Ketamine for Treatment-Resistant Unipolar and Bipolar Major Depression: Critical Review and Implications for Clinical Practice

The paper is a narrative, critical review that synthesises findings from multiple meta-analyses and individual clinical trials of ketamine for treatment-resistant MDD and BP depression. Throughout the text the authors reference pooled analyses, individual randomised controlled trials (including single- and repeat-infusion studies), small open-label series, and case reports to evaluate efficacy, dosing, safety, and comparative data versus electroconvulsive therapy (ECT). The extracted text does not provide a distinct methods section describing systematic search strategies, explicit inclusion/exclusion criteria for studies selected for review, formal risk-of-bias assessment, or quantitative meta-analytic methods applied by the authors themselves. Instead, the review summarises results reported in existing meta-analyses and primary studies, and integrates trial-level findings (for example, number of trials, sample sizes, dosing regimens, and outcome time points) when these details were reported in the source material. Where trial designs are discussed, the review highlights key study features relevant to interpretation: most controlled trials evaluated single low-dose (commonly 0.5 mg/kg IV) or very low-dose ketamine (0.1–0.4 mg/kg IV or intranasal 50 mg) with short-term follow-up (up to about 14 days), and a smaller number of studies examined repeated/serial infusions (typically up to six infusions over 12–14 days). The authors draw on pooled results from up to nine meta-analyses and individual trial reports to comment on effect sizes, response and remission rates, and adverse-event profiles.

Across multiple short-term randomised trials and at least nine meta-analyses, ketamine produced rapid antidepressant effects relative to placebo or active controls, with most consistent efficacy observed within hours to 24 hrs after administration and generally persisting to about 7 days. Pooled effect sizes were large at 24 hrs and smaller-to-moderate by day 7. Reported response rates across controlled and uncontrolled studies were highly variable, ranging from 29% to 90% in cumulative reports. Xu et al. reported pooled response rates of 50% with ketamine versus 13% with control at 24 hrs (risk ratio 2.6, 95% CI: 1.6–4.4), and 31% versus 7% at 7 days (RR 3.4, 95% CI: 1.6–7.1); pooled remission rates were 28% versus 3% at 24 hrs (RR 5.2, 95% CI: 2.1–12.9) and 24% versus 6% at 7 days (RR 2.6, 95% CI: 1.2–5.7). Most single-dose responders relapsed within 2 weeks. Meta-analytic and trial-level data indicate diminished between-group benefit by 14 days in many analyses, although some reports found a persisting effect at 12–14 days. When serial infusion strategies were examined, repeated ketamine infusions (commonly six infusions over 12–14 days) produced larger reductions in depressive symptoms than single infusions at multiple time points (for example, standardised mean differences reported as around -4.2 vs -1.1 at 24 hrs in one pooled comparison). Two controlled trials of approximately thrice-weekly infusions (up to six treatments at 0.5 mg/kg) reported first-infusion response rates of 71–89% that were sustained during the repeated-infusion period; nevertheless, relapse remained common, with mean or median time to relapse of about 18–19 days after the final infusion. Small open-label series reported cumulative remission rates increasing over successive infusions (for example, 10%, 40% and 50% after one, two and four infusions in one study), but many remitters relapsed within weeks. Dose-response evidence was limited but suggestive. A pooled analysis of nine trials (201 subjects) found larger symptom reductions with the conventional low dose (0.5 mg/kg IV) compared with very low doses (0.1–0.4 mg/kg or intranasal 50 mg) at 24 hrs (SMD -1.1, 95% CI: -1.7 to -0.6) and at 3 days (SMD -0.8, 95% CI: -1.4 to -0.3), with a trend favouring low dose at 7 days. Small dose-escalation and open studies showed some responses at lower doses and some greater improvement at higher doses, but all responders in these small samples relapsed within about a week or two. Subgroup analyses reported in multiple meta-analyses showed large reductions in depressive symptoms at 24 hrs in both MDD and BP depressed patients. Some meta-analyses found numerically greater effects in MDD than BP at 24 hrs, with mixed findings at 7 days and no consistent evidence of heterogeneity between diagnostic subgroups at early time points. Regarding suicidality, pooled analyses of seven trials that used suicide items from depression rating scales found significant reductions in suicide item scores with ketamine versus controls at 24 hrs and 3 days, but not at 7 days. One trial using the Beck Scale for Suicidal Ideation reported lower scores with ketamine than midazolam at 48 hrs but not at 72 hrs or 7 days. No studies reported effects on suicide attempts or suicide deaths. Short-term tolerability in controlled trials was generally acceptable: commonly reported adverse effects included transient increases in blood pressure and heart rate, dry mouth, headache, anxiety, confusion and dissociation, typically resolving within about 2 hrs after infusion. A pooled safety dataset (158 ketamine-treated subjects in one meta-analysis) reported study discontinuation rates of 13% for ketamine versus 7% for controls (odds ratio 2.0, 95% CI: 0.9–4.4, P = .11). Psychotomimetic and dissociative symptoms were significantly greater with ketamine than controls: BPRS positive symptom subscale effect (Hedge's g = 0.8, P < .0001) and Clinician-Administered Dissociative States Scale effect (Hedge's g = 1.78, P < .001). A small open study suggested slower infusions (100 min) might reduce hemodynamic changes, but data are preliminary. Direct comparisons with ECT are scarce. One small open parallel-group study (n = 18) reported greater short-term symptom improvement with ketamine than bilateral ECT after the first and second treatments, but no randomised head-to-head trials of ketamine versus an adequate acute ECT course were identified. A meta-analysis of five studies (182 patients) examining ketamine as an augmentation to ECT found no significant advantage in depressive symptom improvement, response, or remission rates, but did find higher rates of disorientation or prolonged delirium with ketamine augmentation (OR 6.59, 95% CI: 1.28–33.82). Longer-term effectiveness and safety data are extremely limited. Evidence for maintenance strategies comes only from case reports and small case series in highly treatment-resistant patients, with no controlled continuation- or maintenance-phase studies identified. Major long-term safety concerns are highlighted by preclinical data and adverse outcomes reported in chronic recreational users, including neurocognitive dysfunction, urinary cystitis, and brain structural changes, but few studies have assessed these outcomes in patients treated for depression. The potential for ketamine abuse or dependence is emphasised; although controlled short-term trials have not reported incident abuse, trial populations were carefully selected and exposed only briefly, limiting inference about long-term addiction risk in routine practice.

