Adverse events in clinical treatments with serotonergic psychedelics and MDMA: A mixed-methods systematic review

Clinical research into serotonergic psychedelics and MDMA has reported promising therapeutic effects for a range of mental disorders, but much of the evidence comes from small, selective samples. Earlier literature described adverse reactions after non-medical use (for example panic reactions, transient psychoses and flashbacks) and reviews up to 1984 characterised harms along a continuum from acute, time-limited reactions to more persistent problems. As clinical trials expand and Phase III studies approach completion, determining the full spectrum of adverse events (AEs) that may occur during or after medically supervised psychedelic treatments is critical for balanced benefit–risk assessments and for identifying patients who may be at greater risk of harm. Breeksema and colleagues set out to systematically review AEs reported in clinical studies of classic/serotonergic hallucinogens (psilocybin, LSD, ayahuasca) and the entactogen MDMA in patients with diagnosed mental disorders. The review sought both quantitative and qualitative evidence on acute (during or on the day of dosing) and late (after the dosing day) AEs, using an exploratory mixed-methods approach to capture the range, timing and contextual interpretation of adverse reactions in clinical settings.

The investigators conducted an exploratory, mixed-methods systematic review following the Cochrane Adverse Effects framework and PRISMA guidance; the review was not preregistered. They defined AEs broadly as any unfavourable or harmful event occurring during or after psychedelic or MDMA administration, and operationally distinguished acute AEs (during or on the day of a session) from late AEs (emerging after the session). The term TEAE was noted but not used as a primary distinction. Selection criteria allowed all quantitative and qualitative clinical studies published since 2000 that described treatment of patients with a mental disorder using classic serotonergic psychedelics or entactogens (including open-label studies, case reports/series and RCTs). Excluded sources were systematic reviews, surveys, secondary analyses and studies with healthy volunteers. Searches of PubMed/Medline, EMBASE and PsycINFO were performed on 28 July 2021, complemented by ClinicalTrials.gov cross-checking, handsearching and reference checking. Detailed search strings were provided in supplemental material (not reproduced here). Rather than a formal overall risk-of-bias assessment, the authors appraised study features relevant to AE reporting: whether AEs were actively monitored or only spontaneously reported, inclusion/exclusion criteria that affect generalisability, and the proportion of participants with prior psychedelic experience. The CASP checklist was used to assess qualitative study rigour. Data extraction captured descriptive and quantitative information on acute and late AEs, timing of assessment and serious AEs from main texts and supplements. AE prevalence was recalculated as a percentage of total participants where possible. Qualitative articles were analysed for themes related to AEs and used to contextualise the quantitative findings. The authors emphasised the exploratory nature of synthesis given heterogeneity in designs, substances, dosages and outcome reporting.

From 5,640 records identified, 44 articles met inclusion (34 quantitative and 10 qualitative), representing 598 unique patients in total, of whom 521 received an active dose. The included literature spanned 2006–2021 and encompassed trials and case series with substantial heterogeneity in design, dose, disorders treated and AE assessment methods; sample sizes ranged from 1 to 105 (mean n = 23). Overall issues in reporting and sample characteristics: Many studies did not systematically assess AEs and relied on spontaneous patient reports or therapist observations. Prior psychedelic experience was common in several study populations (up to 56% in MDMA studies and roughly 50% on average in psilocybin studies). Four quantitative and two qualitative papers did not report on AEs at all. MDMA: Sixteen studies were included (total N = 266 patients treated), of which ten publications described MDMA-assisted PTSD treatment (N = 214) comprising seven RCTs, one open-label trial, one case series and one qualitative study. Active MDMA doses ranged from 50 to 125 mg, sometimes with an optional supplemental half-dose. Eleven MDMA studies reported AEs only when spontaneously reported; two RCTs assessed AEs systematically and one used the UKU side-effect scale (the only study reporting AE severity). Physiologically, nine studies measured blood pressure, heart rate and temperature and found mild transient elevations during sessions; none required medical intervention. One MDMA-related serious adverse event (SAE) was reported: increased premature ventricular contractions necessitating one night of hospital monitoring, which resolved without further evidence of cardiac disease. Acute psychological AEs commonly reported across MDMA studies included anxiety and difficulty concentrating, while common acute physical AEs were fatigue, lack of appetite, jaw clenching, thirst and perspiration. Late AEs reported in PTSD trials included fatigue, headache, low mood, insomnia, irritability and anxiety; a minority reported anxiety, depressed mood or panic up to 2 months post-treatment. Suicidal ideation/behaviour were reported in three MDMA studies, but SAEs were reported in control groups in some instances. Psilocybin: Twenty articles described psilocybin treatment in 257 patients across RCTs, open-label trials, a dose-escalation study and qualitative studies. Active session doses ranged roughly 10–30 mg (or 0.1–0.4 mg/kg). AE ascertainment varied: most studies used spontaneous reporting with some using structured instruments such as the Challenging Experience Questionnaire (CEQ) or post-session interviews; three studies reported AE severity. Physiologically, transient blood pressure elevations were common and significant in three studies; heart-rate increases were inconsistently observed. Acute psychological AEs frequently comprised moderate to severe anxiety, confusion, and transient distress; common physical acute AEs were headache and nausea. Late AEs were dominated by post-treatment headache and, in a small number of cases, deteriorations requiring further care: examples included a participant whose depressive symptoms worsened for several weeks after a session in which confusing traumatic memories surfaced, and three instances of unexpected adverse reactions several weeks after psilocybin (post-traumatic flashback; severe anxiety with substance relapse; exacerbation of depression) that required additional treatment and, in one case, withdrawal from the study. One study reported a suicide 11 days after a very low (1 mg) psilocybin dose; the investigators judged this SAE unrelated to psilocybin or study procedures. Ayahuasca: Three studies covered 48 patients (one RCT in social anxiety disorder, one RCT in TRD, one open-label MDD study). Participants were largely ayahuasca-naïve. No consistent cardiovascular effects were observed. Acute AEs commonly reported were nausea, vomiting, diarrhoea and anxiety; in one TRD study four patients remained hospitalised for seven days described as a 'more delicate condition', but the relationship to ayahuasca effects was not clarified. One participant in a SAD RCT experienced a transient dissociative episode with intense fear and distress that resolved with therapist reassurance. No studies reported late AEs. LSD: Four studies (total N = 27) used active doses of 100–200 μg and included one RCT with nested qualitative work, one observational compassionate-use study (LSD and/or MDMA), and one case report. AE collection relied on spontaneous reporting and therapist observation; AE severity was reported only in the RCT. Acute AEs included illusions, feeling cold or abnormal and anxiety; most events were mild to moderate and often resolved during or after sessions. One participant described a severe panic experience during a first LSD session that later transitioned to a positive, relaxed state. Across compounds, SAEs were uncommon in the included studies. Beyond the single MDMA-related cardiac monitoring incident and the suicide judged unrelated to psilocybin dosing, most physiological abnormalities were transient and self-limiting. The qualitative literature repeatedly illustrated that psychologically challenging experiences—although distressing in the moment—were often retrospectively appraised by patients as meaningful or therapeutically valuable.

