This systematic review (s=8) analyses completed controlled trials of psychedelics for depression, including psilocybin, LSD, ayahuasca, and DMT, all in Phase II or I/II. It evaluates methodological patterns against the draft European Medicines Agency guideline revision, highlighting challenges such as unblinding, expectancy, and adverse event characterisation, while calling for larger studies to assess long-term efficacy and safety.
There is a growing body of clinical research on the therapeutic potential of psychedelics for the treatment of mental health disorders, notably depression. Accordingly, the new revision of the European Medicines Agency guideline on the clinical investigation of products for depression will incorporate a section covering specific regulatory recommendations for the design of studies with psychedelics. The present review investigated the methodological approaches adopted in completed controlled trials of psychedelics for depression in light of initial considerations included in the draft guideline revision. A systematic search conducted on scientific databases (Embase and Medline) and clinical trial registries (clinicaltrials.gov and WHO ICTPR) identified 8 completed trials as of February 2024. The trials tested psilocybin, LSD, Ayahuasca, and DMT, for major depressive disorder or treatment-resistant depression, and were all pahse 2 or 1/2. Patterns in pre-defined methodological variables pertaining to trial design, population, interventions, outcome measures and safety assessments were analysed and collated against considerations on unblinding and expectancy, choice of comparator, the definition of treatment frameworks, the characterisation of the subjective psychedelic experience and the specification of adverse events in relation to subjective psychedelic effects. Areas for future research, including long-term efficacy and safety and the influence of inter-individual differences, can be investigated in larger studies, necessary for marketing authorisation applications. Ultimately, balancing the intricacies of conducting trials with psychedelics with ensuring adherence to regulatory requirements can be facilitated by early dialogue with medicines regulators, and will be essential for the medical development of psychedelics to address unmet patient needs.
Interest in the therapeutic use of psychedelics for mental health disorders, and depression in particular, has expanded rapidly, prompting regulatory bodies to respond. Classic psychedelics are typically defined by agonism at serotonin 2A receptors, and although no classic psychedelic currently has marketing authorisation, regulators in multiple jurisdictions have issued draft guidance or created special access pathways. Against a backdrop of substantial unmet need in depression, including a high prevalence and a large proportion of patients showing inadequate response to first-line antidepressants, the European Medicines Agency (EMA) has proposed a draft revision of its guideline on clinical investigation of products for depression that includes preliminary considerations specific to psychedelic studies. Silva and colleagues set out to systematically evaluate the methodological approaches used in completed randomised controlled trials of psychedelics for major depressive disorder (MDD) or treatment-resistant depression (TRD), and to discuss how those approaches align with initial regulatory considerations in the EMA draft guideline. The review aimed to identify patterns and gaps across trial design, populations, interventions, outcome measurement and safety assessment that are relevant for future clinical development and regulatory submission of psychedelic medicines in the EU.
The investigators conducted a systematic search of scientific databases (Embase and MEDLINE) and clinical trial registries (ClinicalTrials.gov and the WHO International Clinical Trials Registry Platform). The search combined general and substance-specific terms for classic psychedelics and MDMA with a controlled-trials filter adapted from an existing Cochrane filter; the exact search string is available from the authors on request. The initial search identified both publication records and registry entries and was completed on 19 February 2024. Records were screened at title and abstract level against predefined criteria. Included records had to report a human clinical trial that systemically administered a classic psychedelic or MDMA, be conducted after 1995, and test the psychedelic as part of a therapeutic intervention with at least one efficacy endpoint in a patient population. The reviewers grouped records by trial and retained those that targeted a depression indication and were reported as completed at the time of the search. Exclusions applied to non-trial reports, exclusively open-label or non-randomised trials, studies in healthy volunteers only, and trials reporting only safety or mechanistic endpoints. Ongoing or not-yet-started trials were excluded from most analyses because register entries often lacked detailed methodological information. Data extraction focused on a comprehensive set of pre-defined variables informed by the draft EMA guideline and draft FDA guidance. Extracted items covered trial characteristics and design (phase, allocation, blinding, comparators), population selection (indication, psychiatric comorbidity exclusions, concomitant treatments, prior psychedelic experience), intervention details (substance, dose regimen, number of administrations, and the broader treatment framework including practitioner qualifications and presence of preparatory/integration sessions), and outcome assessments (primary efficacy measures and timing, instruments used to characterise subjective psychedelic effects, and safety monitoring). When available, information from trial protocols and registry attachments was included in the extraction.
