Historic psychedelic drug trials and the treatment of anxiety disorders

The paper frames classical psychedelics as three chemical classes—tryptamines (e.g. psilocybin, DMT/ayahuasca), phenethylamines (mescaline) and ergolines (LSD)—and notes contemporary pharmacology links the prototypic psychedelic experience to 5-HT2A receptor agonism. Weston and colleagues situate their review against a backdrop of renewed, rigorous trials from the 2000s showing promise for psychedelics in mood and anxiety related conditions, and emphasise the potential relevance of older literature (1940–2000) to current research questions. The review sets out to systematically examine historic human studies that combined psychedelic drugs with psychological therapies for disorders corresponding to ICD-10 anxiety diagnoses. The investigators aimed to map older diagnostic terminology (notably DSM‑I “psychoneurotic disorder: anxiety reaction”) onto contemporary ICD-10 categories (including generalized anxiety disorder, panic disorder and phobic disorders), to synthesise outcomes across study types, and to identify whether the historic evidence supports further investigation of psychedelics as adjuncts to psychotherapy for anxiety disorders.

Weston and colleagues conducted a systematic search of Ovid MEDLINE, PsycINFO and the Multidisciplinary Association for Psychedelic Studies (MAPS) bibliography for human studies published between 1940 and 2000. Search terms included LSD, lysergic diethylamide, psychedelic, hallucinogen, mescaline, peyote, DMT, dimethyltryptamine, ayahuasca, and psilocybin. The combined searches initially returned several thousand entries; these were manually screened and reference lists of identified articles were examined to find further papers. Each candidate article was read in full by two reviewers (two teams of two authors), and study selection emphasised relevance to ICD-10 anxiety disorders by mapping historic DSM categories (for example DSM‑I psychoneurotic anxiety reaction) onto contemporary ICD-10 diagnoses such as generalized anxiety disorder, panic disorder and phobic disorders. Quality appraisal considered principles for qualitative studies and the STROBE recommendations, but many historic reports lacked sufficient methodological detail. Twenty‑four studies met eligibility criteria for inclusion in the narrative synthesis; however some individual case reports and studies with unusable or pooled outcome reporting were excluded from the quantitative summary. Given heterogeneity in design, populations, interventions, outcome measures and reporting, the investigators used a textual narrative synthesis with tabulation of extracted data rather than a formal meta-analysis or pooled-effect model.

Twenty‑four publications from 1940 to 2000 were included in the review; after excluding individual case reports and studies that did not report outcomes specific to modern ICD-10 anxiety diagnoses, 20 studies contributed to the main summary analysis. Three broad groups of studies were distinguished: Group 1 comprised studies assessing anxiety symptoms across heterogeneous psychiatric cohorts and included several randomized controlled trials (RCTs) using standardised measures; Group 2 comprised extended open‑label case series reporting outcomes for diagnoses corresponding to ICD‑10 anxiety disorders; Group 3 comprised extended series in which anxiety outcomes were pooled with other psychoneurotic diagnoses and so were generally excluded from primary summary calculations. Across Groups 2 and 3 (with certain ambiguous reports excluded), the reviewers report that 94 of 145 (64.8%) cases diagnosed as anxiety reaction were judged to have achieved anything from moderate improvement to full recovery. For phobic reaction the combined figure was 10 of 25 (40%) showing at least moderate improvement. Across all 24 studies the aggregated sample was 805 individuals, the number of psychedelic sessions per person ranged from 1 to 58, and LSD was the most commonly used psychedelic. Within Group 1, three studies reporting on a combined cohort of approximately 95 participants found statistically significant reductions in anxiety symptom measures following LSD‑assisted psychotherapy, and two studies reported dose‑related effects; for example Hausner and Dolezal found a significant improvement in an N5 neuroticism score at 1 year in a 100 µg group versus a 50 µg group (p < .05). Specific open‑label series provided illustrative results: Chandler and Hartman reported a mean clinician rating of 2.5 (“some to considerable” improvement) among 27 patients with anxiety reaction; MacLean reported 15 of 23 anxiety‑reaction patients as ‘‘much improved’’ following a single high dose regimen; Ling and Buckman reported improvement in 27 of 39 patients with tension/anxiety states in an open series. Whitaker’s open trial of 100 patients found among six patients classified with anxiety state, two recovered and two were much improved. Adverse events and safety signals were reported across some series: one historical trial reported that about one in five patients experienced difficult experiences including paranoia and schizophreniform reactions; other series documented severe outcomes including two completed suicides, one attempted suicide and one homicide in pooled reports. The reviewers also highlight substantial heterogeneity in preparation and integration practices (from a few hours of preparation to multi‑week inpatient preparation and many hours of post‑session therapy), variability in outcome measurement (standardised scales in some RCTs versus clinician judgement and collateral reports in many open series), and inconsistent follow‑up timing; these factors limited comparability across studies.

The investigators interpret the historic literature as broadly encouraging but methodologically weak by modern standards. They note that although four of the reviewed studies were higher quality RCTs, most reports suffered from major shortcomings: inadequate assessment procedures, reliance on clinicians’ unblinded judgements, poor or absent follow‑up, lack of attempts to control for intensive patient‑therapist interactions, and diverse, sometimes opaque outcome reporting. Weston and colleagues emphasise that many authors from the historical literature viewed psychedelic compounds as catalysts for psychotherapy rather than as stand‑alone pharmacological cures, and a recurrent theme is the centrality of psychological preparation, integration and the therapeutic “set and setting”. Blinding and placebo control were repeatedly problematic in the historic trials because the subjective and objective effects of psychedelics are typically obvious to both participants and therapists; some investigators therefore randomised between active dose levels rather than active versus inert placebo to mitigate unblinding. The reviewers also highlight documented adverse events in some series and caution that harms were reported, including severe acute reactions and rare serious outcomes in pooled samples. Methodological limitations are reiterated in the dedicated Limitations section: heterogenous populations, variable outcome measures and reporting, and insufficient trial detail precluded formal risk‑of‑bias assessment with Cochrane tools. Finally, the authors conclude that, while historic data cannot substitute for rigorous contemporary trials, the pattern of reported clinical improvements—particularly for generalized or “free‑floating” anxiety—combined with the apparent importance of therapeutic accompaniment, justifies further well‑designed research into psychedelic‑assisted psychotherapy for anxiety disorders. They recommend that future work address blinding challenges pragmatically, standardise preparation and integration procedures, and systematically capture adverse events and longer‑term outcomes.