Breeksema and colleagues conclude that, based on the available clinical literature, MDMA and classic serotonergic psychedelics are generally tolerated in controlled therapeutic settings, but emphasise important caveats. This apparent tolerability is tempered by the small size and heterogeneity of studies, inconsistent and often unsystematic AE assessment, variable blinding and control conditions, and the high prevalence of prior psychedelic exposure among participants, all of which limit generalisability and causal inference. In terms of acute AEs, the review identifies compound-specific patterns: for MDMA, common acute physical effects included fatigue, jaw clenching, thirst and transient elevated vital signs, while acute psychological effects were principally anxiety and concentration difficulties. For psilocybin and LSD, acute psychological reactions (anxiety, paranoia, feeling trapped, confusion) and physical effects (headache, nausea) were common; many of these psychological reactions resolved during sessions with therapist support. Ayahuasca studies frequently reported gastrointestinal AEs such as nausea and vomiting. Late AEs across studies included headache, fatigue, low mood, anxiety and difficulty concentrating, and a small proportion of patients experienced more sustained psychiatric deterioration in the weeks following sessions. The authors discuss how interpretation of psychological 'AEs' is complex: qualitative data show that challenging experiences are sometimes integral to therapeutic change and are retrospectively judged beneficial. They warn, however, that framing all challenging experiences as therapeutic can obscure true adverse outcomes, such as exacerbation of suicidality, dissociation or relapse. Consequently, distinguishing adverse elements that are undesirable and require mitigation (for example severe physiological reactions or persistent worsening of psychiatric symptoms) from transient, potentially therapeutic psychological challenges is vital. Several practical implications are emphasised by the reviewers: rigorous AE monitoring with clear definitions and timing, systematic use of instruments (for example CEQ) to capture challenging experiences, thorough screening and preparation, skilled therapist training, careful attention to set and setting, and longer-term follow-up to detect delayed or persistent adverse reactions. The authors also note systemic concerns raised by case reports of serious boundary violations and therapist misconduct in the broader field, arguing that such transgressions must be addressed and that monitoring adverse effects should include safeguarding for vulnerable patients. Limitations acknowledged by the authors include small and heterogeneous study populations that precluded meta-analysis, inconsistent AE reporting and severity grading, limited placebo-controlled data and challenges with effective blinding, restrictive inclusion criteria that exclude many comorbidities and thus limit external validity, and underrepresentation of psychedelic-naïve and minority populations. They also note the relative scarcity and uneven distribution of qualitative work across compounds, which constrains interpretative nuance for some substances (for example ayahuasca and MDMA). Overall, the authors call for improved, standardised AE reporting in future trials, broader and more representative samples, and integration of qualitative methods to better understand the meaning and consequences of challenging experiences in psychedelic-assisted therapies.

The review summarises that serotonergic psychedelics and MDMA commonly produce transient acute effects such as headache, nausea and anxiety, and that serious adverse events appear rare in the included trials. Challenging psychological experiences are frequent but may be experienced as therapeutically meaningful by many patients; nevertheless, suicidal ideation, suicidality and other serious psychiatric deteriorations, while infrequent, did occur and warrant careful attention. The authors conclude that inconsistent definitions and inadequate assessment of AEs limit current understanding; they recommend systematic, detailed reporting of timing and severity of AEs, longer follow-up, routine use of instruments like the CEQ, and greater transparency to inform safe implementation of psychedelic therapies as research moves into larger and more diverse patient populations.