Ch. VI — Clinical evidence: addiction, PTSD, anorexia, OCD

Chapter VI of Post-2010 Psychedelics: An Expert-Panel Review. For the executive summary and full table of contents, start there.

Abstract. This chapter audits post-2010 psychedelic clinical evidence in indications outside primary mood and anxiety: alcohol use disorder, tobacco cessation, post-traumatic stress disorder, anorexia nervosa, obsessive-compulsive disorder, body dysmorphic disorder, and ibogaine in opioid use and traumatic brain injury. The MDMA-PTSD program (MAPP1/MAPP2) is treated as a clinical-trial case study; sociopolitical narrative and Lykos forensic methodology are deferred to Chapters XI and XII. RoB 2.0 framing is applied where samples are large enough to warrant it; many trials in this chapter are too small for inferential weight and are flagged as such.

6.1 Psilocybin in addiction

Alcohol use disorder

The Bogenschutz program produced two trials that bracket the modern AUD evidence base. The 2015 proof-of-concept (n=10, single-arm, open-label, two psilocybin sessions at 0.3 mg/kg and 0.4 mg/kg embedded in a 14-week Motivational Enhancement Therapy course) reported abstinence and reduced heavy drinking days through 36 weeks¹. The trial is feasibility-only and uncontrolled; effect estimates are descriptive.

The pivotal randomized trial (Bogenschutz 2022, n=93, two psilocybin sessions vs diphenhydramine active control, plus 12 weeks of manualized therapy) is the first adequately-powered psilocybin-AUD RCT². At the primary endpoint, percentage of heavy drinking days over weeks 5–36 was 9.7% in the psilocybin arm vs 23.6% in the diphenhydramine arm (difference: –13.9 percentage points; 95% CI –20.0 to –7.8; p<0.001). Mean drinks per drinking day fell more in the psilocybin arm. The use of diphenhydramine as active control is a methodological strength relative to inert-placebo designs but is unlikely to provide full subjective-effects matching at the doses used; functional unblinding remains substantial. The trial is at low risk for randomization, allocation concealment, and missing-data handling under RoB 2.0, and at high risk on the deviations-from-intended-interventions and outcome-measurement domains for the reasons reviewed in §5.5.

A pilot from the Filament Health PEX010 program reported Phase 2 AUD results in March 2025; details remain in press release form pending peer-reviewed publication³. NIDA-funded follow-up work and the Compass-sponsored COMP004 Phase 2 (psilocybin + naltrexone in AUD) are ongoing.

The mechanistic interpretation of the Bogenschutz signal is the open question for the AUD literature: whether the effect is mediated by drug-cue-related cognitive flexibility (consistent with the entropic-brain prediction; see Chapter III), by reduction of comorbid depression in a population with high baseline depressive symptom burden, or by therapy adherence and engagement effects that scale with the dramatic acute experience. The trial as designed cannot adjudicate among these. A second open question is generalizability: the 2022 sample skewed toward higher socio-economic status and lower comorbidity than the typical US AUD treatment population, and the Bhatt-Ross multisite Phase 3 successor program (not yet started as of the freeze date) will be the first opportunity to test whether the effect replicates in a payer-funded sample.

Tobacco cessation

The Johns Hopkins pilot (Johnson 2014, n=15, open-label, 2–3 psilocybin sessions at 20 mg/70 kg and 30 mg/70 kg embedded in a cognitive-behavioural smoking-cessation protocol) reported a 7-day-point-prevalence abstinence rate of 80% at 6 months (12/15)⁴. The long-term follow-up at 12 months and beyond reported sustained abstinence in 67% (10/15)⁵. These are exceptional rates relative to typical pharmacotherapy + behavioural intervention (typically <35% at 6 months). They are also from a small, single-site, single-arm sample, and the rate at which highly motivated, treatment-seeking volunteers entering a psychedelic trial differ from a general smoking-cessation population is unknown. A larger randomized trial (NIDA-funded multisite, n≈80, psilocybin vs nicotine patch) has reported topline data; the published readout is the meaningful next benchmark.

