Psychiatry has had the same tools for 40 years. Selective serotonin reuptake inhibitors — SSRIs — were revolutionary when fluoxetine launched in 1987. Today, they remain the first-line treatment for 280 million people living with depression worldwide. And they fail roughly half of them.
For patients who cycle through two or more antidepressant trials without adequate response, psychiatry has a name: treatment-resistant depression (TRD). It affects approximately 100 million people globally. The standard options — dose escalation, augmentation, electroconvulsive therapy — work for some and not others. The biological mechanism underlying TRD is still incompletely understood. The treatment gap is real, large, and growing.
Into this gap, psilocybin clinical research has arrived — not with anecdote, but with randomized controlled trials, FDA Breakthrough Therapy Designations, and data that leading psychiatrists describe as unlike anything produced in their careers. What follows is the complete scientific record of what has actually been demonstrated, how it was measured, and what it means for the future of depression treatment.
(COMPASS 25mg vs 9% placebo)
Johns Hopkins MDD trial, week 4
largest recorded in any antidepressant trial
The Depression Treatment Problem, By the Numbers
Major depressive disorder (MDD) carries the heaviest burden of any psychiatric diagnosis by disability-adjusted life years. The World Health Organization estimates 280 million people are affected globally, with a projected increase linked to urbanisation, social isolation, and the aftermath of COVID-19. In the United States alone, 21 million adults experienced a major depressive episode in 2024.
First-line treatment is pharmacotherapy — typically an SSRI or SNRI. These medications work by blocking the reuptake of serotonin (and norepinephrine in the case of SNRIs), increasing synaptic serotonin availability. They require 4–6 weeks to produce clinical effect, and response rates in the STAR*D trial — the largest antidepressant effectiveness study ever conducted — showed only 37% of patients achieved remission on the first medication tried. Cumulative remission across all treatment levels reached only 67% after four sequential trials over 12 months, with progressively lower response rates at each step.
What this means in practice: approximately one-third of depression patients do not achieve adequate response despite multiple medication trials. These patients — the treatment-resistant population — face higher rates of hospitalisation, functional impairment, suicide risk, and healthcare utilisation than any other group in psychiatry. They are the patients for whom something fundamentally new is needed.
Defining Treatment-Resistant Depression
There is no universally standardised definition of treatment-resistant depression, which has complicated clinical trial design. The most widely used definition, and the one adopted by COMPASS Pathways for their Phase 2b trial, is failure to respond to at least two adequate antidepressant trials of different pharmacological classes, at therapeutic doses, for a minimum of six weeks each, during the current depressive episode.
"Adequate response" is typically defined as less than 50% reduction in scores on a validated rating scale such as the Montgomery–Åsberg Depression Rating Scale (MADRS) or the Hamilton Depression Rating Scale (HAM-D). Remission is defined as a score below a threshold indicating minimal to no symptoms — MADRS ≤10 or HAM-D ≤7, depending on the scale used.
COMPASS Pathways: The Phase 2b Landmark Trial
In October 2022, COMPASS Pathways published the results of the largest randomised, controlled, double-blind psilocybin trial ever conducted to that point. The paper appeared in the New England Journal of Medicine. Lead author Guy Goodwin, emeritus professor at Oxford, described it as “the most rigorous evidence base yet generated for any psychedelic treatment.”
The trial enrolled 233 patients with treatment-resistant depression across 22 sites in 10 countries. All participants had failed at least two antidepressant trials. They were randomised to receive a single dose of COMP360 — a pharmaceutical-grade synthetic psilocybin — at one of three doses: 1mg (active placebo), 10mg, or 25mg, administered in a specially designed clinical room with two trained psychological support therapists present throughout the 6–8 hour session.
The primary endpoint was change in MADRS score at week 3. The results were statistically significant for the 25mg group.
