Borderline personality disorder sits near the top of most psychedelic trials’ exclusion lists, alongside psychosis and mania risk: a population defined in part by emotion dysregulation and elevated suicide risk, screened out of studies built around one of the most emotionally intense experiences in clinical medicine. A trial registered September 22 enrolls that diagnosis directly. The design choice worth attention is not the drug. It is the requirement that every participant be established in dialectical behavior therapy before dosing, and stay in it the whole way through.
What was registered
The ADEPT study, registered by the University of Minnesota on ClinicalTrials.gov (NCT07833761) with interventional psychiatrist Ziad Nahas as principal investigator, is a Phase 1, open-label, single-group trial in 18 adults with a confirmed DSM-5 borderline personality disorder diagnosis. Participants receive two 25 milligram doses of psilocybin roughly three weeks apart, each in a supervised session with at least six hours of therapist support, wrapped in three preparation visits and six integration visits across an 18-week study window.
The structural condition is the distinctive part. Eligibility requires active enrollment in the M Health Fairview DBT program for at least a month, with consistent attendance, before a participant can be considered. The full DBT package, weekly individual therapy, group skills training, and phone coaching, continues alongside the psilocybin sessions instead of pausing for them. DBT is the best-evidenced psychotherapy for borderline personality disorder, built specifically around emotion-regulation skills and suicide-risk management. The trial is not testing psilocybin instead of standard care, or beside a generic supportive protocol. It is testing whether dosing can be added on top of the treatment infrastructure this population already relies on.
The registration lists Filament Health, the botanical psilocybin supplier, as industry collaborator. Filament was acquired by Red Light Holland Corp. in April 2026; Red Light Holland renamed itself Rhelion Life Sciences that August. The trial is not yet recruiting, with an estimated start in December 2026 and estimated completion in December 2030.
Why this population has stayed off the map
The rationale for excluding borderline personality disorder from psychedelic trials is not obscure. The concern is that an intense, destabilizing subjective experience could do harm in patients whose core symptoms include affective instability, identity disturbance, and recurrent suicidality, and that a transient post-dosing crisis in this population carries higher stakes than in a typical depression cohort.
The result was, until recently, close to an evidence vacuum. Across ClinicalTrials.gov, the registered clinical evidence for psilocybin in this population still amounts to one completed study: a University of Chicago open-label trial, run with the Usona Institute, that gave a single 25 milligram dose to nine adults with major depressive disorder and co-occurring borderline personality disorder, tracking depression and borderline symptom scales through five weeks. No results were posted in the registry, but the trial’s findings were published in Clinical Neuropharmacology in April 2026: depression symptoms improved significantly from baseline to the study’s endpoint, while borderline personality disorder symptom scores did not change significantly, in a sample of nine participants too small, and a design too unblinded, to establish efficacy for either condition. ADEPT differs from it on both counts that matter: it targets borderline personality pathology as the primary condition, not depression with a comorbidity, and it binds the drug to the disorder’s own standard-of-care psychotherapy rather than administering it on top of study-specific support alone. Within the registry, no earlier psilocybin trial pairs dosing with ongoing DBT.
A trial instrumented to measure the feared harm
Phase 1 means the primary endpoints are safety and feasibility, and the registered outcome list reads like a direct test of the exclusion rationale itself. Alongside standard adverse-event tracking, the primary outcomes include acute suicidal ideation within 24 hours of each dosing session, measured on a dedicated scale, plus a challenging-experiences questionnaire, a psychedelic side-effects inventory, study retention, and DBT session attendance. The trial is explicitly measuring whether dosing destabilizes the thing the field has always worried it would destabilize, and whether patients keep showing up to their existing therapy afterward.
The eligibility criteria carve the acute end of the risk spectrum out, a familiar pattern from other excluded-population trials. Active suicidal ideation with intent or plan is excluded, as is any suicide attempt within the past six months, along with bipolar disorder, psychosis, recent substance use disorder, and an active primary PTSD diagnosis. The PTSD criterion does not amount to a blanket exclusion of everyone with a history of PTSD or a co-occurring PTSD diagnosis. This is a trial in the excluded diagnosis, not in its highest-risk or most complicated presentations.
What 18 people in an open-label design can settle
Not efficacy, and the registration does not claim otherwise. An open-label, single-group study of 18 participants, all of whom know they are receiving psilocybin while continuing an active evidence-based therapy, cannot separate drug effects from DBT effects, expectancy, or attention. Symptom measures, emotion-regulation scales, and quality-of-life instruments all appear in the registration as secondary outcomes, and any movement on them will be uninterpretable as a treatment effect. What the trial can establish is narrower and still consequential: preliminary safety, tolerability, retention, and treatment-engagement findings in a carefully screened borderline personality disorder cohort, and whether the DBT-embedded model holds together operationally. Eighteen participants in an open-label design cannot establish that psilocybin is safe for borderline personality disorder as a whole; the sample is too small, and too carefully screened, to generalize beyond it.
The timeline tempers expectations further. The study opens no earlier than December 2026 and runs on an estimated four-year clock. What it offers in the meantime is a template. If psilocybin research keeps moving into populations the field once ruled out by protocol, the open question has been what the safety architecture for each one should look like. For bipolar II depression, recent trials answered with mood-stabilizer management and crisis-threshold exclusions. For borderline personality disorder, this trial’s answer is to make the existing gold-standard therapy a precondition and a constant, and then measure, within a day of each dose, whether the foundation held.