Abstract
Background:
Results from contemporary clinical trials of serotonergic psychedelic therapies have led to an increasing focus on their potential clinical use across mental disorders. However, studies examining mechanisms of clinical response to psychedelic therapy in psychiatric populations are limited. This review aimed to synthesize evidence from studies examining biomarkers of clinical response to psychedelic therapies.
Data sources and methods:
A systematic search of four databases (MedLine, PsycInfo, EMBASE, and Web of Science) for studies investigating treatment response to psychedelic therapies in psychiatric populations that included both clinical outcomes and a related biomarker was conducted on January 10, 2024. Studies were included if they reported on prospective clinical trials involving the use of a psychedelic in participants diagnosed with any Diagnostic and Statistical Manual or International Classification of Diseases mental disorder, where a biological marker was measured and evaluated in association with treatment response.
Results:
Nine studies investigating the effects of Ayahuasca and psilocybin in major depressive disorder and treatment-resistant depression were included in this review. Several potential biomarkers of response were explored through neuroimaging and blood samples, with significant associations found for serum brain-derived neurotrophic factor, serum C-reactive protein, cerebral activation of the amygdala, and functional connectivity between regions such as the ventromedial prefrontal cortex, anterior cingulate cortex, and posterior cingulate cortex.
Conclusion:
Results of small studies suggest associations between several putative biomarkers and treatment response to psychedelic therapies. Future trials of psychedelic therapies should integrate biomarker assessment in longitudinal designs to advance the understanding of their mechanism of action in mental disorders.
Trial registration:
This study protocol was registered to PROSPERO under the number CRD42021291171.
Plain language summary
Psychedelic therapies, like those using psilocybin and Ayahuasca, are showing promise in treating mental health conditions like depression. However, we still don’t fully understand why these treatments help some people more than others. This review looks at research that has explored biological markers—things we can measure in the brain or body—to see if they can predict who will respond best to these treatments. Studies have found that brain activity and certain proteins in the blood, such as brain-derived neurotrophic factor (BDNF) and C-reactive protein (CRP), may be linked to treatment success. Though the studies are small, they suggest that further research could help us better understand how psychedelics work and improve their use in mental health treatment.
Introduction
There has been converging evidence on the therapeutic potential of serotonergic psychedelic therapies (PT) for mental disorders, including major depressive disorder, post-traumatic stress disorder (PTSD), end-of-life anxiety, and substance use disorders.1 –3 “Classic” serotonergic psychedelics include lysergic acid diethylamide (LSD), psilocybin, ayahuasca, mescaline, and dimethyltryptamine (DMT).4,5 These molecules act on serotonin (5-HT)2A and other receptors to sensory perceptions, altered states of consciousness, alterations in mood, and cognition. 6 When combined with psychological support, there is replicating clinical trial evidence on the safety and efficacy of psychedelics for the treatment of PTSD, substance use disorders, and treatment-resistant major depressive disorder (MDD).4,7 –10
It is hypothesized that through the activation of 5-HT2A receptors, psychedelics increase neural entropy, which allows brain activity to be more flexible and open to new experiences by loosening rigid patterns of thought and perception in the brain. 11 This allows for a bottom-up sensory information to flow more freely, which reduces previously dominant negative narratives that are reinforced in depression or anxiety. This may allow for a reorganization of thought patterns, which can lead to new insights, emotional breakthroughs, and long-term therapeutic benefits when combined with psychological support. 11 Despite the emerging evidence on the efficacy of PT, the mechanisms of these therapeutic effects are not fully understood.
One approach to attempt to delineate the biological mechanisms of therapeutic actions of PT is through the investigation of central and peripheral biomarkers of treatment response. Given the heterogeneity of mental disorders, there is a call for reliable, robust, and valid biomarkers that could aid in the diagnosis, prognosis, and personalized treatment of mental disorders. 12 Integrating biomarker assessment within clinical trials of emerging treatments like PT could provide insight into the mechanisms of their therapeutic action, as well as the biology of stress-related disorders such as MDD. Correlating putative biomarkers with treatment response to PT may help provide data on potential predictors of treatment response to personalize PT treatment. In this context, the current systematic review aimed to synthesize current evidence on biological correlates of treatment response to PT in adults with mental disorders.
