The role of the psychedelic experience in psilocybin treatment for treatment-resistant depression

Previous exploratory trials have suggested that psilocybin, a serotonergic prodrug metabolised to psilocin and acting as a 5-HT2A receptor agonist, can reduce symptoms of major depressive disorder (MDD), including in patients who have not responded to conventional treatments. Neuroimaging and pharmacodynamic studies indicate that 5-HT2A receptor engagement and peak plasma psilocin levels correlate with intensity and timing of the acute psychedelic experience, and that the experience is accompanied by transient increases in brain connectivity. Prior small clinical studies have reported relationships between measures of subjective psychedelic experience (for example Oceanic Boundlessness on the Five-Dimensional Altered States of Consciousness questionnaire and scores on the Mystical Experience Questionnaire) and subsequent clinical outcomes, but those studies were limited in size and varied in blinding methods, leaving uncertainty about which experiential domains are most clinically relevant. Goodwin and colleagues reported analyses from COMP 001, a large, multi-centre, dose-ranging clinical trial designed to probe these questions. The trial compared single oral doses of COMP360 psilocybin at 25 mg, 10 mg and 1 mg given with psychological support in adults with treatment-resistant depression (TRD). Building on the primary finding that 25 mg—but not 10 mg—improved depressive symptoms versus 1 mg at three weeks, the study set out to examine pre-specified exploratory and post-hoc associations between psilocybin dose, the intensity and character of the acute psychedelic experience (measured by the 5D-ASC and the Emotional Breakthrough Inventory, EBI), and change in depression severity, with the hypothesis that greater acute psychedelic intensity would correlate with larger reductions in depressive symptoms.

COMP 001 was a Phase II international, multi-centre, randomised, fixed-dose, parallel-group, double-blind dose-finding trial. Adult outpatients meeting DSM-5 criteria for a single or recurrent episode of moderate-or-greater MDD without psychotic features, who met a definition of TRD (failure to respond to two to four antidepressant trials of adequate dose and duration), were eligible. Key exclusions included recent electroconvulsive therapy for the current episode and serious psychiatric comorbidities; ongoing antidepressants and certain prohibited medications were tapered and discontinued at least two weeks before baseline. Screening included tests for illicit drugs, with limited allowances for cannabis in specified circumstances. Participants provided informed consent and received psychological support from trained therapists. Randomisation allocated 233 participants in a 1:1:1 ratio to single oral doses of 25 mg (n = 79), 10 mg (n = 75) or 1 mg (n = 79) of investigational COMP360 psilocybin. All site staff, participants and remote assessment raters were blinded to allocation; capsules were identical across doses. The mean age was 39.8 years (SD 12.19) and 52% of participants were female. Psychological support comprised three preparatory sessions to build trust and provide psychoeducation, a minimally interactive administration session with non-directive safety support, and two integration sessions after dosing; the number of support sessions was consistent across participants. The primary clinical measure reported here was the clinician-rated Montgomery–Åsberg Depression Rating Scale (MADRS), assessed remotely by raters blind to treatment. MADRS results in these analyses are presented as observed-case means and changes rather than the least-square means used in the primary estimand framework. Acute psychedelic effects were measured using the 5D-ASC (94 items yielding five dimension scores, each item scored on a visual analogue scale 0–100 and dimension scores derived as item means) assessed at the end of the administration session once acute effects had subsided, and the six-item Emotional Breakthrough Inventory (EBI; VAS 0–100, averaged) administered the day after dosing prior to the first integration session. Pre-specified exploratory correlational analyses used Fisher z-transformed Pearson correlation coefficients to relate change from baseline in MADRS total score at Week 3 (the trial primary endpoint) to each 5D-ASC dimension and to the EBI total score; as exploratory analyses, no formal hypothesis significance testing was performed. The investigators also examined the distribution of 5D-ASC dimension scores across doses to assess overlap in subjective effects between groups.

