The Psilocybin Review That Cannot Stop the Approval, and Does Not Need To - PSA's Regulatory Dispatch, Vol. III

The Psilocybin Review That Cannot Stop the Approval, and Does Not Need To

Psychedelic State(s) of America's Regulatory Dispatch, Vol. III:

By Odette Hauke

PSA’s Regulatory Dispatch, Vol. III: October 28th, 2026

ICER votes on Compass Pathways' synthetic psilocybin on May 20, 2027. By then the drug may have been on sale for months. When this same panel judged MDMA, there was no price to argue about. This time there will be. Comments close November 10.

Comments on ICER's draft scope close at 5:00 PM ET on November 10, 2026.1

There is a nonprofit in Boston that decides what a new medicine ought to cost. It cannot keep a drug off the market. It cannot order an insurer to do anything. Insurers read it anyway, and so do the companies that have to sell to them. It is called the Institute for Clinical and Economic Review. Everyone in the drug business calls it ICER.

On October 21, ICER announced that psilocybin is next.¹

The drug is COMP360, the synthetic psilocybin Compass Pathways has spent a decade testing in people whose depression has not lifted after trying everything else. ICER will spend about seven months deciding what it should cost, then put that judgment to a public vote.

The trade press read this as a watchdog circling the psychedelics industry. That is the least interesting thing about it.

The interesting thing is sitting on ICER's own calendar.

Figure 1. The two calendars. ICER publishes its findings on May 5, 2027 and votes on May 20. Compass said in May 2026 that the last pieces of its FDA application were on track to go in by the end of that year, and the voucher it holds can shorten the agency's review to a month or two. The approval and launch window shown is our estimate from those two facts, not an announced date, and no FDA decision date is public. The November 10, 2026 comment deadline is marked because it is the only date on either calendar a reader can still act on. The key separates dates ICER has already set from the Compass dates, which are company statements or our estimate.

Compass has been filing with the FDA in pieces as it goes, rather than all at once at the end, and holds a voucher that can shrink the agency's review to a month or two.3,4 In May it said the last pieces were on track to go in by the end of this year.⁴

So the drug could be approved, priced and on pharmacy shelves before ICER's panel sits down. ICER gets no vote on approval and has never claimed one. But a review that lands after launch is not a weaker review. It is a review with a number in front of it.

What happened to MDMA, and why this is different

The industry's reference point for all of this is MDMA.

In June 2024 this same organization rated the evidence for MDMA-assisted therapy "insufficient," the bottom of its scale. Fourteen of fifteen panelists said it failed to beat no treatment at all.⁵ Two months later the FDA rejected the drug and told Lykos Therapeutics to run another trial.⁶ Nothing in the public record says ICER caused that, and this piece does not claim it did.

What matters is the one thing ICER did not do to Lykos.

It never said what the drug should cost. ICER's signature output is a recommended price, the number it thinks an insurer should be willing to pay. The panel never took its vote on whether the therapy was worth the money either. The reason was procedural: there was "no firm estimate yet of a potential launch price."⁵ No price, no price verdict.

That was a shield, and it belonged to Lykos. A company whose drug has no price cannot be told its price is wrong.

Compass will not have it. ICER's rules let the panel vote on value using a price that has been announced, or one merely assumed.⁸ With a launch in between, nothing has to be assumed. Compass gets the whole apparatus: a recommended price, a spending alert, and a recorded vote.

The Lykos review met a drug whose evidence was in doubt. A Compass review would not doubt the evidence much. It would price it.

Three things to be skeptical about, including in this article

A second published analysis reaches the opposite conclusion from the one this article builds on. Ziadi and Park find psilocybin-assisted therapy both more effective and cheaper than standard care, with the savings holding across the prices they tested. The range they tested was not retrievable, so whether that finding survives at the prices discussed here cannot be checked.¹⁸

The price range quoted throughout is also not a price. It is what Compass executives told an investor conference they were aiming at.¹⁴ Companies move off those numbers all the time.

And the trade story everyone has cited, from Pink Sheet, sits behind a paywall. Its headline says an ICER review of Compass is unlikely to repeat the Lykos damage.⁷

What ICER has already decided, before anyone has commented

Alongside the announcement, ICER posted a draft scope: the document saying what it will measure, and what it will measure the drug against.³ It warrants close reading. Three things in it will matter more than anything Compass says at the meeting.

