MDMA-assisted psychotherapy for treatment of anxiety and other psychological distress related to life-threatening illnesses: a randomized pilot study

Wolfson and colleagues conducted a Phase II, double-blind, randomised, placebo-controlled trial with an open-label crossover to evaluate MDMA-assisted psychotherapy for anxiety related to life‑threatening illnesses (LTIs). The blinded portion comprised two day‑long (approximately 8‑hour) experimental sessions given two to four weeks apart, plus nine non‑drug psychotherapy sessions (three preparatory sessions and six integration sessions). After the primary endpoint (one month after the second blinded session) the blind was broken: participants originally randomised to MDMA received one further open‑label MDMA session and those originally receiving placebo crossed over to receive three open‑label MDMA sessions. All MDMA administrations used a 125 mg initial dose with an optional supplemental dose of 62.5 mg given about 90–150 minutes later. Participants were recruited at an outpatient psychiatric clinic in San Anselmo, CA, between 2015 and 2018. Eligible adults (≥18 years) had a diagnosed life‑threatening cancer or nondementing neurological illness that was ongoing or in remission with risk of recurrence, an estimated life expectancy of at least nine months, and baseline STAI‑Trait scores ≥45 indicating moderate to severe anxiety. Key exclusions included ongoing primary treatment (for example, initial chemotherapy), major medical conditions contraindicated for MDMA (uncontrolled hypertension, significant cerebrovascular or cardiovascular disease, brain tumours, renal disease), certain psychiatric diagnoses (psychotic disorders, bipolar I, dissociative identity disorder), pregnancy, inability to taper psychiatric medications safely, and weight under 48 kg. Participants on prescribed opioid medications were permitted but required to follow tapering instructions around sessions; substance use disorders in remission for at least 60 days were allowed. Randomisation was implemented in an approximately 1:3 ratio (placebo:MDMA) via a web‑based system, with capsules of MDMA or lactose (placebo) prepared to match in appearance and weight. Participants, investigators, and an independent rater were masked until the primary endpoint. Experimental sessions took place in a comfortable, home‑like environment; therapists used a non‑directive approach, provided eyeshades and music for internal focus, and monitored physiological measures (blood pressure, heart rate, temperature) frequently during sessions. Safety monitoring included nightly stay at site after sessions, daily telephone checks for seven days post‑session, the Columbia Suicide Severity Rating Scale at all in‑person visits and selected phone contacts, and systematic collection of adverse events. The primary outcome was change in State‑Trait Anxiety Inventory (STAI)‑Trait scores from baseline to one month after the second blinded session. Secondary outcomes included STAI‑State, Montgomery‑Åsberg Depression Rating Scale (MADRS, rater administered), Beck Depression Inventory‑II (BDI‑II), Pittsburgh Sleep Quality Index (PSQI), Posttraumatic Growth Inventory (PTGI), Global Assessment of Functioning (GAF), FACIT‑Sp, Five Facet Mindfulness Questionnaire (FFMQ), Death Attitude Profile (DAP), Self‑Compassion Scale (SCS), and adverse event reporting. Analyses were conducted on an intention‑to‑treat basis. Given the pilot nature, the study was not powered for definitive hypothesis testing; statistical tests used an alpha of 0.05 two‑tailed, with independent‑samples t tests for the primary outcome, Hedges' g for effect sizes, and repeated‑measures ANOVA for pooled open‑label data. SAS version 9.4 was used for analyses.

