Long-term Follow-Up Outcomes of MDMA-assisted Psychotherapy for Treatment of PTSD: A Longitudinal Pooled Analysis of Six Phase 2 Trials

PTSD is a chronic disorder with substantial prevalence in the general population and markedly higher rates among certain groups such as military veterans and first responders. Existing pharmacological and psychotherapeutic treatments can be effective but many patients either do not respond, discontinue treatment, or relapse; trauma-focused psychotherapies generally produce more durable benefits than medications but a substantial treatment‑resistant population remains. MDMA-assisted psychotherapy, a drug-assisted therapy that pairs MDMA administration with a structured course of psychotherapy that attends to set and setting, was designated a Breakthrough Therapy by the FDA in 2017 on the basis of preliminary phase 2 findings and is hypothesised to work by altering emotional memory processing, increasing self-compassion and empathy, and facilitating fear extinction in a therapeutic context. Jerome and colleagues set out to examine longer-term outcomes after MDMA-assisted psychotherapy by pooling long-term follow-up (LTFU) data from six Phase II trials. The primary aim was to evaluate change in PTSD symptom severity (measured with the Clinician-Administered PTSD Scale for DSM‑IV, CAPS‑IV) from baseline through treatment exit and at least 12 months after the final MDMA session, and to summarise participant-reported benefits and harms using a long-term follow-up questionnaire (LTFUQ). The pooled approach was intended to increase sample size and provide information to inform Phase III assessment planning.

The investigators combined secondary and LTFU data from six Phase II trials conducted between April 2004 and March 2017 at five sites (USA, Canada, Switzerland, Israel). Eligible participants were adults (≥18 years) with chronic PTSD (≥6 months), prior inadequate response to psychotherapy and/or medication, and CAPS‑IV scores meeting study-specific thresholds (≥50 in most studies, ≥60 in one study). Recruitment was by advertisement, referrals and word of mouth. Trial designs included a blinded segment, an open‑label crossover, and a long-term follow-up visit; participants were randomised in the blinded segment to a control (0–40 mg MDMA) or active dose (75–125 mg MDMA), and controls were offered active MDMA during the open‑label crossover so that by treatment exit all participants had received active MDMA. Treatment comprised three 90‑minute preparatory psychotherapy sessions, two (or in most cases three) 8‑hour experimental psychotherapy sessions spaced about a month apart with MDMA dosing (initial dose 75–125 mg with an optional supplemental half‑dose 1.5–2.5 hours later), and three 90‑minute integrative sessions after each experimental session. Participants remained overnight with a night attendant after drug sessions and received brief safety telephone contact for seven days post‑session. Treatment exit was defined as the CAPS‑IV assessment after the last full active MDMA session; long‑term follow-up occurred at least 12 months after that session (MP‑1 had a longer mean interval of ~3.8 years). Primary outcomes were CAPS‑IV total severity scores and diagnostic status; an independent rater blind to therapy sessions conducted CAPS‑IV interviews. The LTFUQ, designed by investigators, assessed perceived benefits and harms (presence, magnitude on a five‑point Likert scale, persistence and trajectory), current therapy and medications, and substance use. Suicidal ideation and behaviour were assessed with the Columbia‑Suicide Severity Rating Scale (C‑SSRS) in four of the six studies. For analysis, a modified intent‑to‑treat (mITT) set included participants who received at least one active MDMA dose and a follow‑up assessment. A mixed‑effects repeated‑measures (MMRM) model compared CAPS‑IV scores at baseline, treatment exit (1–2 months post‑treatment), and long‑term follow‑up (≥12 months); the model included baseline CAPS‑IV, study as a fixed effect, and participant as a random effect. Age, PTSD duration, sex, race and prior self‑reported Ecstasy use were added stepwise to explore associations. Within‑subject Cohen's d effect sizes were calculated and descriptive statistics summarised diagnostic changes, suicidality, and LTFUQ responses.