The drug is listed without the therapy. Under interventions the scope says only this: "Psilocybin (25 mg)." No number of doses. No mention of the trained people who sit with a patient for the hours the drug is working. Compass has published that support model as a formal clinical framework in a peer-reviewed journal.¹¹ The scope does not acknowledge it exists. If the supervision is not in the model, its cost is not in the price, and the thing being priced is not the thing being sold.

Two very different drugs share one line. ICER puts esketamine, the branded nasal spray sold as Spravato, on the same line as intravenous ketamine, a generic anesthetic that doctors use for depression even though it was never approved for it. One costs more than ten times the other to keep a patient on for a year.¹² The whole list is also qualified by "Data permitting," which is ICER reserving the right to drop a comparison if it decides the evidence is too thin. Which of those two the model treats as the thing psilocybin replaces will matter more than every clinical argument Compass makes.

The clock runs five years, not a lifetime. ICER says why: there is not enough evidence yet that the benefit lasts.³ Compass's whole case is that one or two doses produce something durable. The horizon that would reward durability is the one ICER has made conditional, and the condition is the thing in dispute.

There is a fourth, and it is the strangest. ICER says it will calculate the cost per patient put into remission.³ Neither big trial was built to show that. Both measured change on the Montgomery-Asberg Depression Rating Scale, a questionnaire clinicians use to score how severe a depression is, after six weeks.9,10 A score that moves is not a patient in remission. So ICER intends to price a result the trials did not set out to produce, stitching together studies never run against each other. In its own 2019 look at this condition it found no trials at all comparing esketamine to the alternatives it was measured against.¹³

The arithmetic, for readers who want it

The numbers live here. The short version sits on either side of them: what ICER decides to compare the drug against sets the price, and the price barely moves the money.

Start with the only published US study that works out whether psilocybin therapy is worth its cost, and the only one that assumes patients already have access to the expensive nasal spray.¹⁵ Its reported results let the underlying relationship be rebuilt, and the rebuild lands on that paper's own headline figure and on both of the checks it ran against itself.

Compass has described pricing per dose while targeting an annual cost in Spravato's range, normalized to about three doses a year.¹⁴ Priced that way it is a recurring cost by construction, and that decides the review. If a patient pays for this every year, and the alternative is also paid for every year, then how long you follow the patient stops mattering. The benefit the drug has to deliver is just the yearly price difference, divided by what ICER is willing to pay for a year of healthy life.

Figure 2. What price the evidence can carry. For each comparison, the solid bar is the highest yearly price per patient that ICER could call good value on the benefit the trials actually measured. The hatched bar assumes instead that every treated patient is completely cured and stays cured, which no trial supports. No solid bar reaches the price Compass is aiming at. Only against the expensive nasal spray does that price sit below even the best-case bar. Calculated by the author from the published cost study (Avanceña et al. 2025) and what the alternatives cost in ICER's 2019 esketamine review, in 2019 dollars before insurer discounts.

The same model scores untreated major depression at 0.50 where 1.00 is perfect health, and full remission at 0.82.15 That gap is the entire benefit available from curing a depression. At the bottom of the range Compass has described, measured against standard care, the price demands slightly more than the whole gap. It would require lifting the average treated patient into complete remission and holding them there permanently, just to reach the upper edge of what ICER considers acceptable. Against esketamine, which is expensive, the requirement falls by more than half and the price becomes defensible. Against generic ketamine, which is cheap, it does not.

Figure 3. What a year of healthy life would cost. The cost of buying one extra year of healthy life, on the benefit the trials measured, at yearly prices of $50,000 and $65,000 against each alternative. Every combination lands between four and fourteen times the most ICER normally pays, marked by the dashed line near the axis. Even the most favorable case on the chart, $50,000 against the nasal spray, costs $619,355. Author's calculation from Avanceña et al. 2025 and ICER's 2019 comparison costs, in 2019 dollars before insurer discounts.

One choice of what to compare it against is the difference between a price Compass can live with and one it cannot. That choice is being made now, and the window shuts on November 10.

Cost-effectiveness is the main test everyone pays attention to. There is a second test, and it has a hard dollar limit Compass cannot dispute, because ICER publishes the number ahead of time. ICER says a new drug should not add more than $821 million a year to national health spending, down from $880 million in October 2025. The limit comes from how much extra spending the economy can absorb before health costs outrun it. The report will show what share of eligible patients could be treated at different prices without crossing that line.¹⁶

Its 2019 esketamine review shows what that page looks like, and it is not flattering: at the sticker price, only about one eligible patient in six could be treated before the limit was hit.¹² One oddity in that report limits how precisely any of this can be rebuilt, in fairness. The same document credits the identical result to two different spending limits in two different passages.¹²

Figure 4. How far the budget stretches. The curve scales ICER's own published 2019 result by the extra cost per patient, a ratio in which ICER's five-year patient accounting cancels out. ICER's four published figures are plotted as diamonds to check the method: it returns 16% and 21% almost exactly and 32% to within about a point, drifting at the cheapest price, where it returns 58.1% against 62% published. The horizontal scale compresses as it rises. Author's calculation from ICER's 2019 final report and its current spending limit.

