Rapid and sustained antidepressant effects of vaporized N,N-dimethyltryptamine: a phase 2a clinical trial in treatment-resistant depression

Major Depressive Disorder is highly prevalent worldwide and a substantial minority of patients meet criteria for treatment-resistant depression (TRD). Conventional antidepressants typically require several weeks to show benefit, leaving patients vulnerable during that latency period and motivating research into rapid-acting interventions. Serotonergic psychedelics such as psilocybin and ayahuasca have produced rapid antidepressant signals in prior studies, but their multi-hour durations and the requirement for monoamine-oxidase inhibitors in some preparations (e.g. ayahuasca) present practical, safety and cost challenges for broader clinical implementation. This paper reports a Phase 2a, open-label investigation of inhaled (vaporized) N,N-dimethyltryptamine (DMT) in patients with TRD. The authors build on prior phase 1 work that characterised safety and subjective effects of vaporized DMT in healthy volunteers and set out here to evaluate feasibility, safety, tolerability and antidepressant efficacy of a two-step inhalation procedure (15 mg followed by 60 mg) with structured preparation and integration, hypothesising rapid and sustained reductions in depressive symptoms.

Falchi-Carvalho and colleagues conducted a Phase 2a, open-label, single-ascending fixed-order dose-escalation trial in adults with unipolar TRD. The protocol was approved by a university ethics committee and registered on ClinicalTrials.gov. Participants were recruited via advertisements and clinician referrals and had to be aged 18–60 years, meet TRD criteria by psychiatrist assessment supported by the MINI interview, be experiencing a current moderate-to-severe episode (MADRS ≥ 20 for ≥ 4 weeks) and have failed at least two antidepressant treatments during the current episode. Detailed inclusion/exclusion criteria are reported in the Supplementary Materials and Methods. DMT freebase (purity > 96%) was isolated from Mimosa tenuiflora root bark, dissolved in ethanol and administered via a medical-grade vaporizer (Volcano Medic 2). The study took place in a clinical research unit prepared to provide psychological support. After baseline screening and preparation (including an in-person session 7 days before dosing), each participant underwent two inhalation sessions on the dosing day (D0): an initial safety check with 15 mg DMT followed 90 minutes later by 60 mg DMT, provided the patient and clinicians judged it safe to proceed. Sessions included eyeshades, music for the first 15 minutes, and psychological support during acute effects and structured integration afterwards. Blood samples were taken pre-first session (-20 min) and post-second session (+120 min). Outcome assessments occurred at D0, D1, D7, D14, 1 month (M1) and 3 months (M3). The primary outcome was change in depression severity on the Montgomery-Åsberg Depression Rating Scale (MADRS) from baseline to D7. Secondary outcomes included MADRS changes at other time points, response (≥ 50% reduction) and remission (MADRS ≤ 10) rates, changes in PHQ-9, MADRS-SI and Beck Scale for Suicide Ideation (BSI), physiological and biochemical measures during acute effects, and several psychedelic experience scales (VAS, ASC, MEQ, HRS). Statistical analyses used repeated-measures General Linear Models with time as the within-subjects factor and post hoc Dunnett comparisons versus baseline. Effect sizes were reported with Hedge's g. Biochemistry used paired t-tests; subgroup comparisons (antidepressant use versus none) used Mann–Whitney tests. Spearman correlations examined relationships between acute subjective measures and MADRS change, with Bonferroni correction for multiple comparisons. Significance was set at α = 0.05.

