GH001 vs Placebo in Patients With Treatment-Resistant Depression A Randomized Clinical Trial

Treatment-resistant depression (TRD) is common, disabling, and difficult to treat, with fewer than half of patients with major depressive disorder reaching remission on standard antidepressants. The paper notes that after failure of at least two adequate antidepressant trials, many patients continue to have substantial symptoms and functional impairment, and few treatments are approved specifically for TRD. Earlier research has suggested that psychedelics, including mebufotenin (5-MeO-DMT), may produce rapid symptom relief, and an early-phase study of synthetic inhaled mebufotenin (GH001) had already shown promising antidepressant effects. J and colleagues therefore set out to test whether a single-day, individualised dosing regimen of inhaled GH001 would improve depressive symptoms and be safe and tolerable compared with placebo in adults with TRD. The study also aimed to assess remission, other symptom domains, and initial durability of benefit during a longer open-label extension. The paper presents this as a Phase IIb trial designed to address the unmet need for rapid-acting treatments that may induce remission in TRD.

This was a Phase IIb, 7-day randomised, double-blind, placebo-controlled trial conducted at 16 European sites between May 2023 and March 2025, followed by a 6-month open-label extension. Adult patients aged 18 to 64 years with major depressive disorder and TRD were eligible if they had not responded to 2 to 5 oral antidepressant treatments, had a current episode lasting up to 2 years, and had a screening and baseline HAM-D-17 score of at least 20. Diagnostic status was confirmed with the Mini-International Neuropsychiatric Interview, and the current episode was validated using the MGH-SAFER criteria interview. Antidepressants, antipsychotics, and monoamine oxidase inhibitor activity medications were prohibited during the trial and within 2 weeks before baseline; starting or changing psychotherapy during the trial was also not allowed. Participants were randomised 1:1 to GH001 or placebo on day 1. Both groups received an individualised dosing regimen of up to three escalating inhaled doses on the same day: 6 mg, 12 mg, and 18 mg, with 1-hour intervals between doses. Additional doses were given only if the previous dose was well tolerated and the patient had not achieved an intense psychoactive effect, defined by the Peak Experience Scale. The drug was inhaled after vaporisation using the Volcano Medic 2 system. Patients and site staff involved in treatment, assessments, and care were blinded during the placebo-controlled period. Patients were monitored closely after dosing, with efficacy assessed 2 hours after the final dose on day 1 and again on day 2 by telephone and day 8 in person. The primary endpoint was change from baseline to day 8 in MADRS total score. Secondary efficacy measures included remission, defined as MADRS score ≤10, response defined as at least 50% reduction in MADRS, and changes in HAM-A, CGI-S, and Q-LES-Q-SF scores. Safety outcomes included treatment-emergent adverse events, serious adverse events, discontinuations, suicidality, sedation, dissociation, and discharge readiness. Psychoactive effects were measured with the Peak Experience Scale, Mystical Experience Questionnaire, and Challenging Experience Questionnaire. The primary analysis used analysis of covariance adjusting for baseline MADRS, and treatment effect was reported with a 95.18% confidence interval because of an interim-analysis alpha-spending approach. Efficacy and safety analyses included all randomised patients who received at least one dose.

Eighty-one patients were randomised and all completed the placebo-controlled part of the trial: 40 received GH001 and 41 received placebo. Baseline characteristics were similar between groups. In the GH001 arm, 9 patients received only the 6 mg dose, 21 received 6 mg and 12 mg, and 10 received all three doses. The median duration of psychoactive effect was brief, ranging from 9.0 to 14.0 minutes across GH001 doses. Placebo recipients received three placebo doses, and although none had a peak experience, all had some measurable psychoactive-effect score above zero. The primary outcome favoured GH001. Change in MADRS from baseline to day 8 was significantly greater with GH001 than placebo, with a least squares mean difference of -15.5 points (SE 1.7; P < .001; effect size -2.0). Remission at day 8 occurred in 23 of 40 patients (57.5%) in the GH001 group and 0 of 41 patients in the placebo group. The paper also reports improvements in anxiety, global severity, and quality of life, although the extracted text does not provide the full numerical results for all secondary outcomes in the main body here. The authors describe these secondary findings as large. Safety during the placebo-controlled period was broadly acceptable but not free of adverse events. Treatment-emergent adverse events occurred in 29 of 40 GH001-treated patients (72.5%) and 3 of 41 placebo-treated patients (7.3%). The extracted text does not clearly list the full set of most common adverse events, although nausea is mentioned as one of them. There were no severe or serious adverse events in the placebo-controlled period, no suicidality signal judged to be treatment-emergent in the GH001 group, and no suicide behaviours were reported in either group. Most patients were rapidly discharge-ready after dosing, with only one GH001-treated patient not meeting discharge readiness at 1 hour post-dose. In the 6-month open-label extension, all 23 patients who entered remission on day 8 had originally received active GH001 in part 1. Three of these 23 patients (13%) remained in remission without additional treatment across the follow-up. Among the 20 who met retreatment criteria, the median time to first retreatment was 6 weeks, and most required 3 to 4 retreatments. At the final assessment, the mean MADRS score among the 23 day-8 remitters was 7.2. Overall, 20 of 23 patients (87.0%) who remitted in part 1 were in remission at the end of 6 months. No serious treatment-related adverse events were reported during the extension, and the extracted text states there was no evidence of tolerance, although the sentence is truncated.

The authors interpret the placebo-controlled findings as showing that a single-day individualised dosing regimen of inhaled GH001 produced rapid and clinically meaningful improvement in depressive symptoms in TRD. They emphasise the size of the day-8 MADRS advantage over placebo and note that GH001 also improved anxiety, global illness severity, and patient-reported quality of life. They present the magnitude of symptom reduction as comparing favourably with results reported in earlier TRD studies of other interventions, including esketamine, olanzapine/fluoxetine, electroconvulsive therapy, and repetitive transcranial magnetic stimulation, while acknowledging that direct comparisons are difficult because trials differ in design, duration, and patient characteristics. The authors argue that the short psychoactive effect, rapid discharge readiness, and lack of serious adverse events in part 1 support GH001’s tolerability and possible clinical usefulness. They also interpret the open-label extension as suggesting that remission may be sustained over 6 months with relatively infrequent retreatment, although they describe this as an initial finding. They note that the absence of planned psychotherapy may have reduced expectancy and performance bias, and they view the use of the MGH-SAFER interview and the relatively low prior psychedelic exposure in the sample as strengths. Several limitations are highlighted. The psychoactive effects of GH001 would be difficult to mask fully, so functional unblinding may have influenced perceived benefit or response. Although remote MADRS ratings were performed by blinded independent raters who were separated from treatment administration and safety monitoring, the authors say future studies should formally assess blinding integrity. They also note that the sample included relatively few patients with very extensive treatment resistance or long-standing chronic illness, that the Phase IIb sample size was insufficient for robust subgroup analyses, and that the short double-blind period limits comparison with longer TRD trials. The authors conclude that future research should include patients with more advanced illness and should further evaluate durability and safety.