Psilocybin Just Cleared the Highest Bar in Psychiatry — Twice. Now It Faces a Problem No Molecule Can Solve.
Compass Pathways' synthetic psilocybin hit p<0.001 in two Phase 3 trials for treatment-resistant depression. Scaling a therapy that requires 8 hours of trained supervision per dose means the bottleneck isn't the molecule. It's the 22,400 facilitators who don't exist yet.
Negative 3.8 points on the Montgomery-Åsberg Depression Rating Scale. That was the drug-placebo difference in Compass Pathways' COMP006 trial, reported February 17, 2026. Seven months earlier, COMP005 posted -3.6 (95% CI [-5.7, -1.5], p<0.001). Two Phase 3 wins, 500+ patients across 32 U.S. sites, both in treatment-resistant depression (TRD): the population that has already failed at least two rounds of antidepressants.
COMP360, Compass's synthetic psilocybin formulation, is now the first classic psychedelic to clear the highest regulatory bar in psychiatry. An NDA filing is planned for Q4 2026. If approved, a patient who has spent years cycling through SSRIs, SNRIs, and augmentation strategies could receive one or two supervised doses and walk away with six months of sustained benefit.
But between the molecule and the patient sits a workforce problem that no amount of clinical data can solve.
What Two Phase 3 Trials Actually Showed
COMP005 enrolled 258 participants in a randomized, double-blind, placebo-controlled design. A single 25mg dose of COMP360 produced a statistically significant improvement in MADRS score at Week 6 (-3.6 vs. placebo, p<0.001). In COMP006, two fixed doses three weeks apart compared 25mg against a 1mg active control. Result: -3.8 on MADRS at Week 6 (p<0.001), with 39% of the 25mg arm achieving clinically meaningful improvement (≥25% MADRS reduction).
A detail worth pausing on: 40% of patients who responded but didn't achieve full remission after the first dose got there after the second. Week 26 follow-up data from COMP005 showed maintained effects from just one or two administrations. No ongoing dosing. No twice-weekly clinic visits.
That matters because the only other breakthrough TRD therapy approved in the past decade, Johnson & Johnson's esketamine (Spravato), requires exactly that kind of schedule: twice weekly for four weeks, then weekly, then biweekly, potentially indefinitely, each session with two-hour in-clinic monitoring under a REMS program.
How Psilocybin Stacks Against Existing Treatments
Cross-study comparisons carry real caveats (different patient populations, different placebos, different endpoints). With that said, the MADRS effect sizes invite comparison:
| Treatment | MADRS Δ vs. Control | Dosing Schedule | Est. Annual Cost |
|---|---|---|---|
| Standard antidepressants (SSRIs/SNRIs) | -2.1 points1 | Daily, indefinitely | $200–2,400 |
| Esketamine (Spravato) | -3.6 to -4.2 points | Twice weekly → biweekly, ongoing | $6,000–10,000 |
| COMP360 (psilocybin) | -3.6 to -3.8 points | 1–2 doses total | $4,000–8,000* |
*Estimated. Range depends on whether durability is 6 or 12 months per treatment course. Sources: Hengartner & Plöderl 2020, Spravato prescribing information, Compass Pathways investor presentations, Oregon psilocybin service center pricing data.
COMP360's effect size (-3.6 to -3.8) runs about 1.7x the standard antidepressant benchmark and matches esketamine's range. Not dramatically superior on efficacy alone. Where it diverges is the treatment model: one or two sessions versus indefinite maintenance dosing.
Original Analysis: Cost-Per-Remission
No published analysis has directly compared the cost to achieve one remission across these treatment paradigms.
Esketamine: At ~$8,000/year (facility fees included), with a 55% remission rate at 32 weeks (Psychiatric Times review), cost per remission = $8,000 / 0.55 = $14,545. But the patient keeps paying every year.
COMP360: At ~$4,000 per treatment course (1–2 doses, based on Oregon pricing), with an estimated 20% overall remission rate (extrapolating from 39% response and 40% responder-to-remitter conversion), cost per remission = $4,000 / 0.20 = $20,000. More expensive per remission in Year 1.