At week 3, the 25mg group showed a mean MADRS reduction of 12.0 points versus 7.9 points in the 1mg group (p=0.004). More clinically meaningful: remission (MADRS ≤10) was achieved in 29.1% of the 25mg group versus 7.6% in the 1mg group. Sustained response at week 12 was 24.1% for the 25mg group versus 10.1% for the 1mg control.
COMPASS Pathways received FDA Breakthrough Therapy Designation for COMP360 psilocybin therapy in treatment-resistant depression in October 2018 — before the Phase 2b trial was complete. This designation, reserved for treatments showing substantial improvement over existing therapies for serious conditions, allows expedited development and review. It is the same designation that MDMA-assisted therapy held before its eventual New Drug Application. A second BT designation is held by Usona Institute for psilocybin in major depressive disorder.
The 10mg arm did not show statistically significant improvement over the 1mg control on the primary endpoint (p=0.19), which was notable — suggesting a threshold effect rather than simple dose-response linearity. The researchers hypothesised that the 10mg dose may be insufficient to produce the degree of default mode network disruption associated with therapeutic response, while 25mg crosses a neurobiological threshold that enables more substantial network reorganisation.
Adverse events were generally mild to moderate. The most common were headache (56%), nausea (18%), dizziness (14%), and fatigue (17%). Seven participants in the 25mg group experienced suicidal ideation or self-injury in the 12-week follow-up period — a finding the investigators noted was consistent with base rates in the TRD population rather than a drug-specific signal, but one that requires careful monitoring in Phase 3 trials and any potential clinical deployment.
Imperial College London: Head-to-Head Against an SSRI
The 2021 trial published in the New England Journal of Medicine by Robin Carhart-Harris and colleagues at Imperial College London was structurally different from the COMPASS study — and produced results that generated significant debate in the field.
Fifty-nine patients with treatment-resistant depression were randomised to receive either: (A) two sessions of psilocybin (25mg then 10mg, three weeks apart) plus six weeks of placebo capsules, or (B) two sessions of placebo plus six weeks of escitalopram (10–20mg daily, a standard SSRI). The design was notable for maintaining blinding through matched capsule schedules — technically challenging in psychedelic trials where the subjective effects of an active drug are obvious to the participant.
The primary outcome was QIDS-SR-16 (Quick Inventory of Depressive Symptomatology — Self Report) score at six weeks. This showed a mean reduction of 8.0 points in the psilocybin group versus 6.0 in the escitalopram group — a difference of 2.0 points that did not reach statistical significance (p=0.17). The investigators reported that the trial was powered to detect a 4-point difference and was not designed as a non-inferiority trial.
Response vs Remission in Clinical Trials
Two distinct measures are used to assess antidepressant efficacy in clinical trials, and understanding the difference is essential for interpreting psilocybin data. Response is typically defined as a ≥50% reduction from baseline on a validated depression rating scale (MADRS, HDRS, or QIDS-SR-16). It indicates meaningful improvement but not absence of symptoms.
Remission is the higher bar — a total score below a defined symptom threshold (e.g., MADRS ≤10, HDRS ≤7). It indicates that a patient is essentially symptom-free. In pharmaceutical antidepressant trials, remission at week 8 is the gold standard endpoint for regulatory approval. In psilocybin trials, where a single session can produce effects within days, remission at weeks 3–6 has become the primary comparison metric.
However, secondary outcome measures told a different story. On measures of emotional blunting, connectedness, and meaning in life — domains that SSRIs are known to impair through serotonin system-wide suppression — the psilocybin group showed significantly stronger improvements. Specifically:
The Oxford Depression Questionnaire, which measures emotional blunting directly, showed that the escitalopram group experienced significant emotional numbing consistent with SSRI pharmacology, while the psilocybin group did not. The Revised Life Orientation Test and measures of psychological flexibility both favoured psilocybin significantly (p<0.05). Patient-reported experience of connectedness and meaning was also superior in the psilocybin group.