After de-duplication, 7,772 records were screened and 421 records relating to 101 controlled psychedelic trials remained; from these, eight completed trials testing psychedelic interventions for depression met the inclusion criteria. The eight trials were conducted between February 2014 and December 2022, comprised a total of 595 recruited patients, and were all early-phase studies: two Phase I/II and six Phase II trials. Five trials evaluated psilocybin, one evaluated intravenous DMT, one evaluated ayahuasca, and one evaluated LSD. Six trials focused on MDD and two on TRD. Design features varied. Sample sizes per trial ranged from 27 to 233 participants. Control conditions included inactive placebo, active placebo, very low (sub-psychoactive) doses of the active substance, and a wait-list control; half of the trials used a placebo control, three used very low dose comparators, and one used a wait-list design. Of the 595 participants, 525 received a psychedelic in a blinded, randomised context (excluding those assigned to inactive placebo or open-label DMT stages). Randomisation was equal allocation in all trials. Reported blinding levels were: four trials described quadruple blinding (participants, care providers, investigators and outcome assessors), two reported triple-blinding, one double-blinding and the waitlist trial reported single-blinding. None of the trials explicitly reported blinding of data analysts, and no trial formally reported systematic assessment of blinding or expectancy among participants. Population eligibility commonly excluded individuals with current or past psychotic or bipolar disorders (seven trials), and several trials also excluded participants with a family history of those disorders. High suicide risk was an exclusion criterion in seven trials. Substance use disorders were excluded in three trials, and most studies prohibited concurrent antidepressant medication; washout periods typically required at least two weeks off antidepressants. Some trials disallowed antipsychotics, psychostimulants, mood stabilisers or anxiolytics, although use of rescue benzodiazepines or certain antipsychotics during dosing was permitted in some protocols. Three trials limited concomitant psychotherapeutic input and three imposed restrictions based on prior psychedelic experience. Interventions were delivered within broader treatment frameworks rather than as isolated drug administrations. Six trials reported a three-stage model of pre-dosing preparation, dosing sessions and post-dosing integration; five trials applied the same framework to both active and comparator arms when reported. Preparatory phases comprised between two and five visits, dosing sessions were typically one or two administrations lasting about 6–8 hours (except intravenous DMT where acute effects lasted roughly 20 minutes), and integration included at least two post-dosing sessions with some protocols offering up to six. Teams of two practitioners were described in the psilocybin trials for which information was available, with minimum qualifications typically at bachelor level in a mental-health-related field and higher qualifications required in some trials; several trials described bespoke practitioner training. Physical settings for dosing were usually homelike, living-room style environments and six trials used standardised music playlists. Three trials planned audio or video recording of dosing sessions for review or qualitative analysis. Primary efficacy outcomes used validated clinician- or self-rated depression scales: MADRS (n = 4), HAM-D or GRID-HAM-D (n = 2), IDS or Quick IDS (n = 2), and BDI (n = 1). The average timing of the last efficacy assessment was 6.25 weeks after the final psychedelic administration; the shortest primary assessment windows were 1–2 weeks in Phase I/II trials, and two trials had blinded efficacy assessments out to 12 weeks. Two trials included open-label extensions with assessments at 13 and 29 weeks, and two long-term follow-ups assessed participants up to 52 weeks post-dosing. All trials incorporated at least one instrument to characterise the acute subjective psychedelic experience: the Mystical Experience Questionnaire (MEQ) was used in six trials, the Altered States of Consciousness scale (ASC) in four, the Emotional Breakthrough Inventory (EBI) in three, the Ego Dissolution Inventory (EDI) in one, and the Hallucinogen Rating Scale (HRS) in one. Additional post-acute measures included the Persisting Effects Questionnaire (PEQ), the Challenging Experiences Questionnaire (CEQ) and measures of mindfulness, humility and personality (e.g. the Ten Item Personality Inventory). Safety approaches commonly included the presence of monitors during dosing (reported in seven trials), availability of a physician nearby (five trials) and periodic vital-sign monitoring (three trials). Standardised suicidality assessments were included in five protocols, chiefly the Columbia Suicide Severity Rating Scale (C-SSRS; n = 4) and the Suicidal Ideation Attributes Scale (SIDAS; n = 1). Two trials used the CEQ to capture challenging acute experiences. The reviewed trials largely prohibited concurrent antidepressant treatment during the study period, and most excluded participants at high risk of suicidal behaviour.
Silva and colleagues interpret their findings as indicating that clinical research on classic psychedelics for depression is progressing but remains at an early stage, with only Phase I/II and II trials completed as of the search cut-off and no Phase III studies finalised. The authors highlight recurring methodological challenges that map onto concerns raised in draft regulatory guidance from the EMA and FDA: functional unblinding and expectancy, the definition and role of the treatment framework or "set and setting", how to characterise and quantify the acute subjective experience and its relationship to outcomes, and systematic collection and interpretation of safety data including psychologically challenging experiences and suicidality. With respect to unblinding and expectancy, the review notes that the pronounced subjective effects of psychedelics make maintaining conventional blinding difficult, particularly when inactive placebos are used or when participants have prior psychedelic experience. The authors describe several mitigation strategies observed or recommended in guidance: use of active placebos or very low-dose comparators, employing blinded independent raters (including remote raters), systematically assessing blinding and expectancy, and trial designs that offer active treatment to all participants at some point (for example dose-ranging designs, wait-list control or open-label extensions). They caution that registry and publication entries sometimes lack clarity about who was blinded, and that none of the reviewed trials reported formal blinding assessments. The review emphasises the centrality of the treatment framework — preparatory sessions, supervised dosing in a controlled environment, integration work, practitioner qualifications and training — and discusses the regulatory implication that developers must disentangle the specific contribution of the pharmacological intervention from that of adjunct psychological support. The question whether the accompanying psychological intervention should be considered a safety measure or an adjunctive psychotherapy with intrinsic efficacy implications is identified as consequential for trial design and for later clinical translation; to address this, trials should standardise and apply the same treatment framework across active and comparator arms and document adherence, for instance via session recordings where ethically acceptable. On outcomes and timing, the authors note that blinded efficacy assessment windows in completed trials were relatively short — commonly up to 12 weeks — while only a subset of studies reported longer-term follow-up. They recommend that later-phase studies address maintenance of effect and dose-response relationships, and balance the ethical considerations of withholding subsequent treatments against the need to characterise durability of effect. Safety considerations discussed include dose-dependent physiological responses (heart rate, blood pressure, anxiety), potential for acute challenging experiences that may have longer-term psychological consequences, and the need for standardised collection of adverse events tied to subjective effects; the CEQ and other instruments were noted as useful tools already employed in some trials. Finally, the authors acknowledge the strengths of their review — a systematic search and regulatory framing — and its limitations, chiefly the cut-off date of 19 February 2024 and the focus on depression which excluded broader psychedelic research. They recommend early and sustained dialogue between developers and regulators through EMA mechanisms such as scientific advice and qualification procedures to ensure that development programmes meet regulatory expectations and to support the medicalisation of psychedelics for unmet patient needs.