Scaling that published result by cost per patient reproduces ICER's other published figures closely, which is the only reason the next number is worth printing. At the price Compass has described, the share of eligible patients who fit under the limit falls to roughly one in twenty-four.

Figure 5. How many patients fit inside the spending limit. The share of eligible patients who can be treated in a year before ICER's spending limit is reached, at the nasal spray's 2019 price and at COMP360 prices of $50,000 and $65,000. The limit is identical in all three cases; only the extra cost per patient differs, and it is that cost, printed beneath each bar, that does the work. Author's calculation, scaling ICER's published 2019 result.

Which brings up the result that should interest anyone holding the stock.

Revenue is patients multiplied by price. When a spending limit binds, raising the price does not raise revenue, because the number of patients who fit underneath falls by almost exactly as much. The limit cancels out of the calculation, so this holds whatever the limit is, however many patients are eligible, and whatever accounting ICER uses.

Figure 6. Why price is not the lever. The money Compass can reach is the spending allowance multiplied by price, divided by price minus the costs the drug saves, so the size of the allowance cancels out and the shape holds whatever the limit is, however many patients are eligible, and whatever patient accounting ICER uses. Above roughly $25,000 a year the line is almost flat: another $65,000 of price moves it by about 4%. Author's calculation.

Raising the annual price by 30% moves reachable US revenue by about half of one percent, in the wrong direction. The pricing debate that has occupied investors is, on these numbers, close to irrelevant. What moves revenue is what the drug displaces. Replace branded esketamine and the cost offset is large, the number of patients who fit under the limit multiplies, and revenue moves in a way no price increase can match. The commercial case for COMP360 is a substitution case, not a pricing case. What it replaces is exactly what that list of comparisons decides.

The quieter invitation

The scope also invites anyone with an interest to name "wasteful or lower-value services" in depression care that could be "reduced, eliminated, or made more efficient" to free money for something better.³ Patient organizations are among those invited to answer. They are being asked to nominate parts of current depression care for defunding, in a proceeding whose other product is a ceiling price for the thing that would replace them.

What would prove this wrong

Four ways the argument above fails.

1. The approval slips past the meeting. If the review stalls, ICER arrives with no launch price and the Lykos shield reappears.
2. Patients get a course, not a subscription. If real-world dosing settles at one or two treatments rather than a yearly repeat, the price is paid once and the five-year clock starts working for Compass instead of against it. The required benefit falls by roughly a factor of five.
3. Esketamine anchors the model. Every ceiling in these figures rises if the comparison runs against the expensive branded drug at today's costs rather than 2019's.
4. Lost work time counts. Roughly two-thirds of depression's cost to the country is paid in workplace productivity, and ICER's framework can promote that accounting into the main analysis.3,8 It is where this therapy's case is strongest.

Why this one sets the pattern

The list of organizations ICER consulted includes a company whose drug is not under review at all: Definium Therapeutics, which is developing a form of LSD for other conditions.2,17 An assessment of one product does not usually consult the maker of another. Whatever the panel adopts on May 20, 2027 becomes the starting frame for every supervised psychedelic that follows.

File by November 10

Here is the part that is actually actionable, and the window is short.

Anyone can write to ICER about this scope. Not a lawyer, not a lobbyist: anyone. It costs nothing but an afternoon, and it is the cheapest influence anyone in this field will ever get. Three pages maximum, 12-point Times New Roman, to publiccomments@icer.org, received by 5:00 PM ET on November 10, 2026. Received, not postmarked. Miss it by an hour and it does not count.¹

Two asks are worth more than all the rest. Make ICER say whether esketamine or generic ketamine is the thing psilocybin replaces, and object to their sharing a line. Make it say whether a patient gets a course of treatment or a yearly repeat, because that one assumption moves the whole calculation by a factor of five.

After that the schedule hardens: a revised scope on or about November 18, the modeling plan in January, the draft report in March.² By the time there is something to argue with, the argument will largely be over.

