Psychedelic Trials Are Producing Stunning Numbers. The Methodology Underneath Them May Not Hold.

Ben Spielberg spent years planning for a single moment. His post this week announced that the first participant in a psychedelic clinical trial had been successfully randomized at Stella Mental Health in Santa Monica — a site that had undergone DEA Schedule 1 registration, endless GCP certifications, staff training in research protocol, and, memorably, multiple meetings about a single ensuite toilet required to clear the dosing room for regulatory compliance. One participant. Years of infrastructure. And Spielberg described it, plainly, as “a Herculean achievement.” That framing deserves to stay with you as you read what follows, because it tells you something real about the distance between the field’s ambition and its operational readiness.

That same week, Phase 2 data on luvesilocin — a synthetic psilocybin derivative developed for postpartum depression — circulated widely, with a reported 70% remission rate following a single 30mg dose. Neuroscientist Matthew Wall publicly declared that psilocybin combined with cognitive behavioral therapy represents “the future,” and that the older non-directive therapy model “is going to be replaced.” And Susanne Prinz, returning from ICPR 2026, issued a pointed reminder that menstrual cycle phase, hormonal status, contraceptive use, and sex-specific pharmacokinetics are still not being systematically considered in psychedelic trial design. Three data points, three practitioners, one week. The field is moving fast. The question is whether it is moving in the right direction.

The 70% Problem

Start with luvesilocin. A 70% remission rate in postpartum depression after a single dose is, as John James Staniszewski noted in his post, “the kind of result that gets the entire psychiatric community paying attention.” He is correct. It is also the kind of result that, without careful scrutiny of the methodology underneath it, risks becoming the kind of result that collapses in Phase 3 and sets the field back five years.

PPD affects the health of both mother and child, and speed of effect matters enormously — traditional antidepressants can take weeks to reach full efficacy, per Staniszewski’s post, while psychedelic-assisted approaches are being studied for faster meaningful improvement. That clinical logic is sound. But Prinz’s post identifies a structural gap that should be on every sponsor’s protocol review checklist right now: women in psychedelic trials are often included, but the trial design stops there. Menstrual cycle phase is not tracked. Hormonal status at baseline is not stratified. Contraceptive use — which affects serotonergic pharmacokinetics — goes unrecorded. Menopausal status, which fundamentally alters how a serotonin-adjacent compound behaves, is not systematically captured.

For a drug being developed specifically for a female-only indication — postpartum depression — this gap moves from methodological sloppiness into something closer to a foundational validity problem.

Prinz draws directly on two papers from ICPR 2026: Grace Blest-Hopley’s “The 51% blind spot: Psychedelics, women’s health and the opportunity we must not miss,” and a multi-author design paper by Blest-Hopley, Haijen-Bongers, Kuypers, Aicher, and Mason on psychedelic trial design and coordination in women. The argument in both, per Prinz’s characterization, is that sex-specific factors shape subjective experience, safety, tolerability, and outcomes — and that clinical settings need to account for gender-specific themes in preparation, informed consent, therapeutic framing, and group composition, not just pharmacokinetics. This goes well beyond adding a sex covariate to a regression model. It means rethinking what a protocol collects, what a session looks like, and what integration support is offered.

Sponsors running female-predominant or female-exclusive psychedelic trials who have not read Blest-Hopley’s work are designing blind.

The Therapy Model Fork

Wall’s declaration about CBT-based psychedelic therapy is the kind of confident prediction that earns engagement precisely because it cuts against established practice. The non-directive model — grounded in the assumption that the compound itself, given in a supportive but minimally structured environment, drives the therapeutic effect — has been the dominant framework through the MAPS MDMA trials, through Imperial College’s psilocybin depression work, through most of the foundational literature. Wall is now arguing, based on a pilot study combining CBT with psilocybin, that the structured approach will prove more powerful.