Across the six studies, 107 participants were enrolled; 100 participants received an active MDMA dose (75–125 mg) in two to three sessions and formed the primary treatment sample. The mITT sample for the long‑term CAPS‑IV assessment comprised 91 participants, and 83 participants completed the LTFUQ. Participants were on average 40.5 years old (SD 10.63), 57.9% female, 89.7% white/Caucasian, with a mean PTSD duration of 214.1 months (SD 189.32). On the primary outcome, CAPS‑IV total severity scores fell significantly from baseline to treatment exit: least‑squares mean change (SE) = -44.8 (2.82), p < .0001, corresponding to a within‑subject Cohen's d = 1.58 (95% CI 1.24, 1.91). Scores decreased further from treatment exit to long‑term follow‑up: LS mean change (SE) = -5.2 (2.29), p < .05, with Cohen's d = 0.23 (95% CI 0.04, 0.43), indicating additional but smaller average improvement after treatment exit. The proportion of participants who no longer met CAPS‑IV diagnostic criteria increased from 56.0% at treatment exit to 67.0% at long‑term follow‑up. Clinically significant improvement (≥15‑point reduction on CAPS‑IV) was observed in 82.0% at treatment exit; 26.4% showed a ≥15‑point decrease from treatment exit to long‑term follow‑up. Eleven participants (12.1%) met the study's relapse definition (≥15‑point drop at exit followed by ≥15‑point increase by LTFU). Suicidality data (n = 68 in the four studies using C‑SSRS) showed high lifetime prevalence of ideation and behaviour (86.8% reported lifetime ideation), baseline positive ideation in 60.3% and positive behaviour in 1.5%. At LTFU, 24.2% (15/62) reported positive ideation since treatment exit, 1.6% (1 participant) reported serious ideation, and no participants reported suicidal behaviour. LTFUQ results indicated that 97.6% of respondents reported at least one benefit from study participation; among those, 92.2% reported that some to all benefits persisted at LTFU and 53.2% indicated large benefits that lasted or continued to grow. Seven participants (8.4%) reported any harms, two (3.1%) reported harms that persisted to LTFU, and no harms were rated as severe. The most common harm was worsened mood (n = 3). Nine participants reported a relapse of PTSD symptoms since active treatment; all nine reported one or more significantly stressful life events. At LTFU, 45.8% reported taking any medications and 40.0% reported being in therapy (40.0% specifically for PTSD). Self‑reported Ecstasy/MDMA use since treatment exit was reported by 8 of 83 participants (9.6%); six of these had prior Ecstasy use before the study. Alcohol consumption decreased for 22 participants (40.0%) and increased for 2 (3.6%). The MMRM covariate analysis indicated a significant effect of study, suggesting outcome changes varied across the constituent trials.

Jerome and colleagues interpret the pooled data as showing large reductions in PTSD symptom severity by 1–2 months after MDMA‑assisted psychotherapy that were largely sustained and in some cases continued to improve at least 12 months later. The authors highlight that 82% of participants achieved a clinically meaningful ≥15‑point CAPS‑IV reduction at treatment exit, and the proportion free of PTSD diagnosis rose to 67% at long‑term follow‑up. They also note reductions in self‑reported suicidal ideation and that participants overwhelmingly reported perceived benefits that often persisted or grew over time, while reported harms were infrequent and not rated as severe. The authors situate these findings alongside prior phase 2 reports and other longitudinal PTSD treatment studies, suggesting durability of effects comparable to several established treatments. They propose possible mechanisms for sustained improvement, including enduring psychological and interpersonal changes (for example, increased openness, empathy, spiritual life and posttraumatic growth) and improved therapeutic alliance facilitated by MDMA's acute effects. Key limitations acknowledged by the investigators include the absence of a long‑term control group (all participants had received active MDMA by LTFU), heterogeneity across pooled studies in number of sessions and timing of follow‑up, differences in site populations and languages, and potential contribution of post‑study treatments (therapy, medications) to long‑term outcomes. The open‑label nature of much of the pooled data and the small number of participants reporting harms limited causal inference and statistical comparison for harms. The authors also note the possibility—though they consider it unlikely given participant reports and therapist impressions—that participants might have been motivated primarily by seeking MDMA's pleasurable effects. On the basis of these results and their limitations, the investigators conclude that MDMA‑assisted psychotherapy merits further investigation in larger, controlled Phase III trials to clarify durability, safety, optimal dosing and session number, mechanisms of benefit, and the role of any additional post‑treatment interventions.

The authors conclude that pooled Phase II data support MDMA‑assisted psychotherapy as an efficacious treatment for PTSD with symptom improvements sustained from 1 to 3.8 years post‑treatment. Self‑reported benefits outweighed harms, and no signals of post‑trial substance abuse emerged in available data. These findings corroborate earlier Phase II results and, according to the investigators, justify further controlled research to assess long‑term benefits and safety.