Therapeutic effect of psilocybin in addiction: A systematic review

Substance use disorders (SUDs) remain common and burdensome globally, with alcohol use disorder the most prevalent non-tobacco SUD and high relapse rates across disorders despite available treatments. The authors frame a substantial unmet need for improved interventions: approved pharmacotherapies are limited to a few SUDs, relapse often exceeds 65% within 12 months, and non-substance-related disorders such as gambling disorder show even higher recurrence. Interest has returned to psychedelic-assisted therapies, which in most trials combine an acute psychedelic administration with structured psychological support (preparation, the psychedelic session, and integration). Psilocybin, a tryptamine that acts principally at 5-HT2A receptors, is one candidate of interest given historical and more recent studies suggesting potential therapeutic effects across psychiatric conditions. Blest-Hopley and colleagues set out to determine whether psilocybin has efficacy for SUDs or non-substance-related disorders. Noting that previous systematic reviews tended to restrict searches to the last 25 years and may have missed mid-20th-century trials, they performed a date-unrestricted systematic review of clinical trials evaluating at least one dose of psilocybin in adults with a diagnosed SUD or related disorder, and included a formal quality assessment of the identified trials.

The review followed PRISMA guidance and searched seven electronic databases from the earliest records up to 2 September 2022, with no publication date restrictions. Clinical trial registries were also searched; the authors reported identifying registered trials in addition to published articles. Inclusion criteria required clinical trials (open-label pilot, single-blind, or double-blind placebo-controlled) that administered at least one dose of psilocybin to adult patients (≥18 years) diagnosed with a SUD or non-substance-related disorder using DSM or ICD criteria, reported outcomes assessing substance use severity or abstinence, enrolled ≥10 participants, and were published in one of several permitted languages. Exclusions covered animal studies, healthy volunteer experiments, observational designs, reviews, qualitative studies, opinion pieces, conference abstracts, and case reports. Data extracted from included studies encompassed authorship, year, study design, participant numbers and characteristics, details of the psilocybin and psychotherapeutic interventions, and the SUD- or disorder-related outcomes. For risk-of-bias assessment, the investigators used the ROBINS-I tool for non-randomised studies and the revised Cochrane RoB 2 tool for the randomized trial; assessments focused on the primary efficacy outcome and followed the standard domain-based signalling-question approaches of each tool. The methods section also notes screening and selection steps leading from an initial large record set to the final included articles, and that ongoing registered trials were recorded.

The systematic search yielded 6,832 unique records; after title/abstract screening and full-text review of 36 articles, six articles (representing four distinct clinical trials, two of which produced long-term follow-up articles) met inclusion criteria. All included trials were conducted in the USA or Poland. Across the four trials, 151 patients received psilocybin, with reported psilocybin doses ranging approximately from 6 to 40 mg in the studies that reported fixed or weight-adjusted doses. The authors also identified 130 registered clinical trials in trial registries and reported 11 ongoing trials as of 2 September 2022. Three trials addressed alcohol use disorder. A small open-label pilot by Bogenschutz et al. enrolled 10 patients with DSM-IV alcohol dependence and reported a statistically significant within-subject reduction in percentage heavy drinking days between baseline and weeks 5–12 (mean difference 26.0, 95% CI = 8.7–43.2, p = 0.008). A larger double-blind, placebo-controlled randomized trial (n = 95) compared psilocybin to the active placebo diphenhydramine over a 32-week double-blind period. Psilocybin recipients had a lower mean percentage of heavy drinking days [mean 9.7 (SD = 26.2)] versus diphenhydramine [mean 23.6 (SD = 26.1)], yielding a mean difference of 13.9 (95% CI = 3.0–24.7), Hedges g = 0.52, p = 0.01. Complete abstinence during the double-blind period did not differ significantly between groups (psilocybin 22.9% vs diphenhydramine 8.9%; OR = 3.1, 95% CI = 0.9–10.4, p = 0.06). Blinding was substantially compromised: nearly all participants and therapists correctly guessed treatment assignment at both sessions. No serious adverse events were reported in the psilocybin arm. In a single-arm study by Rydzyński et al. (n = 31) patients received psilocybin (6–30 mg, mean 12 mg) and initially also received LSD before switching to psilocybin; at a mean follow-up of 6 years, 32% (10/31) were completely abstinent, 32% had been abstinent for 6–12 months, and 58% achieved what the authors termed a “satisfactory therapeutic effect” (not further defined). No treatment-related serious somatic or psychiatric adverse effects were reported. One trial evaluated tobacco dependence. Johnson et al. enrolled 15 smokers with multiple prior quit attempts and delivered three psilocybin sessions (0.3 mg/kg then two 0.4 mg/kg sessions) alongside cognitive behavioural therapy. Seven-day point prevalence abstinence was 80% (12/15) at 26 weeks, 67% (10/15) at 52 weeks, and 60% (9/15) at long-term follow-up (mean interval 30 months, range 16–57 months). Reported treatment-related adverse effects were generally mild; transient anxiety or fear during sessions and headaches were the most commonly noted effects. No eligible trials were identified for opioid, cocaine, methamphetamine, cannabis, benzodiazepine/hypnotic, or gambling/gaming disorders, although several registered and ongoing trials for opioid, cocaine, and methamphetamine use disorders were noted. Quality assessment rated the three non-randomised trials as having serious to critical risk of bias and the randomized trial as having some concerns of risk of bias.

Blest-Hopley and colleagues interpret the body of evidence as cautiously encouraging: all four included clinical trials, which combined psilocybin with psychotherapy, reported beneficial effects on alcohol or tobacco-related outcomes, but the evidence base is small and heterogeneous. The authors emphasise that three of the four trials were single-arm pilot studies with substantial risk of bias, while the single randomized trial, although positive on its primary drinking-days outcome, had blinding difficulties and only some concerns on risk of bias. These limitations mean efficacy is not yet established and larger, rigorous randomised controlled trials are required. The discussion situates the findings within historical and mechanistic contexts. Earlier mid-20th-century research produced many trials with LSD for alcohol and opioid disorders, but psilocybin trials before the 1980s were scarce. Psilocybin and LSD share primary 5-HT2A receptor activity, and proposed mechanisms for therapeutic effects include neurobiological changes (e.g. increased neuroplasticity possibly via BDNF) and psychological processes, notably mystical-type experiences that may mediate enduring behavioural change in SUD. Dose and dosing strategy remain open questions; historical LSD trials often used very high single doses, whereas contemporary psilocybin trials have varied doses and frequently adjust dose by body weight despite limited evidence that weight-adjustment improves outcomes. Key methodological challenges are highlighted. Maintaining masking in psychedelic trials has proved difficult—participants and therapists commonly identify active treatment—introducing expectancy effects that may inflate observed effects. Concomitant psychotherapy in most trials complicates attribution of benefit specifically to psilocybin, and the authors note that future factorial or multi-arm designs could help disentangle pharmacological and psychotherapeutic contributions. Generalisability is another concern because strict inclusion criteria enhance safety but may limit external validity. The authors note ongoing and planned randomised trials, including some comparing psilocybin to approved pharmacotherapies, which should clarify comparative efficacy, optimal dosing regimens, and whether psychotherapy is an essential component of effective treatment.