Arketamine as adjunctive therapy for treatment-resistant depression: A placebo-controlled pilot study

Ketamine at subanaesthetic doses has been shown across many trials to produce rapid, though often transient, antidepressant effects in major depressive disorder (MDD), including treatment-resistant depression (TRD). Racemic ketamine contains two enantiomers, (R)-ketamine (arketamine) and (S)-ketamine (esketamine); esketamine is typically considered more active clinically because of higher NMDAR affinity and greater anaesthetic potency. Nonetheless, preclinical work has suggested that arketamine may have more potent and longer-lasting antidepressant effects with fewer psychotomimetic or abuse-related properties, and human data remain very limited to date. Leal and colleagues designed the present study to explore the feasibility of conducting controlled clinical trials of arketamine in patients with TRD and to provide an initial, exploratory comparison of its efficacy and safety against placebo. As the first randomised, double-blind, placebo-controlled trial using arketamine in depressed patients, the study aimed to test procedures, tolerability, and to generate preliminary efficacy signals that could inform future, better-powered trials.

This was a randomised, double-blind, placebo-controlled crossover pilot trial with ten participants conducted at a university hospital in Salvador, Brazil. No formal sample size calculation was performed because prior controlled human data for arketamine were lacking. Adults aged 18–65 years with DSM-5 major depressive disorder without psychotic features, a Montgomery–Åsberg Depression Rating Scale (MADRS) score ≥25 at screening, and a history of failure of at least two adequate antidepressant trials in the current episode were eligible. Participants were required to be on stable antidepressant medication for at least two weeks prior to enrolment. Exclusion criteria included psychotic disorders, current substance use disorders, significant cognitive impairment, pregnancy or breastfeeding, unstable medical illness, or depression better accounted for by another disorder. In a crossover format all participants received both intravenous interventions separated by a one-week interval: arketamine at 0.5 mg/kg and placebo (saline), each infused over 40 minutes. Randomisation determined the order of interventions and drug preparation was handled by a single investigator who did not participate in clinical assessments to preserve blinding. The arketamine used was produced from racemic ketamine to yield (R)-ketamine hydrochloride of approximately 100% purity and an R/S enantiomeric ratio >99.5:0.5. The primary efficacy measure was change in MADRS scores, assessed pre-infusion, at 60, 120 and 240 minutes after infusion start, and on days 1, 3 and 7 after each infusion. Response was defined as ≥50% reduction in pre-infusion MADRS and remission as MADRS ≤10. Secondary efficacy measures included Clinical Global Impression–Severity and –Improvement (CGI-S, CGI-I). Safety assessments included the Clinician-Administered Dissociative States Scale (CADSS) before and 40 minutes after infusion start, vital signs and sedation monitoring during and after infusion, and the Brief Psychiatric Rating Scale if psychotic symptoms emerged. Efficacy analyses used linear mixed effects (LME) models with individuals as random effects and fixed effects for treatment, time and order where appropriate. A carryover effect (treatment in the first period influencing the second) was explicitly tested with an LME model including treatment, order and their interaction. Post hoc pairwise comparisons used Tukey correction. Descriptive statistics summarised participant characteristics and safety data. Analyses were performed in R with two-sided tests at a 0.05 significance level.