Bobo and colleagues interpret the accumulated evidence as supporting measured enthusiasm for ketamine's therapeutic potential in treatment-resistant unipolar and bipolar depression, particularly for very short-term relief. They underscore that ketamine produces a clear and clinically significant antidepressant effect that emerges within hours and commonly lasts up to about 7 days after a single dose. The authors note that responsiveness appears broadly similar across MDD and BP depression subgroups in the short term, and that the possibility of rapid reductions in suicidal ideation has received attention, although evidence for sustained antisuicidal effects is lacking. Important limitations and uncertainties are emphasised. Generalisability is a major concern because most trials were conducted in resource-rich research settings with extensive monitoring and excluded patients with substance-use disorders, psychotic features, unstable medical illness, or imminent suicide risk; between 21% and 67% of screened subjects were excluded in the trials reviewed. The brief duration of benefit is identified as a critical limitation: most responders relapse within 2–3 weeks and even serial infusion strategies have not yet demonstrated durable maintenance of benefit. The authors stress that there are no established, evidence-based methods to sustain ketamine's initial antidepressant effects over the longer term, and that adjunctive strategies tested to date (for example, riluzole or D-cycloserine following ketamine) have only preliminary proof-of-concept data and are not consistently effective. Safety uncertainties constitute another central concern. Short-term trials show mainly mild, transient adverse effects, but the overall safety dataset is small (for example, pooled analyses involving a few hundred short-term exposures) and insufficient to detect less frequent or delayed harms. The authors draw attention to risks suggested by preclinical studies and findings in chronic nonmedical ketamine users—neurocognitive impairment, urinary tract pathology and adverse brain changes—and note the real potential for abuse, dependence and tolerance. They caution that exclusion of higher-risk patients from clinical trials and the small number of exposures per participant limit the ability to detect these problems. Finally, Bobo and colleagues raise concern about the lack of formal regulatory or systematic pharmacovigilance monitoring in many settings where off-label ketamine therapy is already being provided, and they argue that passive, voluntary adverse-event reporting is inadequate for a treatment with uncertain long-term risks. As implications for practice and research, the authors recommend caution: routine adoption outside research should be measured, informed consent should explicitly address limited long-term evidence and potential risks, and careful screening, monitoring and follow-up are necessary. They call for continuation- and maintenance-phase trials and for studies of ketamine in naturalistic cohorts with standardised protocols to generate preliminary longer-term effectiveness and safety data.

The authors conclude that there is insufficient empirical support for the early, widespread adoption of ketamine into routine clinical practice for depression. Despite clear potential for significant short-term benefit in treatment-resistant patients, longer-term effectiveness and safety remain inadequately characterised, and ketamine is not FDA‑approved for depression. Given current off-label use, clinicians should apply careful screening, management and follow-up, obtain informed consent that addresses uncertain longer-term risks and possible nonresponse, and consider participation in or collection of standardised naturalistic cohorts or formal trials to better define maintenance strategies, safety, and effectiveness over time.