Methods
Literature search
This study protocol was registered to PROSPERO under the number CRD42021291171 prior to commencing the review. The literature search following PRISMA 2020 guidelines 13 was initially conducted on November 19, 2021 and was updated on January 10, 2024, using the following databases: MedLine, PsycInfo, EMBASE, and Web of Science. Key search terms used were (psychedelic* OR (lysergic acid diethylamide OR LSD) OR mescaline OR psilocybin OR Ayahuasca OR (dimethyltryptamine OR DMT) OR ibogaine) AND (neural correlate* OR peripheral correlate* OR marker* OR (mechanism* OR mechanism of action) OR correlate OR biomarker OR (FMRI OR PET OR EEG OR imaging) OR (inflammation OR inflammatory marker)). The full MedLine search is in the Supplemental Figure 1.
Eligibility criteria
Studies included met the following criteria: (1) Included adults aged 18 years or older with mental diagnoses according to any version of standardized DSM or International Classification of Diseases criteria; (2) studies administered at least one dose of a serotonergic psychedelic drug; (3) studies must have included both a measure of clinical response and a biological marker or correlate, such as imaging, peripheral biomarkers, or neurophysiology; (4) studies were required to include a comparator arm, which could include inert placebo, active placebo, treatment as usual, other pharmacological or psychological treatments, or pre- and post-psychedelic dosing measures; (5) studies had a sample size of greater than 5 in each treatment arm, and; (6) study designs must have been prospective studies or secondary analysis of prospective studies and excluded chart reviews. Excluded studies included: (1) studies which used non-serotonergic psychedelics, such as cannabis, ketamine, or 3,4-methylenedioxymethamphetamine; (2) pediatric or non-psychiatric populations; (3) animal studies; and (4) review papers or meta-analyses. Only studies published in English in a peer-reviewed journal were included.
Outcome measures
The main outcomes of this review were statistical associations between treatment response to psychedelic therapy (as measured by changes to standardized clinical rating scales) and changes to biological markers. Neural correlates were defined as associated brain changes after administration of a psychedelic therapy, quantitatively assessed via tests including functional magnetic resonance imaging (fMRI), positron emission tomography (PET), and electroencephalogram (EEG). Peripheral correlates were defined as associated biochemical changes after administration of a psychedelic therapy, quantitatively assessed via blood tests.
Study selection
Abstracts were screened based on predetermined inclusion criteria by two independent reviewers (CT, BDMJ, SW, and SGS) on Covidence. Relevant full texts were then reviewed by two independent reviewers (SW and MTT). Conflicts were discussed and resolved by consensus.
Data extraction
Data extraction was performed by two independent reviewers (SW and MTT) using a Microsoft Excel sheet created a priori. The extraction categories included study design characteristics (including observation window, diagnostic criteria, and diagnosis of the population), descriptive statistics of the study sample, treatment intervention substance and dosing information, adverse events, clinical outcome findings, and biological outcome findings.
Quality assessment
The included studies were assessed for quality using the Cochrane Risk of Bias Tool 2 14 independently by two reviewers (SW and MTT). Five key domains were evaluated: random sequence generation (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), and selective reporting (reporting bias). Each domain was rated as low, moderate, or high risk. Overall study risk was summarized as follows: studies were rated as “Low” if most domains were judged to be low risk with no major concerns in any critical areas; “Medium” if there was a mixture of low and moderate risk ratings or a single moderate concern in a less critical domain; and “High” if one or more critical domains were rated as high risk or if multiple domains raised substantial concerns.
Data synthesis and analysis
Where possible, when three studies reported a sufficiently similar outcome, pair-wise meta-analysis was performed. However, this was limited given the substantial heterogeneity of the included studies. Subgroup analyses based on the biological marker utilized and the psychiatric population were completed where possible.
