| Trial / Study | Year | Compound / Class | Population | N | Comparator | Endpoint | MADRS Δ vs Control | Effect Size | Response (≥50%) | Remission (MADRS ≤10) | Key Context |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Standard-of-Care Benchmarks | |||||||||||
| Cipriani et al. Lancet, 2018 | 2018 | 21 Antidepressants (network meta-analysis) | MDD | 116,477 | Placebo | 8 wks (typical) | — | SMD 0.30 | ORs 1.37–2.13 | — | Largest AD meta-analysis. Modest average effect; high heterogeneity across compounds. |
| STAR*D Step 1 Rush et al., AJP, 2006 | 2006 | Citalopram (SSRI) | MDD | 3,671 | None (open-label) | 12–14 wks | — | — | ~49% | 36.8% | Real-world pragmatic trial. No placebo arm. 2023 reanalysis (Pigott et al.) estimated true cumulative remission across all 4 steps at ~35% (HRSD) to ~41% (QIDS-SR), not 67% as originally reported. |
| STAR*D Steps 3–4 Rush et al., AJP, 2006 | 2006 | Various switches & augmentations | TRD-equivalent | ~310 | None (open-label) | 12–14 wks | — | — | — | 13–14% | After 2 failed treatments — functionally TRD. Relapse rates 40–71% at 12-month follow-up. Dropout increased to ~46% by Step 3. |
| Rapid-Acting Interventions (FDA-Approved) | |||||||||||
| TRANSFORM-2 Popova et al., AJP, 2019 | 2019 | Esketamine nasal spray (Spravato) + new oral AD | TRD | 236 | Placebo nasal spray + new oral AD | Day 28 | –4.0 95% CI: –7.31 to –0.64; p=0.020 | ~0.30 | — | — | Only 1 of 3 short-term pivotal esketamine trials achieved statistically significant separation on primary endpoint (per FDA review). Both arms improved substantially due to new AD initiation. Twice-weekly dosing under supervision required. |
| ESCAPE-TRD Reif et al., NEJM, 2023 | 2023 | Esketamine nasal spray vs quetiapine XR (+ SSRI/SNRI) | TRD | 676 | Quetiapine XR (active comparator) | Wk 8 | — | — | — | 27.1% vs 17.6%p=0.003 | Open-label, rater-blinded. First head-to-head vs active TRD treatment. 21.7% sustained remission without relapse through Wk 32. |
| IV Ketamine meta-analysis Kryst et al., Pharmacol Rep, 2020 | 2020 | Ketamine IV (0.5 mg/kg) | MDD + Bipolar | ~500 | Placebo (saline) | 24 hrs – 7 days | — | SMD 0.70 | — | — | Pooled across multiple RCTs. Rapid onset; effects typically wane within 1–2 weeks without repeated dosing. Single-infusion design. |
| Psilocybin-Based Interventions (Investigational) | |||||||||||
| COMP360 Phase 2b Goodwin et al., NEJM, 2022 | 2022 | COMP360 psilocybin 25mg (Compass Pathways) | TRD | 233 | 1mg psilocybin (active control) | Wk 3 | –6.6 25mg vs 1mg; p<0.001 | d = 0.55 | 37% vs 18% | 29% vs 8% | Three-arm dose-finding (25/10/1mg). Suicidal ideation in 25mg group drew scrutiny. Single supervised session (~6–8 hrs). |
| Usona Phase 2 Raison et al., JAMA, 2023 | 2023 | Psilocybin 25mg (Usona Institute) | MDD | 104 | Niacin 100mg (active placebo) | Wk 6 | –12.3 95% CI: –17.5 to –7.2; p<0.001 | d = 0.90 | — | Sustained: 25% vs 9.1%p=0.05 | MDD, not TRD. Largest psilocybin between-group MADRS Δ in RCT. Single 25mg dose. Non-industry sponsor. Sustained remission = MADRS ≤10 maintained across visits. |
| COMP005 Phase 3 2026 Compass Pathways (June 2025); durability update (Feb. 17, 2026) | 2025–26 | COMP360 psilocybin 25mg | TRD | 258 | Placebo (inert) | Wk 6 | –3.6 95% CI: –5.7 to –1.5; p<0.001 | — | Not disclosed | Not disclosed | First Phase 3 for a classic psychedelic. Single dose. Wk 26 durability in responders. Compass reports 25% achieved ≥25% MADRS reduction — a non-standard threshold, roughly half the conventional ≥50% response bar. |
| COMP006 Phase 3 Today Compass Pathways, Feb. 17, 2026 | 2026 | COMP360 psilocybin 25mg (×2 doses) | TRD | 581 | 1mg psilocybin (active control) | Wk 6 | –3.8 95% CI: –5.8 to –1.8; 25mg vs 1mg; p<0.001 | — | Not disclosed | Not disclosed | Largest psychedelic RCT (N=581). Two doses 3 wks apart. Active comparator (1mg) may have some effect. Compass reports 39% achieved ≥25% MADRS reduction — a non-standard threshold. Standard response/remission rates not released. NDA planned Q4 2026. |
