Safety, Pharmacokinetics, and Pharmacodynamics of a 6-h N,N-Dimethyltryptamine (DMT) Infusion in Healthy Volunteers: A Randomized, Double-Blind, Placebo-Controlled Trial

Van D Heijden and colleagues frame the study around the problem of post-stroke disability, emphasising that cortical neuroplasticity is a key mechanism underlying motor recovery and that pharmacological agents which promote plasticity during a putative early “critical period” after stroke could be therapeutically valuable. They note preclinical and some human evidence that serotonergic psychedelics (acting at 5-HT2A receptors and Sigma1R) can enhance dendritic complexity and synaptogenesis, and that prolonged exposure in vitro (for example 6 h) produced more sustained neuroplastic changes than shorter exposures. DMT is highlighted as a candidate because it is a potent 5-HT2A and Sigma1R agonist and, when given intravenously, can be maintained at target plasma concentrations despite rapid metabolic clearance when given orally. The paper sets out to characterise the clinical pharmacology of a prolonged intravenous DMT regimen in healthy volunteers before moving to patient studies. Specifically, the investigators tested a 30-s bolus followed by a 6-h infusion at dose levels predicted to remain below a putative psychedelic plasma threshold, with the principal aims of assessing safety, pharmacokinetics (PK), and pharmacodynamics (PD) to inform future proof-of-mechanism work in patient populations such as people recovering from ischaemic stroke. The study thus focuses on tolerability, systemic exposures, subjective and objective CNS effects, and exploratory biomarkers relevant to neuroplasticity.

This was a randomised, double-blind, placebo-controlled, single ascending dose study conducted in three planned cohorts of ten healthy volunteers each (mixed psychedelic-experienced and -inexperienced, aged 18–60). Inclusion and exclusion criteria were standard for safety (for example excluding hypertension, cardiovascular disease, seizure disorders, recent substance abuse, current MAO inhibitor use, pregnancy), and participants underwent screening and a preparatory information session with a psychedelic guide prior to dosing. Randomisation within cohorts was computer-generated in an 8:2 ratio (DMT:placebo). The extracted text reports 29 participants randomised (12 female, 17 male); one participant in the highest planned dose cohort was excluded prior to administration and not replaced. DMT fumarate was administered as a 30-s intravenous bolus immediately followed by a continuous 6-h infusion. Planned dose levels were simulated from prior population PK work to achieve sub-psychedelic exposures relative to an assumed threshold at the time (about 45 ng/mL following a 0.2 mg/kg IV bolus). The final dosing regimens reported were cohort 1: 1.5 mg bolus + 0.105 mg/min infusion; cohort 2: 7.5 mg bolus + 0.525 mg/min infusion; cohort 3: 5.0 mg bolus + 0.7875 mg/min infusion (the bolus was adjusted downward for cohort 3 based on interim safety/PD data while maintaining the planned infusion rate). Safety assessments included vital signs (SBP, DBP, pulse, temperature, respiratory rate), 12-lead ECG, and clinical ratings for suicidality and psychopathology. PD and subjective measures comprised a broad battery: the real time intensity scale (RTIS), three visual analogue scales (VAS Bond and Lader; VAS Bowdle; VAS drug rating), the 11-Dimensions Altered States of Consciousness (11D-ASC) and the Hallucinogen Rating Scale (HRS). Neurophysiological and functional CNS domains were assessed with the NeuroCart battery (saccadic peak velocity, smooth pursuit, postural stability/body sway, adaptive tracking) and EEG. Blood sampling captured PK time courses, serum/plasma BDNF, cortisol and prolactin, and pre-dose MAO-A enzyme activity. PK was analysed by noncompartmental methods (PKNCA in R v4.0.3), with AUC calculated using linear-up/log-down and half-life derived by linear regression in the terminal phase. Repeated PD measures up to 24 h were analysed with mixed effects models (treatment, time, treatment-by-time fixed factors; participant random factor; pre-dose average as covariate) in SAS9.4. Single-measure PD scales were analysed by ANCOVA. Exploratory post hoc analyses examined sex differences descriptively. The investigators did not correct p-values for multiple testing and considered p < 0.05 statistically significant.

