Safety Across Classic Psychedelics in GAD: What Three Investigational Programs Show So Far

The Emerging Safety Profile

For patients with generalized anxiety disorder (GAD) who ask about psychedelic-assisted treatments under investigation, the safety question spans a class rather than a single drug. Three investigational classic serotonergic psychedelics are now in GAD trials, lysergide, psilocybin, and a deuterated form of DMT, and their early safety data can be considered in the context of a common receptor mechanism and a supervised delivery model.

One class, multiple compounds

Classic serotonergic psychedelics act as agonists at the 5-HT2A receptor, and their safety considerations are linked to its known pharmacology. FDA’s 2026 guidance on psychedelic clinical investigations identifies considerations relevant across the class, including intense acute perceptual effects; mechanism-based cardiac risk tied to 5-HT2B activity and its association with cardiac valvulopathy; interactions with antidepressants, lithium, monoamine oxidase inhibitors and other concomitant medications; and abuse-potential assessment for Schedule I agents.¹ It also describes a supervised delivery model, with two monitors through the dosing session, where the lead monitor does not hold medical credentials but an on-call licensed physician is able to reach the site within 15 minutes.¹ None of these agents is approved; each is being studied under the FDA framework.

Lysergide (DT120)

The most developed GAD program, DT120 (lysergide) ODT, studied as MM120 in the published Phase 2b trial, is now in Phase 3. In a pooled analysis of four lysergide studies (three Phase 1 and one Phase 2b; 257 participants), treatment-emergent events were predominantly mild to moderate and occurred mostly on the dosing day, with no treatment-related serious adverse events, no hallucinogen persisting perception disorder, and no active or worsening suicidal ideation on the C-SSRS.² In the Phase 2b GAD trial (n=80 across the relevant arms), one participant discontinued for an adverse event; one serious adverse event, a panic attack 97 days after dosing, was reported and adjudicated unrelated to treatment.³ The acute effects of a single 100-microgram dose were dose-related, with visual perceptual changes in 92.5% of patients, nausea in 40%, and headache in 35%, with these effects resolving by the end of the dosing session.³

Psilocybin (PSX-001)

Incannex has reported Phase 2 results for PSX-001, psilocybin-assisted psychotherapy delivered across two 25-milligram sessions in 73 patients with GAD. In company-reported results that have not been peer reviewed, no serious adverse events occurred, and the majority of treatment-emergent events were transient, mild to moderate, and consistent with the expected effects of psilocybin. One of 73 participants withdrew during the seven-week program, and there were no signs of increased suicidality, psychosis, or prolonged psychological distress.⁴ That trial, Psi-GAD-1, is registered on the Australian New Zealand Clinical Trials Registry.⁵

Deuterated DMT (HLP004)

HLP004, a deuterated form of DMT (N,N-dimethyltryptamine) developed as CYB004, is in Phase 2 for GAD, with the trial using two intramuscular doses three weeks apart.⁶ Its citable safety data so far come from Phase 1 studies in healthy volunteers. In company-reported Phase 1 results that have not been peer reviewed, the compound was well tolerated, with no serious adverse events, and a majority of adverse events that were mild to moderate and self-limiting. Its acute effects were short, persisting for roughly 40 minutes after an intravenous bolus.⁷ Detailed safety data from the GAD trial itself have not yet been reported.

Shared safety profiles despite different delivery approaches

Across the programs, the acute-safety picture shares similarities: 1) effects are predominantly mild to moderate, transient, and concentrated in the supervised session, and 2) no treatment-related serious adverse events in the lysergide and psilocybin GAD datasets (the single lysergide serious event was adjudicated unrelated). In those datasets, where these outcomes were assessed, there was no indication of persistent perceptual disorder or worsening suicidality; the deuterated DMT GAD trial has not yet reported safety data. The cardiac (5-HT2B) and abuse-potential questions that apply to the class are not directly addressed by the early clinical safety data reported here, and they remain open across the three programs.

The programs differ most in delivery. Lysergide is a single orally disintegrating tablet given in a session lasting several hours; psilocybin is administered in two oral dosing sessions paired with psychotherapy, and the deuterated DMT GAD trial uses two intramuscular doses approximately three weeks apart. Acute-effect duration by the intramuscular route has not been reported for HLP004. These are early-phase datasets, several are company topline results that have not yet peer reviewed, and the trials differ in design, dose, and population. The comparison is therefore qualitative rather than a head-to-head assessment of safety.

