Therapeutic effects of classic serotonergic psychedelics: A systematic review of modern-era clinical studies

Andersen and colleagues frame the review in the context of a long history of psychedelic research beginning in the 1950s, when LSD, psilocybin, DMT and mescaline were investigated both as tools for modelling psychosis and as potential therapeutics across a range of psychiatric conditions. Interest in these compounds waned after restrictive legislation in the 1970s and methodological shortcomings in many early studies, but more recent decades have seen renewed scientific work using modern imaging, epidemiology and controlled experimental designs. Naturalistic and experimental studies have suggested associations between past psychedelic use and reduced psychological distress, and laboratory studies in healthy volunteers report sustained improvements in well-being and personality measures well beyond the drugs' short pharmacokinetic presence. This paper sets out to collate modern-era (post-2000) clinical studies that administered classic serotonergic psychedelics (psilocybin, LSD, ayahuasca/DMT, mescaline) in controlled settings for psychiatric indications. The stated aim is to provide an overview of preliminary evidence for efficacy and safety, the duration of therapeutic effects, and the relationship between the subjective psychedelic experience and clinical outcomes. The authors follow PRISMA-guided systematic review methods to identify and synthesise relevant trials from 2000 to May 1 2020.

The investigators performed a systematic literature search of PubMed and PsycINFO for English-language studies published from 1 January 2000 to 1 May 2020, using a search string combining diagnostic terms (e.g. depression, anxiety, addiction) with drug terms (psilocybin, DMT, LSD, mescaline, ayahuasca). The start date reflected prior reviews showing an absence of clinical trials between 1982 and 2000. PRISMA principles guided the search and selection. Eligibility criteria included randomized single- or double-blind placebo-controlled trials, open-label trials and proof-of-concept trials in adult psychiatric populations, plus follow-up papers reporting long-term data from such trials. Excluded were case reports, observational and cross-sectional studies. Participants were adults (18+) with psychiatric diagnoses established by clinicians or structured DSM/ICD interviews. Interventions were limited to administration of classic serotonergic psychedelics in controlled laboratory settings. Outcomes had to be assessed with validated clinician-rated or self-report symptom scales. Study selection screened 671 references (14 duplicates removed), yielding 16 full texts for eligibility; 16 papers passed eligibility and corresponded to ten unique clinical trials. Because of substantial heterogeneity across trial designs, dosing regimens, drug types and target populations, the authors judged a quantitative meta-analysis inappropriate and instead performed a narrative synthesis. Data were grouped into three diagnostic categories for presentation: mood and anxiety disorders, cancer-related anxiety and depression, and substance use disorders. Trial characteristics reported include design (four randomized crossover trials, four open-label trials, others with mixed reporting), country, sample sizes and concomitant non-drug psychotherapy, which varied considerably between studies. The authors extracted adverse-event reporting and measures of subjective experience when available; heterogeneity and missing descriptive statistics prevented pooling of most effect estimates.