Randomisation yielded 233 participants across the three dose groups with baseline demographic and clinical characteristics reported as comparable across arms. Mean intensity ratings on all 5D-ASC dimensions increased with dose, indicating a dose-related effect on average; however, substantial inter-individual variability produced considerable overlap in 5D-ASC scores between the 1, 10 and 25 mg groups. For example, of 74 participants reporting Oceanic Boundlessness scores below 12.5, 32.4% were from the 10 or 25 mg groups, and of participants reporting Oceanic Boundlessness scores above 87.5, 33.3% were from the 10 mg group, illustrating overlap across dose groups. Correlation analyses relating the primary clinical endpoint (change from baseline in MADRS total score at Week 3) to acute subjective measures showed that higher scores on certain 5D-ASC dimensions and on the EBI were associated with greater reductions in depressive symptoms. Specifically, moderate negative correlations (meaning greater improvement in MADRS with higher experiential scores) were observed overall for Oceanic Boundlessness, Visual Restructuralization, Auditory Alterations and for the EBI total score. By contrast, the Reduction of Vigilance and Anxious Ego Dissolution dimensions showed small or negligible correlations with MADRS change, particularly within the 25 mg group. Scatterplots and density plots (reported in the paper) indicated similar linear relationships across doses for Oceanic Boundlessness and the EBI, whereas the Anxious Ego Dissolution dimension showed no clear relationship with clinical change. The extracted text does not report numerical correlation coefficients or confidence intervals in the sections provided, nor does it report exact effect size estimates for MADRS change by dose within these extracted passages; the primary trial publication contains the detailed efficacy statistics referenced by the authors.

Goodwin and colleagues interpret the findings as evidence that specific domains of the acute psychedelic experience are associated with subsequent clinical improvement in TRD. Although mean psychedelic effects increased with dose, overlap in individual subjective reports reduced the likelihood that participants could reliably infer their assigned dose; the authors suggest that a multiple-dose design may therefore mitigate some blinding challenges seen in two-arm comparisons with inert placebo. Oceanic Boundlessness, Visual Restructuralization and EBI scores showed the strongest relationships to Week 3 MADRS change, consistent with prior smaller studies linking positive experiential dimensions to therapeutic outcome. The authors propose two non-mutually exclusive interpretations: first, that experiential domains such as Oceanic Boundlessness may reflect pharmacodynamic target engagement (for example 5-HT2A receptor activity and associated changes in brain connectivity) necessary for clinical benefit; second, that aspects of the subjective experience itself—particularly emotional breakthrough as measured by the EBI—may be mechanistically important for clinical improvement. The dissociation between experiential domains is emphasised: dimensions associated with more negatively valenced or challenging experiences, exemplified by Anxious Ego Dissolution, showed weaker or absent correlations with therapeutic outcome, a result the authors note does not support claims that challenging experiences per se drive long-term benefit. They acknowledge several limitations: correlation does not establish causation; acute mood at the time of reporting may influence retrospective experiential ratings; the EBI was administered the day after dosing and may be confounded by early antidepressant response; expectation effects remain a possible source of bias though prior prospective work found no effect for expectation in psilocybin studies, and the present trial did not formally assess the integrity of the blind by asking participants to guess allocation. Finally, the correlational analyses were exploratory and require independent confirmation. The authors suggest that if the psychedelic experience functions as a pharmacodynamic biomarker it could index individual variability in drug absorption, metabolism, brain penetration and receptor responsivity, thereby explaining why participants receiving the same oral dose may effectively experience different biologically active exposure.

COMP 001, the largest randomised, double-blind, multi-dose trial of psilocybin to date, found dose-related mean clinical and experiential effects and substantial individual variability in acute subjective response. The intensity of particular experiential domains—notably Oceanic Boundlessness, Visual Restructuralization and emotional breakthrough—predicted clinical improvement at three weeks across oral doses. The authors conclude that these associations support the notion that the psychedelic experience may serve as a pharmacodynamic marker of psilocybin action and that the overall findings bolster the validity of claims from randomised, blinded trials that psilocybin treatment has potential utility in treating treatment-resistant depression.