The authors conclude that methodological issues are a prominent theme in the early-phase clinical development of psychedelics for depression, but that these challenges are recognised and can be addressed through carefully designed development programmes aligned with regulatory frameworks. They note that EMA platforms exist to support developers, and that multi-stakeholder engagement will be important to advance scientifically robust and regulatorily acceptable trials with the aim of delivering innovative treatments to patients.
The clinical application of psychedelics has received a surge of interest in recent years, marked by a proliferation in academic research, development initiatives and clinical trials investigating the therapeutic potential of these substances in a range of conditions. Psychedelics are substances producing markedly altered perceptual states, with the term often referring to the so-called "classic" psychedelics, which are those acting on serotonin 2A receptors, though other compounds with distinct mechanisms of action, namely 3,4-methylenedioxymethamphetamine (MDMA), are sometimes also categorised as psychedelics. Against a backdrop of growing needs in mental health care, psychedelics have sparked hope and expectations amongst patients and clinicians. While currently no classic psychedelic substance has received marketing authorisation, medicines regulators have responded to this surge of research initiatives. The U.S. Food and Drugs Administration have published draft guidance for developersand Canada and Australia have implemented special access regulatory frameworks that allow prescription under extraordinary situations or for particular indications. While the first application for the approval of a psychedelic treatment in the US, MDMA-assisted psychotherapy for post-traumatic stress disorder, resulted in a negative outcome in August 2024, the concerns identified during the evaluation can provide further insights for medicines developers intending to market psychedelic treatments to treat mental disorders. Psychedelic treatments are of major interest for the treatment of depression. Affecting an estimated 5% of adults and determined as one of the most debilitating mental disorders globally, depression presents a major burden at individual, societal and economical levels, and its incidence has further increased during the Covid-19 pandemic. Additionally, up to two thirds of patients do not respond sufficiently to current first line intervention with antidepressants. Such treatment-resistant depression is strongly associated with greater individual disability, cost of illness and suicide. This underscores the critical need for continued research on and development of innovative therapeutic approaches for depression. To support the investigation of the therapeutic potential of psychedelics in depression, the European Medicines Agency (EMA) has incorporated a preliminary section on this topic in its draft third revision of the guideline on the clinical investigation of medicinal products in depression. EMA scientific guidelines aim to reflect the current state-of-the-art expertise in a field, as informed by regulatory procedures such as scientific advice, and are updated as scientific evidence becomes well-established. The guideline aims to support medicines developers intending to apply for marketing authorisation in the EU, presenting specific considerations when developing products for the treatment of depression, and includes specific considerations for the design of studies on psychedelics. These address methodological concerns for clinical evidence generation associated with psychedelics research. In a complementary action, the EMA also organised a multi-stakeholder workshop on psychedelics to seek direct input from stakeholders and experts and create a channel of communication with regulators (European Medicines Agency, 2024b). In this paper, we present a systematic evaluation of methodological approaches in clinical trials that have explored the therapeutic potential of psychedelics in the treatment of major depressive disorder and discuss these features in light of initial regulatory considerations for this field presented in the draft EMA guideline.
Randomised controlled clinical trials using classic psychedelics (psilocybin, LSD, DMT, Ayahuasca, DMT, 5-meO-DMT) or MDMA were found by searching scientific databases (Embase and Medline) and trial registry platforms (Clinicaltrials.gov and the WHO International Clinical Trials Registry Platform). The search string (available upon request to the authors) was designed by combining terms for psychedelics, both broader class terms and specific keywords for each substance, with terms to identify controlled trials applied to the scientific databases, adapted from an existing Cochrane filter.
The results of the search were screened at the title and abstract level, applying the following criteria: the record was included if it reported on a study which 1) involved the systemic administration of a classic psychedelic substance or MDMA to human participants, or on the long-term follow-up of such a study, 2) was conducted after 1995 (roughly marking the beginning of the "second wave" of psychedelic research), and 3) tested the psychedelic as part of a therapeutic intervention (i.e. at least one of the primary endpoints was a measure of efficacy, and the study participants belonged to a patient population). The resulting records were grouped by trial, and the publications which 4) investigated a depression indication and 5) were reported as completed at the time of the search were retrieved and included in the review. When additional publications or full protocols were annexed to the eligible trials' registries, these were also retrieved for data extraction. Records were excluded if they did not involve a clinical trial, if the trial was exclusively open-label, non-randomised, or on healthy participants only, or if endpoints pertained solely to safety, tolerability or exploration of a mechanism of action. Given that ongoing trials or trials that had not started at the time of analysis did not provide methodological information on most selected variables, these trials were pragmatically excluded from further analyses.