Eighteen participants were enrolled and randomised between May 2015 and February 2017: 13 to MDMA and 5 to placebo. Most screened individuals were excluded at telephone screening (92 of 110) for reasons such as not living locally or being too medically unwell to participate. The sample had a mean (SD) age of 54.9 (7.9) years, was 77.8% female and 83.3% White/Caucasian. All participants had a prior LTI diagnosis; 94.4% had neoplasms as the index diagnosis. High rates of prior psychiatric diagnoses were reported: anxiety 83.3%, major depression 77.8%, PTSD 72.2%, and insomnia 61.1%. Baseline mean (SD) STAI scores were 61.1 (7.0) for Trait and 57.4 (10.9) for State. For the primary outcome, mean (SD) change in STAI‑Trait from baseline to one month after the second blinded session was greater for the MDMA group, −23.5 (13.2), than for the placebo group, −8.8 (14.7). This between‑group difference did not reach conventional statistical significance (p = 0.0558). The reported Hedges' g was 1.03 (95% CI: −5.25, 7.31). The authors noted a potential placebo‑group outlier with a change score of −35; removal of that single case produced a statistically significant between‑group difference (p = 0.0066). Secondary outcomes showed significant between‑group benefits for posttraumatic growth (PTGI: MDMA Δ = 12.9, SD = 23.3 vs placebo Δ = −2.6, SD = 6.1; p = 0.04, g = 0.50) and mindfulness (FFMQ: MDMA Δ = 0.4, SD = 0.6 vs placebo Δ = 0, SD = 0.2; p = 0.04, g = 0.67). Measures of STAI‑State, depression, sleep quality and global functioning trended toward greater improvement in the MDMA group but did not reach statistical significance in the blinded comparison. After unblinding and crossover, all participants eventually received three MDMA‑assisted psychotherapy sessions. Combined analyses across baseline, treatment exit (after three MDMA sessions), 6‑month and 12‑month follow‑ups using repeated‑measures ANOVA indicated statistically significant improvements across most outcome domains compared with baseline; pairwise tests supported durability of gains at 6 and 12 months. Within the MDMA group, an additional unblinded third session yielded a further mean decrease in STAI‑Trait of about 4 points from the last blinded measurement. Safety findings indicated MDMA was generally well tolerated in this sample. Expected acute reactions during blinded MDMA administrations included thirst, jaw clenching/tight jaw, dry mouth, headache and perspiration. In the seven days after administration the most frequent reactions were fatigue, needing more sleep, insomnia and low mood, with these reactions diminishing over the week. Treatment emergent adverse events were infrequent and similar between groups; most other adverse events were attributed to participants' LTIs. Serious adverse events occurred that were related to cancer progression in three participants (one participant discontinued after primary endpoint and later died from cancer‑related complications). Vital signs (systolic and diastolic blood pressure, heart rate, body temperature) generally rose more in the MDMA group; only peak body temperature differed significantly (mean [SD] MDMA 37.3 °C [0.7] vs placebo 36.9 °C [0.3], p < 0.0001). No medical interventions were required for vital sign elevations and values returned toward baseline by session end. Suicidal ideation or behaviour per the C‑SSRS was not observed. Blinding appeared limited: investigators correctly guessed drug condition in 32 of 36 (88.9%) blinded sessions and participants in 31 of 36 (86.1%) sessions; in two of the three participant incorrect guesses the participant believed they had received MDMA when they had received placebo.

Wolfson and colleagues interpret the results as preliminary evidence that MDMA‑assisted psychotherapy may reduce anxiety and improve related outcomes in people with LTIs, but they emphasise important caveats. The primary blinded comparison showed a large numerical advantage for MDMA on STAI‑Trait scores but did not reach statistical significance (p = 0.056). The investigators highlight that a single placebo‑group outlier with a pronounced improvement substantially influenced the result, and exclusion of that case rendered the between‑group difference statistically significant (p = 0.0066). The possibility of placebo effects was also acknowledged because some placebo participants believed they had received MDMA. At the primary endpoint the MDMA arm showed statistically significant improvements in mindfulness (FFMQ) and posttraumatic growth (PTGI). After crossover, when all participants had received three MDMA sessions, the pooled sample demonstrated significant and durable improvements across anxiety, depression, sleep, global functioning, wellbeing domains, self‑compassion, mindfulness and attitudes toward death that persisted at 6‑ and 12‑month follow‑ups. The authors note those long‑term within‑subject gains lack a concurrent control group after crossover, limiting causal inferences about durability. Regarding mechanisms, the discussion references prior literature suggesting MDMA can reduce amygdala reactivity and modulate frontal and insular activity, increase oxytocin and serotonergic signalling, and foster self‑compassion and prosociality—mechanisms that could plausibly facilitate therapeutic processing of trauma or illness‑related fears. The authors point out a high prevalence of prior PTSD diagnoses (72.2%) in their sample and propose that overlapping neural and psychological processes may have been engaged. On safety, the investigators report that MDMA was well tolerated in this clinical context: adverse reactions were generally mild to moderate, short‑lived and typical for MDMA, with no treatment discontinuations attributable to MDMA and no suicidal behaviour detected. They caution that the small sample limits precision in estimating risk and recommend continued safety evaluation in larger samples. Key limitations explicitly acknowledged include the small, pilot sample not powered for definitive hypothesis testing, potential bias from an influential outlier and imperfect blinding, and limited generalisability because the sample was predominantly female and White. The authors conclude that larger, more diverse trials are needed to confirm efficacy, clarify mechanisms and identify mediators or moderators of response.

The authors conclude that these pilot data provide preliminary support for the safety and feasibility of MDMA‑assisted psychotherapy as a potential treatment for anxiety and other psychiatric symptoms associated with life‑threatening illnesses. They state the findings justify further clinical trials with larger, more diverse samples to evaluate efficacy, durability and safety more definitively.