Treatment-Resistant Depression (TRD)

This report summarises what Blossom’s database shows about psychedelic and rapid-acting treatments for treatment-resistant depression (TRD), and what it does not show. The short version: TRD is the indication where this field has gone furthest. Esketamine is approved, a synthetic psilocybin has met a Phase 3 endpoint, and a wave of fast-acting tryptamines is reporting striking early results. The progress is real and, unusually for psychedelic medicine, regulator-validated. It also comes with modest effect sizes, short-lived benefits, a serious unblinding problem and a long list of expensive failures. Both halves matter.

A note before the evidence

This page is a research summary, not medical advice, and nothing here is a treatment recommendation. Of the treatments discussed, only esketamine (and ketamine, off-label) is in routine clinical use; the classic psychedelics are investigational. Decisions about treatment for depression belong with a qualified clinician.

Two framing points. First, this page is scoped to treatment-resistant depression specifically, not to depression in general; where a study really enrolled broader major depression, that is flagged. Second, on the numbers: Blossom currently tracks 349 papers and 210 trials tagged to this topic, and those counts appear on this page. The tag is leaky, pulling in general-depression studies, ECT-anaesthesia ketamine work, biomarker papers and non-drug interventions, so the genuinely TRD-specific core is smaller. Read the counts as database coverage, not as a clean tally of TRD treatments.

What counts as treatment-resistant depression?

TRD is usually defined as depression that has not responded to at least two adequate courses of antidepressants. It is the hard end of a common illness: around a third of people with depression meet this bar[1]BMJ state-of-the-art review (2026). That is the unmet need driving everything below, and it is why the regulators and the money have concentrated here rather than on first-line depression, where cheaper options already work for most people.

Two paradigms: rapid-acting versus psychedelic

The treatments split into two families with different logics. The glutamatergic, rapid-acting drugs, ketamine and esketamine, block the NMDA receptor and trigger fast AMPA-driven synaptic plasticity[2]Molecular Psychiatry (2026), lifting mood within hours but fading within weeks, so they are dosed repeatedly. The serotonergic psychedelics, psilocybin and the 5-MeO-DMT and DMT agents, act on the 5-HT2A receptor and pair a single dose with psychological support, aiming for a durable effect from one or a few sessions. The newest tryptamines try to combine the two: a psychedelic mechanism with a rapid-acting, minutes-long experience. For all their differences, the two families appear to converge downstream on the same thing, a burst of synaptic plasticity and neurotrophic signalling in a depressed brain whose circuits have grown rigid, which is part of why a glutamate drug and a serotonin drug can both lift mood when older antidepressants have not.

Esketamine: the approved benchmark

Esketamine is the yardstick because it is the only approved drug here. Its programme is deep: a monotherapy trial beating placebo at four weeks[3]JAMA Psychiatry (2025), higher remission than quetiapine in ESCAPE-TRD[4]CNS Spectrums, ESCAPE-TRD (2025), and good two-year retention in responders[5]Eur. Neuropsychopharmacology, ESCAPE-LTE (2026). In January 2025 it became the first medicine approved as a stand-alone treatment for TRD.

The honest read is that approval does not mean a large effect. A PRISMA meta-analysis put esketamine’s benefit over placebo as modest, on a par with adding an antipsychotic, with no significant effect on suicidality[6]Am. J. Psychiatry, PRISMA meta-analysis (2025), and its delivery is demanding: every dose is given in a certified clinic with hours of monitoring, and dissociation and sedation are common[7]Am. J. Psychiatry (2025). It is a real, new option, not a breakthrough cure.

Psilocybin: the first Phase 3 win, and the asterisks

Psilocybin had its landmark moment in 2025, when COMPASS’s synthetic COMP360 became the first classic psychedelic to meet a Phase 3 endpoint in depression[8]COMPASS Phase 3 COMP005 topline, 2025, a single 25 mg dose beating placebo by 3.6 MADRS points at six weeks, with no concerning suicidality signal. For a field long on Phase 2 and short on confirmation, this mattered.

The qualifications are substantial and worth stating plainly. The effect was modest, similar to the earlier Phase 2b. The confirmatory second Phase 3 (COMP006) has since also met its primary endpoint (February 2026), and with 26-week durability data reported in July 2026 COMPASS has begun a rolling FDA submission targeting a possible 2027 launch, though the modest effect size and the unblinding critique still apply. An independent academic trial, EPIsoDE, missed its primary endpoint in 2026[9]JAMA Psychiatry, EPIsoDE (2026), a reminder that an industry win is not the whole literature. And a control-group meta-analysis concluded psilocybin’s efficacy is overestimated[10]JAMA Network Open (2025) because its placebo patients, unlike those in antidepressant trials, can usually tell they did not get the drug. Even the encouraging single-dose results, such as a severe-TRD open-label trial[11]Am. J. Psychiatry (2025), sit inside that unblinding caveat, and real-world response rates have run below trial figures[12]Lancet Reg. Health Europe (2026).

