Long-term follow-up of psilocybin-assisted psychotherapy for psychiatric and existential distress in patients with life-threatening cancer

Cancer diagnoses commonly provoke substantial psychiatric and existential distress, with anxiety and depressive disorders reported in up to 40% of patients in hospital settings. Earlier research has found limited and inconsistent benefits from conventional psychotropic medications in this population, and psychosocial interventions targeting existential distress have shown only modest efficacy and methodological limitations. Historically, classical psychedelics were investigated for cancer-related distress in mid-20th century trials, and more recent randomized crossover trials of psilocybin and LSD with psychological support (total N≈104 across trials) have reported preliminary safety and medium-to-large acute therapeutic effects, but long-term outcome data remain sparse. Agin-Liebes and colleagues conducted a long-term follow-up (LTFU) of participants from a prior randomized, placebo-controlled crossover trial that compared a single dose of psilocybin (0.3 mg/kg) with a single dose of niacin (250 mg) combined with psychotherapy in patients with cancer-related psychiatric and existential distress (parent trial N=29). The present study aimed to determine whether the symptom reductions and improvements documented at parent study completion were maintained at two extended follow-up points conducted on average 3.2 years and 4.5 years after participants' psilocybin dosing date.

This LTFU used a within-subjects design involving participants who had completed the parent crossover trial. Of the original 29 participants, 13 were deceased by the time of LTFU; the study team contacted the remaining 16, and 15 agreed to participate at the first LTFU. One of those participants died before the second LTFU, leaving 14 participants for the later time point. The first LTFU occurred on average 3.2 years (range 2.3–4.5 years) and the second on average 4.5 years (range 3.5–5.5 years) after each participant's psilocybin dose. In the parent trial participants had been randomly assigned to receive either psilocybin first and niacin second, or niacin first and psilocybin second, with nine preparatory and integration psychotherapy sessions provided by a dyadic therapy team. The original trial used a crossover at seven weeks, and final parent-study assessment occurred at 6.5 months after the second drug administration. For the LTFU, the investigators re-administered a range of primary and secondary measures drawn from the parent protocol, including anxiety and depression scales (HADS, BDI-II, STAI), death anxiety (DAS), hopelessness (HAI), demoralization (DS), quality of life (WHOQOL-BREF), spiritual well-being (FACIT-Sp-12), selected items from the Persisting Effects Questionnaire, and the Mystical Experience Questionnaire (MEQ-30) scores collected at the time of dosing in the parent study. Analyses collapsed the original dose-sequence groups into a single cohort because the crossover design precluded valid long-term between-group comparisons and to maximise power given the modest sample. Long-term effects were evaluated using mixed-effects repeated measures models (MMRM) and planned within-subject t-tests (with Tukey post-hoc) comparing baseline to 6.5 months and to the two LTFU points; comparisons were also made between the 6.5-month endpoint and each LTFU. Remission status (partial/complete remission versus active cancer) was entered as a covariate to test its impact. Clinical response was defined as a ≥50% reduction from baseline; remission required ≥50% reduction plus HADS-D≤7 or BDI≤12 where applicable. Spearman rank correlations examined associations between MEQ-30 scores and long-term change, and between elapsed time since the psilocybin session and symptom change. Participants also reported any psychotherapy or psychotropic medication received during the follow-up period and any lasting adverse effects.

