studies. For crossover trials, only the first-period data were used; when response rates were not reported, validated imputation methods for continuous-to-dichotomous conversion were applied. Publication bias was assessed via a weighted linear regression of treatment effect on the inverse of total sample size and adjusted using the trim-and-fill method when appropriate. Subgroup and meta-regression analyses examined moderators including formulation (intravenous racemic versus intranasal esketamine), treatment-resistance status, unipolar versus bipolar diagnosis, trial design (crossover versus parallel), monotherapy versus adjunctive use, and comparator type.

The search yielded 2,494 records; after deduplication 1,972 records were screened and 361 full texts reviewed, leaving 24 randomised controlled trials for inclusion. Across the 24 trials there were 1,877 participants with depression. Seven trials were crossover designs and the remainder were parallel. Most trials originated from the United States (71%) or Taiwan (13%). The vast majority of participants had unipolar major depression (n = 1,836; 97.8%), with 41 participants (2.2%) having bipolar-spectrum depression. Sample sizes varied widely (range 9–342), mean ages ranged from 35.9 to 70.0 years, and females comprised 60.7% (1,139/1,877). Most trials enrolled treatment-resistant depression (TRD) participants, though definitions of TRD varied across studies. Exclusion criteria commonly omitted psychotic disorders, serious medical conditions, severe substance use disorders and pregnancy or breastfeeding. Pooling across trials, ketamine (all formulations combined in the overall pairwise analysis) demonstrated a significant benefit versus control on key outcomes. Response rate was higher with ketamine, RR = 2.0382 (95% CI 1.5748 to 2.6380), and remission rate was also increased, RR = 2.0029 (95% CI 1.5005 to 2.6735). Depression severity decreased more with ketamine, SMD = -1.1430 (95% CI -1.4613 to -0.8247), and suicidality scores showed a small-to-moderate reduction, SMD = -0.3867 (95% CI -0.7082 to -0.0653). Study completion and overall drop-out rates were similar between groups: 147/1,011 (14.5%) dropped out from ketamine arms versus 141/980 (14.4%) from controls, RR = 0.97 (95% CI 0.72 to 1.29, p = 0.82). Across trials, 52 adverse-event discontinuations were recorded (37 in experimental arms, 15 in control arms); severe psychotic symptoms were not reported and cardiovascular side-effects were rare. When comparing formulations, intravenous racemic ketamine showed greater overall response and remission rates and fewer drop-outs due to adverse events relative to intranasal esketamine in subgroup/moderator analyses. Trial design influenced effect estimates: crossover trials tended to show larger response and remission effects, while parallel trials showed greater reductions in depression scores. No significant moderator effects were found for TRD status versus non-TRD, unipolar versus bipolar diagnosis, monotherapy versus adjunctive use, comparator type (placebo versus active), mean age or proportion female. Efficacy persisted across early timepoints up to four weeks post-treatment for response, remission and depression score change, though reductions in suicidality were not statistically significant at two- and four-week timepoints and appeared to wane over time. Publication bias assessments indicated significant small-study/publication bias for response, remission and depression rating scores; after trim-and-fill correction effect sizes were attenuated but remained in favour of ketamine: corrected response RR = 1.4209 (95% CI 1.0950 to 1.8438), remission RR = 1.5521 (95% CI 1.1472 to 2.1000), and depression SMD = -0.4832 (95% CI -0.8453 to -0.1212). Suicidality effect increased slightly after correction (SMD = -0.5034, 95% CI -0.8180 to -0.1888). Overall trial quality was judged to be high with only a few domains at high risk of bias.

Bahji and colleagues interpret their findings as indicating that intravenous racemic ketamine demonstrates greater short-term efficacy than intranasal esketamine for depression, with higher response and remission rates and fewer drop-outs due to adverse events in the pooled randomised-trial evidence available up to December 2019. They note that these superiority signals were most evident in early follow-up, with effects diminishing after approximately four weeks for some outcomes. The authors position their review as the first systematic comparison of racemic ketamine and intranasal esketamine, and they contrast the broader regulatory approval and larger long-term evidence base for esketamine with the off-label status and shorter-term data supporting racemic ketamine. Mechanistically, the discussion reiterates the central role attributed to NMDAR antagonism and downstream AMPA-mediated signalling and neuroplasticity, while acknowledging ongoing debate about contributions from opioid receptors and other systems; preclinical data on the R-enantiomer (arketamine) are noted as promising but clinical evidence remains limited. The investigators emphasise the clinical relevance of rapid anti-suicidal effects observed early after administration, while also recognising uncertainty about durability. Key limitations identified by the authors include evidence of publication bias that inflated some effect estimates, limited follow-up duration in included trials (typically up to four to eight weeks), unrepresentative trial populations that excluded common psychiatric and medical comorbidities, heterogeneity in patient samples (unipolar versus bipolar, TRD definitions), variation in dosing regimens (single versus repeated doses) and in whether ketamine was used as monotherapy or adjunctive therapy, and inconsistent reporting of specific adverse events such as dissociation. They acknowledge that subgroup and meta-regression methods can only partially account for heterogeneity and that head-to-head randomised trials are needed to confirm comparative efficacy and clarify safety profiles. The authors conclude that while racemic ketamine shows greater short-term efficacy in these pooled data, the lack of FDA approval for racemic ketamine, limited long-term evidence and practical considerations around route of administration mean more randomised controlled trials and longer follow-up are required before definitive clinical recommendations can be made.

Relative to intranasal esketamine, intravenous racemic ketamine demonstrated greater overall response and remission rates and fewer drop-outs due to adverse events in short-term randomised-trial data up to eight weeks. However, racemic ketamine remains off-label for depression in contrast to FDA-approved intranasal esketamine, and evidence on long-term benefits and safety is insufficient. The authors call for more randomised, head-to-head trials and longer-term studies to resolve comparative efficacy and safety questions across ketamine formulations.