8.1 Points on the Depression Scale That Clinicians Can Actually See. One Pill. One Dose. Six Hours.
Definium Therapeutics' DT120, a single orally disintegrating tablet containing 100 micrograms of pharmaceutical-grade LSD, just posted the largest placebo-adjusted improvement in a Phase 3 depression trial ever recorded. It is also the first antidepressant to cross the threshold where a doctor can tell, by observation alone, that the drug worked.
Negative 8.1 points on the Montgomery-Åsberg Depression Rating Scale. That is the placebo-adjusted improvement Definium Therapeutics reported today from its Phase 3 Emerge trial, a randomized, double-blind study of 149 adults with major depressive disorder across 20 clinical sites. One tablet, one visit, discharged the same day, no ongoing prescriptions, no weekly clinic appointments, no maintenance dosing.
To understand why that number matters, you first need to know what it replaced.
The Depression Scale Nobody Can See
SSRIs have dominated depression treatment for three decades. Their drug-placebo difference on the MADRS, the gold-standard clinician-rated scale for depression severity, lands between 2 and 3 points across 131 randomized controlled trials enrolling over 27,000 patients. That is not a controversial number; it is a meta-analytic consensus published in BMC Psychiatry, replicated by Jakobsen et al. at 1.94 HDRS-equivalent points across an even larger dataset.
It is also a number that sits below the clinical detection threshold. Far below.
When researchers at the Technical University of Munich calibrated MADRS scores against the Clinical Global Impression-Improvement scale (the instrument where a clinician simply rates whether a patient got better), they found that "minimal improvement" requires roughly 8 points of change. Below that threshold, clinicians performing blinded assessments cannot reliably distinguish a treated patient from an untreated one sitting in the same waiting room, meaning the gold-standard measure of whether an antidepressant works at the population level is functionally invisible to the doctors prescribing it. SSRIs at 2 to 3 points fall so far beneath that line that a psychiatrist evaluating the patient without knowing their medication status would, on average, call the outcome "no change."
Every antidepressant approved in the past 30 years sits below this threshold. Every single one.
What DT120 Actually Showed
Emerge enrolled 149 adults aged 18 to 74 with DSM-5 major depressive disorder and baseline MADRS scores of 26 or above, meaning moderate-to-severe depression. Participants received either DT120 (a 100-microgram lysergide orally disintegrating tablet) or placebo in a single supervised visit, then returned for follow-up assessments at weeks 1, 2, 4, 6, 8, and 12.
Results at each measured timepoint, from the company's press release and confirmed in an 8-K SEC filing:
| Timepoint | MADRS Δ vs. Placebo | p-value |
|---|---|---|
| Week 1 | -14.2 points | <0.0001 |
| Week 6 (primary) | -8.1 points | <0.0001 |
| Week 12 | -7.3 points | <0.0001 |
Week 1: negative 14.2 points, not a typo. Within seven days of a single dose, the separation between drug and placebo was nearly twice the effect seen at the primary endpoint. By Week 6 it had settled to -8.1, and by Week 12 it held at -7.3, suggesting durability through at least three months without redosing.
Safety data was clean: of reported adverse events, 99% were mild to moderate, with zero serious adverse events and no suicidality signal on the Columbia Suicide Severity Rating Scale. Average time to discharge from the supervised setting: 5.8 hours, median 5.1, with every participant cleared by hour 8. Nobody was hospitalized.
Nobody relapsed acutely.