What it adds up to

Read together, the early data describe a class whose principal safety considerations are mechanism-based. FDA directs sponsors to address these considerations through eligibility screening, medication restrictions and washout, session monitoring, and abuse-potential assessment, and sets a high evidentiary bar for development. Acute risk concentrates in a supervised session, and interactions with concomitant psychotropic medication are a practical matter for GAD patients already on treatment.¹ Whether these early profiles hold in larger, confirmatory trials is the open question.

DT120 (lysergide), PSX-001 (psilocybin), and HLP004 (deuterated DMT) are investigational agents. Their safety and efficacy have not been established. None are approved by the FDA for any indication.

One Label, Different Stages of Development

The phrase “psychedelic pipeline” groups compounds at very different stages of development, and important distinctions get lost in the shared label. What separates these agents is where each one stands. Some are approved for one condition, while others remain investigational. The supporting evidence also varies, ranging from published trials to company-reported results. For generalized anxiety disorder (GAD), the development has moved into later-stage trials for some agents, while others remain in earlier phases of clinical development.

Developmental Stages in GAD Treatment

Two investigational psychedelic agents, DT120 and PSX-001, are now in GAD trials, at different stages of clinical development. DT120 (lysergide) ODT, the compound published as MM120, is being evaluated in Phase 3 trials for GAD,¹ while psilocybin-assisted therapy PSX-001 (Psi-GAD) remains in Phase 2 development.² The DT120 program includes randomized, placebo-controlled clinical studies, including the Phase 2b dose-finding study of MM120 published in JAMA.³ The PSX-001 program has been evaluated in a randomized, placebo-controlled trial with results reported by the study sponsor.²

Esketamine Is Approved for Depression, Not GAD

No psychedelic is approved for GAD, and the one approved agent in this group, esketamine, is indicated for depression. Esketamine is an NMDA-receptor antagonist rather than a classic serotonergic psychedelic and carries an FDA-approved indication for treatment-resistant depression, including use as monotherapy.⁴ Its approval does not extend to GAD.

Reading by Stage, Not by Label

Stage of development is an important distinction when assessing the psychedelic pipeline in GAD. The programs now span different phases, with DT120 in Phase 3 development and PSX-001 in Phase 2. The lysergide program’s two Phase 3 trials in GAD, Voyage (NCT06741228) and Panorama (NCT06809595), are at different points: Voyage has company-reported results and Panorama results are pending.¹ PSX-001 has an open Investigational New Drug application with the FDA and is preparing to initiate a multi-jurisdiction Phase 2 trial.² For GAD, the pipeline is real and growing, but its data are early, and no psychedelic is yet approved for anxiety.

A Developing Regulatory Framework

The regulatory framework for psychedelic development is also continuing to evolve. In July 2026, the FDA finalized guidance for sponsors conducting clinical investigations of psychedelic drugs, addressing considerations for developing these treatments and conducting clinical studies.⁵ These considerations are real. In the Phase 2b GAD study of MM120, approximately 85% of participants correctly identified their treatment assignment, a limitation relevant to interpreting trials of agents with perceptible acute effects.³ Regulatory decisions have also shaped the field. In 2024, the FDA declined to approve MDMA-assisted therapy for post-traumatic stress disorder, a different indication, citing unresolved questions about study conduct and durability of effect.⁶ For GAD, this evolving framework provides additional context as investigational approaches move through different stages of clinical development, while regulatory approval remains a separate milestone from advancing a compound into a later-stage trial.

Two Analyses of Treatment Limitations

Generalized anxiety disorder (GAD) and treatment-resistant depression (TRD) are chronic psychiatric conditions in which many patients do not achieve durable relief. A real-world claims analysis presented at the American Psychiatric Association (APA) Annual Meeting characterized treatment patterns in newly diagnosed GAD.¹ Baseline data presented at the American Society of Clinical Psychopharmacology (ASCP) Annual Meeting, reported as preliminary interim results, described the population enrolled in two Phase 3 trials in TRD.² The analyses are not directly comparable, but each documents challenges in current pharmacotherapy: limited treatment durability in newly diagnosed GAD, and a severe, treatment-resistant course in depression.

Treatment Instability in Newly Diagnosed GAD

Among 259,158 adults newly diagnosed with GAD, 76% received pharmacotherapy within 12 months, most often an SSRI (45%), a benzodiazepine (22%), or an atypical antidepressant (21%).¹ In a treatment-patterns sub-cohort of 59,275 patients followed from treatment initiation, 18% remained on their initial therapy through the year. Fifty-eight percent discontinued, defined as an interruption exceeding 42 days, and 24% switched or added a medication. Among those who discontinued, 42% restarted after a median of 145 days, and 76% of patients who changed treatment underwent a further change within the study period.