| DMT-Based Interventions (Investigational) | |||||||||||
| SPL026 Phase 2a Today Erritzoe et al., Nature Medicine, Feb. 2026 | 2026 | DMT fumarate 21.5mg IV (Helus Pharma / Cybin) | MDD | 34 | Placebo (saline) | Wk 2 | –7.35 95% CI: –13.62 to –1.08; p=0.023 | d = 0.82 | 35% vs 12% | 29% vs 12% | First placebo-controlled DMT RCT. 10-min IV infusion — far shorter sessions than psilocybin. Wide CIs reflect small N. MDD (not TRD). Helus not advancing IV SPL026. |
| Inhaled DMT Phase 2a Falchi-Carvalho et al., Neuropsychopharmacology, 2025 | 2025 | Vaporized DMT (15mg + 60mg) | TRD | 14 | None (open-label) | Day 7 | —Open-label; no between-group Δ | g = 2.20 (within-subj) | 85.7%open-label | 57%open-label | GH Research. No control arm — within-subject data not comparable to between-group RCTs. N=14. Dose-escalation design. |
Methodological note — why these numbers cannot be directly ranked: This table presents efficacy signals across trials that differ fundamentally in design, and readers should resist the temptation to rank-order treatments by any single column. Key confounds include:
Population severity. TRD patients (who have failed ≥2 prior treatments) are inherently harder to treat than general MDD populations. Smaller effect sizes in TRD trials do not necessarily indicate inferior pharmacology — they reflect a more refractory patient population. STAR*D Steps 3–4, where patients had already failed multiple treatments, showed remission rates of just 13–14%.
Comparator choice. Between-group effect sizes are determined as much by the comparator as by the active drug. Inert placebos (COMP005, SPL026) produce larger Δ values than active comparators (COMP006's 1mg psilocybin, TRANSFORM-2's new oral AD). A –3.8 MADRS difference against an active comparator that may itself have some therapeutic effect is not equivalent to a –3.8 against saline. Open-label trials with no comparator (GH Research, STAR*D) produce within-subject changes that cannot be compared to between-group differences.
Endpoint timing. Trials measure different phases of the antidepressant response. A Week 2 assessment (SPL026) captures acute response; Week 6 (COMP360, Usona) captures subacute consolidation; Day 28 (esketamine) and 12–14 weeks (STAR*D) measure different therapeutic windows entirely. Early responders may regress; slow responders may not yet have peaked.
Blinding integrity. Classic psychedelics produce unmistakable subjective effects, making true double-blinding essentially impossible. Functional unblinding inflates expectancy effects in the active arm and may deflate placebo response. This is a known limitation across all psychedelic trials and has not been resolved by any design to date.
Sample size & confidence. Small Phase 2a trials (SPL026, N=34; GH Research, N=14) produce wide confidence intervals and should be interpreted as proof-of-concept signals, not definitive efficacy estimates. The Phase 3 COMP360 trials (N=258 and N=581) provide substantially more statistical power.
MADRS scale: Montgomery–Åsberg Depression Rating Scale. 10 items, scored 0–60 (higher = more severe). | Response (≥50%): Standard threshold across the depression literature — a ≥50% reduction in MADRS from baseline. | Remission (MADRS ≤10): Near-absence of symptoms; considered the clinical gold standard. | Not disclosed indicates the trial sponsor reported neither standard response (≥50%) nor remission (MADRS ≤10) rates. Compass Pathways reports a non-standard ≥25% MADRS reduction threshold, which is noted in the Key Context column but excluded from the Response column to preserve cross-trial comparability. | SMD / Cohen's d / Hedges' g: Standardized effect size measures; values of 0.2, 0.5, and 0.8 are conventionally considered small, medium, and large, respectively. Between-group and within-subject effect sizes are not interchangeable.