Participants and dosing: Twenty-nine participants (12 female, 17 male) were randomised; one subject in cohort 3 was excluded prior to administration. Cohorts were dosed with the three ascending regimens described above. No participants dropped out following dosing. Safety and tolerability: All adverse events (AEs) were self-limiting and generally mild to moderate. The most frequent AEs with DMT were headache, nausea and fatigue; fatigue was the most common AE under placebo. One 18-year-old psychedelic-naïve female experienced anxiety and panic attacks the day after the cohort-2 infusion (7.5 mg bolus + 0.525 mg/min); symptoms subsided after counselling over a period of up to 6 months. Dose-dependent increases in mean systolic blood pressure of approximately 10 mmHg (dose 2) and 20 mmHg (dose 3) were observed, with maximum individual increases of 39 and 25 mmHg respectively; mean diastolic pressure increased by 9 mmHg for dose 3. All blood pressure elevations returned to baseline within 1 hour after stopping the infusion. No serotonergic toxicity, clinically significant ECG changes, suicidality or treatment-emergent psychopathology (BPRS, CSSRS) were reported. Pharmacokinetics: DMT plasma exposure generally increased with dose. Mean AUC_last (±SD) was 14.2 (6.8), 85.9 (21.6) and 157.6 (55.3) h*ng/mL for doses 1–3, respectively. Mean C_max (±SD) was 4.6 (2.9), 25.3 (24.1) and 35.9 (34.0) ng/mL for doses 1–3. Mean terminal half-life ranged from 0.2 to 0.3 h. The apparent volume of distribution during terminal elimination was large (431–616 L) and steady-state volumes were very large (3,834–6,516 L). T_max showed interindividual variability (mean T_max 0.9 h for doses 1 and 2, 1.9 h for dose 3, with wide absolute ranges). Coefficients of variation for C_max, AUC_last and t_1/2 spanned approximately 25%–62%, indicating moderate to high intersubject PK variability. No robust sex differences were observed overall, though females had slightly higher plasma concentrations than males at dose 3 in descriptive analyses. Subjective pharmacodynamics: Compared with placebo, dose 3 produced statistically significant increases on RTIS dimensions of bodily, emotional/metacognitive and visual intensity; RTIS effects peaked rapidly after the bolus and generally decreased over time except for RTIS emotional intensity which remained elevated during infusion. Dose 2 produced minimal RTIS effects. On VAS measures, dose 3 increased VAS Bond and Lader mood and calmness and decreased alertness; dose 2 showed smaller increases on some VAS Bowdle and drug-liking items. On the 11D-ASC, dose 2 produced low mean scores (7.0%–13.1% across subscales, higher for elementary imagery), whereas dose 3 elicited higher mean scores for complex imagery (39.8%), elementary imagery (37.9%) and blissful state (31.9%) with low anxiety and disembodiment scores. HRS subscales showed minimal effects at dose 1 and small-to-moderate increases at doses 2 and 3, with intensity ratings highest for the two higher exposures. Objective neurophysiological and performance measures: DMT at doses 2 and 3 decreased adaptive tracking performance (mean changes −4.2% and −4.9% respectively) and increased body sway substantially (+59.9% and +56.6% over baseline for doses 2 and 3). Pupil iris ratio increased by +0.07 for dose 3 and remained elevated throughout the infusion. EEG showed statistically significant suppression of parieto-occipital alpha power for dose 3 and an increase in central gamma power for dose 2; the gamma finding was judged possibly spurious or muscle-confounded. Saccadic peak velocity and simple reaction time were unaffected. Serum cortisol, prolactin and plasma/serum BDNF did not differ from placebo at any dose. MAO-A and PK correlations: Pre-dose MAO-A activity did not correlate significantly with DMT half-life, dose-normalised AUC_inf or dose-normalised AUC_last (Pearson r values close to zero; p-values not significant), though the authors note limited sampling and sample size as constraints.

Van D Heijden and colleagues interpret the findings as evidence that a 30-s bolus followed by a 6-h intravenous DMT infusion at the tested regimens is generally safe and well tolerated in healthy volunteers, producing mean peak plasma exposures from roughly 5 to 36 ng/mL. They note that negligible subjective psychedelic effects were observed at mean C_max ≈ 25 ng/mL, whereas mild but detectable effects accompanied EEG alpha suppression at ≈ 36 ng/mL. All AEs were self-limiting except for one participant who developed anxiety requiring integration counselling; this case emphasises the need for careful screening and follow-up. The investigators highlight that cardiovascular effects were mild (mean SBP increases of 10–20 mmHg at higher exposures) and reversible within an hour of infusion cessation; no serotonergic toxicity or clinically meaningful ECG changes occurred. They also discuss the moderate-to-high interindividual PK variability observed (%CVs ≈ 20%–60%) and suggest possible contributors including metabolic variation (MAO-A and CYP2D6), differences in body composition, assay variability and procedural factors; however, pre-dose MAO-A activity did not correlate with PK in this dataset, and CYP2D6 genotyping was not performed. The authors propose that the observed subjective effects and their time courses suggest that the bolus produced rapid peaks associated with larger transient subjective intensity, and that an infusion-only regimen without a loading bolus might better achieve a sustained sub-psychedelic exposure in future patient studies. Regarding biomarkers and mechanisms, the lack of change in peripheral BDNF is emphasised as not conclusive evidence that DMT cannot promote plasticity at these exposures, because peripheral BDNF has methodological limitations and may not reflect central TrkB engagement. The authors therefore recommend additional work to characterise CSF and peripheral PK, and to align human exposures with preclinical models of neuroplasticity. They acknowledge important limitations: the small sample size, the controlled research setting (which may modulate subjective effects), heterogeneity in psychedelic experience among participants, limited MAO-A sampling, and absence of CYP2D6 genotyping. Finally, they suggest the present safety, PK and PD data justify progression to proof-of-mechanism studies in patient populations, while urging exploration of sources of PK variability and consideration of effect-based administration strategies (for example titration or clamping) if variability cannot be reduced.