Carhart-Harris, now at UCSF, has argued that these secondary outcomes point to a qualitatively different mode of antidepressant action — one that does not suppress emotional range but expands psychological flexibility. He describes the psilocybin mechanism as “lifting the lid on suppressed emotional processing,” in contrast to the SSRI mechanism of broadly dampening serotonergic tone.
vs 28% escitalopram (QIDS-SR-16)
(psilocybin: days vs SSRIs: 4–6 weeks)
Johns Hopkins MDD trial
Johns Hopkins: Psilocybin in Non-Resistant Depression
While COMPASS and Imperial focused on treatment-resistant populations, researchers at Johns Hopkins University School of Medicine examined a different and arguably more significant question: does psilocybin work in standard major depressive disorder — the much larger population of patients with depression who have not yet failed multiple treatments?
The 2021 trial, led by Alan Davis and Matthew Johnson and published in JAMA Psychiatry, enrolled 24 adults with MDD (not TRD). Participants received two psilocybin sessions (20mg/70kg and 30mg/70kg, four weeks apart) combined with supportive psychotherapy. The control group received a delayed-treatment waitlist design.
The results were striking. At the primary endpoint (week 4 post-second-session), 71% of participants showed clinically significant response (≥50% reduction in HDRS-17 scores), and 54% met full remission criteria. The between-group effect size was Cohen’s d=2.5 — a magnitude that no conventional antidepressant has produced in controlled trials. By comparison, SSRI trials typically show effect sizes of d=0.3–0.5 versus placebo.
Follow-up data at 3 months showed maintained effects in 75% of responders, without additional treatment. Johnson has noted that the absence of daily dosing requirements represents a fundamentally different treatment paradigm: rather than requiring continuous pharmacological maintenance, a single treatment episode appears to produce a durable neurobiological reset.
The Mechanism: Why Psilocybin Works Where SSRIs Fail
Understanding why psilocybin produces such different outcomes — in effect size, speed of onset, and qualitative texture — requires understanding its neurobiological mechanism in contrast to conventional antidepressants.
SSRIs work proximally: they block the SERT (serotonin transporter) protein, preventing serotonin reuptake and increasing synaptic serotonin concentration. Over weeks, this produces downstream adaptations in serotonin receptor density and sensitivity. But SSRIs do not directly change neural connectivity patterns, alter default mode network dynamics, or produce the structural neuroplastic changes now documented with psilocybin. They manage symptoms through ongoing pharmacological presence — which is why discontinuation typically leads to relapse.
Psilocybin acts directly as a 5-HT2A receptor agonist — binding to postsynaptic receptors and triggering intracellular signalling cascades that SSRIs never reach. The 5-HT2A receptor is the most densely expressed in the prefrontal cortex and is the primary binding site for psychedelics. Activation triggers:
1. Default Mode Network suppression. The DMN — the neural network associated with self-referential thinking, rumination, and autobiographical memory — shows marked hyperactivity in depression. The insula, medial prefrontal cortex, and posterior cingulate cortex all show elevated baseline activity in depressed patients. Psilocybin produces rapid, dose-dependent suppression of DMN activity measured via fMRI — the same suppression associated with ego dissolution and the reported cessation of ruminative thought.
2. BDNF upregulation and neuroplasticity. Within 24 hours of a psilocybin dose, brain-derived neurotrophic factor (BDNF) levels surge significantly in the hippocampus and prefrontal cortex. BDNF is the primary growth factor for neuronal maintenance and synaptic formation. Depression is associated with chronically low BDNF — particularly in hippocampal regions responsible for emotional regulation. The psilocybin-triggered BDNF surge initiates a window of enhanced structural plasticity during which new synaptic connections form more readily than baseline.
3. Increased global connectivity. Resting-state fMRI studies show that psilocybin transiently disrupts normal within-network connectivity patterns while dramatically increasing cross-network communication — the state described phenomenologically as ego dissolution and mechanistically as “network entropy.” Post-session, normal connectivity returns but with measurably different organisation — rigid fixed attractors replaced by more flexible neural dynamics.
