Single-dose psilocybin therapy for alcohol use disorder: Pharmacokinetics, feasibility, safety and efficacy in an open-label study

Interest has grown in the therapeutic potential of classic psychedelics that act as serotonin 2A receptor agonists, and psilocybin in particular has shown promise for treating alcohol use disorder (AUD). Earlier work includes a meta-analysis of single-dose LSD trials from the 1960s–1970s suggesting long-term benefits for alcohol misuse, and recent studies with psilocybin have demonstrated efficacy using two dosing sessions. However, several clinical studies in depression have reported positive outcomes after a single dose, and pharmacokinetic/pharmacodynamic (PK/PD) data in patients with AUD are sparse, despite concerns that chronic alcohol use may alter drug metabolism and subjective responsiveness. This study by Jensen and colleagues set out to evaluate a single fixed 25 mg oral dose of psilocybin given within a structured therapeutic course for people with severe AUD. The investigators aimed to characterise feasibility, safety, pharmacokinetics and pharmacodynamics, and to assess short- and medium-term clinical outcomes over 12 weeks following a single dosing session, addressing the gap in single-dose trials and PK/PD data in an AUD population.

This was an investigator-initiated, open-label, single-group 12-week study conducted at the Copenhagen University Clinic for Psychedelic Research. Ten treatment-seeking adults (eight men, two women; median age 44 years; median body weight ~80.0 kg) were enrolled between March and June 2023 and followed until September 2023. Recruitment was by self-referral and screening reduced 37 initial enquiries to 10 participants. All participants provided informed consent and the trial had ethics and regulatory approvals. Inclusion required DSM-5 AUD and ICD-10 alcohol dependence, AUDIT ≥ 15, at least five heavy drinking days in the prior 28 days (defined as ≥ 60 g for men or ≥ 48 g for women on any day), and body weight 60–95 kg. Key exclusions were a personal or first-degree family history of psychotic disorders, current suicidal ideation or past suicide attempt, other drug use disorders (except nicotine), current medications for AUD or serotonergic drugs, significant cardiovascular disease (including prolonged QTc or uncontrolled hypertension). Participants received no monetary compensation. The intervention was a five-session psilocybin therapy course delivered by a psychotherapist and a medical doctor: two preparation sessions, one dosing session with a fixed oral 25 mg psilocybin capsule, and two post-dose integration sessions (day after and one week later). Therapeutic elements included motivational interviewing, acceptance and commitment therapy, and guided imagery and music. On dosing day participants arrived after a light low-fat meal and 24-hour abstinence from caffeine and alcohol; medical checks (BrAC, urine drug screen, CIWA-R) were performed, a peripheral venous catheter inserted, and a standardised 360-minute music programme used during the session. Subjective drug intensity (SDI) was rated every 20 minutes on a 0–10 Likert scale. Mystical-type experience was assessed post-session using the MEQ30. Pharmacokinetic sampling collected plasma before dosing and at 40, 60, 80, 100, 120, 140, 180, 240 and 360 minutes post-dose. Free, unconjugated psilocin was measured by ultra-high-performance liquid chromatography–tandem mass spectrometry and reported in µg/L. Noncompartmental PK parameters (Cmax, Tmax, half-life, AUC) were calculated. The PPC–SDI relationship was modelled using a sigmoid Emax model with SDI bounded 0–10 and Spearman correlations used to quantify associations. Clinical outcomes included alcohol consumption by Timeline Followback at weeks 4 and 12, AUDIT at week 12, craving by the Penn Alcohol Craving Scale (PACS) and measures of self-efficacy, mindfulness, psychological flexibility and depressive symptoms at weeks 1, 4 and 12. Statistical analysis used linear mixed models with unstructured covariance for repeated measures, Cohen's d for effect sizes, and exploratory correlations and group comparisons. Missing clinical outcome data due to relapse were replaced with the participant’s worst observed value; missing PPC and SDI values were handled by last observation carried forward. Analyses were performed in R using relevant PK and mixed-model packages.