Results
The literature search yielded 6807 studies, 3960 of which were duplicates and removed prior to further evaluation. A total of 3268 studies were excluded for not meeting inclusion criteria at the abstract screening stage, and a further 299 studies were excluded for not meeting inclusion criteria at the full paper review stage (Figure 1). Consequently, nine studies were included in this review (Table 1).
Systematic article identification process.
Summary of included articles.
ACC, anterior cingulate cortex; ASL, arterial spin labeling; BDI, Beck Depression Inventory; BDNF, brain-derived neurotrophic factor; BOLD, blood oxygen level dependent; CBF, cerebral blood flow; CRP, C-reactive protein; EEG, electroencephalogram; fMRI, functional magnetic resonance imaging; GRID-HAMD, GRID-Hamilton Depression Rating Scale; HAMD, Hamilton Depression Rating Scale; IL-6, interleukin-6; MADRS, Montgomery–Åsberg Depression Rating Scale; MDD, major depressive disorder; PCC, posterior cingulate cortex; PFC, prefrontal cortex; QIDS-SR16, Quick Inventory of Depressive Symptomatology—Self-Report; RSFC, resting-state functional connectivity; SHAPS, Snaith-Hamilton Pleasure Scale; STAI, State-Trait Anxiety Inventory; TRD, treatment-resistant depression; vmPFC, ventromedial prefrontal cortex.
Study populations and characteristics
All the studies focused on depression with either a diagnosis of MDD (seven studies) or MDD (three studies). Two studies15,19 included control participants in addition to a patient population, and one study 17 included an escitalopram comparator group. As such, these patient populations had severe levels of depression and prior nonresponse to traditional antidepressant medication.15,17,18 Mertens et al. 20 classified the severity of depression using the Quick Inventory of Depressive Symptomatology—Self-Report (QIDS-SR16). That study had a predominantly severe or very severe population, with patients diagnosed with major depression for an average of 17.7 years and having tried an average of 4.6 antidepressant medications. In Almeida et al., the patient sample had a moderate to severe depression with a HAMD score ⩾17 at the time of the study. The duration of illness ranged from 1 to 10 years. All the patients in that study had previous pharmacotherapy with at least two antidepressants, with 75% having used two to four different antidepressants and 25% having used five or more. All had a history of treatment with serotonin-selective reuptake inhibitors (SSRIs), 64.3% with tricyclic antidepressants (TCAs), 53.6% with serotonin and norepinephrine reuptake inhibitors (SNRIs), and 25.0% with norepinephrine and dopamine reuptake inhibitors (NDRIs). The MDD patients in Skosnik et al. 23 had a mean of 20 years since diagnosis and a mean of 4.7 trials of different antidepressant medications prior to enrollment in the study. Over half (57.8%) had a comorbid anxiety diagnosis.
Almeida et al. 15 was also one of two studies to list sociodemographic and education characteristics, with most participants having secondary education, but with control participants having higher levels of education and income. Skosnik et al. 23 found that the majority (89.3%) of their patient population had greater than high school level education, with 57.8% completing college level education and 26.3% completing a Master’s degree. The mean age ranged from the lowest at 39.8 years to the highest at 44.7 years.20,21 Doss et al., 18 Almeida et al., 15 and Skosnik et al. 23 had a higher proportion of women participants (66.7%, 63%, and 68%, respectively), while the remaining studies had over two-thirds of their sample size consisting of male participants.
All studies’ exclusion criteria included current medical disease, current or previous history of neurological disorders, and personal or family history of schizophrenia or bipolar affective disorder. Most studies excluded participants with previous use of psychedelics and previous substance abuse; however, Skosnik et al. 23 included participants with prior psychedelic use if they had a minimum of 3 years since their last use of hallucinogens. The psilocybin studies provided psychological support prior to and surrounding the dosing sessions to ensure participants were safe and informed.