6.2 MDMA-assisted therapy for PTSD: a clinical-trial case study

The MAPS / Lykos MDMA-PTSD program is the single most consequential clinical-trial sequence in modern psychedelic psychiatry. As clinical evidence, the trials are positive on their primary endpoints; as a regulatory submission, they failed. Both findings are real and need to be held in view together. This section treats only the clinical-evidence side; the methodology forensics, ethics, and the Buisson case are covered in Chapter XII (Agent G), and the regulatory narrative is in Chapter XI.

MAPP1 (Mitchell 2021, n=90, randomized 1:1 to three MDMA sessions vs three placebo sessions plus manualized therapy, 18-week primary endpoint) reported a least-squares-mean difference on the CAPS-5 of –13.9 points (95% CI –20.6 to –7.2; p<0.001) favouring MDMA⁶. Response was 67% in the MDMA arm vs 32% in placebo; clinical remission was 33% vs 5%. The effect size is large by PTSD-pharmacology standards.

MAPP2 (Mitchell 2023, n=104, identical design and protocol) replicated the primary endpoint with a CAPS-5 difference of approximately –11 points favouring MDMA, with 71% no-longer-meeting-PTSD-criteria in the MDMA arm vs 48% in placebo⁷. The internal replication was a key element of the regulatory submission.

The trial design is technically rigorous on randomization, allocation concealment, and outcome measurement (CAPS-5 is clinician-administered by raters separate from dosing sessions). It is at high risk under RoB 2.0 for the deviations-from-intended-interventions domain, because MDMA’s acute subjective effects are functionally unblinding to essentially all participants. The functional unblinding rate in MAPP1/MAPP2 was reported as ~90% in the MDMA arm. It is at some-concerns-to-high risk for outcome-measurement bias because the manualized therapy explicitly encouraged participants to interpret MDMA-induced emotional experiences as therapeutically meaningful — a framing that interacts with patient-reported aspects of the CAPS-5 even when the rating is clinician-administered.

The ICER Final Evidence Report on midomafetamine-assisted psychotherapy (27 June 2024) is the most important third-party clinical-evidence audit of the program⁸. The independent appraisal committee voted 14–1 that current evidence was not adequate to demonstrate superior net health benefit over no MDMA-AP, and 15–0 that the evidence was not adequate to demonstrate superior benefit over short-term trauma-focused psychotherapies (e.g., Cognitive Processing Therapy, Prolonged Exposure). The report’s stated reasoning: functional unblinding, adverse-event reporting gaps, expectancy effects, and questions about the generalizability of the MAPS site network. ICER did not produce a health-benefit price benchmark because the evidence was rated insufficient.

The FDA Psychopharmacologic Drugs Advisory Committee met on 4 June 2024 and voted 2–9 against effectiveness and 1–10 against benefit-over-risk⁹. The FDA issued a Complete Response Letter on 9 August 2024 declining approval and requesting an additional Phase 3 trial¹⁰. The CRL was made public on 4 September 2025 under FDA’s transparency initiative.

For the clinical evidence question — does MDMA-AT work in PTSD — the panel’s reasonable read is: yes, on the primary endpoints, with effect sizes meaningfully larger than the SSRIs approved for PTSD (sertraline, paroxetine). The reason MAPS / Lykos did not get an approval is not that the efficacy data were null; it is that the methodology and conduct concerns were severe enough that FDA judged the data insufficient to establish substantial evidence of effectiveness in the regulatory-statute sense. The clinical-evidence case and the regulatory-evidence case can diverge — Chapter XII is where the methodology forensics is unpacked.

Three Lykos-affiliated papers were retracted from Psychopharmacology in August 2024 for issues related to adverse-event reporting and protocol-deviation disclosure. Those retractions are confined to ancillary publications and do not retract MAPP1 or MAPP2 themselves; they do bear on the audit trail.

6.3 Psilocybin in anorexia nervosa

The Johns Hopkins / UCSD pilot (Peck 2023, n=10, single 25 mg dose of psilocybin plus psychological support, Phase 1, open-label, all-female anorexia-nervosa cohort) reported feasibility, tolerability, and acceptability as primary outcomes¹¹. Secondary outcomes: reduction in eating-disorder symptom scores, with substantial inter-individual variability. The cohort included patients with anorexia-nervosa restrictive subtype with mean BMI 19.7 kg/m². No clinically significant changes in ECG, vital signs, or suicidality were observed. The published findings are explicit that the trial is feasibility-only and does not support efficacy claims.