Comorbidity Differs by Treatment Persistence

Patients who did not persist on treatment differed from those who did. Psychiatric comorbidity was more common in the non-persistent group (26.4% vs 20.4%), including major depressive disorder (17.3% vs 12.6%).¹ Medical and neurologic conditions and overall symptom burden showed the same direction, with Medicaid coverage more common among non-persistent patients (17% vs 9%). These comparisons were unadjusted, and claims data cannot establish why treatment changed, but they identify a subgroup with greater psychiatric and medical comorbidity.

Severity and Chronicity in the TRD Trials

The TRD trials enrolled patients who had already failed multiple treatments (839 participants across the two trials). Entry required failure to respond to two, three, or four pharmacological treatments, each taken at an adequate dose for at least eight weeks in the current episode; approximately one-third of participants had failed three or more.² The data indicate a severe and chronic population, with 58% to 62% meeting the threshold for severe symptoms at baseline, the current episode had lasted a mean of 36 to 42 months with more than half exceeding two years, and mean lifetime depressive episodes of 7.5 and 6.2 across the two trials. Lifetime suicidal ideation was recorded in 9.7% and 15.1% of participants, and a lifetime attempt in 9.7% and 12.9%, across the two trials. The investigators noted that counting only the failures that met the trials’ entry criteria understates the actual degree of treatment resistance because many participants had also failed other treatments, drug and non-drug, that did not meet those criteria.

What the Two Analyses Indicate About Unmet Need

The two analyses describe different challenges in current care. Newly diagnosed GAD patients frequently left their initial treatment within the first year, and non-persistent patients had higher rates of psychiatric comorbidity. The depression-trial population had already failed multiple adequate treatments and was severely and chronically ill at entry. The findings are consistent with the broader literature, which describes GAD as a chronic disorder in which many patients do not respond to initial pharmacotherapy, with limited options thereafter,³ and treatment-resistant depression as a severe, chronic condition that carries a heavy burden.⁴ While the two datasets cannot directly be compared, they reinforce that GAD and TRD are chronic conditions marked by substantial unmet need.

How the Receptor Switches On

Classic psychedelics act as agonists at the serotonin 2A (5-HT2A) receptor rather than the serotonin transporter blocked by SSRIs and SNRIs. A 2020 structural study captured the active-state 5-HT2A receptor bound to a hallucinogen and coupled to its Gq protein, showing how ligand binding stabilizes the active conformation that drives Gq-mediated intracellular signaling.¹

A Receptor That Holds On

A crystal structure of LSD bound to the 5-HT2B receptor, together with binding-kinetics experiments, showed that LSD dissociates exceptionally slowly from both 5-HT2B and 5-HT2A, the receptor central to psychedelic effects.² Simulations attributed the slow kinetics to a lid formed by an extracellular loop over the binding pocket. Loosening that lid by mutation accelerated LSD’s dissociation and selectively reduced its β-arrestin2 recruitment.² The authors noted that a long receptor residence time could contribute to LSD’s long duration of action.²

Intracellular 5-HT2A Drives Plasticity

5-HT2A activation alone is not the full story. Serotonin activates the same receptor but does not reproduce the growth-promoting effects of psychedelics. In rat and mouse cortical neurons and in cell culture, Vargas and colleagues showed that the plasticity-promoting pool of 5-HT2A is intracellular, and serotonin, unlike lipophilic psychedelics, cannot readily cross the membrane to reach it.³ This difference in receptor location may help explain why psychedelics and endogenous serotonin can produce different effects at the same receptor.

Not the Only Door

The 5-HT2A receptor is not the only fast route to change. Esketamine, approved for treatment-resistant depression, acts through non-competitive NMDA-receptor antagonism on inhibitory interneurons, disinhibiting glutamate and engaging downstream plasticity pathways.⁴ This increases glutamatergic signaling and activates pathways involving AMPA receptors, BDNF, and mTORC1 that have also been linked to synaptic plasticity.⁴ Two different receptor targets converge on overlapping signaling, underscoring that the receptor engaged, not the overall level of serotonin, defines these approaches.

What It Means for GAD

In GAD, the property that would fit this receptor biology is durability. Clinical research in GAD is exploring whether the effects of one or two doses of investigational 5-HT2A agonists can persist beyond the initial treatment period. If they do, that would be consistent with slow 5-HT2A engagement and intracellular signaling producing change that outlasts the drug’s exposure. How receptor kinetics, signaling, and cellular location contribute to any longer-lasting effects is still being studied, and further clinical research will determine how consistently such effects can be sustained in GAD.

Three targets, one downstream destination: How current and investigational approaches engage the brain, by mechanism