The peak BDNF elevation following a psilocybin session creates what researchers describe as a “critical period of plasticity” — a window of 48–72 hours during which the brain is exceptionally receptive to new learning, relational processing, and emotional integration. The quality of this integration period — including psychological support, reflective practices, somatic work, and nutritional support — is increasingly recognised as a primary determinant of long-term therapeutic outcome. This is not a passive recovery period. It is the phase during which the neuroplastic gains of the session are consolidated into lasting change.
The FDA Pathway: What Phase 3 Must Prove
COMPASS Pathways initiated Phase 3 trials for COMP360 in treatment-resistant depression in 2023, following the Phase 2b results. Phase 3 represents the regulatory standard for approval — requiring replication of Phase 2 findings in a larger, more diverse population with a standardised treatment protocol.
The Phase 3 design includes several features that distinguish it from Phase 2b and will determine the regulatory outcome. First, sample size: Phase 3 will enrol over 500 participants across approximately 40 sites in multiple countries. Second, standardisation: the psychological support protocol has been formalised into a manual that allows consistent delivery across facilitators and sites. Third, duration: the primary endpoint extends to week 6, with long-term follow-up to one year.
The FDA has also issued guidance requiring that any approved psilocybin therapy include mandatory therapist training and certification as a condition of the drug’s use — acknowledging that the therapeutic effect is not separable from the psychological support context. This creates a unique regulatory paradigm: a pharmaceutical treatment that comes with a required human delivery system.
The primary challenges for Phase 3 approval include: blinding integrity (participants generally know they received an active substance), replication of the 25mg dose response finding, demonstration of acceptable safety in a larger and more heterogeneous population, and establishing the durability of effect beyond three months. The MAPS MDMA precedent — where Phase 3 data showed strong efficacy but the FDA required an additional clinical trial due to blinding concerns — illustrates that approval is not guaranteed by positive Phase 2 data alone.
What This Means for Treatment-Resistant Patients Now
For patients with treatment-resistant depression, the clinical trial data represents something that has not existed before in psychiatry: robust evidence that a single-dose treatment can produce rapid, durable remission in a population for which current pharmacotherapy has failed. The effect sizes, onset speed, and safety profile differ categorically from anything in the existing formulary.
Regulatory approval is not yet achieved. The FDA pathway requires Phase 3 completion, and timeline projections place potential approval between 2027 and 2029 pending trial results. However, the scientific consensus has shifted substantially — from skepticism about psychedelic research as fringe science to active engagement from mainstream academic centres including Harvard, Stanford, NYU, Oxford, and Cambridge.
One of the most significant findings across all psilocybin depression trials is the safety record when administered under controlled conditions. The physiological risk profile of psilocybin is well-established from decades of pharmacological research: no known lethal dose in humans, no organ toxicity, no addiction potential, no withdrawal syndrome. The primary risks are psychological — challenging experiences that require skilled support to navigate, and contraindications including personal or family history of psychotic disorders. Across more than 1,000 psilocybin sessions in clinical research, no serious adverse events attributable to psilocybin pharmacology have been reported when proper screening and support protocols are followed.
For patients navigating their options today, ketamine and its FDA-approved derivative esketamine (Spravato) represent the currently available rapid-acting treatment for TRD. Esketamine received FDA approval in 2019 and is administered intranasally in clinical settings. While ketamine shows rapid antidepressant effects — often within hours — the effect duration is shorter than psilocybin data suggests, with most patients requiring repeated infusions. The mechanism is also different: NMDA receptor antagonism rather than 5-HT2A agonism, without the neuroplastic signature documented with psilocybin.
The arrival of multiple converging lines of Phase 2 evidence for psilocybin, across different populations, institutions, and trial designs, represents a scientific moment without clear precedent in psychiatry. The question is no longer whether psilocybin produces antidepressant effects — the data is too consistent to dismiss. The question is how to translate that evidence into a safe, scalable, and accessible treatment system.