DT120: Voyage evidence and Panorama's masking test

Serotonergic intervention in GAD

DT120 is lysergide tartrate, developed in an orally disintegrating tablet formulation. It is a serotonergic psychedelic with partial agonist activity at 5-HT2A receptors. The intended clinical pattern is prolonged anxiety improvement after a supervised administration, rather than daily symptom suppression through continuous drug exposure. Analytical judgment: persistence of clinical benefit beyond acute psychoactive effects is central to the hypothesis, but does not by itself identify the neural mechanism responsible. Expectancy, treatment setting and patient selection can also influence psychiatric ratings. Generalized anxiety disorder should remain separate from the sponsor's major-depression program because HAM-A and MADRS measure different syndromes; success in depression cannot substitute for a second adequate anxiety trial. [2] [4]

Dose and supervised-session burden

The pivotal regimen is a single 100-microgram dose. The ODT formulation is intended to modify absorption and the practical dosing session, while clinical efficacy is assessed over twelve weeks. Analytical judgment: a formulation claim about faster absorption does not establish that every patient can leave supervision earlier or that anxiety benefit is stronger. Dosing-day tolerability, medical monitoring and discharge criteria are part of the intervention. Comparisons with the earlier MM120 dose-finding program should account for formulation and protocol changes. A durable response cannot be inferred simply from a long acute subjective experience, and repeated open-label dosing in an extension tests a different treatment strategy from the blinded single-dose primary period. [1] [2] [4]

Phase 2b dose-finding context

The Phase 2b study randomized 198 participants across 25, 50, 100 and 200 micrograms or placebo, with arm sizes 39, 40, 40, 40 and 39 respectively. The study's primary timing was week four, whereas the pivotal GAD endpoint is week twelve. The original presentation reported 65% response at week twelve for 100 micrograms. Analytical judgment: selecting a dose using a small multi-arm experiment is appropriate for development, but its observed response rate may regress in a larger confirmation study. Different baseline severity, rating conduct and missing-data treatment further limit direct comparison. The Phase 2b study is supportive dose-selection evidence; the later randomized Voyage dataset is the more direct benchmark for what Panorama is designed to replicate. [3] [4]

Voyage Phase 3 benchmark

Voyage randomized 214 adults, 107 per arm, to 100 micrograms or placebo. At week twelve, HAM-A least-squares mean changes were −11.6 and −6.2, an adjusted difference of −5.4 points, p<0.0001. Response, defined as at least fifty-percent HAM-A improvement, was 43% versus 16%; remission was 14% versus 4%. Analytical judgment: these categories answer different questions from mean change and should not be conflated. The result supports a controlled average benefit but does not mean most patients entered remission. Percentage response counts should not be reverse-engineered from rounded values. Confidence intervals and full disposition remain important, particularly when comparing the substantially different headline response percentage with the earlier small Phase 2b study. [1]

Panorama and functional unblinding

Panorama completed enrollment of 245 participants randomized 2:1:2 to 100 micrograms, 50 micrograms or placebo. Its primary comparison remains 100 micrograms versus placebo for week-twelve HAM-A change; the low-dose arm is intended to complicate recognition of assigned treatment. The company guides September 2026 topline. Analytical judgment: an active low dose can improve masking but does not guarantee it, because subjective intensity may still differ and the low dose may have real efficacy. Interpretation should include participant and assessor blinding information, rescue medication and the prespecified analysis hierarchy. A replicated effect with complementary masking design would be more persuasive than one trial alone, while a smaller effect would require examination of confidence intervals rather than an automatic declaration of contradiction. [1] [2] [4] [5]

Safety and funding context

Voyage described mainly transient mild-to-moderate dosing-day events and no new suicidality signal; this is not a complete arm-level psychiatric safety table. Supervised administration still requires assessment of acute distress, cardiovascular changes, delayed symptoms and discontinuations. Analytical judgment: screening out unstable psychiatric or medical conditions may make trial safety less generalizable to unselected practice. Twelve-week follow-up after one dose cannot define the safety of indefinite repeat treatment. Definium reported approximately $1.1 billion in cash and investments at June 2026 following substantial equity financing, with operating runway projected into 2030. That resources the program but also reflects dilution and an operating assumption. The clinical conclusion should turn on reproducible blinded anxiety benefit, durability and session burden, not the size of the financing. [1] [2] [4]

