A pharmaceutical-grade version of LSD, taken as a dissolving tablet, significantly eased anxiety in a large late-stage trial — the latest sign that psychedelics are edging toward mainstream medicine.
Definium Therapeutics’ drug DT120 is an orally disintegrating tablet delivering a controlled dose. In its Phase 3 Voyage study in generalized anxiety disorder, announced August 12, 2026, anxiety scores on a 56-point scale fell by an average of 11.6 points at 12 weeks — a 5.4-point separation from placebo.
What the numbers mean
An 11.6-point improvement on a 56-point scale is a substantial change, and the figure that matters is the 5.4-point separation from placebo, since the placebo arm also improved.
That gap beat Wall Street’s expectations, and it sits within the range generally considered clinically meaningful on this measure — large enough that a patient and clinician would notice, rather than a statistically detectable difference with no practical import.
Effects were seen as early as day two, which distinguishes this sharply from conventional anxiety treatment where antidepressants take weeks to work.
The blinding problem nobody has solved
Psychedelic trials face a methodological difficulty that deserves stating plainly: participants know whether they received the drug.
A psychedelic dose produces unmistakable perceptual and cognitive effects over hours. Whatever the protocol says, essentially everyone in both arms correctly guesses their assignment.
That matters because expectation powerfully influences self-reported symptoms, and anxiety scores are self-reported. A participant who knows they received an experimental treatment they hope will work, after an intense experience, may report improvement partly for that reason.
The field is aware of this and has no good solution. Active placebos producing some noticeable effect have been tried and do not fully close the gap. A 5.4-point separation is real; how much of it survives the unblinding problem cannot be determined from the trial.
The 6.4-hour session
Treatment sessions lasted about 6.4 hours on average, and that number is the practical crux of whether this becomes a usable treatment.
Psychedelic therapy is not self-administered. Patients are dosed in a clinical setting with trained personnel present throughout, and a session consuming most of a working day requires a room, staff and scheduling.
Scaling that to a condition as common as generalized anxiety disorder is a genuine health system problem. The economics depend on how many sessions are required and how durable the effect proves — a single session producing a year of benefit is deliverable; repeated sessions are considerably harder.
Safety
The company reported no new safety concerns and no signals of suicidal thoughts or behaviours.
That second point is specifically important. Psychiatric drugs are scrutinised closely for suicidality signals, and psychedelics carry an additional concern about precipitating psychological distress or, rarely, psychotic episodes in susceptible individuals. Trials exclude people with personal or family histories of psychosis for that reason, which is also a limit on generalisability.
The commercial reaction
Shares rose about 15% at the open. Leerink’s Marc Goodman called the data “about as good as we could have hoped for.”
The result builds on positive Phase 3 depression data reported in June, and analysts project multibillion-dollar peak sales across anxiety, depression and PTSD if approvals follow.
Two positive Phase 3 readouts in different indications within months is what moves a drug from interesting to expected, and it substantially de-risks the platform question of whether the effect is real and reproducible.
Why generalized anxiety disorder is a notable target
Most psychedelic development has focused on treatment-resistant depression and PTSD — conditions where existing options fail conspicuously and the unmet need justifies an unusual intervention.
Generalized anxiety disorder is more common and better served by existing treatment, which raises the bar. A drug requiring a supervised six-hour session must offer something antidepressants and therapy do not — speed, durability, or efficacy in patients those have failed.
How psychedelic approval would actually work
If a drug like this is approved, the delivery arrangements would look unlike anything in conventional psychiatry, and that has become the field’s central practical question.
These are not medicines dispensed at a pharmacy. Approval would come with a restricted distribution programme specifying that dosing occurs at certified sites, with trained personnel present throughout, monitoring until the effects resolve, and defined procedures for managing acute distress during the session.
That creates an infrastructure requirement rather than a prescribing decision. A health system adopting the treatment must establish physical spaces, train and certify staff, and schedule day-long appointments — before treating anyone.
The economics are correspondingly unusual. Most of the cost is not the drug but the supervised session, which means reimbursement must cover clinician time on a scale psychiatry rarely bills for. How payers respond to that is arguably a bigger determinant of whether these treatments reach patients than any remaining question about whether they work.
Psychedelic medicines remain investigational and are given under close supervision, not self-administered. This describes an investigational drug and clinical-trial results, and is not medical advice. Do not attempt self-treatment.