Thirteen outpatients were screened between January 2020 and December 2021; ten met eligibility criteria, were randomised and received both interventions. The sample had chronic and severe illness: mean baseline MADRS was 36.1, mean CGI-S 5.1, the current episode duration averaged 31.0 months (SD 36.9), total illness duration 8.6 years (SD 13.2), and participants had failed a mean of 4.2 (SD 4.0) adequate antidepressant trials. Two participants had prior suicide attempts and none had a history of substance use disorder. An LME test for carryover detected a significant interaction between treatment and order (F[1, 118] = 10.54; p < 0.001), interpreted as evidence of a carryover effect; inspection suggested this arose largely from a single participant who maintained low MADRS scores after receiving arketamine in Phase 1. Because of this, the primary efficacy analysis was limited to Phase 1 only, giving five participants per arm, although analyses including both phases are also reported. In the Phase 1 (first-period) analysis the LME model showed a main effect of time on MADRS (F[6, 48] = 2.43; p = 0.038). Post hoc testing identified a significant reduction from baseline (mean = 36.1, 95% CI 29.9–42.3) to day 3 post-infusion (mean = 26.4, 95% CI 20.2–32.6), p = 0.01. There was no main effect of treatment (arketamine versus placebo; F[1, 8] = 0.78; p = 0.40) and no treatment-by-time interaction (F[6, 48] = 0.28; p = 0.95), indicating both groups improved over time to a similar degree. Effect size estimates (Cohen’s d) were large at some time points (notably days 3 and 7) but had wide confidence intervals that included negative values. When both phases were analysed together (all 10 participants receiving both interventions) results were similar and effect sizes were smaller. CGI-S and CGI-I showed no significant effects for time, treatment, or their interaction (CGI-S: time F[4, 81] = 0.921, p = 0.456; treatment F[1, 81] = 0.027, p = 0.870; CGI-I: all p > 0.10). Regarding categorical outcomes across both phases, no responders were observed at 60 or 120 minutes; at 240 minutes two participants responded to arketamine and none to placebo. On days 1 and 3 two responders were observed in the arketamine condition and one in placebo; at day 7 there was one responder in the arketamine group and none in placebo. Overall, five distinct participants met response criteria at least once after arketamine, one of whom also met response after placebo. One participant achieved remission after arketamine; another participant achieved response and remission after both interventions. Safety signals were minimal. Mean CADSS scores at the end of infusion were 0.6 for placebo (median 0) and 1.6 for arketamine (median 0); the highest CADSS scores recorded were 10 and 4, both following arketamine. Sixty per cent of participants had a CADSS score of zero after each intervention. Mean maximal systolic blood pressure increase from baseline was +5.3 mmHg after arketamine and +8.7 mmHg after placebo; diastolic increases were +8.0 mmHg after arketamine and +6.7 mmHg after placebo. Heart rate and oxygen saturation remained stable, and no clinically relevant sedation was observed. Four participants reported mild, transient adverse events during arketamine infusion (headache, numbness, somnolence, dizziness); two reported adverse events during saline (somnolence; anxiety, nausea and dizziness). No serious adverse events occurred.

Leal and colleagues conclude that it is feasible to conduct clinical trials of arketamine in academic settings, but that a crossover design may be suboptimal because of potential carryover and the uncertain duration of effect. Their exploratory analyses did not find evidence that a single 0.5 mg/kg intravenous infusion of arketamine produced greater antidepressant benefit than placebo in this small TRD sample; the primary Phase 1 analysis showed improvement over time but no treatment difference, and pooled Phase 1 and 2 analyses were consistent with that finding. The authors contrast these null findings with robust antidepressant effects seen in animal models and with results from their previous open-label pilot, which showed larger MADRS reductions and higher response rates; they attribute much of that discrepancy to expectancy and open-label effects. They also note relevant clinical differences between samples: the present participants were generally more chronic and had more psychiatric comorbidity (70% had comorbidities such as obsessive–compulsive disorder or anxiety disorders), factors known to worsen treatment outcome. Several methodological explanations for the negative result are discussed. The dose chosen (0.5 mg/kg) mirrored common racemic-ketamine practice, but the authors acknowledge that arketamine may require different dosing in humans to reproduce preclinical effects; higher doses could increase efficacy but might worsen tolerability and abuse-related risks. Repeated dosing might be necessary for response in some individuals, and crossover designs complicate evaluation of multi-dose regimens. The small sample means the study was underpowered to rule out clinically meaningful effects; the authors emphasise that such pilot data cannot establish absence of efficacy. Other limitations they note include lack of pharmacokinetic measurements, non-standardised concomitant medications, and absence of a formal assessment of the adequacy of blinding. They also raise the possibility that the relatively benign acute side-effect profile of arketamine could reduce unblinding and thus reduce placebo-controlled effect sizes compared with agents that produce more obvious subjective effects. Implications for future research include favouring parallel-group designs, exploring higher or flexible dosing and repeated administrations, and incorporating pharmacokinetics and blinding assessments. The authors point to ongoing and planned trials internationally that will address some of these questions and to the need for adequately powered Phase II studies to clarify arketamine’s antidepressant potential.

In this small pilot crossover trial involving ten patients with treatment-resistant depression, a single 0.5 mg/kg intravenous infusion of arketamine was not superior to placebo in reducing depressive symptoms. The study was underpowered to definitively assess efficacy but found an encouraging safety and tolerability profile with minimal dissociative or haemodynamic effects. The investigators recommend further adequately powered trials, likely using parallel designs, that consider higher or flexible doses and repeated administrations to determine whether arketamine has clinically meaningful antidepressant effects in humans.