The Efficacy Gap Nobody Has Published
Cross-trial comparisons carry real caveats, which the limitations section below addresses. With those caveats stated up front, no published analysis has placed all four major antidepressant paradigms on one scale with their dosing burdens, so here is that table.
| Treatment | MADRS Δ vs. Placebo | Doses Needed | Onset | Monitoring per Dose | Population |
|---|---|---|---|---|---|
| SSRIs/SNRIs | ~2–3 pts | Daily, indefinite | 4–6 weeks | None | MDD |
| Esketamine (Spravato) | ~4 pts | Repeated nasal spray + oral antidepressant | 24 hours | 2+ hrs/session (REMS) | TRD |
| COMP360 (psilocybin) | 3.6–3.8 pts | 1–2 doses | 1–3 weeks | ~8 hrs/session | TRD |
| DT120 (LSD) | 8.1 pts | Single tablet | 1 week | ~6 hrs | MDD |
Sources: Hengartner & Plöderl 2020, Spravato TRANSFORM-2, Compass Pathways COMP005/006, Definium Emerge press release. TRD = treatment-resistant depression (failed 2+ prior treatments). MDD = major depressive disorder.
DT120's 8.1-point separation makes it the first compound in any drug class to breach the 8-point clinician-detection threshold in a Phase 3 trial. That sentence requires a caveat the size of a footnote: it tested in MDD, not the harder-to-treat TRD population where psilocybin and esketamine were studied, and that population difference inflates the apparent gap, as the limitations section details below.
The Math: Why the Threshold Matters
Twenty-one million American adults experience a major depressive episode each year, per NIMH data. Roughly 30% of those, about 6.3 million, will not respond adequately to their first two antidepressants and meet the clinical definition of treatment-resistant depression. Standard of care for the remaining 14.7 million is an SSRI that, on average, produces an improvement indistinguishable from placebo to their own doctor.
None of this means SSRIs do nothing; individual patients clearly benefit, and the population-level effect is statistically significant. But when the average treated patient looks the same as the average untreated patient to a trained clinician, the treatment has a measurement problem that compounds across the healthcare system: doctors cannot tell who is responding without lengthy titration, patients cycle through three or four medications over months, and the economic cost of failed trials runs to $29,000 per patient per year in excess healthcare utilization.
A drug that clears the detection threshold changes the clinical workflow from start to finish, because it means a psychiatrist can determine whether the treatment is working at the very first follow-up visit rather than prescribing blind for months while the patient suffers. Week 1 results that separate from placebo by 14.2 points mean a clinician can assess response at the first follow-up visit rather than waiting six weeks for an SSRI to maybe work.
Regulatory Tailwinds
DT120 holds FDA Breakthrough Therapy Designation, which guarantees intensive regulatory guidance and a rolling review process. The broader policy environment has shifted since April 2026, when the Trump administration signed an executive order directing the FDA to expedite psychedelic research for mental health. Three psychedelic compounds have received Rare Pediatric Disease Priority Review Vouchers: COMP360 (psilocybin for TRD), Usona's psilocybin (for MDD), and Transcend Therapeutics' methylone (for PTSD), and the VA launched its first MDMA trial.
Wall Street noticed immediately: DFTX shares jumped roughly 52% on the Emerge data, and the company reported $411.6 million in cash. Analyst estimates for peak sales range from $620 million (RBC Capital Markets) to $1.5 to $2 billion (Leerink Partners) for the MDD indication alone, with potential expansion into anxiety and PTSD still unlocked.
Limitations
Typical Phase 3 depression studies enroll 300 to 500 patients. This one enrolled 149. Definium's second Phase 3 trial remains ongoing, and until it reads out, the 8.1-point result stands on a single dataset. The history of psychiatry is littered with impressive Phase 3 results that failed replication. It happens constantly.
Population differences distort the comparison table above. Dramatically, jarringly so.
DT120 was tested in MDD patients with baseline MADRS scores above 26, a population that includes anyone with moderate-to-severe depression regardless of treatment history, which makes these patients substantially easier to move on any depression scale than the TRD cohorts where psilocybin and esketamine were tested. COMP360 and esketamine targeted TRD, a population defined by having already failed multiple treatments. TRD patients are, by definition, harder to move on any depression scale, and effect sizes in TRD consistently run smaller than in MDD for the same drug. Comparing DT120's 8.1 to COMP360's 3.8 without noting this is misleading, which is why the table explicitly labels each population.