Data extraction was conducted on a comprehensive set of variables that aimed to capture features of the trials' applied methodologies. These variables were pre-defined based on considerations highlighted in the draft EMA guideline on investigation of products for depression, as well as the draft FDA guidance on the clinical investigation of psychedelic drugs. They included general trial characteristics and design features (e.g. trial identifiers, sponsor information, countries of trial sites, trial phase, type of design, allocation, blinding, choice of comparators), population characteristics (e.g. indication, exclusion criteria pertaining to comorbidities, additional treatments, and prior experience with psychedelics), intervention details (e.g. substance, dosing, number of administrations, and treatment framework, i.e. "set and setting" information, including the provision of psychological support or psychotherapy, and the qualification of involved practitioners) and outcome assessments (e.g. primary endpoint measures, efficacy assessment timepoints, assessment of subjective psychedelic effects).
The initial systematic search was completed on 19 February 2024. The search identified 7691 records from publication records and 1700 from clinical trial registries. After removal of duplicates, 7772 records were screened (Fig.). After screening steps 1-3, 421 records presenting information from 101 controlled trials testing the therapeutic efficacy of a psychedelic remained; a full list of these records is available in the online Appendix. Of these 101 clinical trials, 8 completed trials investigating psychedelic treatments for a depression indication were reviewed.
All trials were conducted between February 2014 and December 2022 (Table). Three trials had commercial sponsors, and five were conducted in an academic setting. Two trials were phase 1/2 and six were phase 2 trials. The number of participants across studies ranged from 27 to 233, with a total of 595 recruited patients across all trials. Six trials evaluated the efficacy of a psychedelic substance in the treatment of major depressive disorder (MDD) and two focused on treatmentresistant depression (TRD). Five studied psilocybin, one looked at DMT, one at Ayahuasca, and one at LSD. Half of all trials had a placebo-control group (either an inactive or active placebo), whilst three trials compared the active treatment group (s) to a control group receiving a very low treatment dose (with negligible psychoactive effect). One trial compared immediate treatment with psilocybin with a wait-listed psilocybin control. Two studies -the psilocybin waitlist-control trial, and the DMT trial -offered an active dose to every participant, but the DMT one did so in an open-label extension stage. Thus, of the total 595 participants included in the trials, 525 were treated with a psychedelic in a blinded, randomised setting (excluding participants receiving placebo or open-label DMT). All the trials randomised with equal allocation across study arms. In terms of blinding, four studies reported quadruple blinding, defined as the blinding of study participants, care providers, investigators and outcomes assessors; two studies reported triple-blinding; one reported double-blinding; and one, the waitlist-control trial, reported singleblinding. Blinding of data analysts was not specifically addressed in any of the studies. Notably, while the blinding of "outcomes assessors" was reported in five studies on clinicaltrials.gov, only one explicitly described the use of independent remote raters, as well as other strategies, to further reduce the risk of unblinding. No trials formally reported the assessment of blinding or expectancy amongst participants.
Seven of the reviewed trials excluded participants with comorbid psychiatric conditions, specifically past or present psychotic or bipolar disorders; four trials also excluded individuals with a close family member diagnosed with these conditions (Table). Additionally, three trials ruled out participants with alcohol or other substance use disorder, with one of these also excluding individuals with moderate to severe tobacco use disorder. Notably, high suicidality was a basis for exclusion in seven trials. Regarding medication use, all trials excluded concurrent antidepressant medication of any class during the study (Table). While discontinuation protocols varied by study and medication, they generally mandated a medication-free period of at least two weeks prior to participation. Furthermore, five trials explicitly excluded participants on antipsychotics and psychostimulants, four excluded those on mood stabilisers and three excluded those on anxiolytics (although this did not forbid the potential use of certain antipsychotics or benzodiazepines as rescue medications during the psychedelic dosing session if deemed medically necessary, which was mentioned as an option in five trials). Three studies also imposed constraints on the degree of psychotherapeutic intervention participants could receive alongside the trial. Three trials imposed restrictions on participants' psychedelic use history; participants were excluded if they had used a psychedelic substance within a certain number of months before the trial or more than 10 times in their life, and the Ayahuasca trial fully excluded individuals with experience of this particular substance.
