Psilocybin-assisted psychotherapy for depression and anxiety associated with life threatening illness: A phase 2b randomized controlled trial

A life-threatening diagnosis commonly produces profound psychological and existential distress, worsening symptoms such as pain, nausea and insomnia and reducing quality of life. Previous research has shown only modest benefit from existing pharmacological and psychological interventions in palliative populations, and spiritual well-being has been identified as an important target for reducing existential distress. Early trials of psilocybin in cancer patients reported rapid reductions in depression, anxiety and death-related distress, but evidence remains limited for people with non-malignant life-threatening illnesses and for mixed diagnostic samples. Ross and colleagues designed a Phase 2b, double-blind randomized controlled trial to test whether a single 25 mg dose of psilocybin, delivered alongside manualised preparatory and integration psychotherapy, reduces depression and anxiety in adults with life-threatening illness compared with an active placebo (100 mg niacin) plus the same psychotherapy. The trial also included an open-label extension in which all participants received psilocybin, enabling comparison of one versus two psilocybin doses and assessment of durability of effects up to 26 weeks. Secondary and exploratory aims addressed quality of life, death attitudes, mystical experiences and other measures of spiritual and existential well-being.

The study was a single-site, double-blind, placebo-controlled RCT with an open-label extension, conducted at St Vincent's Hospital Melbourne between January 2020 and October 2023 under Good Clinical Practice standards. Ethical approval was obtained and written informed consent provided by participants. Eligibility criteria required adults (18–80) with an advanced life-threatening illness (malignant or non-malignant) and clinically significant depression and/or anxiety; screening included phone triage, SCID-5 screening questions and a semi-structured DSM-5 interview by a study psychiatrist. Participants needed a minimum score of 8 on either HADS subscale. Key exclusions included personal or first-degree family history of psychosis or bipolar disorder. Participants on antidepressants tapered under supervision with a two-week washout where required. After baseline electronic self-report measures and at least 6 hours of preparatory psychotherapy, participants were randomised 1:1 to receive a single double-blind dose of either 25 mg psilocybin or 100 mg niacin (active placebo) in identical capsules. Randomisation used permuted blocks (sizes 2–6) generated by an unblinded monitor; allocation was concealed by prelabelled containers. Niacin was chosen to produce flushing and therefore help preserve blinding. The dosing protocol implemented a standardised set-and-setting approach: 2–3 preparatory sessions, a supervised dosing day with eye-shades and music, and three integration sessions after each dose; a consistent therapist dyad supported each participant. Therapists were experienced clinicians trained in a manualised protocol combining psychodynamic therapy with acceptance and commitment therapy (ACT) techniques, somatic resourcing and creative arts methods. Qualitative interviews were collected but are not reported here. Primary efficacy assessments targeted change on the Hospital Anxiety and Depression Scale (HADS) depression and anxiety subscales from baseline to 6–7 weeks post-dose 1, with primary timepoints at Day -1, Day +1, 3 weeks and 6/7 weeks. Key secondary measures were the Beck Depression Inventory-II (BDI-II) and the State-Trait Anxiety Inventory—state version (STAI-s); additional instruments included the Death Attitude Profile–Revised (DAP-R), WHOQOL-BREF, Mystical Experience Questionnaire (MEQ) and Persisting Effects Questionnaire (PEQ). Safety was monitored using CTCAE v5 criteria with systematic recording of treatment-emergent adverse events, and cardiac observations were performed on dose days. The open-label extension used the 6/7-week assessment as the new baseline and repeated the same schedule after a second psilocybin dose. Statistical planning aimed for a sample of 40 to detect a 50% reduction in primary outcomes, but the study enrolled 35 participants. Primary outcomes were hypothesis-tested with a Bonferroni-adjusted α of 0.025 to control type I error across two primary endpoints. Analyses employed marginal mean models with robust standard errors and exchangeable correlation structures for longitudinal data; MEQ was analysed with linear regression. Missing data were handled with multiple imputation using MICE with predictive mean matching under a missing-at-random assumption. Analyses were performed in Stata v17 and reported as point estimates with 95% confidence intervals; secondary and exploratory outcomes were described without multiplicity-adjusted hypothesis testing.

