MM120 (LSD) Phase 3 for anxiety: what the trials actually show.

The Phase 2b GAD trial, properly framed

The Robison et al. paper published in JAMA in October 2025 [Robison 2025] covers MindMed's MMED008 study (NCT05407064). The design: 198 adults with moderate-to-severe generalized anxiety disorder (Hamilton Anxiety Rating Scale [HAM-A] ≥20 at screening), randomised to one of four single doses of MM120 (25, 50, 100, or 200 µg) or placebo, administered in a clinical setting on a single dosing day. Primary endpoint was the dose-response relationship for change in HAM-A at Week 4, assessed via MCP-Mod modelling.

In the published trial, the 100 µg and 200 µg doses both reached statistical significance versus placebo on the primary analysis. At Week 4, the least-squares mean placebo-adjusted HAM-A difference was −5.0 points (95% CI −9.6 to −0.4) for 100 µg and −6.0 points (95% CI −9.8 to −2.0) for 200 µg. The 25 µg and 50 µg arms did not separate from placebo. (Earlier MindMed press materials in 2024 had cited a larger −7.6-point placebo-adjusted effect from interim analyses — the peer-reviewed JAMA values are the figures to anchor to.)

MindMed's 12-week durability disclosures (~65% response and ~48% remission at Week 12 in the 100 µg arm, from a single dose) come from sponsor press releases rather than the peer-reviewed primary publication, and should be read accordingly. Either way, the profile — a single supervised dose with durability measured in weeks — is not what existing GAD pharmacotherapy looks like. SSRIs require chronic daily dosing for months to reach numerically smaller effects, with discontinuation syndromes on the back end.

Two caveats to hold while reading those numbers. First, n=198 across five arms means the per-arm sample is small (~39–40 per arm); the point estimates are credible but the confidence intervals are wide, and the 100 µg confidence interval reaches almost to zero. Second, classical psychedelics produce strong subjective effects that participants and raters notice — visual perceptual changes occurred in 92.5% of the 100 µg arm versus 10.3% on placebo in this trial — which means functional unblinding is a real concern. The field has not yet figured out how to credibly blind a participant against having taken a psychedelic.

What MM120 actually is — formulation matters

"LSD" in a pharmaceutical context isn't the same compound as the blotter-acid LSD that has informal associations. MindMed's MM120 is lysergide D-tartrate formulated as a sublingual lyophilized wafer — a freeze-dried tablet that dissolves under the tongue at a tightly controlled dose. The formulation choice gives MindMed two things: a defensible IP position around the specific drug product (the LSD molecule itself is generic), and reproducible pharmacokinetics that a random blotter does not deliver.

Dose is also worth being precise about. "Low-dose LSD" in the consumer microdosing context typically means 5–20 µg — sub-perceptual doses taken on a recurring schedule. MM120 at 100 µg is not microdosing. It is a fully psychedelic dose given in a supervised clinical setting over a six-to-eight hour window, with prepared, monitored, and post-session integration support. The "low-dose" framing in some coverage confuses the consumer-microdose conversation with a quite different clinical product.

The two Phase 3 trials — Voyage and Panorama

The Phase 3 program for GAD consists of two pivotal trials.

Voyage (NCT06741228): 214 participants, first patient dosed December 2024, primary completion targeted May 2026. The trial design is a 12-week double-blind Part A comparing single-dose MM120 to placebo on the HAM-A, followed by a 40-week open-label extension (Part B) measuring durability, re-dosing options, and longer-term safety. Topline expected in the first half of 2026.

Panorama (NCT06809595): 245 participants, started January 2025, primary completion November 2026. Same design structure as Voyage. Topline expected in the second half of 2026.

Both trials are sized to detect the Phase 2b effect with adequate statistical power. The FDA will want to see replication across both — single positive Phase 3 in psychiatry is not usually sufficient for approval, especially in a class with no regulatory precedent. MindMed is also running Emerge, a parallel Phase 3 in major depressive disorder, with first patient dosed April 2025 and topline targeted mid-2026.

The "first FDA-approved psychedelic" claim, audited

The framing that turns up in nearly every consumer-facing story on MM120 is "could be the first FDA-approved psychedelic." That framing is loose enough to be technically defensible and misleading enough to be worth taking apart.

Spravato (esketamine, Janssen) was FDA-approved in 2019 for treatment-resistant depression and again in 2023 for major depressive disorder with acute suicidal ideation. It is a scheduled, regulated, in-clinic dissociative drug delivered via nasal spray under a Risk Evaluation and Mitigation Strategy (REMS) protocol — pharmacologically not a classical serotonergic psychedelic, but operationally it is the psychiatry-product template that every classical-psychedelic program is following. The "first" framing requires a careful qualifier: classical 5-HT2A agonist psychedelic, of which there is currently none on the US market.

Even with that qualifier, MM120 is not necessarily first in line. COMPASS Pathways' COMP360 psilocybin for treatment-resistant depression hit its first Phase 3 primary endpoint in June 2025 with a –3.6 MADRS placebo-adjusted reduction [COMPASS 2025], and the second Phase 3 trial (COMP006) has a 26-week readout in the second half of 2026. COMPASS is targeting NDA submission in late 2026. MM120's NDA on its current schedule comes after that — meaning if the field gets a first-approved classical psychedelic, it is at least as likely to be psilocybin as it is LSD.

The competing queue — COMP360 is ahead

Three psychedelic NDA candidates are credibly in the FDA queue in the next twelve to twenty-four months.