The administration of the psychedelic substance was not done in isolation but rather fit into a broader treatment framework. Most (n = 6) trials reported a three-stage model with pre-dosing, dosing and postdosing phases (Table). For 5 trials, it was explicitly presented that the three-stage model was applied for both the treatment and comparator arms. For the trials for which further data was available, the pre-dosing phase involved between two to five visits, ranging from one to five and a half hours long each, commonly including screening visits and at least two so-called "preparatory" or "safety" sessions. One study specified that their preparation period specifically ranged from 2 to 8 h in total. The dosing phase consisted of one (n = 5) or two (n = 4) planned psychedelic dosing sessions (the DMT trial offered an optional second dose in an open-label phase, resulting in one or two dosing sessions per participant), each spanning 6-8 h, except for the sessions involving intravenous DMT administration, in which the subjective effects were expected to subside after 20 min. When more than one dosing was planned, they were usually spaced 1-3 weeks apart. Following dosing, there were a minimum of two "integration" sessions, with some protocols involving up to six visits encompassing additional evaluations or further integration, particularly in repeated-dosing frameworks, according to the needs of participants. Moreover, six studies incorporated extended follow-up procedures, which entailed additional assessments conducted either in person or over the phone. When looking at the content of the sessions accompanying psychedelic administration, the terms "treatment framework" and "practitioner" will henceforth be used to refer to the treatment model and to the individuals delivering support in such models, independently of the terms employed in each individual study (which will be signalled by quotation marks). The reviewed studies revealed slight variations in the applied treatment frameworks and their respective terminologies, although it is relevant to note that these may be influenced by the particular psychedelic substance, the study phase, and the type of publication reporting the information (studies with only trial registries as data sources usually contained less information). For instance, the LSD trial did not define any particular treatment framework. The Ayahuasca trial mentioned "support" being provided during the session, and the DMT study referred that "therapists" would be available for "support and guidance" with the outcomes of the psychedelic experience. The psilocybin studies employed a range of terms for their treatment frameworks, such as "psychotherapy", "psychological support" and "psychological counselling", as well as when referring to the involved practitioners, "facilitators" or "therapists", sometimes using these terms seemingly interchangeably even within the same trial. Three trials explicitly mentioned the delivery of "psychotherapy", with one detailing a treatment framework that allowed practitioners to utilise various psychotherapeutic modalities tailored to each participant's needs. A fourth trial, while not using the term "psychotherapy" but "psychological support", described the implementation of a manualised "Set and Setting" protocol. The four psilocybin trials for which information on the treatment framework was available described teams of two practitioners, with some distinguishing between "lead" and "assistant" or "cofacilitator" roles. They specified the practitioners' required qualifications, with all possessing at least a university degree in a mental healthrelated field and some necessitating higher qualifications for at least one of the practitioners, for example, a master's or doctoral level degree, or a doctoral degree with additional experience in treating MDD. Three of these trials also explicitly alluded to specially designed training programmes for their practitioners. Often, external guidance and prior research on psychedelic-assisted therapy was referenced for the development of the applied treatment frameworks and practitioner training and best practices. Regardless of the adopted treatment framework and terminologies, descriptions of the purpose of each phase of the treatment model and the role of the professionals presented some common themes across the trials. In the pre-dosing phase, sessions aimed to establish trust and rapport in order to minimise the risk of adverse effects on the day of dosing (n = 4), set intentions and expectations (n = 3), as well as reviewing information and educational materials on the psychedelic experience (n = 2). During dosing sessions, the two professionals were present to monitor the participant (n = 2) and respond to their needs (n = 4). Post-dosing "integration" sessions facilitated the participant's inner processing of the experience (n = 5) and addressed safety questions (n = 2). Notably, three studies reported that video and audio of dosing sessions would be recorded, provided the participant had consented, for review, qualitative analysis, and verification of protocol compliance. In one case, post-dosing interviews were also planned to be recorded. In relation to the setting, for five of the reviewed trials information was available about the physical environment for the dosing sessions; these were consistently described as spaces designed to emulate a homely living-room ambiance with the aim of promoting calm and comfort. Six trials also noted the implementation of a standardised music playlist during the psychedelic dosing session. Treatment framework described 6 (75) 1 (12.5) 1 (12.5) Same treatment framework in comparator arm 5 (62.5) 0 3 (37.5) Described as "psychotherapy" 2 (25) 5 (62.5) 1 (12.5) Qualifications for practitioners described 3 (37.5) 4 (50) 1 (12.5) Training for practitioners described 3 (37.5) 4 (50) 1 (12.5) Setting for dosing described 6 (75) 1 (12.5) 1 (12.5) Use of music during dosing 6 (75) 1 (12.5) 1 (12.5) Recording of dosing sessions 3 (37.5) 4 (a Could not be assessed from the information available
Trials used at least one validated tool to assess changes in disease severity as a primary endpoint. Tools included the Montgomery-Asberg Depression Rating Scale (MADRS) (n = 4), the Hamilton Depression Rating Scale (HAM-D or GRID-HAM-D) (n = 2), the Inventory of Depressive Symptomatology (IDS or Quick IDS) (n = 2), Beck Depression Inventory (BDI) (n = 1) (Table). On average, the time of the last efficacy assessment occurred at 6.25 weeks after the last administration of the psychedelic. The shortest efficacy assessment timepoints were at one and two weeks, in the two phase 1/2 trials, and the most extended efficacy assessment timepoint took place at 12 weeks post-dosing, in two trials. These figures do not account for further assessments during open-label extension periods; two trials included such an extension, with evaluations conducted at 13 weeks (three months) and 29 weeks (six months) after the last dosing session. Additionally, there were two long-term follow-up analyses which assessed participants for up to 52 weeks (one year) following the psychedelic dosing period. In addition to evaluating changes in symptom severity, all reviewed trials incorporated at least one tool to characterise the psychedelic experience. Tools to assess this included the Mystical Experiences Questionnaire (MEQ) (n = 6), the Altered States of Consciousness Scale (ASC) (n = 4), the Ego Dissolution Inventory (EDI) (n = 1), the Hallucinogen Rating Scale (HRS) (n = 1), and the Emotional Breakthrough Inventory (EBI) (n = 3) (Table). For one trial, the protocol included interviewing of participants about their acute subjective effects. Except for two studies, these assessments were conducted immediately following the psychedelic experience. Furthermore, five trials utilised tools to evaluate the post-acute psychological effects of the psychedelic administration, such as the Persisting Effects Questionnaire (PEQ) (n = 1), the Challenging Experiences Questionnaire (CEQ) (n = 2), and various other metrics that assess shifts in mindfulness and humility (n = 1), and personality traits, with the Ten Item Personality Inventory (TIPI) (n = 1). One trial employed the Geneva Emotional Music Scales (GEMS) to gauge the emotional impact of the music participants were exposed to during the psychedelic dosing sessions.