Of 23 patients screened, 14 were enrolled and treated between 9 October 2023 and 30 March 2024. No participant experienced critical physiological changes during the initial low dose and all chose to proceed to the higher 60 mg inhalation. Primary psychiatric outcomes: There was a significant main effect of time on MADRS (F(5,65) = 13.04, p < 0.0001, ηp² = 0.50). Pairwise comparisons indicated significant reductions versus baseline at D1 (mean change 8.93 ± 10.45, p < 0.001; Hedge's g = 2.76), D7 (11.07 ± 12.33, p < 0.001; g = 2.20), D14 (11.00 ± 10.68, p < 0.001; g = 2.47), M1 (14.21 ± 12.08, p < 0.001; g = 1.90) and M3 (18.79 ± 13.07, p < 0.001; g = 1.33). Reported response rates (≥ 50% MADRS reduction) were 78.57% (11/14) at D1, 85.71% (12/14) at D7 and D14, 64.29% (9/14) at M1 and 57.14% (8/14) at M3. Remission rates (MADRS ≤ 10) were 71.43% at D1, 57.14% at D7 and D14 and M1, and 35.71% at M3. Subgroup and self-report outcomes: PHQ-9 scores followed a similar time course with a significant main effect of time (F(5,65) = 10.78, p < 0.001). The investigators compared patients taking antidepressants (AD) with those not taking them and found marginally larger MADRS decreases in the AD group at D7 (Mann–Whitney U = 7.5; p = 0.07). Response/remission at D7 were 100% and 70% respectively in the AD group versus 50% and 25% in the no-AD subgroup. Suicidality: MADRS-SI showed a significant main effect of time (F(5,65) = 9.97, p < 0.001, ηp² = 0.43) with significant reductions at all post-dose time points (e.g. D1 mean 0.36 ± 0.63, p < 0.001; g = 2.61). At baseline 43% of patients had severe suicidal ideation and 43% had ideation of lesser severity; by D1 no patient had severe ideation, and at M3 only 21% had any suicidal ideation. BSI scores also decreased significantly over time (F(5,65) = 3.70, p = 0.005). Correlations linked reductions in suicidality to improvements in non-suicide MADRS items: MADRS-SI reductions correlated with MADRS-nonSI (r = 0.84, p = 0.0004) and BSI reductions correlated with MADRS-nonSI (r = 0.75, p = 0.004). Subjective experience and correlations: High-dose subjective intensity on a VAS averaged 84.37 ± 21.49 with positive valence noted; ASC total averaged 31.64 ± 14.55 and MEQ total 2.41 ± 1.13. There was no difference in subjective intensity between patients on antidepressants and those not (p = 0.98). A marginal positive correlation was observed between MADRS improvement at D7 and the ASC Complex Imagery factor (r = 0.72, p = 0.057); no correlations survived correction for multiple comparisons beyond this trend. Physiology and biochemistry: Acute increases in systolic and diastolic blood pressure and heart rate occurred after the second administration but were transient: SBP increases were significant from +2 to +8 min, DBP from +2 to +6 min, and HR from +2 to +4 min, with values returning to baseline by +30 and +120 min. SpO2 and respiratory rate did not change significantly. Two participants were excluded from blood analyses due to venous access problems; among the remaining 12, small but statistically significant post-dose changes included increases in ALT, AST, total cholesterol, HDL, LDL and estimated GFR, and a decrease in triglycerides; mean values remained within population reference ranges. Safety and adverse events: No serious adverse events were reported. Across time points 77 distinct medical concepts were recorded, predominantly mild and transient. The most frequent acute complaints were pharyngeal discomfort (reported in 57.1% of sessions on D0), cough and headache. Most adverse events were confined to the acute dosing phase and resolved in follow-up assessments.

Falchi-Carvalho and colleagues interpret their findings as evidence that a single-day, two-step inhaled DMT regimen produced rapid and sizeable antidepressant and antisuicidal effects in this small open-label TRD sample, with significant reductions apparent at D1 and sustained, on average, to the 3-month follow-up. They note large effect sizes on MADRS at early time points and response/remission rates that compare favourably with some published results for other rapid-acting treatments, while acknowledging differences in design, dosing and measures across studies. The authors situate their results alongside prior reports of IV DMT and vaporized 5-MeO-DMT, observing similar rapid-onset antidepressant signals. They highlight several potential reasons for the magnitude of effects observed: the higher inhaled dose used here compared with some IV studies, the structured preparation and integration, and setting-related factors that could augment outcomes. Subjective experiences were intense but generally well tolerated; mystical-type experiences were uncommon, and imagery-related phenomena showed a marginal association with clinical improvement. The investigators point out that the observed reductions in suicidal ideation were largely correlated with overall mood improvement, suggesting a mood-dependent mechanism rather than a distinct antisuicidal effect. With respect to safety, the authors report transient, modest increases in blood pressure and heart rate during the acute phase without clinical sequelae and no clinically meaningful biochemical abnormalities. Pharyngeal irritation related to inhalation was the most frequent adverse event but was short-lived. Nonetheless, they acknowledge substantial limitations: the open-label design and small sample size preclude causal inference and accurate estimation of effect magnitude; lack of pharmacokinetic data prevents dose–concentration interpretation; TRD characterisation relied in part on self-report for some participants; the sample was demographically homogeneous and limited to ages ≤ 60 years; and cardiac safety monitoring was not comprehensive (for example no serial 12-lead ECG QT assessments were reported). The authors recommend that future research employ larger, diverse, randomised placebo-controlled designs, include pharmacokinetic and more comprehensive cardiac safety evaluations, assess repeated or cumulative dosing strategies to prolong or enhance response, extend integration support, and investigate predictors of both beneficial and adverse outcomes. They conclude that inhaled DMT represents a non-invasive, time-efficient intervention with potential for outpatient delivery and scalability, which may be practical within interventional psychiatry frameworks, but underscore the need for confirmatory randomised trials to validate efficacy and safety.