COMPASS Pathways COMP360 psilocybin (TRD): Phase 3 COMP005 met its primary endpoint in June 2025 with –3.6 MADRS placebo-adjusted difference (p < 0.001, n=258). Phase 3 COMP006 reads out 2H 2026. NDA targeted late 2026. Currently the leading classical-psychedelic NDA candidate.

MindMed MM120 (GAD): Phase 3 Voyage 1H 2026, Phase 3 Panorama 2H 2026. NDA earliest 1H 2027.

GH Research GH001 5-MeO-DMT (TRD): Phase 2b hit in February 2025 with –15.5 MADRS placebo-adjusted reduction — the largest depression effect size in any published psychedelic trial — but Phase 3 has not yet started dosing. GH001 is an inhaled, ultra-short-duration psychedelic (effect window ~20 minutes) that solves the Lykos-style chair-time problem but introduces its own regulatory novelty as a vaporised formulation.

The first FDA-approved classical psychedelic is more likely to be psilocybin for depression than LSD for anxiety. MM120 has the cleaner anxiety signal; COMPASS has the earlier filing.

What the Lykos rejection means for MM120

In August 2024, the FDA issued a Complete Response Letter to Lykos Therapeutics on its NDA for MDMA-assisted therapy for PTSD [Lykos 2024]. The letter demanded another Phase 3 trial and explicitly flagged concerns around functional unblinding (participants knew when they had MDMA), the integrity of the psychotherapy-as-product wrap-around, and the rigor of the clinical-trial data overall. The decision reset the psychedelic-medicine field's regulatory assumptions.

MM120 inherits some Lykos problems and sidesteps others.

Sidestepped: MM120 is a drug-only label. MindMed is not asking the FDA to approve a psychotherapy protocol as part of the product. The dosing setting includes monitoring and preparation but the regulatory ask is a pharmacological intervention, not a combined drug-plus-therapy intervention. That removes one of the structural objections the FDA raised against Lykos.

Inherited: The functional-unblinding problem is identical. A participant who receives 100 µg of LSD will know within ninety minutes that they did not receive placebo. The bias this introduces in primary endpoints is real and unsolved. The FDA will look at the Phase 3 data with the post-Lykos lens — sensitivity analyses on unblinded vs. blinded subgroups, objective endpoints where they exist, blinded outcomes assessors, and the data integrity of the trial sites. MindMed's program has been planned with that lens in mind, but the regulatory environment is unmistakably harder than it was before August 2024.

Safety — what the Phase 2b actually showed

Treatment-emergent adverse events in the 100 µg arm hit 97.5% versus 56.4% in placebo. Almost all were dose-day events — illusions, hallucinations, euphoria, anxiety, headache, nausea, transient blood pressure elevation, and mydriasis — and almost all were mild to moderate and resolved within the dosing window. There were no serious adverse events, no cases of hallucinogen persisting perception disorder (HPPD), and no treatment-emergent suicidality in the double-blind period.

Two longer-term safety questions remain open. The first is cardiac valvulopathy risk: LSD is a 5-HT2B receptor agonist, and chronic 5-HT2B agonism is the mechanism by which fenfluramine, pergolide, and cabergoline produced valvular heart disease. A single 100 µg dose is far below the cumulative exposure that causes valvulopathy with those drugs, and Phase 2b did not show cardiac signal. The question for the longer term is what happens if re-dosing becomes part of the maintenance protocol — a question the open-label extension parts of Voyage and Panorama are designed to inform.

The second is abuse and diversion. LSD is Schedule I in the United States. Approval would require rescheduling to Schedule III or IV through DEA action, and the in-clinic-only delivery format (sublingual wafer dissolved under supervision) is part of the abuse-mitigation case. None of that is unprecedented — the same playbook covers Spravato and would cover psilocybin — but the rescheduling step adds 12 to 18 months between FDA approval and commercial availability.

Timeline — the honest version

Realistic dates to anchor.

1H 2026: MM120 Voyage Phase 3 topline. The single most important readout for the program.

2H 2026: Panorama Phase 3 topline. Emerge MDD topline. COMPASS COMP006 readout.

Late 2026 / 1H 2027: COMPASS files NDA for COMP360. MindMed files NDA for MM120 in GAD if both Phase 3s replicate.

2H 2027 / 2028: FDA review periods. Advisory committee meetings if convened.

2028–2029: Credible window for FDA approval and DEA rescheduling completion, assuming clean reviews. Commercial launch follows rescheduling.

Approval is not around the corner. Any timeline framing this as a 2026 or even 2027 product is reading the trial readouts as approvals, which they aren't.

What this means for patients

For someone with GAD now, MM120 is not yet a treatment option outside a clinical trial. The honest reading of where the program stands today.

Clinical trial access: Voyage and Panorama are both still recruiting at multiple US sites. ClinicalTrials.gov is the canonical place to check eligibility and active sites. Inclusion typically requires HAM-A ≥18, no recent psychedelic use, no cardiac disease, no personal or first-degree-relative history of psychotic disorders. Discuss with your prescriber before applying.

Standard-of-care alternatives: SSRIs (sertraline, escitalopram), SNRIs (venlafaxine), buspirone, and CBT remain the evidence-based first-line options for GAD. The numerically larger effect sizes in MM120 Phase 2b do not yet justify abandoning standard care — they justify watching the Phase 3 readouts.

Off-label / unregulated access: Unsupervised recreational LSD is not the same intervention as MM120 in a clinical setting. The dose, the setting, the preparation and integration support, and the safety screen are all part of the clinical effect. Wellness Radar does not recommend pursuing psychedelic experiences outside a clinical trial or, where available, a regulated jurisdiction (Oregon Measure 109, Colorado Proposition 122) with trained facilitators.