The presence of monitors during dosing sessions was explicitly mentioned in seven of the reviewed trials, and for three it was specified that the monitors were the same practitioners who accompanied the participant throughout the treatment framework (Table). Additionally, five trials stated the availability of a physician nearby (who could have been one of the practitioners monitoring the dosing session) and another three also reported the ongoing monitoring of vital signs. Five studies reported conducting suicidality assessments using the Columbia Suicide Severity Rating Scale (C-SSRS) (n = 4) and the Suicidal Ideation Attributes Scale (SIDAS) (n = 1), both before and after treatment, and often continuously throughout the study.
Eight completed randomised clinical trials investigating the therapeutic potential of psychedelics in a depression indication were identified, and several variables pertaining to their methodologies were extracted and reviewed. There were five trials of psilocybin, and the remaining three tested LSD, Ayahuasca and intravenous DMT. Notably, all these are classic psychedelics, and while MDMA was included in the search, no trials using it for depression had been completed. Most trials were phase 2, except the Ayahuasca and DMT studies which were classified as phase 1/2. There are no completed phase 3 trials to date, although two were ongoing at the time of writing (NCT06308653 and NCT05624268). Although important strides are being made in this field, it is recognised by stakeholders and regulators that a number of knowledge gaps remain to be addressed to generate clinical evidence sufficient to support a marketing authorisation application for psychedelic medicines. These include the impact of unblinding and expectancy, the definition of "set and setting" and importance of applied treatment frameworks, the characterisation and role of the acute psychedelic experience, and the systematic collection of safety data. These questions are reflected in FDA and EMA draft guidance documents on the use of psychedelics to treat depression (FDA, 2023; European Medicines Agency, 2024a), and were explored in an EMA multi-stakeholder workshop on psychedelics (European Medicines Agency, 2024b). Still, since the majority of completed work has been early phase, it should be noted that certain characteristics of the reviewed studies fundamentally diverge from the more extensive phase 3 studies usually required to support a marketing authorisation application, for example, in terms of sample size and population diversity and comprehensive efficacy and safety assessments. Questions such as the influence of inter-individual differences, the impact of psychiatric comorbidities, the possible use of concomitant medications and characterisation of long-term safety can and should be addressed in subsequent, larger clinical trials and potentially supported In terms of targeted indications, the majority of trials focused on MDD, while two specifically opted to focus on TRD. The EMA draft depression guideline highlights that beginning development on more severely affected populations, such as those with TRD, might be favoured due to the marked perceptual and behavioural alterations induced by psychedelics and the associated difficulties in defining a benefit-risk balance (European Medicines Agency, 2024a). While some stakeholders may defend that a treatment with the potential to alter the disorders' trajectory could be a valid first intervention (European Medicines Agency, 2024b), the subjective effects of psychedelics and psychological mechanisms of these treatments, involving profound and perhaps challenging emotional experiences, and associated safety considerations may favour initially focusing on populations not responsive to other treatment approaches. Downstream issues related to the implementation of these treatments in the healthcare system, are further arguments against initial development of psychedelics as a first-line approach. The reviewed trials adopted randomised, controlled, parallel-arm designs. Control groups varied from the typical, inactive placebo control to active placebos or low doses of the active substance, to a waitlistcontrol. Data pertaining to blinding strategies, primarily extracted from trial registries, indicated that all but one study concealed treatment allocation both from the participants and study personnel, with the remaining trial adopting an alternative waitlist-control design wherein only participants could be blinded. The randomised double-blind placebo-controlled clinical trial is considered the gold standard for establishing the efficacy of an investigational medicinal product. Yet, this model must be adapted to the particular features of the investigational product and its intended use. It is recognised that some inherent characteristics of psychedelics treatments pose challenges in upholding the conventional trial model. For example, while not unique to psychedelics, the adequate blinding to treatment allocation is an exacerbated challenge in psychedelic trials due to the unique subjective effects of these substances. Functional unblinding is particularly prominent in trials with inactive placebo controls and in trials with psychedelic-experienced participants, and has the potential to introduce bias by creating expectancy effects. In addition to unblinding, such expectancy effects may also stem from other trial-related factors, such as the information provided by the trial team or external factors like the public perception and media representation of psychedelic treatments. Expectancy can impact outcomes in either direction: positive expectancy may enhance therapeutic benefits not attributable to the active substance (placebo effect), while excessive expectations can result in disappointment and clinical worsening (nocebo effect). While challenges pertaining to unblinding, expectancy and choice of an adequate comparator are recognised and should be considered in clinical trials of psychedelics, they are not an insurmountable obstacle to obtain a marketing authorisation. As highlighted in the new EMA guideline revision, there are strategies to help address these issues (European Medicines Agency, 2024a). These include the use of blinded independent raters and the incorporation of systematic assessments of blinding and expectancy in trials, the use of active placebo or low doses of psychedelic substances as comparators, and the offering of psychedelic to every participant. Most trials did specify blinding of outcomes assessors, but discrepancies and ambiguities in the methods of reporting blinding approaches across different sources (e.g. scientific publication, protocol and trial registry entry) should be taken into account. For example, a trial entry on the trial registry may not specify whether the roles of "investigators" compared to "care providers" or "outcomes assessors" are held by the same individuals. Additionally, publications