The clinical interpretation: psilocybin is feasible in a medically-stable anorexia population at moderate dose, and the safety signal is initially reassuring at n=10. The unresolved concerns the panel should track: (i) the cardiovascular and electrolyte profile of severely underweight anorexia patients during peak psychedelic experience is not well-characterized at population scale; (ii) the rate at which psilocybin-induced ego-dissolution interacts with the obsessional body-image structures of anorexia in non-therapeutic directions is unknown; (iii) the dropout and discontinuation patterns in larger samples will be informative. The Compass-sponsored Phase 2 anorexia trial is ongoing, and additional academic-led trials are enrolling.

6.4 Psilocybin in obsessive-compulsive disorder

The foundational modern OCD trial is the Moreno / Delgado pilot at the University of Arizona (n=9, modified double-blind within-subject design, four single doses spanning sub-hallucinogenic to hallucinogenic ranges)¹². The trial reported acute reductions on the Yale-Brown Obsessive Compulsive Scale (Y-BOCS) at every dose level, lasting up to 24 hours. The sample is too small for inferential weight; the trial established feasibility, safety in OCD, and a signal of acute Y-BOCS reduction sufficient to justify further work.

Modern Phase 2 efforts include an ongoing Wessa / Imperial-led trial and a Johns Hopkins-led RCT, with full readouts pending as of the freeze date. The clinical-evidence case for psilocybin in OCD remains: small early signal, no adequately-powered modern RCT yet published. The condition is mechanistically interesting because OCD is one of the most well-characterized illnesses involving compulsive cognitive and behavioural rigidity; the entropic-brain / REBUS framework (Carhart-Harris & Friston 2019, see Chapter III) predicts plausibility but the trial evidence is not yet there.

6.5 Body dysmorphic disorder and other indications

The Schneier group at Columbia University has reported a small (n≈12) open-label trial of psilocybin in body dysmorphic disorder. The trial is feasibility-and-tolerability scale and is too small for inferential weight; the primary methodological challenge is that BDD shares features with OCD and with non-suicidal self-injurious presentations where psychedelic-induced cognitive flexibility may be desired but psychedelic-induced perceptual distortion may be iatrogenic in a population in which appearance-related perceptual distortion is the core symptom. The trial enrolled patients with moderate-to-severe BDD on the BDD-YBOCS scale and reported reductions in obsessional-appearance preoccupation and ritual frequency at 12-week follow-up; the open-label design and small sample preclude inferential weight.

Other indications with reported pilot or Phase 2 data as of the freeze date include: postpartum depression (Reunion Neuroscience RE104 / luvesilocin met primary endpoint Phase 2 on 18 August 2025, with a placebo-adjusted MADRS reduction of –5.8 points at day 7 in n=84; FDA Breakthrough Therapy Designation granted February 2026; Phase 3 planned)¹³; phantom limb pain and chronic pain (early-stage exploratory work, no Phase 2 RCT to date); demoralization in HIV and palliative care (Anderson et al. open-label pilots at NYU and UCLA); migraine prophylaxis (Schindler et al. small RCTs at Yale showing reduced headache frequency at low doses, though the design has been criticised); and adjustment disorder. None of these — except the Reunion postpartum signal, which is uniquely positioned because the active control was a sub-perceptual dose of the same compound — have yet generated the kind of Phase 2b dataset that warrants regulatory-grade audit. The panel should treat them as exploratory, with the postpartum program as the next watch.

6.6 Ibogaine: acknowledgment and scope

Ibogaine and its principal metabolite noribogaine sit outside the classical serotonergic psychedelic class and are formally out-of-scope for deep coverage in this report (see Chapter XIII for the scope statement). Two brief notes are warranted because the 2024 work has been widely cited.