All ten participants completed the five-session therapy course. Pharmacokinetic analysis revealed marked between-participant variability in plasma psilocin concentrations after the single 25 mg dose: peak concentrations (Cmax) ranged approximately from 14 to 59 µg/L. Two participants were highlighted for atypical kinetics—one with a high Cmax and another with delayed absorption—yet neither showed divergent clinical outcomes. Temporal profiles of SDI closely tracked plasma psilocin concentration curves across participants, and the investigators modelled this relationship with a sigmoid Emax approach; the extraction does not report parameter estimates such as PPC50 or γ. Exploratory analyses suggested that mystical-type experiences were associated with sustained reductions in heavy drinking days, whereas plasma psilocin concentrations themselves did not appear to predict clinical improvement. Primary clinical outcomes showed statistically and clinically meaningful reductions in alcohol use over 12 weeks. Percentage heavy drinking days fell by 37.5 percentage points from baseline to week 12 (95% CI −61.1 to −13.9, p = 0.005; Cohen's d = 1.02). Drinks per day decreased by 3.4 units (95% CI −6.5 to −0.3, p = 0.035; Cohen's d = 0.71). Alcohol craving measured by PACS decreased markedly at week 1 (mean change −8.9 points, 95% CI −14.0 to −3.7, p = 0.003, Cohen's d = 0.97), with sustained reductions at week 4 (−5.9, 95% CI −10.56 to −1.23, p = 0.018, d = 0.77) and a borderline change at week 12 (−5.1, 95% CI −10.2 to 0.06, p = 0.052, d = 0.61). Self-efficacy increased substantially: +17.6 points at week 1 (95% CI 4.3 to 30.8, p = 0.015, d = 2.39), +18.7 at week 4 (95% CI 9.44 to 27.96, p = 0.001, d = 2.72), and +18.5 at week 12 (95% CI 9.34 to 27.65, p = 0.001, d = 2.67). Measures of trait mindfulness, psychological flexibility and depressive symptoms showed no significant change from baseline to week 12. Individual-level data indicated that nine of ten participants had significantly reduced drinks per day at week 4 and seven of ten maintained significant reductions at week 12. Safety and feasibility outcomes were favourable: there were no serious adverse events reported, cardiovascular monitoring (blood pressure and heart rate) alongside repeated blood sampling was feasible and well tolerated in this psychiatric patient group. The authors note that missing clinical outcome data were imputed using worst-observed values for those who relapsed, and missing PK/SDI values were carried forward, as reported in the methods.

Jensen and colleagues interpret the findings as evidence that a single fixed 25 mg psilocybin dose delivered within a structured therapeutic course is feasible, well tolerated and associated with reductions in alcohol consumption and craving over a 12-week follow-up in this small open-label sample of treatment-seeking adults with severe AUD. Pharmacokinetic analyses revealed notable inter-individual variation in psilocin exposure, but these differences did not appear to compromise safety or overall clinical response. The close temporal correspondence between plasma psilocin concentrations and subjective drug intensity supports the expected PK–PD coupling; however, plasma concentrations were not clearly linked to medium-term drinking outcomes in the exploratory analyses. The investigators situate their results within accumulating evidence that subjective, including mystical-type, experiences during psychedelic sessions may mediate therapeutic benefit; their exploratory analyses suggested an association between profound mystical-type experiences and sustained reductions in heavy drinking days. They also highlight the practical importance of demonstrating that repeated blood sampling for PK/PD work is feasible in an AUD clinical population, opening opportunities to investigate biological mechanisms such as neuroplasticity and inflammation in future work. Key limitations acknowledged by the authors include the small sample size, limited representation of women (two of ten), the open-label single-arm design without a control or placebo, potential selection bias from self-referral and elevated expectations due to media coverage, and consequent inability to draw causal inferences about efficacy. The authors therefore call for larger, randomised, placebo-controlled single-dose trials to confirm these preliminary findings and to clarify the roles of pharmacokinetics and subjective experience in clinical outcomes. The extracted text notes that such trials are underway.