Psilocybin as a New Approach to Treat Depression and Anxiety in the Context of Life-Threatening Diseases-A Systematic Review and Meta-Analysis of Clinical Trials

The authors searched PubMed, Web of Science, Scopus and SciELO in January 2020 using a Boolean strategy combining terms for depression, anxiety or PTSD with hallucinogens/psychedelics and psilocybin, and filtered for clinical trials or human studies. Reference lists of relevant papers were also screened. Titles and abstracts were screened and full texts assessed by two reviewers independently, with a third reviewer adjudicating disagreements. Inclusion criteria required human clinical trials with a true control group, enrolment of patients with depression and/or anxiety associated with a life-threatening disease, administration of psilocybin, and use of BDI and/or STAI to measure outcomes. Data extracted included publication year, sample size, study design, medical condition, participant age and gender, intervention and control conditions, dosing and duration; extraction was performed independently by two authors with a third consulted for inconsistencies. Risk of bias for included randomized trials was assessed with the Cochrane Collaboration tool (seven domains: random sequence generation, allocation concealment, blinding of participants/personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other bias), with two reviewers assigning judgments. Statistical synthesis used weighted mean differences (WMD) of pre-to-post change between intervention and control groups, weighting by inverse variance, implemented in Comprehensive Meta-Analysis v2.0. Heterogeneity was quantified with Higgins' I2 and subgroup analyses examined dose and follow-up time as moderators. Publication bias was investigated via funnel plots, Egger's regression test and Duval and Tweedie’s Trim and Fill method. Sensitivity analyses used a leave-one-out approach to assess robustness. Study selection followed PRISMA flow procedures: the search initially identified 722 records (including eight from other sources); after duplicate removal 670 records remained, 32 were screened in full, seven underwent full-text eligibility assessment and three trials satisfied inclusion criteria. Those three trials were split into multiple effect sizes by follow-up time, producing 11 effect sizes for the meta-analysis.

Three randomized, double-blind trials published between 2011 and 2016 met the inclusion criteria, contributing a total of 92 patients who received psilocybin across dose ranges reported as 0.2–0.4 mg/kg; psilocybin was synthesised and administered orally in capsules. Variation in dose reporting and units complicated standardisation across trials. Follow-up assessments were performed at multiple time points up to about three months in some cases, yielding 11 effect sizes for the primary psychometric outcomes. For depression measured with BDI, the meta-analysis pooled 11 effect sizes representing 92 patients and found a statistically significant benefit favouring psilocybin versus control: WMD = -4.589 (95% CI = -4.207 to -0.971; p = 0.002). The authors applied a fixed-effects model because heterogeneity was reported as I2 = 0%. Anxiety measured by STAI-Trait was also analysed across 11 effect sizes (92 patients). The pooled effect favoured the intervention: WMD = -5.906 (95% CI = -7.852 to -3.960; p < 0.001), with I2 = 0% and a fixed-effects model. STAI-State results were derived from nine effect sizes including 41 patients and showed a significant benefit for psilocybin: WMD = -6.032 (95% CI = -8.900 to -3.164; p < 0.001), again with I2 = 0%. Subgroup analyses examined dose and follow-up time. Psilocybin reduced BDI and STAI-Trait scores across the doses tested but the effect was not clearly dose-dependent. Statistical significance versus control was observed at 0.4 mg/kg for BDI and at 0.3 and 0.4 mg/kg for STAI-Trait. Time-wise, significant reductions were reported at follow-up windows of 38–189 days for BDI and 14–189 days for STAI-Trait. The sensitivity (leave-one-out) analyses reportedly did not materially change pooled effects, suggesting result stability. Publication-bias assessments indicated funnel-plot asymmetry; the authors applied Trim and Fill and Egger’s regression test, and report that Egger’s test indicated evidence of publication bias with a p-value > 0.05 (results and adjusted estimates were not shown in the extracted text). Risk-of-bias evaluation judged the three trials to satisfy the seven Cochrane domains; all trials were described as randomised and double-blinded and outcome accounting was complete. Secondary physiological outcomes showed transient sympathetic effects after psilocybin. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) increased significantly post-dose, with SBP increases significant up to about 6 hours (p < 0.017) and DBP increases significant up to about 5 hours (p < 0.001). Reported DBP increases ranged between 1.194 and 11.381 mmHg with a mean increase of 7.741 mmHg. Heart rate rose significantly, peaking on the third and fourth hours after administration, and values tended to return to baseline after six or more hours. No serious adverse events were reported in the included trials; transient nausea, episodes of psychological discomfort and transient anxiety were noted, and no cases of hallucinogen-persisting perception disorder or prolonged psychosis were reported in these studies.

The authors interpret their findings as preliminary evidence that psilocybin can produce sustained reductions in symptoms of depression and anxiety among patients with life-threatening illnesses when administered in controlled settings. They highlight plausible neurobiological mechanisms including 5-HT2A receptor agonism, acute deactivation of the medial prefrontal cortex as seen on fMRI, attenuation of amygdala responsiveness to threat-related stimuli and decreased connectivity within the default mode network; these effects are proposed to underlie altered perception, emotion regulation and self-related processing that may mediate therapeutic benefit. Beyond pharmacology, the investigators emphasise the experiential component of psychedelic therapy: supportive 'set and setting' and guided sessions may foster mystical or reconciling experiences, emotional processing and meaningful cognitive–affective change that contribute to lasting improvement. The transient sympathetic physiological effects observed (increased blood pressure and heart rate, pupil dilation) are consistent with prior literature, while serious physiological disturbances, ECG changes, persistent psychosis or hallucinogen-persisting perception disorder were not observed in the included trials. The discussion acknowledges practical constraints and limitations. Legal restrictions on psilocybin complicate clinical trials and add cost. The authors also note the small number of eligible trials and modest sample sizes, heterogeneity in dosing units and follow-up schedules, and identified publication-bias signals; they nevertheless state that the evidence remains promising but requires cautious interpretation. Sensitivity analyses were reported to support robustness, but the extracted text indicates that some publication-bias assessments and adjusted results were not shown. As implications, the paper calls for further translational and mechanistic research, larger and rigorously designed clinical trials, and carefully defined clinical conditions and precautions for any therapeutic use. The authors suggest that, should a psilocybin medicine be approved, regulatory scheduling and oversight would be appropriate to mitigate risks and ensure safe clinical deployment.

Vargas and colleagues conclude that psilocybin may be effective in reducing symptoms of depression and anxiety in the context of life-threatening diseases. Although the meta-analysis detected publication-bias signals, the authors contend this does not invalidate the overall conclusions. They emphasise that the results are promising and recommend further translational, clinical and mechanistic studies to clarify efficacy, safety and mechanisms of action.