The review included ten clinical trials reported across 16 papers, involving 201 recruited patients (188 received at least one active psychedelic dose). Diagnostic groups comprised cancer-related anxiety and depression (C-RADD; n=89), other illness-related anxiety and depression (I-RADD; n=12), depressive disorders (n=66), substance use disorders (SUD; n=25) and obsessive–compulsive disorder (OCD; n=9). Trial designs varied: four randomized placebo (including active placebo) crossover trials, four open-label trials and other designs treated as open-label due to incomplete reporting. Because of heterogeneity, results are synthesised narratively. Mood and anxiety disorders: An open-label ayahuasca study in 17 recurrent major depressive disorder patients reported significant reductions in HAM-D and MADRS scores at multiple post-treatment time points (p<0.001), though there was no control condition. In a randomized trial of freeze-dried ayahuasca for treatment-resistant depression (n=29), the ayahuasca group showed greater improvement than placebo at day seven (p=0.019, d=0.98); response rates at day seven were 57% versus 20% (p=0.04), while remission differences were not statistically significant (43% vs 13%, p=0.07). Carhart‑Harris and colleagues conducted an open-label trial in 20 treatment‑resistant depressed patients receiving two psilocybin doses (10 mg then 25 mg) and observed large short‑term reductions in QIDS-SR16 (d range = 2.1–2.3, p<0.001) with sustained effects at 3 months (d=1.5, p<0.001) and 6 months (d=1.6, p=0.004). For OCD, Moreno et al. (n=9) reported reductions in YBOCS scores ranging 23–100% on one or more sessions, with eight of nine showing ≥25% reduction and six ≥50% at 24 hours; one participant reported sustained remission at 6 months. End-of-life/cancer-related anxiety and depression: Grob et al.'s double-blind crossover trial (psilocybin vs niacin) found no significant between-group differences, but pooled within-subject analyses after all participants received psilocybin showed reductions in trait anxiety at 1 and 3 months and depression at 6 months. Griffiths et al.'s randomized crossover study of high-dose versus low-dose psilocybin in psychologically distressed cancer patients found substantial between-group benefits five weeks after the first dose: clinical antidepressant response 92% (high dose) vs 32% (low dose), and anxiolytic response 76% vs 24%. Effects were sustained at 6 months across the cohort, with average response rates of 83% for anxiety and 78% for depression and remission rates of 57% and 65%, respectively. Ross et al. (randomized double-blind crossover with niacin control) observed acute and maintained within-group reductions after psilocybin, with 83% antidepressant response at 7 weeks (BDI) and 58% anxiolytic response (HADS-A) versus 14% in the control group; effects persisted at 8 months and, in a 4.5-year follow-up of survivors, 60–80% retained clinically significant response. Substance use disorders: In an open-label alcohol dependence trial (psilocybin), Bogenschutz et al. reported significant reductions in days of drinking and heavy drinking across a 32-week follow-up compared with baseline and the pre-treatment period. Johnson et al.'s smoking cessation study (n=15) combined psilocybin sessions with cognitive behavioural therapy and achieved 80% abstinence at 6 months (12/15), 67% abstinence at 12 months (10/15), and long‑term follow-up showed 9 participants abstinent at a mean of 30 months, with 7 reporting continuous abstinence since the first psilocybin session. Subjective experience and predictors: Nine trials collected measures of the acute subjective psychedelic experience; seven reported analyses relating those measures to clinical outcomes. In six studies (psilocybin for TRD, two psilocybin C-RADD trials, two psilocybin SUD trials and one ayahuasca depression trial), the magnitude or quality of mystical/peak experiences during dosing significantly predicted short-term (1–8 weeks) clinical benefit, and in one trial (smoking cessation) predicted long-term outcome (30 months). One trial (psilocybin for OCD) found a nonsignificant correlation. Setting and adverse events: Nine of ten trials used a common controlled session setting: individual sessions in an isolated room, staff present throughout, standardised music, dimmed lights or eyeshades, and non-directive psychosocial support. Cardiovascular responses (blood pressure, heart rate) rose transiently around 2–3 hours post-dose and returned to baseline without serious cardiovascular events. Reported non‑cardiovascular adverse events varied in reporting methods; in ayahuasca trials 31 doses to 31 patients produced purging in 16 (52%), nausea in 10 (32%) and transient anxiety in 7 (23%). Across psilocybin trials (145 patients, 268 doses) common side-effects included transient anxiety/fear (39, 27%), headache (32, 22%) and nausea/purging (18, 12%). In the single LSD study (11 patients, 22 doses) frequent adverse events were illusions (72.7%), feeling cold (45.4%) and a feeling of abnormality (40.9%). The authors report that overall adverse events were mild and transient.

The authors conclude that modern-era clinical trials of classic serotonergic psychedelics demonstrate feasibility, an acceptable safety profile under controlled conditions, and promising early evidence of efficacy across depression, anxiety (including cancer-related), OCD and substance use disorders. They note that no severe adverse events were reported in the reviewed trials and that most observed side-effects were transient. Caution is urged, however, because many trials were small, open-label and heterogeneous in design and psychosocial support, limiting firm efficacy conclusions. Controlled crossover trials in cancer-related anxiety and depression provide stronger evidence than open-label studies; Griffiths and Ross both reported large between-group effects favouring higher-dose psilocybin over low-dose or niacin controls with sustained improvements at months to years' follow-up. The authors highlight ongoing larger multicentre trials (COMPASSPathways, Usona Institute) as likely to provide more definitive evidence. For addictive disorders, early open-label findings are encouraging: the alcohol trial showed sustained reductions in drinking up to 8 months and the smoking cessation study reported abstinence rates substantially higher than typical pharmacotherapies, though the open-label design tempers conclusions. The discussion also places these findings alongside older-era meta-analytic evidence for LSD in alcoholism (reported OR ~2.0) and compares these effect sizes with those of established pharmacotherapies. Regarding mechanisms, the paper emphasises 5-HT2A receptor agonism as central: receptor occupancy correlates with subjective intensity and 5-HT2A blockade attenuates both subjective and mood effects. The authors describe possible downstream effects including increased cognitive flexibility, trait increases in openness, alterations in default mode network connectivity and amygdala reactivity, and transient windows of neural plasticity that, coupled with psychotherapeutic support, may enable lasting psychological change. They also discuss how the acute mystical or peak experience appears specifically predictive of positive outcomes, more so than basic perceptual changes, and suggest that therapeutic alliance and empathogenic effects of psychedelics (e.g. increased closeness, trust) may contribute to outcomes. Key limitations acknowledged are heterogeneity across trials (design, dose, population, psychosocial support), small sample sizes, incomplete reporting that precluded meta-analysis, and variable measurement and analysis of the subjective experience. The authors call for larger, rigorously controlled trials, standardisation of psychosocial procedures, investigation of optimal dosing and re-dosing strategies, and mechanistic work—including neuroimaging and measures of therapeutic alliance—to clarify how and for whom psychedelic-assisted therapy is effective. They also note unresolved practical questions about cost-effectiveness and whether shorter or group-based session formats could be implemented if boosters are required.