generally do not specify who was responsible for data analysis and whether they were blinded to treatment conditions of trial participants. Yet, it should also be considered that, due to the inherent challenges in maintaining blinding with psychedelics, scientific papers and study protocols may refrain from claiming certain groups were blinded. One trial described the use of remote, blinded raters for primary outcome measures, but none reported the systematic assessment of blinding and expectancy. With regard to the choice of comparator to minimise unblinding and negative expectancy effects, six out of eight trials opted for using active comparators or lower psychedelic doses. Only phase 1/2 trials used inactive placebo, which might be more fitting to characterise initial safety and tolerability. Offering the active substance to every participant (as was done in the trials testing different dose levels of the psychedelics, in the waitlist-control trial and the open-label stage of the DMT trial) may be a way to counter negative expectancy effects arising from the participant realising they were assigned to the placebo group. Additionally, the management of expectancy effects should be sought through the rational and careful delivery of information on the therapeutic potential of the psychedelic substance, and the adequate training of practitioners to unbiasedly deliver such information. It was observed specifically in the Ayahuasca trial that choosing to exclude experienced users may be another approach by sponsors to minimise unblinding, but this might not be preferred in larger later phase trials because it could limit sample representativeness. Yet, it raises the valuable question of what impacts, if any, previous experience of participants and study personnel with psychedelic substances may have in this context, for instance in terms of expectancy effects. As shown in the reviewed trials, psychedelic medicines are being developed as treatments requiring only one or two administrations in a controlled setting, which represents a paradigm shift from most other antidepressants. Importantly, in these trials the delivery of the psychedelic treatment not only occurred in a controlled environment with ongoing monitoring, but often involved a broader treatment framework involving the provision of psychological support or psychotherapy. This included pre-dosing preparation and post-dosing integration sessions with a trained practitioner, as well as their possible involvement during the dosing sessions. This is in part because the subjective effects of psychedelics are known to be largely impacted by the "set" and the "setting", respectively referring to the individual's mindset going into dosing and the environment in which the psychedelic is taken. The importance of the treatment framework is further emphasised in light of the hypothesis that the effect of psychedelics treatments cannot be entirely reduced to the pharmacological effect of the substance but rather can be understood as a catalysed process of inner healing and ongoing personal transformation, with many trial participants perceiving the psychedelic experience as the start of a therapeutic journey. From a medicines' regulation perspective, it is imperative to quantify the specific contribution of the administration of the psychedelic, disentangling it from the potential effect of any additional interventions. It is currently debated whether the treatment framework accompanying the administration of psychedelics is a safety measure or an adjunct therapeutic intervention that might enhance efficacy (as the term "psychotherapy" indicates), and each has distinct implications for medicine development. In any case, this highlights the importance of clearly describing and applying the same treatment framework in both active and comparator arms in clinical trials with psychedelics. While not always explicit, this appeared to be the case in all the reviewed studies for which details on the treatment framework were available. Adequate adherence to treatment framework protocols could be checked through the audio and video recording of study sessions, which was done in three of the reviewed trials, albeit implications for confidentiality and the potential for impacting behaviours and biasing results should be considered. Additionally, the clear definition of the treatment framework will be of importance for later translation into clinical practice. There is a wide demand for transparent communication of applied treatment frameworks and training requirements for practitioners from trials and for the setting of global gold standards without compromising the ability meet individual needs. In this vein, it was interesting to see that, despite discrepancies in the applied treatment frameworks in the reviewed trials, some common themes in their objectives and roles of practitioners across the pre-dosing, dosing and post-dosing phase, as well as in features of the setting, could be distinguished. These themes included providing a calm and comfortable physical environment, building trust with practitioners and providing clear and thorough information about the treatment prior to administration, addressing participants' emerging needs during dosing, and supporting the integration of the experience afterward, efforts aimed at ensuring participant welfare. Other areas highlighted as requiring further research to meet regulatory requirements are the maintenance of therapeutic effects and the further characterisation of dose-response (European Medicines Agency, 2024a). In the reviewed trials, the latest efficacy timepoint was one year after dosing in a long-term follow up, with 12 weeks as the longest blinded period. It should be noted that all reviewed studies forbid the use of concomitant antidepressants and some even limited the amount of psychotherapy participants could receive alongside the trial. Thus, if long-term efficacy is seen as something that must be characterised in absence of subsequent treatments, a balance between this characterisation and the ethical consideration of withholding a patient from seeking further treatment during a longer trial period arises. However, in light of the evolving way of describing treatment effects in clinical trials, it may be possible to include further treatments in the definition of the efficacy question of interest, or at least allow them in the study to be dealt with modelling methods in the analysis. It is also important to further elucidate the relationship between the dose of psychedelic, the degree of subjective effects or particular features of the psychedelic experience, and therapeutic outcomes. There are a range of standardised questionnaires designed to measure features of the subjective psychedelic experience which may be helpful in this regard. All of the reviewed trials used at least one of such tools. As mentioned, the dosing sessions took place in a controlled