September 14: PANORAMA primary endpoint met

The company published PANORAMA topline results on September 14 at 07:00 EDT. The prior estimated window is superseded by an observed disclosure date and this event moves to History. The Phase 2b and VOYAGE sections above remain separate evidence; the new PANORAMA results are not replacements for those study populations. [6]

PANORAMA randomized 245 adults aged 18–74 with DSM-5 generalized anxiety disorder and HAM-A scores of at least 20 at screening and baseline. Allocation was 2:1:2 to one dose of DT120 ODT 100 micrograms (96), 50 micrograms (52), or placebo (97). The blinded 12-week Part A is followed by a 40-week open-label extension allowing up to four repeat doses. The interim release does not establish durability across that extension.

Baseline HAM-A was 28.3 for 100 micrograms and 28.0 for placebo. The primary week-12 difference was −5.1 points, with p<0.0001 and Cohen's d 0.64. Lower scores mean less anxiety. The company identified the CGI-S week-12, HAM-A week-1 and CGI-S day-2 comparisons below as key multiplicity-controlled secondary endpoints. Other responder endpoints should not automatically inherit that statistical protection.

Responder outcomes and the lower dose

At week 12, at least 50% HAM-A improvement occurred in 32% versus 14% (reported difference 18 percentage points; p<0.05); remission defined as HAM-A≤7 was 15% versus 4% (reported 12 points; p<0.05); and mild symptoms defined as HAM-A<16 were 35% versus 17% (reported 19 points; p<0.01). These published contrasts retain source rounding. Patient counts, confidence intervals and a full analysis plan were not supplied in the extracted topline tables. [6]

The 50-microgram arm was not powered for formal comparisons. Its placebo-adjusted HAM-A change was −3.5 points at week 4 and −3.6 at week 12. Absolute arm means were not disclosed in this summary. A numerical difference between 50 and 100 micrograms is not a demonstrated dose-comparison result.

PANORAMA safety and observation burden

The percentages below are reported values. Safety-population denominators were not explicitly supplied alongside these percentages, so randomized counts must not be used to manufacture n/N. Overall discontinuation is distinct from discontinuation caused by an adverse event. The release reported no drug-related serious adverse events and no new suicidality signal; that is not a statement that all-cause serious events or deaths were zero. [6]

Dosing-day illusion was reported in 68% and 61%, nausea in 37% and 26%, and headache in 24% and 28% for 100 and 50 micrograms, respectively. These specific-event placebo rates were not supplied. The distinct acute perceptual effects also make functional unblinding a relevant interpretation question for subjective symptom scales; no formal blinding-integrity analysis was recovered.

Observation lasted at least eight hours. For 100 micrograms, mean and median time to end-of-session criteria were 6.2 and 6.0 hours, and 94% met those criteria by eight hours. This does not mean participants left before the minimum observation period. The separate program-wide figure of more than 1,000 sessions through September 10, with 97% meeting criteria by eight hours, counts sessions rather than unique people.

Regulatory implications and remaining evidence

Analytical judgment: replication in a randomized Phase 3 trial reduces uncertainty about a single-dose effect over 12 weeks. It does not resolve the repeat-treatment schedule, long-term tolerability, missing-data sensitivity or how an eight-hour supervised treatment will operate outside trials. Full protocol and statistical analysis plan remain necessary to assess estimand choices and robustness. [6]

Management plans a pre-NDA meeting in the fourth quarter of 2026 and an NDA submission in the first half of 2027. Neither is an accepted application or an approval. Regulatory review will need the integrated efficacy, safety and administration package; the September 14 event is a clinical readout.

Sources / References

1. Definium: Voyage Phase 3 results, August 12, 2026
2. Definium Q2 2026 SEC update
3. MindMed: Phase 2b GAD original APA 2024 presentation
4. Definium 2025 Form 10-K: design and adaptive sample-size plan
5. Definium August 2026: September Panorama discussion
6. Definium · PANORAMA Phase 3 topline · September 14, 2026