Ayahuasca
Two studies based on the same randomized double-blinded placebo-controlled trial with parallel arm design (NCT02914769) in Brazil explored the effects of Ayahuasca treatment.15,19 Both studies included 28 patients with MDD and 45 control participants. The placebo formulation was matched for the brownish color and sour and bitter taste of Ayahuasca. Zinc sulfate was incorporated to produce mild gastrointestinal upset, a common side effect of Ayahuasca. Ayahuasca was delivered in a single liquid dose of 1 mL/kg. Dosing sessions in a controlled environment lasted 6 h for Galvao-Coelho et al. 19 and 4 h for Almeida et al. 15 Brain-derived neurotrophic factors (BDNF), the biomarker studied by Almeida et al., 15 were elevated in the participants in the Ayahuasca condition compared to the participants in the placebo condition. Ayahuasca treatment was associated with a reduction in depression symptoms in patients as measured by MADRS 48 hours after the dosing session, and this association was significantly negatively correlated with serum BDNF levels 48 h after the dosing session. Thus, improvement in depression symptoms post-treatment was associated with increased post-treatment BDNF levels. Of note, baseline serum BDNF was not a predictor of MDD. Galvao-Coelho et al. 19 included the biological correlates of C-reactive protein (CRP), cortisol, BDNF, and interleukin 6 (IL-6); however, only CRP demonstrated a significant association, with decreased post-treatment plasma CRP in both patient participants and the control group. In the patient group, the Ayahuasca-associated reduction in depression scores as measured by the MADRS was significantly correlated with the reduction in CRP levels. In summary, these two studies reported associations between Ayahuasca treatment response and an increase in BDNF and a reduction in CRP, respectively.15,19
Psilocybin
The remaining seven studies explored the effects of psilocybin administered with psychological support. Four studies16,20 –22 from the United Kingdom were registered under the open-label clinical trial ISRCTN14426797. All these studies involved a 10 mg test dose and 25 mg therapeutic dose administered 1 week apart with functional magnetic resonance imaging (fMRI) as the main biological correlate in 19 adults with MDD.
Carhart-Harris et al. 16 examined the effect of psilocybin therapy on cerebral blood flow (CBF) and resting-state functional connectivity (RSFC) at 1 day and 5 weeks post-treatment. Psilocybin therapy was associated with an acute reduction in depression symptoms 1 day post-treatment and a sustained reduction in depression symptoms at 5 weeks post-treatment, as measured by the QIDS-SR16. Almost half (47%) of patients demonstrated a ⩾50% reduction in QIDS-SR16 score at 5 weeks post-treatment. There was a significant association between psilocybin therapy’s treatment response and decreased CBF in the bilateral amygdala 1 day post-treatment. The study also reported increased resting-state functional connectivity (RSFC) between the ventromedial prefrontal cortex (vmPFC) and the bilateral inferior-lateral parietal cortex 1 day post-treatment predicted treatment response at 5 weeks. Conversely, decreased bilateral parahippocampus RSFC with the lateral and medial prefrontal cortex (PFC) 1-day post-treatment predicted treatment response at 5 weeks.
Using data from the same trial, Mertens et al. 20 examined the effect of psilocybin therapy on PFC functional connectivity during processing of emotional faces. Findings showed that psilocybin therapy was associated with a reduction in depression score according to the Beck Depression Inventory (BDI) and that approximately 58% of patients met criteria for remission (BDI ⩽ 9) at one-week post-treatment. They found that psilocybin therapy was associated with increased functional connectivity between the amygdala and several visual areas (intracalcarine cortex, supracalcarine cortex, cuneus, precuneus, and right lateral occipital cortex) during face processing 1-day post-treatment. They also found an associated increased functional connectivity of the vmPFC with left occipital and parietal lobe regions (supracalcarine cortex, intracalcarine cortex, precuneus, and lingual gyrus) during face processing 1 day post-treatment. In particular, functional connectivity between the vmPFC and the occipital-parietal cluster 1 day post-treatment was significantly correlated with the change in BDI scores at 1 week post-treatment. Roseman et al. 21 also examined psilocybin’s effect on emotional face processing. They reported that state depression ratings showed a significant reduction 1 day post-treatment and that the change in BDI scores at 1 week was significantly associated with post-treatment increases in amygdala responses for fearful > neutral face conditions. Similarly, the QIDS response at 1 day, 1 week, and 3 weeks post-treatment was also associated with increases in amygdala responses for fearful > neutral face conditions. Lastly, Shukuroglou et al. 22 explored the relationship between music-evoked emotion with psilocybin therapy. Psilocybin was administered as a low dose (10 mg) on the first session and at a higher dose (25 mg) at the second session, 1 week apart. Psilocybin therapy was associated with reduced anhedonia scores as measured by the Snaith-Hamilton Pleasure Scale (SHAPS) one day after the second dosing session. In both music and no music condition scans, psilocybin therapy was associated with reduced nucleus accumbens (NAc)-default mode network (DMN) coupling 1 day after the second dosing session.