environment, with monitoring by trained practitioners and the nearby availability of a physician explicitly reported for most reviewed trials, and the assessment of vital signs explicitly mentioned in three trials. During the acute experience, increases in heart rate, blood pressure and anxiety levels, alongside the marked alterations in the perception of reality and the possibility of challenging psychological experiences, underscore the need of administering these substances in a controlled environment. Existing evidence appears to support that the incidence of these effects is dose-dependent. Notably, there is an ongoing debate on how best to frame the psychologically challenging experiences that may arise during the acute psychedelic effects. Although these can be emotionally distressing, their intensity, duration and possible resolution of therapeutically-relevant content may have varying impacts on clinical outcomes. While all clinical trials must collect and report data on adverse events (AEs), the need to define and systematically monitor them in psychedelics trials, particularly those associated with the subjective effects of these substances, has been emphasised by various stakeholders, in order to build a robust safety knowledge base and help better inform clinicians and patients involved in psychedelic treatments. Interestingly, two of the reviewed studies applied the standardised Challenging Experiences Questionnaire, positing it as a potentially useful measure to support the capturing of these effects. Because the psychedelic experience can have enduring psychological impact and, if challenging, potentially cause distress long after the acute subjective effects have subsided, the need for long-term safety surveillance needs to be considered. This also underscores the importance of the provision of continued care post-dosing, both in the form of integration sessions and the availability of support systems post-trial. Notably, the EU Horizon-funded, multi-national PsyPal trial, focusing on the therapeutic application of psilocybin in palliative care for progressive, incurable illnesses, aims to implement this feature (European Commission EU Funding & Tenders Portal). Five of the reviewed trials specifically included assessments of suicidality with standardised tools in their protocols (although it should be noted that high risk of suicidal behaviour was an exclusion factor in most trials). Given the focus on depression, the need for careful monitoring and investigation of suicidality has been highlighted in the draft EMA depression guidelines as a safety-related concern requiring attention in further studies of psychedelics (European Medicines Agency, 2024a). Another vital consideration for safety is mitigating the risks associated with the state of heightened vulnerability and suggestibility experienced by individuals under the subjective effects of psychedelics. For this purpose, the meticulous drafting of informed consent procedures, informing participants as accurately as possible about what they may experience, and the proper training and appropriate professional oversight of the practitioners accompanying them is paramount. The strengths of this paper lie in the use of a robust systematic search strategy and its unique combination of a detailed review of psychedelics clinical trials' methodologies with regulatory insights important for the advancement of psychedelic medicalisation in the EU. Limitations of this work include: the limited scope on trials done by the date of the search, February 19, 2024, excluding eligible studies that may have been completed thereafter, and focusing on depression, (although a complete list of all trials found in the search is available in the Appendix). Specific recommendations tailored to the particular development programme can be sought through engagement with medicines' regulators. EMA platforms and tools, such as scientific advice, qualification procedures and the innovation task force, exist for developers to seek direct input from regulators on different aspects of their development programmes, and can help ensure that psychedelics trials are methodologically sound and meet regulatory requirements. Furthermore, the EMA offers additional support and incentives to registered small and medium sized enterprises through its SME office. Ultimately, dialogue and collaboration between developers, medicines regulators, and other key stakeholders, will be essential for the development and application of psychedelic medicines.
Although the development and clinical application of psychedelics encompass a spectrum of challenges, methodological issues arising from the design of clinical studies consistently emerge as a prominent theme. While the trials completed in depression examined in this paper are early phase studies and more work is needed, the present review highlights that existing knowledge gaps and methodological complexities are being acknowledged and can be addressed with carefully conceived development programmes adhering to existing regulatory frameworks. The EMA has systems in place to assist developers in these matters and support the medical development of psychedelics. Multi-stakeholder engagement can ensure that meaningful progress is being made toward bringing innovative treatments to patients in need.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Create a free account to open full-text PDFs.
Papers cited by this study that are also in Blossom
Breeksema, J. J., Kuin, B. W., Kamphuis, J. et al. · Journal of Psychopharmacology (2022)
Butlen-Ducuing, F., McCulloch, D. E-W., Haberkamp, M. et al. · Lancet (2023)
Carbonaro, T. M., Bradstreet, M. P., Barrett, F. S. et al. · Journal of Psychopharmacology (2016)
Carhart-Harris, R. L., Giribaldi, B., Watts, R. et al. · New England Journal of Medicine (2021)
Davis, A. K., Barrett, F. S., May, D. G. et al. · JAMA Psychiatry (2021)
Goodwin, G. M., Aaronson, S. T., Alvarez, O. et al. · New England Journal of Medicine (2022)
Goodwin, G. M. · Journal of Affective Disorders (2023)
Davis, A. K., Streeter Barrett, F., Cosimano, M. P. et al. · Journal of Psychopharmacology (2022)
Jacobs, E., Murphy-Beiner, A., Rouiller, I. et al. · Neuroethics (2023)
Palhano-Fontes, F., Barreto, D., Onias, H. et al. · Psychological Medicine (2018)
Peill, J. M., Marguilho, M., Erritzoe, D. et al. · Journal of Psychopharmacology (2024)
Raison, C. L., Sanacora, G., Woolley, J. D. et al. · JAMA (2023)
Roseman, L., Nutt, D. J., Carhart-Harris, R. L. · Frontiers in Pharmacology (2018)
Neil, J. C., Nutt, D. J. · Journal of Psychopharmacology (2022)
Papers in Blossom that reference this study
Yacoub, J., Bray, E., Bayyat, J. et al. · Biorxiv (2026)
Studerus, E., Gamma, A., Kometer, M. et al. · PLOS ONE (2012)
Schindowski, E. M., Jungwirth, J., Schuldt, A. et al. · EClinicalMedicine (2023)
Weiss, B., Leor Roseman, •., Giribaldi, B. et al. · International Journal of Mental Health and Addiction (2024)