Ketamine: the off-label workhorse

Intravenous racemic ketamine is the pragmatic mainstay: far cheaper than esketamine, widely used in specialist clinics, and supported by a large trial cluster including evidence that it is non-inferior to ECT for non-psychotic TRD[1]BMJ state-of-the-art review (2026). Its weakness is the same as esketamine’s, only starker: the benefit fades within weeks, so it becomes an indefinite maintenance treatment, and with no company owning the molecule, access, dosing and quality vary widely.

The next wave: 5-MeO-DMT and DMT

The most striking recent numbers come from ultra-short-acting tryptamines. Inhaled 5-MeO-DMT (GH001) reported a Phase 2b with a large effect and 57% day-8 remission versus zero on placebo[13]JAMA Psychiatry, GH001 (2026); intranasal 5-MeO-DMT (BPL-003) reported rapid, durable responses with patients dischargeable in roughly 98 minutes[14]J. Psychopharmacology, BPL-003 (2026); and vaporised DMT reported 86% response in a small open-label trial[15]Neuropsychopharmacology (2025). If they hold up, a psychedelic experience lasting minutes rather than hours could make this kind of therapy far easier to scale, which is exactly the bottleneck psilocybin faces.

The caution is proportionate to the excitement. These are small, early, often open-label studies, and the unblinding and durability questions that complicate psilocybin apply here with even less controlled data. Large early effects in open-label trials are precisely the pattern that later, blinded trials so often shrink.

Who is developing what

TRD has the most concentrated commercial pipeline in psychedelic medicine. Johnson & Johnson’s Janssen owns esketamine, the only approved product[16]FDA esketamine monotherapy approval, Jan 2025; COMPASS Pathways leads the classic psychedelics with COMP360 psilocybin and its Phase 3 win[8]COMPASS Phase 3 COMP005 topline, 2025; and the non-profit Usona Institute advances psilocybin on a less commercial track. The fast-follower wave, GH Research (GH001) and atai Life Sciences with Beckley Psytech (BPL-003 and DMT formulations), is betting on short-duration tryptamines, while several groups reformulate ketamine into oral and single-enantiomer products.

The failures are part of the honest picture. Relmada’s esmethadone failed three Phase 3 trials and was dropped[17]FierceBiotech, Relmada Phase 3 failures (2022), and Biogen and Sage’s zuranolone was approved only for postpartum depression after an FDA rejection for major depression[18]FDA CRL for zuranolone in MDD (2023). A well-capitalised, crowded pipeline is a sign of momentum and of how hard the indication is, not of a solved problem.

Durability and the redosing problem

Across every compound, durability is the quiet weakness. Ketamine and esketamine fade within weeks and are dosed indefinitely. Even psilocybin, sold on the promise of lasting change from a single dose, shows waning benefit over time: a veteran cohort saw response fall from 80% at six months to 40% at twelve[19]J. Affective Disorders (2025), and the intensity of the acute experience, not just the drug, tracks with who improves[20]J. Affective Disorders (2025). Whether the answer is periodic redosing, better integration therapy, or accepting that these are maintenance rather than one-shot treatments is still unresolved.

Cost, access and the scale problem

A theme that rarely makes the headlines runs underneath all of this: delivery. Esketamine must be given in a certified clinic with hours of monitoring under a REMS programme, which is expensive and limits where it can be offered. Off-label ketamine is cheaper per dose but is delivered through a patchwork of largely unregulated, often out-of-pocket clinics, so quality and access vary widely. Psilocybin therapy, as trialled, pairs a single dose with a full day of supervised support plus preparation and integration sessions, which is effective but labour-intensive and hard to scale.

This is the real reason the minutes-long tryptamines attract so much investment: a treatment that frees up the clinic in under two hours is far easier to deliver to the millions of people with treatment-resistant depression than one that occupies a therapist and a room for an entire day. Whether the rapid agents can match the durability of the longer sessions is the open question on which much of the field’s practical future depends.

Reading this honestly

So where does TRD sit? Further ahead than any other psychedelic indication, and that lead is genuine: an approved drug, a real Phase 3 win, and a pipeline of credible next-generation agents. But the honest reading refuses two temptations at once, the dismissive one that says none of this works, and the breathless one that says depression is solved. The effects are real and modest, the durability is short, the blinding is imperfect, and the graveyard of failed programmes is long. For people with treatment-resistant depression, that combination, real options that are not miracles, is both the most hopeful and the most honest thing the evidence will support.