Studying the Effects of Classic Hallucinogens in the Treatment of Alcoholism: Rationale, Methodology, and Current Research with Psilocybin

Classic hallucinogens — here defined as serotonergic 5-HT2A agonists such as psilocybin, mescaline, and LSD — have re-emerged as candidates for therapeutic use in a range of psychiatric conditions, including addictions. Earlier clinical work from the 1950s–1970s, basic science linking these drugs to 5-HT2A-mediated cortical and intracellular signalling, and observational findings from religious groups that use sacramental hallucinogens have all been invoked as indirect or direct support for their potential to affect addictive behaviour. Renewed, rigorous human research with psilocybin and improved neuroscientific tools have intensified interest in re-evaluating these agents for alcohol dependence. Bogenschutz frames this paper as a consolidating review that (1) summarises the preclinical, neuroimaging, observational, and clinical evidence relevant to using classic hallucinogens in addictions, (2) presents a rationale for clinical investigation, and (3) outlines methodological issues and design options for future randomised trials — with particular attention to psilocybin. The aim is practical as well as conceptual: to identify what is known, what remains uncertain, and how trials should be designed to answer the key efficacy and mechanistic questions.

The extracted text does not present a formal methods section describing systematic search strategies, databases searched, or explicit inclusion/exclusion criteria. From the content, the paper is best characterised as a narrative review and methodological discussion drawing on diverse sources: preclinical pharmacology, human neuroimaging and laboratory studies, cross-sectional observational data from sacramental contexts, historical clinical trials (notably LSD trials from the 1950s–1970s), recent experimental studies in healthy volunteers, and early/open-label clinical work with psilocybin. Rather than reporting a systematic synthesis, the authors assemble key lines of evidence and then move to translational and trial-design issues. Topics covered include likely neurobiological mechanisms (5-HT2A receptor pharmacology and downstream effects), acute human neuroimaging findings, historical clinical trial outcomes and a recent meta-analysis of LSD trials for alcoholism, contemporary psilocybin studies in healthy volunteers and small clinical samples, and a detailed consideration of clinical-trial design elements specific to hallucinogen-assisted therapy (choice of drug and dose, session number and structure, psychosocial platform, blinding and control conditions, outcome measures, sample selection, sample size, and candidate mediators and moderators).

Preclinical and pharmacological evidence: The paper summarises animal and molecular findings linking classic hallucinogens to 5-HT2A receptor activation as the primary mediator of their subjective and some cellular effects. Antagonism at 5-HT2A (e.g. by ketanserin) blocks subjective effects in humans. Hallucinogenic 5-HT2A agonists engage distinct intracellular signalling cascades compared with non-hallucinogenic agonists, implicating pathways involving Gi/o proteins and Src for LSD. Preclinical data also indicate drug-induced down-regulation of cortical 5-HT2A receptors after administration, regionally specific changes in expression of neurotrophic factors (BDNF, GDNF), and transient dendritic spine remodelling. The authors note there is, however, no direct evidence yet that these mechanisms mediate anti-addictive effects. Neuroimaging and acute human effects: Human PET, SPECT, and fMRI studies of acute psilocybin and ayahuasca effects report heterogeneous results. Some PET/SPECT studies showed regionally increased glucose metabolism or blood flow (e.g. anterior cingulate, frontomedial cortex, anterior insula, amygdala/parahippocampal gyrus), whereas an intravenous psilocybin fMRI study reported decreases in regional cerebral blood flow and BOLD signal in medial prefrontal and cingulate regions and reduced functional coupling between medial prefrontal and posterior cingulate cortices. Another report found increased activation to autobiographical recollection after intravenous psilocybin. Importantly, no neuroimaging studies of persisting (post-intoxication) brain changes are reported in the extracted text. Psychological/behavioural effects in humans: Studies in healthy volunteers show that high doses of psilocybin reliably occasion profound, often “mystical” experiences with persisting positive changes in attitude, mood, behaviour, and reported altruism. For example, in one double-blind study 22 of 36 participants met a priori criteria for a ‘‘complete mystical experience’’ after a high-dose session; 14 months later 67% rated the session among their five most significant spiritual experiences and 61% reported moderate to extreme positive behavioural change. Dose–response trends for persisting positive effects have been reported, with higher acute adverse effects at higher doses (severe anxiety and transient paranoia more common at 30 mg/70 kg than 20 mg/70 kg), but no persisting adverse effects observed in those volunteer samples. Observational and historical clinical evidence: Cross-sectional studies of religions that sacramentally use serotonergic hallucinogens (e.g. Native American Church peyote use, ayahuasca-utilising groups) show consistently lower rates of alcohol dependence, though cultural/contextual factors likely contribute. Historically, at least a dozen controlled clinical trials (primarily with LSD) examined treatment of alcohol dependence; a recent meta-analysis of six randomised trials (five double-blind) pooled data from 325 participants receiving LSD and 211 receiving control treatments. All trials used a single high-dose LSD session; doses varied widely (roughly 210–800 mcg). At the first post-treatment follow-up (1–12 months) the pooled odds ratio for improvement was 1.96 (95% CI 1.36–2.84, Z = 3.59, p = .0003). Among five trials reporting dichotomous outcomes, 59% of LSD-treated participants were classified as significantly improved versus 38% of controls, giving a pooled absolute benefit difference of 16% (95% CI 8%–25%, p = .0003), corresponding to a number needed to treat of 6. Treatment effects attenuated over time but remained significant at 6 months. Recent and ongoing clinical work with psilocybin: Small studies have explored psilocybin in obsessive–compulsive disorder and anxiety related to advanced cancer; the latter pilot study (psilocybin 0.2 mg/kg versus placebo) showed trends but no definitive effect, possibly limited by dose. Open-label pilot studies of psilocybin for smoking cessation and for alcohol dependence have been initiated, and additional trials are underway or planned (e.g. at Johns Hopkins and NYU for cancer-related anxiety). No randomised, definitive psilocybin trials for alcohol dependence are reported in the extracted text. Dosing, session structure and safety: Historical addiction trials favoured a single high-dose (“psychedelic”) session, aiming to elicit a peak-mystical experience. Reported effective LSD doses in the historical trials ranged roughly 3–11.4 mcg/kg; psilocybin dosing for research contexts has ranged up to about 0.43 mg/kg (≈30 mg/70 kg). The authors note evidence for increased persisting positive effects with dose, but also for more acute adverse reactions at higher doses, and suggest that a pragmatic approach might be 1–3 sessions or a small series (2–4) with dose escalation to balance efficacy and tolerability. Severe acute anxiety and transient paranoia have been observed at higher psilocybin doses in volunteers, but persisting adverse effects were not seen in those studies. Psychosocial platform and moderators: The paper emphasises that hallucinogen-assisted interventions are best viewed as combined drug-plus-psychosocial packages. Psychosocial preparation, support during sessions, and integration afterwards are considered critical; therapeutic models vary from psycholytic (low/moderate repeated doses with active psychotherapy during intoxication) to psychedelic (higher dose, non-directive support, focus on peak experiences). Candidate moderators include history of psychedelic exposure, family/personal history of psychosis (considered a contraindication), alcoholism subtype, severity, co-occurring disorders, personality, and genetic variants; the authors highlight the importance of assessing these factors in trials. Blinding, controls and trial design: Maintaining double blinding is challenging because psychoactive effects are obvious; strategies discussed include active comparators (e.g. stimulants such as methylphenidate), low-dose hallucinogen controls, complex blinding schemes (partial disclosure of possible treatments), and multi-arm designs with “distractor” groups to preserve uncertainty. The extracted text describes examples where methylphenidate served as an active control and complex blinding reduced correct identification of condition by session monitors. Cross-over designs are considered problematic for alcohol treatment because clinical status can change over months; other designs (e.g. initial double-blind phase followed by open-label crossover) are discussed. Outcomes, mediators and sample size: Recommended outcomes include dimensional measures commonly used in alcohol trials (percent days abstinent, percent heavy drinking days, drinks per drinking day) plus time-to-event measures and broader psychosocial and quality-of-life indices. Candidate proximal mediators include the phenomenology of the drug experience (measured by instruments such as the Five-Dimensional Altered States of Consciousness questionnaire and the Hallucinogen Rating Scale), enduring changes in mood, craving, motivation, self-efficacy, personality, and values. The authors caution against underpowered trials: extrapolating from historical effect sizes, a two-group trial may require over 150 participants per arm to achieve 80% power; they cite the Vivitrol pivotal trial (624 participants across three arms) as an example of trial scale needed to detect modest but clinically meaningful effects.