A separate open-label clinical trial 18 (NCT03181529) examined the effect of psilocybin therapy on cognitive flexibility through performance on the Penn Conditional Exclusion Test (PCET). Twenty-four patients with MDD completed this waitlist-controlled trial that included two dosages of psilocybin (20 mg/70 kg as the first dose and 30 mg/70 kg as the second dose) and 8 h of preparatory therapy sessions over 2 weeks. Psilocybin therapy was associated with a significant reduction in depression scores as measured by the GRID-Hamilton Depression Rating Scale (HAMD) and improvements in cognitive flexibility at 1 week and 4 weeks post-treatment, although the change in depression scores was not associated with improved cognitive flexibility. Magnetic resonance imaging (MRS) was used to quantify in vivo the neurometabolites of glutamate and N-acetylaspartate. Reductions in both the anterior cingulate cortex (ACC) were associated with psilocybin therapy after 1 week. Increases in the dynamics of functional connectivity between the ACC and posterior cingulate cortex (PCC) were also associated with psilocybin therapy after 1 week.
Daws et al. 17 pooled data from two studies: an open-label clinical trial in 16 adults with MDD, and a double-blinded randomized controlled trial in 59 adults with MDD with escitalopram as the comparator arm. In the open-label MDD trial, psilocybin therapy was associated with reduced brain network modularity at 1 day after the second dosing session, which suggests greater increases in functional connectivity between brain networks, as there is a reduction in within-module connections. Psilocybin therapy was also associated with a reduction in the DMN recruitment and increased between-network integration between the DMN and the executive network and the salience network at 1 day after the second dosing session. There was also a significant association between decreased brain network modularity and sustained improvements in depression severity at 3 weeks post-treatment (as measured by the BDI, a secondary outcome in the parent clinical trial). In the randomized controlled trial comparing psilocybin therapy with escitalopram, there was an association between a reduction in depressive symptoms at 3 weeks post-treatment (as measured by the BDI, a secondary outcome in the parent trial) with reduced brain network modularity in the psilocybin arm but not the escitalopram arm. They were not able to replicate the findings of the open-label trial that DMN recruitment had associations with treatment response to psilocybin therapy. However, an exploratory analysis did report that psilocybin therapy increased dynamic flexibility with an association between changes in network flexibility and BDI scores at 1 day after the second dosing session, which was not seen with escitalopram.
Finally, Skosnik et al. 23 explored the impact of psilocybin therapy on EEG correlates in a double-blind, placebo-controlled, within-subject, fixed-order design (NCT03554174). The analysis included 19 patients with MDD who were given a placebo at the first dosing session and psilocybin 0.3 mg/kg to a maximum dose of 35 mg at the second dosing session. Results reported a significant effect of time and of drug on depression symptoms, as measured by the GRID-HAMD, but did not find a drug × time interaction. Psilocybin therapy was associated with increased auditory evoked EEG theta power 2 weeks after dosing. The researchers purported that theta oscillations are involved in synchronizing connectivity that can be related to neuroplasticity. Notably, increased EEG theta power was associated with decreases in depression scores.