Of 106 individuals prescreened, 55 consented and 35 were randomised to the double-blind phase (17 to psilocybin, 18 to niacin). Baseline characteristics showed a mean age of 60.9 (sd 11.5) years in the psilocybin arm and 51.3 (sd 13.9) in the niacin arm; 54.3% of the total sample were female. Most participants were Australian-born, none identified as Aboriginal or Torres Strait Islander, and 11 reported prior psychedelic use. The majority had moderate functional impairment and 46% had comorbid depression and anxiety. Thirteen participants tapered antidepressants prior to enrolment. During the double-blind phase three participants in the psilocybin arm did not complete the 6/7-week follow-up (one due to disease progression, one withdrew because of anxiety during dosing, and one died from their terminal illness); no participants withdrew from the niacin arm during the double-blind phase. All placebo participants proceeded to the open-label psilocybin dose; four participants from the psilocybin arm did not proceed to the open-label dose for medical, family or COVID-related reasons. By 26 weeks, two participants had died and one was lost to follow-up, leaving 25 participants (9 originally randomised to psilocybin and 16 originally randomised to niacin) who completed both doses and follow-up. On primary outcomes at 6–7 weeks post-dose 1, the psilocybin group showed significantly greater reductions in HADS depression compared with control (β = -2.49, p = .02) with a large effect size (d = 1.12). HADS anxiety reductions did not reach statistical significance between groups (p = .07). Complementary measures demonstrated robust effects: BDI-II scores reduced significantly in the psilocybin arm (β = -7.56, p = .004; d = 2.87), and STAI-state anxiety showed a significant reduction (β = -12.59, p = .005; d = 4.51). The authors report that pre-treatment mean BDI-II was approximately 21 (moderate severity) and decreased to a non-clinical mean of about 9.23 by weeks 6–7 post-dose 1. Secondary and exploratory outcomes indicated meaningful improvements in death-related attitudes, demoralization and hopelessness, and increases in quality of life and spiritual well-being. Following their open-label psilocybin dose, former placebo participants showed significant within-group increases in PEQ positive mood (β = -1.65; 95% CI -2.68 to -0.63; p = .002; d = -0.87) and improvements in altruism and positive attitudes with similar effect sizes. At 6/7 weeks after the open-label dose and at the 26-week follow-up, there were no significant differences between participants who had received two psilocybin doses and those who had received one dose; therapeutic gains were generally sustained through 26 weeks. Safety data showed no treatment-emergent serious adverse events attributed to psilocybin. Eight serious medical events unrelated to the investigational product occurred during the study period, including three deaths expected from terminal illness and five hospitalisations for intercurrent medical issues. Treatment-emergent adverse events were reported more frequently in the psilocybin arm (100%) than in the niacin arm (61.1%); common events included transient hypertension during dosing, headache and nausea. Short-lived psychiatric events in the psilocybin group included transient anxiety (23.5%), transient paranoia (5.9%) and passive suicidal ideation (5.9%), all of which resolved within the session or within 48 hours with clinical support. Cardiac observations noted expected, transient, asymptomatic increases in blood pressure on dose days.

Ross and colleagues interpret their findings as showing rapid and durable reductions in depressive and some anxiety symptoms following psilocybin-assisted psychotherapy in people with life-threatening illness. Improvements were detectable within a day and persisted up to six months. The BDI-II demonstrated particularly large reductions in depressive symptoms, with more than 50% of participants in the psilocybin arm achieving ≥50% BDI-II score reductions; HADS-Depression also showed significant between-group benefit while HADS-Anxiety trended in the same direction but did not reach the corrected significance threshold. The authors note that applying a Bonferroni correction across two primary outcomes, while conservative, may have reduced power to detect anxiety effects on HADS, and that scale differences (HADS-A’s one-week recall versus STAI-s’s focus on current state) may explain discrepant findings. Beyond symptom scores, the study observed reductions in death anxiety, demoralisation and hopelessness, alongside increases in spiritual well-being, quality of life and reports of mystical-type experiences. The investigators highlight that many participants reported enlivening, meaningful psychological states rather than sedation, which aligns with the goals of palliative care to address psychological and spiritual needs. The open-label phase showed that participants initially allocated to placebo improved after receiving psilocybin, and no sustained differences were found between one- and two-dose groups at 26 weeks, suggesting most gains may accrue after a single session for this sample. The authors acknowledge several important limitations. The study enrolled 35 participants rather than the planned 40, leaving it underpowered for some comparisons and precluding detailed subgroup analyses (for example, examining effects of antidepressant tapering). Blinding was difficult to maintain: retrospective file audits suggested many participants receiving psilocybin correctly inferred their allocation, and no formal blinding integrity assessment was performed, raising the possibility of expectancy effects. The small sample size may have inflated effect size estimates for some measures, and limited demographic diversity and an uneven mix of malignant and non-malignant diagnoses constrain generalisability. Practical considerations for wider implementation are noted: the intensive 2:1 therapist model may be resource-intensive and future work should explore scalable delivery options, recruitment of more diverse populations, and formal evaluation of dose-frequency effects with larger samples. Despite these caveats, the authors conclude that psilocybin-assisted psychotherapy had an acceptable safety profile in a medically unwell cohort and produced clinically meaningful improvements in psychological and existential outcomes.

This Phase 2b randomised controlled trial with an open-label extension found that a single 25 mg dose of psilocybin administered with structured preparatory and integration psychotherapy produced rapid, clinically meaningful reductions in depressive symptoms and improvements in spiritual well-being among people with life-threatening illness. Benefits were maintained up to 26 weeks, no treatment-emergent serious adverse events were attributed to psilocybin, and adverse effects were transient and manageable. The investigators conclude that psilocybin-assisted psychotherapy shows safety and therapeutic promise as an intervention for end-of-life psychological and existential distress, while emphasising the need for larger, more diverse trials and exploration of scalable therapeutic models.