The counterintuitive read here is worth sitting with. The conventional assumption is that psychedelics work by dissolving the cognitive rigidity that makes depression and trauma so treatment-resistant — and that layering structured CBT on top of that state would interfere with the mechanism rather than enhance it. Wall is arguing the opposite: that CBT provides the scaffolding that converts a pharmacologically induced neuroplastic window into durable behavioral change. If he is right, the entire therapy model underlying most existing trial data needs to be reconsidered as a comparator problem. Studies run under non-directive protocols are not measuring the same intervention as studies run under CBT-integrated protocols. They are not directly comparable. And the field is about to have both kinds of data in circulation simultaneously, which will create an interpretation crisis that regulators will need to navigate without precedent.

This is not a minor academic dispute. The FDA’s 2023 MDMA rejection — the Complete Response Letter to Lykos Therapeutics — cited concerns about functional unblinding and the inseparability of drug effect from therapy effect as core evidentiary problems. If the therapy model is now actively contested within the field, sponsors submitting NDAs or pursuing breakthrough therapy designation need to pre-specify their rationale for protocol selection with a precision that current guidance does not require and most current protocols do not achieve.

Safety, Speed, and What the Field Keeps Not Saying

Rosana Freitas shared a paper this week on psychedelic-induced hypomania and mania, with a data point worth embedding in every risk section of every active IND in this space: approximately 5.8% of cases in randomized controlled trials experienced transient euphoria or dysphoria in controlled clinical settings, with higher risk in naturalistic and unsupervised settings, particularly in bipolar I disorder. No evidence of de novo bipolar induction. Symptoms acute and self-limited. That is a relatively reassuring safety profile — in controlled settings, with screened populations, administered by trained staff in rooms that required their own regulatory approval process.

The Stella randomization story makes that condition concrete. Spielberg’s team trained clinical staff without research backgrounds, obtained DEA Schedule 1 registration after a physical DEA site visit, and built out physical infrastructure to meet dosing room requirements. That is the compliance floor for a single site to run a single arm of a single trial. Now scale that to the kind of multi-site, decentralized, or community-integrated delivery model that every health system wants to deploy if these drugs clear approval. The 5.8% adverse event rate in controlled trials is not the rate you will observe in routine clinical practice. The gap between trial conditions and post-approval deployment is where psychedelic therapy’s real safety story will be written — and no one is designing for it yet.

Meanwhile, Doug Drysdale’s post on Autobahn Therapeutics’ elunetirom data — a 16.8-point improvement on the HAMD-17 at week six, 75% response rate, 50% remission in bipolar depression through a neuronal energy restoration mechanism — is a reminder that the competitive pressure on psychedelic sponsors is not purely from each other. Non-psychedelic CNS innovation is generating rapid, durable efficacy signals in overlapping indications. Luvesilocin’s 70% PPD remission is striking. So is elunetirom’s week-two onset. The regulatory and payer audience is watching both.

The field has the efficacy signals it needed to be taken seriously. What it builds around them — the sex-stratified design standards, the therapy protocol pre-specification, the post-approval safety infrastructure — will determine whether those signals translate into durable approvals or become another chapter in the FDA’s growing list of promising CNS programs that couldn’t survive the transition from Phase 2 to regulatory reality. Spielberg’s team held many meetings about a toilet. The field now needs to hold many more meetings about the design assumptions sitting quietly inside every protocol that nobody has yet thought to question.

References

1. Ben Spielberg — First participant randomized at Stella Mental Health psychedelic trial, LinkedIn, 2025
2. Susanne Prinz — Women in psychedelic-assisted therapy and research, LinkedIn, 2025
3. Matthew Wall — Psilocybin-assisted CBT pilot results, LinkedIn, 2025
4. John James Staniszewski — Luvesilocin Phase 2 data, 70% PPD remission rate, LinkedIn, 2025
5. Rosana Freitas — Psychedelic-induced hypomania and mania, 5.8% transient adverse event rate in RCTs, LinkedIn, 2025
6. Doug Drysdale — Autobahn Therapeutics elunetirom Phase 2 bipolar depression data, LinkedIn, 2025