Blinding in psychedelic trials is structurally compromised. Patients know. That matters here. A patient who receives 100 micrograms of LSD knows it within 90 minutes. Expectation effects in depression are large, and functional unblinding can inflate both drug-arm improvement and placebo-arm underperformance, a methodological vulnerability that has plagued every psychedelic trial ever conducted because there is no inert substance on Earth that mimics a twelve-hour psychedelic experience without producing one. Definium has not published its blinding assessment data.
LSD remains Schedule I under federal law. FDA approval does not override DEA scheduling, and commercial manufacturing of a Schedule I substance requires a separate DEA rulemaking process that has no guaranteed timeline. Psilocybin faces the same hurdle; neither compound has cleared it yet, and no psychedelic compound has ever completed the DEA rescheduling process required for commercial sale.
Durability data extends to just 12 weeks, and depression is a chronic, relapsing condition. Whether a single dose of DT120 holds at 6, 12, or 24 months determines whether it is a cure or a periodic treatment, and the difference matters enormously for cost-effectiveness modeling.
Against
The strongest case against reading DT120 as a breakthrough is that the trial design made it easy to look like one. MDD patients above MADRS 26 are sick enough to show large placebo-adjusted effects but have not yet demonstrated treatment resistance, meaning they might have responded to a second or third SSRI trial given time. The comparison that matters is not DT120 versus a sugar pill but DT120 versus sequenced SSRI therapy in the same population, a study that has not been run, and DT120's monitoring requirement of roughly 6 hours per dose makes it meaningfully more expensive and logistically demanding than writing a prescription. If you compare DT120 only to esketamine and psilocybin, it looks transformative, a potential first-line psychedelic for depression that beats everything in the comparison table by a wide margin. If you compare it to venlafaxine titrated over 8 weeks at $4 per month with zero monitoring, the value proposition becomes a question of how much faster onset and larger effect size are worth to a healthcare system already struggling with psychiatrist shortages.
What You Can Do
If you or someone you know has depression: do not seek illicit LSD. Street doses are uncontrolled, the compound is illegal, and the clinical context of DT120 (pharmaceutical-grade purity, trained supervision, structured follow-up) cannot be replicated outside a trial. Talk to a psychiatrist about existing options, including esketamine if prior medications have not worked, because the distance between a controlled clinical trial and a street dose of LSD is the distance between treatment and roulette.
If you are a clinician: watch for the second Phase 3 readout (expected late 2026 or early 2027), which will determine whether the 8.1-point effect replicates. If it does, NDA filing could follow within 12 to 18 months given Breakthrough Therapy Designation, which means the earliest a prescription pad could write "DT120" is sometime in 2028 or 2029 assuming the DEA cooperates on rescheduling. Start planning for the clinic-flow implications of 6-hour supervised sessions.
If you are an investor: DFTX's 52% single-day move priced in a best-case scenario on every axis simultaneously. The stock trades on two binary events: the second Phase 3 readout and DEA scheduling. A replication failure or rescheduling delay would unwind most of that premium. Act accordingly.
Bottom Line
For thirty years, the standard treatment for depression has produced an effect so small that the prescribing doctor cannot tell it worked by looking at the patient. DT120 just posted an 8.1-point placebo-adjusted MADRS improvement in a Phase 3 trial, crossing a clinical detection threshold that no antidepressant has ever reached. A single tablet, a single visit, 5.8 hours of monitoring, measurable improvement within a week. The caveats are real: 149 patients, one study, an MDD population that is easier to treat than the TRD cohorts where competing psychedelics were tested, and a Schedule I compound that cannot be prescribed until the DEA acts regardless of what the FDA decides. But if the second trial confirms, the question facing American psychiatry will shift from "do psychedelics work for depression" to "how quickly can the healthcare system absorb a treatment that actually clears the bar."