Bogenschutz and colleagues conclude that existing evidence warrants renewed clinical investigation of classic hallucinogens for alcoholism and other addictions. They interpret the preclinical data, historical clinical trials, observational associations in sacramental contexts, and contemporary human experimental studies as converging lines of support, while acknowledging major uncertainties. In particular, they note that mechanistic links between the molecular and cellular effects documented in animals and any anti-addictive behavioural effects have not been demonstrated, and that no studies have yet characterised persisting neurobiological changes in humans after therapeutic dosing. The authors stress several practical and conceptual challenges for future research. Foremost among these are the difficulty of blinding in trials with pronounced subjective drug effects, the need to integrate psychosocial preparation and integration with pharmacological treatment, and the likelihood that psychological state before, during, and after drug administration moderates or mediates outcomes. They recommend that future trials systematically measure candidate mediators (including the character of the acute experience) and moderators (clinical, psychological and biological), include neurobiological endpoints where practical (e.g. neuroimaging of cue-reactivity, inhibitory control, receptor occupancy), and be adequately powered to avoid type II errors that undermined earlier conclusions about LSD’s efficacy. Limitations and uncertainties acknowledged in the paper include heterogeneity and methodological limitations of the historical trials, attenuation of treatment effects over time (with prior studies suggesting diminished effects after 3–6 months), the absence of data on persisting brain changes, and the practical and ethical constraints on trial designs (e.g. providing adequate psychosocial support even in control conditions). For trial design, the authors recommend considering an active but putatively ineffective psychoactive control or low-dose control to reduce expectancy effects, ensuring the psychosocial treatment is credible for participants who do not receive a strong drug experience, and exploring designs that minimise demoralisation and differential dropout. They also call for exploratory work to establish optimal dosing and session number (noting strongest historical support for a high-dose, single-session model but also making a case for a few-session, dose-escalation approach). Finally, the authors outline implications for future research: prioritise early-phase studies to refine dose, session structure and psychosocial platforms; incorporate neuroimaging and other biological measures to probe mechanisms; measure mediators and moderators systematically; and plan adequately powered randomised controlled trials only once proof-of-concept and feasibility questions are addressed. Overall, the paper frames classic hallucinogen-assisted therapy as a promising but still nascent therapeutic approach that requires careful, methodologically rigorous clinical investigation before definitive efficacy or implementation claims can be made.