The largest psilocybin depression trial yet missed its primary endpoint

Several psilocybin depression trials have reported positive results in recent years, but limitations in trial design — small samples, imperfect blinding, single-dose protocols — have left open questions the field has called for larger, more rigorous studies to answer. The EPISODE trial, published this week in JAMA Psychiatry, is one such effort — and its result is more complicated than a simple positive or negative headline would suggest [s1].

The design

This two-center, triple-blinded (investigator, participant, and rater all blinded — a stricter standard than the double-blinding common in psychedelic trials, where participants often correctly guess their assignment) phase 2b trial was active placebo-controlled, recruiting adults aged 25 to 65 with treatment-resistant depression who had been withdrawn from antidepressant medication, predominantly from two outpatient settings in Germany [s1]. Participants were randomized in a 2:2:1:1 ratio across four sequences, receiving two doses six weeks apart: placebo (100 mg nicotinamide) then 25 mg psilocybin; 5 mg then 25 mg psilocybin; or 25 mg psilocybin twice, or 25 mg then 5 mg, each embedded in psychotherapeutic sessions [s1]. The primary endpoint was treatment response — at least a 50% reduction on the Hamilton Rating Scale for Depression (HAMD17) — at week 6, before the second dose [s1].

A total of 144 participants were randomized, with 142 included in the primary efficacy analysis: 47 in the 25 mg psilocybin group, 48 in the 5 mg group, and 47 in the nicotinamide placebo group [s1].

What it found

Response rates at the primary endpoint were 17.0% for 25 mg psilocybin, 12.5% for 5 mg psilocybin, and 10.6% for nicotinamide placebo [s1]. The first prespecified statistical comparison — 25 mg psilocybin versus placebo — was not significant (adjusted odds ratio 1.73, 95% CI 0.53–6.23, p = .19 against a one-sided alpha of .03) [s1]. Because the trial used a hierarchical testing procedure — a prespecified statistical safeguard where subsequent comparisons are only formally tested if the first one clears its bar — no further formal statistical testing was performed after that first comparison failed to reach significance [s1].

Despite the primary endpoint result, analyses of key secondary endpoints — mean changes from baseline on HAMD17 and the Beck Depression Inventory II — provided what the trial describes as "exploratory evidence of a clinically meaningful effect" of 25 mg psilocybin [s1]. On safety, 25 mg psilocybin was linked to adverse events, predominantly occurring acutely around dosing, and to higher reports of suicidal ideation on dosing days specifically — 4% versus 1–2% in comparator conditions [s1]. Two serious adverse reactions were reported following 25 mg psilocybin, including one case of hallucinogen persisting perception disorder (HPPD), a condition involving lasting visual disturbances after psychedelic use [s1].

Why the hierarchical testing detail matters

The hierarchical testing procedure used here is a legitimate and common statistical safeguard against false-positive findings from testing too many comparisons — but its consequence in this trial is that once the primary 25 mg-versus-placebo comparison failed to reach its prespecified threshold, the secondary endpoint results, however encouraging they looked descriptively, are formally "exploratory" rather than confirmed findings [s1]. That's a meaningfully different evidentiary status than a trial that hits its primary endpoint and then reports secondary findings as confirmatory support — this trial's secondary results are suggestive, not statistically validated by the trial's own prespecified design.

The safety signals deserve equal weight to the efficacy story

A reported case of hallucinogen persisting perception disorder — a recognized but uncommon and potentially lasting adverse effect of psychedelic use — is a serious finding in a trial of this size and rigor, and the elevated rate of dosing-day suicidal ideation (4% versus 1–2%) in the 25 mg group is a safety signal that a purely efficacy-focused reading of this trial would risk underweighting [s1]. Both findings occurred within a carefully controlled, monitored clinical trial setting with psychotherapeutic support — underscoring that even in that context, psilocybin at 25 mg carried real, documented risks in this population specifically.

What this trial adds to the field, honestly

EPISODE is among the more rigorously designed psilocybin depression trials published to date — triple-blinded, active placebo-controlled, with 142 participants completing primary analysis, larger than many prior psilocybin depression trials. That it did not confirm its primary hypothesis, even while producing encouraging exploratory secondary results, is itself a genuinely important data point for a field that has, at times, been criticized for smaller trials producing more uniformly positive headlines. The authors' own conclusion is carefully calibrated: describing a "clinically meaningful reduction in depressive symptoms" from secondary analyses, while explicitly stating the trial "did not show a significant effect on the primary outcome" [s1].

What to watch

Whether larger confirmatory trials, potentially with an even larger sample size than EPISODE's 142, can resolve whether the secondary-endpoint signal reflects a real effect that this trial was underpowered to confirm on its primary measure, and closer characterization of the HPPD and dosing-day suicidal ideation signals. Psilocybin is not approved for depression; this article describes investigational drug trial results and is not medical advice.

Sources

1. Efficacy and Safety of Psilocybin in Treatment-Resistant Major Depression: The EPISODE Randomized Clinical Trial — JAMA Psychiatry, 1 May 2026