Discussion
This systematic review synthesized current evidence of biomarkers of treatment response to PT in adults with mental disorders. Included studies recruited participants with MDD who were treated with either psilocybin or ayahuasca. Potential peripheral biomarkers included BDNF and plasma CRP that correlated with treatment response. Neuroimaging studies reported potential involvement and activity of the amygdala, vmPFC, ACC, and PCC, which correlated with treatment response following psychedelic administration. The observed physiological changes following psychedelic therapy are similar to studies on biomarkers of treatment response in antidepressant studies.24 –26 While these biomarkers are not specific to PT and have also been implicated in response to traditional antidepressants, these consistent findings may suggest a shared neurobiological pathway underlying the pathology of MDD and underlying pathways for treatment response. Almeida et al. 15 reported increased BDNF (a marker of neuroplasticity) post-ayahuasca treatment that correlated with a reduction in depressive symptoms, which speaks to compromised synaptic and structural plasticity associated with depression. 27 The increased vmPFC resting-state functional connectivity reported by Carhart-Harris et al., 28 the decreased DMN recruitment reported by Daws et al., 17 and increased functional connectivity between the ACC and PCC suggest that the antidepressant effects of psilocybin therapy are occurring on a network integration level. These findings require replication in independent samples.
Limitations
The results of this review must be interpreted with caution due to the limitations of the current literature. Given the heterogeneity in the study types and outcomes reported in the included studies, a meta-analysis could not be performed. Relatedly, the limited data from small studies of only two psychedelics (Ayahuasca and psilocybin) in depression does not allow for the evaluation of measures such as time-group interactions or the establishment of treatment-specific effects. The small sample sizes limit the statistical power to reliably detect generalizable biomarker associations. As previously discussed, 29 functional unblinding and expectation bias are considerable limitations of the therapeutic effects reported in published trials. Future studies with an adequate sample size and active controls (e.g., comparing one psychedelic with another) may improve the validity and replicability of the findings. Furthermore, integrating multiple biomarkers into a composite measure, for example, correlating peripheral biomarkers such as BDNF with central measures such as functional connectivity alongside clinical symptom changes, may lead to more reliable predictors of treatment response to complex interventions like psychedelic therapy.
Conclusion
The current systematic review identified potential peripheral (e.g., BDNF, CRP) and central (activity changes in amygdala, vmPFC, DMN) biomarkers of treatment response to psychedelic therapy in individuals with depression. However, the current evidence is limited by small sample sizes, lack of replication, and a narrow range of biomarker modalities evaluated in one clinical population. These limitations underscore the need for further biomarker integration in clinical trials of PT. This may advance an understanding of the mechanisms of therapeutic action of PT and the biology of mental disorders.
Supplemental Material
sj-docx-1-tpp-10.1177_20451253251384513 – Supplemental material for Biological markers of treatment response to serotonergic psychedelic therapies: a systematic review
Supplemental material, sj-docx-1-tpp-10.1177_20451253251384513 for Biological markers of treatment response to serotonergic psychedelic therapies: a systematic review by Stanley Wong, Brett D. M. Jones, Mathura T. Thiyagarajah, Sami G. Sabbah, Chase Thompson, Marco Solmi, Madeha Umer, Christoph Zrenner, Daphne Voineskos, Joshua D. Rosenblat, Benoit H. Mulsant, Daniel M. Blumberger and Muhammad Ishrat Husain in Therapeutic Advances in Psychopharmacology
Supplemental Material
sj-docx-2-tpp-10.1177_20451253251384513 – Supplemental material for Biological markers of treatment response to serotonergic psychedelic therapies: a systematic review
Supplemental material, sj-docx-2-tpp-10.1177_20451253251384513 for Biological markers of treatment response to serotonergic psychedelic therapies: a systematic review by Stanley Wong, Brett D. M. Jones, Mathura T. Thiyagarajah, Sami G. Sabbah, Chase Thompson, Marco Solmi, Madeha Umer, Christoph Zrenner, Daphne Voineskos, Joshua D. Rosenblat, Benoit H. Mulsant, Daniel M. Blumberger and Muhammad Ishrat Husain in Therapeutic Advances in Psychopharmacology
Footnotes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