Method, and what is not established here

Verified against primary records: every ICER date, quotation, comparison and rule above comes from ICER's own posted documents, and what the trials set out to measure comes from the public trial registry. The regulatory status and the fourth-quarter filing target come from Compass's first-quarter 2026 results release, which also reported a cash runway "well beyond launch and into 2028."⁴ A second-quarter release was not retrievable, so no later financial figure is asserted here.

Company guidance, labeled as such: the $50,000 to $65,000 range is what Compass executives told an investor conference they were aiming at. It is not a price the company has set, and it was checked in earlier work rather than re-retrieved for this piece.¹⁴ No FDA decision date is public.

Own derivations: the figures and every threshold in them are the author's reconstruction from one published model, checked against that model's own published tests of itself. They are not a prediction of what ICER will find, since ICER will use the full trial data and will publish its own assumptions in January 2027.2 What the alternatives cost comes from ICER's 2019 estimates, in 2019 dollars and before the discounts insurers negotiate. The direction of every finding holds if those costs move; the exact dollar ceilings do not.

Not assessed: whether the hours a patient spends supervised get billed as monitoring or as therapy, which no insurer has yet said in public. Approval is also not sale. Psilocybin remains in Schedule I, and federal law gives the Attorney General ninety days after approval to issue the rule that must precede any lawful sale.¹⁹

