Psilocybin for Opioid Use Disorder (OUD)

This is a double-blind, adaptive, 2-stage, multi-site, phase 2 randomized controlled clinical trial designed to evaluate effects of moderate and high dose psilocybin, relative to low-dose psilocybin control, in OUD patients who continue to use illicit opioids in spite of adherence to standard-of-care treatment with methadone. Up to 480 participations will be consented to yield 240 randomized participants. This study is part of the NIH HEAL Initiative (https://heal.nih.gov/). In Stage 1, subjects will be randomly assigned to one of three groups: psilocybin 30 mg (high dose), psilocybin 20 mg (medium dose), and psilocybin 1 mg (control condition). By the end of Stage 1, an interim statistical analysis will be performed. The study will proceed to Stage 2 if at least one of the active dosages of psilocybin demonstrates 1) acceptable safety, based on analysis of safety data from Stage 1; and 2) conditional power of at least 25%, based on effect size estimates for the primary opioid use outcome (weeks of biologically-verified abstinence during 24 weeks of follow-up). Using a priori decision rules, the interim analysis will determine which of the active treatment groups (30 mg, 20 mg, or both) will be retained in Stage 2 of the trial. Stage 2 will continue the study, using the same treatment and assessment protocols, but retaining only the active dosage or dosages with a high probability of demonstrating efficacy relative to the psilocybin 1 mg control condition. The primary aims are to 1) Evaluate safety and efficacy outcomes in Stage 1 subjects in order to optimize design of the Stage 2, 2) Determine whether treatment with a single high (30 mg) or medium (20 mg) dose of psilocybin improves OUD treatment outcomes, relative to psilocybin 1 mg (control condition), in patients who continue to use illicit opioids despite adherence to methadone treatment, 3) Evaluate the effects of high-dose psilocybin and medium dose psilocybin on self-reported OUD-related neuropsychopathology, and 4) Identify likely responders to psilocybin treatment by using machine learning to model post-treatment OUD outcomes, based on pretreatment characteristics including all relevant clinical data, evaluations, and questionnaires.