Fifteen participants completed the first LTFU and 14 completed the second. No participants reported lasting negative or adverse effects attributed to the psilocybin-assisted therapy. During follow-up, five participants (39%) reported receiving some form of psychotherapy and three participants (23%) reported psychotropic drug treatment; only one participant (8%) received psychotherapy specifically targeting cancer-related distress. At the second LTFU, 71% of participants reported partial or complete remission of their cancer, and 29% had active disease. Mixed-effects repeated-measures analyses indicated statistically significant reductions relative to baseline on all primary measures of anxiety and depression at 6.5 months, and at both LTFU points. Effect sizes were large: mean Cohen's d at 6.5 months=1.90 (range 1.27–2.67), at the first LTFU mean d=1.30 (range 0.93–1.97), and at the second LTFU mean d=1.41 (range 0.86–1.89). The analysts noted sustained reductions at the first LTFU compared with the parent-study 6.5-month endpoint for most primary measures, with exceptions for HADS-A and STAI-T. At the second LTFU, 57% of participants met the criterion for a clinically significant anxiolytic response on HADS-A, 71% showed clinically significant reductions on HADS-T (combined anxiety and depression), and depression response rates on HADS-D and BDI ranged from 57% to 79%. Depression remission rates ranged from 50% to 79% at the second LTFU. Secondary outcomes also showed durable improvements. Hopelessness, demoralization and death anxiety were significantly reduced relative to baseline at 6.5 months and at both LTFU points; reported effect sizes were large (e.g. secondary outcomes mean Cohen's d at 6.5 months=1.39, first LTFU=1.39, second LTFU=1.60 with ranges reported). Spiritual well-being and faith domains (FACIT-Sp-12) improved at the second LTFU. Quality-of-life findings were mixed: psychological and environmental domains improved at the first LTFU, but the psychological gains were not sustained at the second LTFU. MEQ-30 scores (mystical-type experience intensity assessed on dosing day) did not significantly correlate with long-term change on primary anxiety or depression measures. On measures of persisting subjective effects, participants overwhelmingly attributed positive long-term changes to the psilocybin experience: 71% rated it among the single or top-five most personally meaningful experiences, 96% rated it among the single or top-five most spiritually significant experiences, 86% reported increased life satisfaction or well-being, and 100% reported "moderate," "strong" or "extreme" positive behavioural change. Cancer remission status did not significantly interact with primary or secondary outcome scores. Finally, elapsed time between the psilocybin session and the second LTFU correlated positively with change scores for depression and hopelessness—HAD-D r=0.70, p<0.01; HAI r=0.69, p<0.01—an association the investigators interpreted as greater elapsed time being associated with larger reductions in these domains.

Agin-Liebes and colleagues interpret their findings as the first evidence that psilocybin-assisted psychotherapy may be associated with sustained reductions in anxiety, depression, hopelessness, demoralization and death anxiety, and with improved spiritual well-being at an average of 3.2 and 4.5 years after a single psilocybin session accompanied by psychotherapy. The magnitudes of change relative to baseline were large, and approximately 60–80% of participants continued to meet criteria for clinically significant antidepressant or anxiolytic responses at the second LTFU. Participants also reported enduring, subjectively meaningful and spiritually significant changes attributable to the intervention. The investigators caution that causal attribution is limited by the parent trial's crossover design and by naturalistic factors: 71% of participants had entered partial or complete cancer remission by the second LTFU, and spontaneous resolution of cancer-related adjustment disorders could account for some improvement. The small sample size and limited demographic diversity (predominantly non-Hispanic White, well educated) reduce power and limit generalisability. The authors note that few participants reported receiving additional cancer-focused psychiatric treatments during follow-up, which they argue makes it less likely that post-trial interventions fully explain the observed improvements. Regarding mechanisms, the earlier parent-trial finding that mystical-type experiences partially mediated short-term anxiolytic and antidepressant effects was not replicated for long-term outcomes in this LTFU, possibly owing to reduced power or differential mechanisms over longer intervals. The authors propose psychological mechanisms such as increases in trait openness and cognitive flexibility, and neurobiological mechanisms linked to serotonin 2A receptor signalling and changes in neural connectivity, but emphasise these remain exploratory. They also stress the importance of psychotherapeutic context, arguing that set, setting and integration are likely crucial to durable therapeutic benefit and that neurobiological accounts alone are insufficient. Finally, the authors suggest clinical implications—psilocybin-assisted psychotherapy could address under-recognised existential distress in cancer care and might offer a protective strategy against suicidality—while calling for adequately powered, parallel-group randomized trials without crossover, broader and more diverse samples, and studies designed to probe psychological and neurobiological mechanisms. They also highlight practical considerations for future phase III–style research, including trial design elements and funding challenges.

The authors conclude that this long-term follow-up provides the first suggestion that a single psilocybin-assisted psychotherapy session, embedded within a structured psychotherapeutic context, can be associated with durable reductions in psychiatric and existential distress up to an average of 4.5 years post-treatment. They reiterate that efficacy cannot be definitively established from these data because of the parent trial's crossover design, small and demographically limited sample, and the possibility of naturalistic recovery. The conclusion calls for larger, randomized parallel-group trials with adequate controls, diverse and representative samples, and mechanistic investigations, noting that federal funding and regulatory support would be important steps toward establishing whether psilocybin-assisted psychotherapy can become an approved clinical option for patients with life-threatening cancer.