Sources

1. Institute for Clinical and Economic Review. "Institute for Clinical and Economic Review to Assess Therapy for Treatment-Resistant Depression." Press release, October 21, 2026. https://icer.org/news-insights/press-releases/institute-for-clinical-and-economic-review-to-assess-therapy-for-treatment-resistant-depression/ ↩
2. Institute for Clinical and Economic Review. "Depression: Treatment-Resistant" assessment page, full timeline and stakeholder list (New England CEPAC main review May 20, 2027), accessed October 25, 2026. https://icer.org/assessment/treatment-resistant-depression-2027/ ; and "Psilocybin for Treatment-Resistant Depression: Public Meeting," draft virtual meeting agenda, May 20, 2027. https://icer.org/wp-content/uploads/2026/10/ICER_TRD-Draft-Agenda_102126.pdf ↩
3. Institute for Clinical and Economic Review. "Psilocybin for Treatment-Resistant Depression: Draft Background and Scope," October 21, 2026. https://icer.org/wp-content/uploads/2026/10/ICER_TRD_Draft-Scope_For-Publication_102126-1.pdf ↩
4. Compass Pathways plc. "Compass Pathways Announces First Quarter 2026 Financial Results and Business Highlights," May 13, 2026 (FDA granted NDA rolling submission and review; CNPV awarded; rolling submission underway with sections already submitted; final NDA submission on track for Q4; $466 million cash at end of first quarter, runway well beyond launch and into 2028). https://s204.q4cdn.com/927483861/files/doc_news/Compass-Pathways-Announces-First-Quarter-2026-Financial-Results-and-Business-Highlights-2026.pdf ↩
5. Institute for Clinical and Economic Review. "Report at a Glance: Post-Traumatic Stress Disorder," June 2024 (MDMA-AP rated insufficient; votes 14-1 and 15-0; no Health-Benefit Price Benchmark; no vote on long-term value for money absent a firm launch price estimate). https://icer.org/wp-content/uploads/2024/06/PTSD_RAAG_June-2024.pdf ↩
6. Lykos Therapeutics. "Lykos Therapeutics Announces Complete Response Letter for Midomafetamine Capsules for PTSD." Press release, August 9, 2024 (CRL; FDA requested an additional Phase 3 study). Corroborated in Mind Medicine (MindMed) Inc., Form 10-K for FY2024. ↩
7. Pink Sheet (Citeline). "ICER Review Of Compass Pathways' Psilocybin Unlikely To Replicate Lykos Impact," October 2026. Headline cited as a headline; the article is paywalled and was not retrieved. https://insights.citeline.com/pink-sheet/market-access/health-technology-assessment/ ↩
8. Institute for Clinical and Economic Review. "2023-2026 Value Assessment Framework," updated July 1, 2026 (threshold range $100,000-$150,000 per QALY and per evLY; estimated or placeholder pricing for products not yet approved; long-term value for money voted at current or assumed prices). https://icer.org/wp-content/uploads/2025/12/ICER_2023_2026_VAF_070126.pdf ↩
9. ClinicalTrials.gov. COMP005, NCT05624268, "Efficacy, Safety, and Tolerability of COMP360 in Participants With TRD" (n=258 actual; primary outcome: COMP360 25 mg versus placebo, change from baseline in MADRS total score, week 6).
10. ClinicalTrials.gov. COMP006, NCT05711940, "Efficacy, Safety, and Tolerability of Two Administrations of COMP360 in Participants With TRD" (n=572 actual; primary outcome: COMP360 25 mg versus COMP360 1 mg, change from baseline in MADRS total score, week 6).
11. Kirlić N, Lennard-Jones M, Atli M, Malievskaia E, Modlin NL. "Compass Psychological Support Model for COMP360 Psilocybin Treatment of Serious Mental Health Conditions." American Journal of Psychiatry. 2025;182(2):126-132. doi:10.1176/appi.ajp.20230884. PMID 39741434. ↩
12. Institute for Clinical and Economic Review. "A Look At: Esketamine for Treatment-Resistant Depression," report at a glance, 2019 ($198,000 per QALY and $2,592,000 per life-year gained; annual list price $32,400 with benchmark prices of $17,700 at $100,000 per QALY and $25,200 at $150,000, a 25%-52% discount; continuing annual costs $30,800 esketamine versus $2,500 IV ketamine excluding lost work time; unanimous vote that evidence was insufficient to distinguish esketamine from ketamine, TMS, ECT or olanzapine). https://icer.org/wp-content/uploads/2020/10/TRD-RAAG_061919.pdf ↩
13. Institute for Clinical and Economic Review. "Esketamine for the Treatment of Treatment-Resistant Depression: Final Evidence Report," June 20, 2019 (twelve rTMS randomized trials reviewed, eleven sham-controlled and one head-to-head versus ECT; no trials found comparing esketamine to rTMS, ECT or augmentation with olanzapine). https://icer.org/wp-content/uploads/2020/10/ICER_TRD_Final_Evidence_Report_062019.pdf ↩
14. Compass Pathways management remarks, Deutsche Bank 2026 Healthcare Summit, September 2026, as reported (per-dose pricing targeting an annualized cost in the approximate range of Spravato's stated $50,000-$65,000 per year; average two to four doses annually, normalized to three; Spravato administered approximately 25-50 times per year). Company guidance, not a published price; not re-retrieved for this piece. ↩
15. Avanceña ALV, Vuong L, Kahn JG, Marseille E. "Psilocybin-assisted therapy for treatment-resistant depression in the US: a model-based cost-effectiveness analysis." Translational Psychiatry. 2025;15:330. doi:10.1038/s41398-025-03556-4. PMID 40883271. ↩
16. Institute for Clinical and Economic Review. "Value Assessment Framework" methods page, "Potential Budget Impact Threshold Updated for 2025-2026," accessed October 25, 2026 (annual budget impact threshold for prescription drug therapies lowered from $880 million to $821 million in October 2025; threshold calculated as double the average net budget impact that would contribute to health care cost growth exceeding growth in the US economy; reports include the maximum percentage of eligible patients treatable at multiple price points without exceeding the threshold). https://icer.org/our-approach/methods-process/value-assessment-framework/ ↩
17. ClinicalTrials.gov. DT120 trial records listing sponsor Definium Therapeutics US, Inc. and intervention DT120 with other names MM120, lysergide tartrate and LSD D-Tartrate (NCT06741228; NCT05407064).
18. Ziadi Y, Park T. "Cost-Effectiveness of Psilocybin-Assisted Therapy Versus Standard of Care for Patients With Treatment-Resistant Depression." Value in Health Regional Issues. 2026;56:101605. doi:10.1016/j.vhri.2026.101605. PMID 41842876. ↩
19. 21 U.S.C. 811(j)(1)-(2) (interim final rule not later than 90 days after the later of the HHS scheduling recommendation or notification that the Secretary has approved an application). https://www.govinfo.gov/content/pkg/USCODE-2023-title21/html/USCODE-2023-title21-chap13-subchapI-partB-sec811.htm ↩

Author disclosure. Odette Hauke is an epidemiologist and regulatory scientist in independent practice, advising sponsors developing oncology, rare disease and psychedelic drugs, including LSD, MDMA, DMT and ibogaine. She was previously Associate Director of Regulatory Affairs at AtaiBeckley and has served as a regulatory consultant to its ibogaine, R-MDMA and DMT programs. With more than thirteen years of experience before FDA, DEA, EMA, MHRA and Health Canada, she has authored hundreds of investigational new drug applications in oncology and rare disease.