The Stanford / Williams Lab work (Cherian 2024, n=30, prospective observational, magnesium-ibogaine protocol in male Special Operations Forces veterans with predominantly mild traumatic brain injury) reported significant improvements in disability (WHODAS), PTSD (CAPS-5), depression (MADRS) and anxiety symptoms both immediately and one month post-treatment¹⁴. The magnesium co-administration was intended to mitigate ibogaine’s known QT-prolongation and cardiac-arrhythmia risk; no clinically meaningful QT prolongation or hemodynamic instability was observed in this sample, though the sample is small and patients underwent intensive cardiac monitoring. The study is uncontrolled, observational, with significant selection (veteran population, self-funded retreat-setting protocol in Mexico, pre-screening for cardiac fitness). The effect-size estimates are striking but should not be extrapolated to the general TBI or PTSD population.

The clinical development of ibogaine is fragmented: DemeRx IB Inc. (atai-affiliated) has been advancing DMX-1002 oral ibogaine HCl for opioid use disorder; the historical ibogaine cluster includes Howard Lotsof’s 1962 self-administration observation and the Deborah Mash lab’s work in the Bahamas (St. Kitts) from the 1990s onward. The compound’s cardiac risk profile and the failure of multiple commercial development efforts have kept it on the periphery of FDA-trackable trials. Colorado’s Proposition 122 will permit licensed ibogaine use from June 2026 onward, which will produce real-world safety data outside the controlled-trial setting.

The 2025 Nature Mental Health follow-up paper from the Stanford group reported EEG and MRI biomarkers of treatment response in the same MISTIC cohort, which is mechanistically interesting but adds little to the clinical-evidence case.

References

1. Bogenschutz MP, Forcehimes AA, Pommy JA, Wilcox CE, Barbosa PCR, Strassman RJ. Psilocybin-assisted treatment for alcohol dependence: a proof-of-concept study. J Psychopharmacol 2015;29(3):289–299. PMID: 25586396.
2. Bogenschutz MP, Ross S, Bhatt S, et al. Percentage of heavy drinking days following psilocybin-assisted psychotherapy vs placebo for AUD: a randomized clinical trial. JAMA Psychiatry 2022;79(10):953–962. PMID: 36001306.
3. Johnson MW, Garcia-Romeu A, Cosimano MP, Griffiths RR. Pilot study of the 5-HT2AR agonist psilocybin in tobacco addiction. J Psychopharmacol 2014;28(11):983–992. PMID: 25213996.
4. Johnson MW, Garcia-Romeu A, Griffiths RR. Long-term follow-up of psilocybin-facilitated smoking cessation. Am J Drug Alcohol Abuse 2017;43(1):55–60. PMID: 27441452.
5. Mitchell JM, Bogenschutz M, Lilienstein A, et al. MDMA-assisted therapy for severe PTSD: phase 3 study (MAPP1). Nat Med 2021;27(6):1025–1033. PMID: 33972795.
6. Mitchell JM, Ot’alora G M, van der Kolk B, et al. MDMA-assisted therapy for moderate-to-severe PTSD: phase 3 trial (MAPP2). Nat Med 2023;29(10):2473–2480. PMID: 37709999.
7. ICER. Midomafetamine-Assisted Psychotherapy for PTSD: Final Evidence Report. 27 June 2024.
8. FDA Psychopharmacologic Drugs Advisory Committee. Meeting on midomafetamine, 4 June 2024.
9. FDA Complete Response Letter to Lykos Therapeutics, 9 August 2024 (publicly released 4 September 2025).
10. Peck SK, Shao S, Gruen T, et al. Psilocybin therapy for females with anorexia nervosa: phase 1 feasibility. Nat Med 2023;29(8):1947–1953. PMID: 37488291.
11. Moreno FA, Wiegand CB, Taitano EK, Delgado PL. Safety, tolerability, and efficacy of psilocybin in 9 patients with OCD. J Clin Psychiatry 2006;67(11):1735–1740. PMID: 17196053.
12. Reunion Neuroscience Inc. RECONNECT Phase 2 RE104 luvesilocin postpartum depression press release, 18 August 2025. https://reunionneuro.com/2025/08/18/.
13. Cherian KN, Keynan JN, Anker L, et al. Magnesium-ibogaine therapy in veterans with TBI. Nat Med 2024;30(2):373–381. PMID: 38182784.

← Ch. V · Overview · Ch. VII →