Break-even: If psilocybin's effect lasts 12 months, annual cost is $4,000 versus esketamine's $8,000. Psilocybin wins from Year 1. If durability is only 6 months, annual cost matches esketamine at $8,000 but with fewer clinic visits.
What neither calculation captures: TRD patients consume ~$25,000/year more in healthcare than non-TRD patients (ER visits, hospitalizations, disability). If psilocybin reduces even a fraction of that utilization, the economics shift considerably. Nobody has measured this yet.
22,400 Facilitators Who Don't Exist
Here is a calculation that Compass's investor presentations do not include.
Each COMP360 session requires approximately 8 hours of trained facilitator supervision: preparation, the dosing session itself, and integration. Roughly 2.8 million American adults meet the clinical definition of TRD (about 30% of the 21 million with major depressive disorder). If every TRD patient sought psilocybin therapy with two sessions each:
2.8 million patients × 2 sessions × 8 hours = 44.8 million facilitator-hours per year.
At 2,000 clinical hours per facilitator per year, that requires 22,400 dedicated facilitators. For context, the U.S. has approximately 45,000 practicing psychiatrists total (Merritt Hawkins).
Oregon's legal psilocybin framework already allows non-physician facilitators with specialized training. Compass's model would likely require a similar expansion: trained psychologists, nurse practitioners, clinical social workers. Building that workforce takes years and state-by-state regulatory battles that have nothing to do with what happens inside a Phase 3 trial.
This is the problem no molecule can solve. COMP360 proved it works in a controlled clinical setting with trained staff and careful monitoring. Proving it can reach even 10% of the people who need it requires building a delivery system that doesn't exist.
The MDMA Shadow
Lykos Therapeutics' MDMA for PTSD hit a similar wall in August 2024, despite positive Phase 3 results. An FDA advisory committee voted 9-2 against approval, citing functional unblinding (patients knew they got the real drug because they felt high), data integrity questions, and trial design limitations.
Functional unblinding is structurally unsolvable in psychedelic trials. Participants receiving 25mg of psilocybin know they received it. COMP006's use of 1mg as an "active placebo" was meant to address this, but 1mg may produce subtle effects, muddying the comparison. FDA will scrutinize this.
Strongest Counterargument
The -3.6 to -3.8 MADRS improvement, while achieving p<0.001, falls below the 5-point threshold many psychometricians consider the minimum clinically important difference. A 2020 BMC Psychiatry analysis of Cipriani et al.'s full antidepressant dataset found that standard antidepressants produce only a 2.1-point mean difference, and virtually no individual drug clears the 5-point bar either. Critics can argue that COMP360's significance reflects adequate sample size (large N, small effect) rather than clinically meaningful benefit. A patient scoring 34 on the MADRS (severe depression) who improves by 3.8 points still scores above 30. Compass must address this with responder analyses showing meaningful categorical shifts, not just mean differences.
Limitations
Several blind spots deserve explicit acknowledgment. All efficacy comparisons are cross-study; no head-to-head trial between psilocybin and esketamine exists. Cost estimates use Oregon's legal psilocybin market, not Compass's actual pricing. Insurance coverage pathways remain unknown, and REMS requirements will add unestimable costs. Durability beyond 26 weeks is unconfirmed in the Phase 3 population. "Treatment-resistant" definitions vary across trials (two versus three failed treatments), complicating cross-study comparisons. And the facilitator bottleneck calculation assumes full TRD prevalence seeking treatment, when actual adoption will be constrained by stigma, geography, and insurance long before hitting workforce limits.
The Bottom Line
Two Phase 3 trials have demonstrated what smaller studies spent a decade suggesting: synthetic psilocybin produces statistically significant, durable improvements in treatment-resistant depression with just one or two doses. For 2.8 million Americans who have already failed multiple rounds of conventional medication, that sentence alone is worth something. But the distance between a successful NDA and a scaled treatment reaching patients in Topeka and Tallahassee runs through workforce development, insurance negotiation, and regulatory frameworks still being written state by state. Compass solved the pharmacology. Everything else remains unsolved.