Psychedelic therapy in the treatment of addiction: the past, present and future

Zafar and colleagues frame addiction as a chronic, relapsing condition with large and growing global prevalence, substantial mortality and a major socioeconomic burden. They note current treatments—primarily psychosocial interventions with limited pharmacological adjuncts—have low reach and modest effectiveness (for example, high relapse rates within months and few licensed medications for alcohol use disorder). The introduction highlights renewed scientific interest in psychedelic therapies, both ‘‘classic’’ serotonergic agents (LSD, psilocybin, DMT/ayahuasca, 5-MeO-DMT, mescaline) and ‘‘non-classic’’ compounds (ketamine, MDMA, ibogaine), and indicates that modern work ranges from observational, naturalistic studies through to contemporary randomized controlled trials and mechanistic neuroimaging studies. This review sets out to summarise the historical and contemporary clinical evidence for psychedelic therapies in addiction, to synthesise real-world and trial data from the mid‑20th century to the present, and to outline translational human neuropsychopharmacology techniques (fMRI, PET, SPECT, EEG and molecular imaging) that could illuminate mechanisms of action. The authors emphasise that a mechanistic, biomarker-driven understanding could help optimise drug development and personalise treatment for people with substance use and behavioural addictions.

The paper synthesises evidence across three eras: historical studies (mid‑20th century), recent real‑world observational data, and modern clinical trials. Key empirical points reported in the review include: Historical and early randomized data. A 2012 meta-analysis of six randomized trials (536 participants) administering a single LSD dose for alcoholism found a beneficial effect versus placebo (odds ratio (OR) 1.96, 95% CI 1.36–2.84, p = 0.0003); 59% improved in the LSD groups versus 38% in controls at first follow‑up, and effects were significant at 3 and 6 months but not at 12 months. From these pooled results the authors note a number needed to treat (NNT) of six for LSD in alcoholism, compared with higher NNTs for existing medications. In a randomized study from the early 1970s in incarcerated individuals with heroin addiction, one high‑dose LSD session (300–350 μg) together with inpatient care showed higher abstinence rates at 12 months (33% abstinent in LSD group vs 5% in controls), although the groups differed in non‑drug aspects of care. Non‑classic agents and historical case series. Ketamine combined with psychotherapeutic approaches was associated in Russian non‑randomized work with higher 12‑month abstinence (66% vs 24% in self‑selected controls) after three intramuscular doses of 2.5 mg/kg. Ibogaine case series and open‑label reports in opioid dependence describe rapid resolution of withdrawal signs (for example, 25 of 33 patients showed withdrawal resolution to 72 h post‑dose in one series at 19 ± 7 mg/kg). A NIDA‑funded Phase I ibogaine study was halted due to cardiac safety concerns; no randomized controlled trial data for ibogaine are yet published. Observational and survey evidence. The authors summarise multiple retrospective and prospective population studies reporting associations between classic psychedelic use and lower odds of substance problems. Examples: an analysis of ~44,000 people with prior illicit opioid use found psychedelic use associated with a 27% reduced risk of past‑year opioid dependence; a representative U.S. survey analysis reported lifetime peyote use associated with approximately 50% lower odds of past‑year cocaine use disorder (OR ~0.47). In a prospective cohort of 1,093 daily illicit opioid users, psychedelic use remained independently associated with reduced odds of subsequent daily opioid use (OR 0.45, 95% CI 0.29–0.70) after adjusting for confounders. Retrospective surveys of moderate lifetime classic psychedelic use also reported substantial proportions of people reducing or quitting smoking and other substance use. Modern clinical trials. Contemporary trials have produced preliminary positive signals across substances: - Psilocybin for alcohol use disorder: a pilot study (n = 10) combining motivational enhancement therapy with two high doses (0.3 and 0.4 mg/kg) reported rapid and sustained reductions in percentage of drinking days and heavy drinking days up to 36 weeks, and a 50% reduction in craving at 36 weeks. A Phase II double‑blind RCT (n = 93) found the psilocybin arm had 10% heavy drinking days versus 24% in an active control (diphenhydramine) during a 32‑week double‑blind period, a mean difference of 14% (95% CI 3.0–24.7; p = 0.01); mean daily alcohol consumption was also lower. - Psilocybin for tobacco use disorder: an open‑label, single‑arm study of 15 treatment‑seeking smokers reported biochemically confirmed abstinence in 12/15 (80%) at 6 months, substantially higher than typical smoking cessation interventions; higher mystical‑experience scores correlated with better outcomes. - Ketamine for alcohol dependence: a US trial in 40 outpatients randomised to a single ketamine infusion (0.71 mg/kg) versus midazolam plus motivational therapy reported that ketamine increased likelihood of abstinence and delayed relapse, though the extracted text is incomplete on effect sizes. A UK double‑blind phase II trial (n = 96) randomised patients to ketamine or placebo infusion and to psychotherapy (mindfulness) versus psychoeducation; the most favourable outcomes were in patients receiving three infusions of 0.8 mg/kg ketamine plus psychotherapy, who had more abstinent days at 6 months than placebo plus psychoeducation. - MDMA and ibogaine: small proof‑of‑concept work of MDMA in alcohol use disorder (n = 14) reported tolerability and marked reductions in average weekly alcohol consumption at 9 months versus pre‑trial. Multiple ibogaine observational studies in opioid and cocaine dependence report reductions in withdrawal and craving scores and in some cases sustained reductions up to 12 months, but safety concerns and the absence of RCT evidence are emphasised. Several ibogaine trials are registered but results are pending. Neuroscience and biomarker evidence. The review notes that human mechanistic data in addiction are sparse but growing. Relevant findings summarised include: - fMRI: limited data indicate low doses of LSD can increase reward‑related brain activity in healthy volunteers. Psilocybin in depression and healthy samples alters amygdala connectivity (for example, decreased amygdala–visual cortex coupling during threat processing), increases emotional empathy and sociality, and sub‑acute changes in amygdala responsivity have correlated with clinical outcomes. - Resting‑state connectivity: psychedelics produce acute disruption of default mode network (DMN) connectivity and increases in cross‑network coupling (for example between the salience network (SN) and DMN) in healthy subjects; interpretation in addiction is speculative but the authors propose that re‑broadening of salience and DMN modulation may underlie therapeutic effects. - PET and molecular imaging in addiction (not yet applied to psychedelics in human addiction): studies report decreased dopamine D2/3 receptor availability in alcohol and stimulant use disorders, upregulated mu‑opioid receptor availability in cocaine and alcohol dependence, and GABA α5 alterations in alcohol and heroin dependence. Novel radiotracers (e.g. [11C]PHNO, [11C]Cimbi‑36, [11C]UCB‑J) permit more specific investigation. Preclinical evidence suggests psilocybin can increase synaptic markers (SV2A), but human PET studies of psychedelic impact on neuroreceptor availability, neurotransmitter release capacity or synaptic markers in addiction populations have not been reported. Safety signals. The authors reiterate cardiac safety concerns with ibogaine that halted earlier trials, and note that modern trials have generally reported tolerability in small samples (for example, no severe adverse events in the small MDMA AUD study). They caution that many positive signals derive from open‑label, retrospective or small studies and that larger controlled trials are needed.

Zafar and colleagues interpret the assembled literature as indicating promising clinical signals for psychedelic therapies across a range of addictions, from observational associations to pilot and Phase II trial data. They stress that both classic serotonergic psychedelics and non‑classic compounds have shown reductions in substance use, craving, and relapse‑related behaviours in varying study designs, and that several larger Phase III programmes are underway which could support regulatory approval within the coming years. The authors position these clinical signals against substantial gaps: much of the epidemiological and naturalistic evidence is observational and subject to confounding; many historical studies lacked modern methodological rigour; sample sizes in modern trials are often small; and there remains a paucity of mechanistic human data linking molecular and circuit changes to clinical outcomes. They highlight specific safety concerns that have impeded development in some compounds (notably ibogaine's cardiotoxicity) and advocate for careful safety‑first progression in clinical development. A central thrust of the discussion is a call to deploy translational neuropsychopharmacology—multimodal fMRI, PET/SPECT molecular imaging, neurotransmitter release paradigms, EEG plasticity assays, and synaptic markers—to establish ‘‘molecular‑functional‑clinical’’ bridges. The authors propose specific experimental approaches (task‑based and resting‑state fMRI, PET receptor and release studies, SV2A imaging) to test hypotheses about reward‑circuit restoration, neurotransmitter deficits (dopamine, opioid, serotonin), and neuroplastic changes following psychedelic therapy. They argue such biomarker work could enable prognostic stratification, precision medicine approaches, and faster, more efficient drug development. Key limitations acknowledged by the authors include the limited quantity and quality of mechanistic studies in addiction populations, reliance on self‑report and observational designs in some evidence streams, and the need to tailor imaging and molecular approaches to specific addiction subgroups. They recommend longitudinal, pre‑versus‑post designs with multimodal imaging and appropriate psychometric batteries to determine whether acute and sub‑acute brain changes predict longer‑term clinical outcomes and relapse risk.

The authors conclude that psychedelic therapy has shown therapeutic promise in treating a variety of substance and behavioural addictions across real‑world and controlled settings. While early and contemporary clinical data are encouraging, a substantial gap remains in understanding the biological and psychological mediators of response. Zafar and colleagues therefore advocate integrating established biomedical and neuropsychopharmacological techniques into future clinical programmes to identify biomarkers of treatment response, to inform patient stratification, and to accelerate the safe and evidence‑based translation of psychedelic therapies into precision‑oriented addiction care.