Human Cortical Serotonin 2A Receptor Occupancy by Psilocybin Measured Using [11C]MDL 100,907 Dynamic PET and a Resting-State fMRI-Based Brain Parcellation

Barrett and colleagues frame psilocybin as a 5-HT 2A receptor agonist with emerging therapeutic promise for mood and substance use disorders, while noting large between-subject variability in both acute subjective effects (for example, mystical or peak experiences, visual imagery, fear or anxiety) and longer-term clinical outcomes. Previous human neuroimaging and pharmacological antagonist studies implicate 5-HT 2A receptor signalling in those acute and downstream neural effects, and recent PET work has linked plasma psilocin, neocortical 5-HT 2A occupancy, and subjective measures. However, prior molecular imaging reports have tended to characterise whole-neocortex binding rather than regional distributions of occupancy, leaving unresolved whether and where regional 5-HT 2A receptors are engaged by psychedelics in humans. This open-label pilot study set out to measure regional cortical 5-HT 2A receptor occupancy by psilocybin in healthy volunteers. Using [11C]MDL 100,907 PET scanning, the investigators tested the hypothesis that binding of this selective 5-HT 2A inverse agonist would decrease across a broad set of cortical regions after oral psilocybin, with region definitions derived from subject-specific resting-state functional connectivity parcellations so that occupancy could be interrogated in functionally homogeneous, study-relevant ROIs.

Participants were recruited from volunteers who had completed a prior blinded pharmacology study of oral psilocybin and dextromethorphan; all had at least 25 lifetime exposures to classic psychedelics. Six individuals were initially enrolled, but two (one female, one male) withdrew prior to PET or drug procedures, leaving four completers (2 male, 2 female; mean age 28 years, range 27–30; all reported Caucasian ethnicity). At least one month elapsed between the parent study and the present procedures. Screening included medical and psychiatric evaluation, ECG, routine blood tests and the SCID-IV; exclusion criteria included current Axis I disorder, substance dependence (excluding nicotine and caffeine), pregnancy or nursing, a personal or immediate family history of psychotic or bipolar-spectrum disorders, current serotonergic or antipsychotic medications, intolerance of prior psilocybin in the parent trial, and MRI/PET contraindications. Written informed consent was obtained under institutional oversight. Procedures combined pre-drug MRI and two PET scans per participant: a baseline PET with [11C]MDL 100,907 and a blocking PET after oral psilocybin. Resting-state fMRI and structural MRI were acquired before PET or drug administration to create subject-specific functional parcellations. For the blocking condition, participants received 10 mg/70 kg oral psilocybin administered 80 minutes prior to radiotracer injection to approximate peak plasma levels; during PET scans participants listened to a standard music playlist, were instructed to attend inward, and remained in continuous contact with study monitors. Pre-scan urine drug and pregnancy testing and brief physical exams were performed; monitors from the parent study were present throughout. MRI acquisition used a 3T Siemens Skyra with a 32-channel coil. A high-resolution MPRAGE (1 mm slices) and resting-state BOLD EPIs (180 volumes; 3 × 3 mm in-plane; TR = 2.46 s; total ~7 min 22 s) were preprocessed (slice-timing, motion correction, co-registration, normalization to MNI space, smoothing 6 mm FWHM, scrubbing, detrending, nuisance regression including 6 motion parameters and five principal components from ventricles/white matter, and band-pass filtering 0.008–0.09 Hz). Preprocessed resting-state data were clustered into 200 parcels using spectral clustering of the between-voxel gray-matter correlation matrix; these parcels served as ROIs for PET analysis. PET methods: [11C]MDL 100,907 was synthesised and injected (mean activity ~20.46 mCi ± 0.55 mCi) with dynamic acquisition for 90 minutes using a 30-frame protocol. Images were reconstructed with attenuation and decay correction; dynamic PET images were registered to baseline PET and normalised to MNI space via the structural MRI. ROI time–activity curves were obtained by applying the resting-state-derived parcels to the dynamic PET data. Binding potential (BP_ND), an index of specific tracer binding, was estimated by fitting a 3-parameter simplified reference tissue model (SRTM: R1, k2REF, BP_ND) to all ROI TACs with k2REF coupling, using cerebellum as the reference region. ROIs with low BP (< 0.2), low distribution volume ratio (DVR < 1.2), or impossible occupancy values (< 0% or > 100%) were discarded; after this quality control 137 ROIs remained and BP at baseline and percent occupancy by psilocybin were calculated for each retained ROI. The extracted text does not report further details of statistical testing or modelling of inter-individual effects.

The extracted text does not include a dedicated Results section, but outcome data are reported within the Discussion and Methods. After exclusion of ROIs that failed quality criteria, 137 functionally defined cortical ROIs were retained for analysis. The investigators report substantial 5-HT 2A receptor occupancy by psilocybin across the cortex: average occupancy values across retained ROIs ranged from 19.84% to 74.72% (per-ROI averages), a magnitude stated to exceed the 7–11% test–retest variability previously reported for [11C]MDL 100,907. Three ROIs with the greatest occupancy were located within the default mode network (the subgenual anterior cingulate cortex and bilateral angular gyri). The authors also emphasise marked inter-individual variability in regional occupancy; for example, occupancy in the left angular gyrus is given as ranging between 48.5% and 90.5% across individuals (the extracted text contains both the 19.84–74.72% ROI-average range and this wider individual-range example, which appear inconsistent). Subcortical regions with low tracer binding (including striatum and basal ganglia) were discarded from the PET analysis because they failed the binding or DVR criteria. The extracted materials do not present formal group-level statistical tests, confidence intervals, p-values, or voxel-wise maps within the provided text, nor do they report plasma psilocin levels or correlations between occupancy and subjective measures in this extract. The sample available for analysis comprised four completers.

Barrett and colleagues interpret their findings as evidence that oral psilocybin produces widespread and substantial cortical 5-HT 2A receptor occupancy in humans, measured with [11C]MDL 100,907 and functionally derived cortical ROIs. They position these results as complementary to prior PET work and to fMRI studies showing psilocybin-related alterations in primary sensory cortices and default mode network regions, noting that the regions with high occupancy are those that both densely express 5-HT 2A receptors and have been implicated in mood disorders and addiction. The use of a different 5-HT 2A ligand from previous studies is highlighted as an additional contribution. The authors acknowledge important limitations. Foremost is the small sample size (four completers), which limits generalisability and statistical inference. They note alternative explanations for the observed inter-individual variability in occupancy that were not tested here, including genetic polymorphisms affecting 5-HT 2A receptor expression/function and inter-individual differences in psilocybin/psilocin bioavailability. The investigators recommend future studies that include controlled measurement of plasma psilocin and broader brain-derived measures (for example, functional connectivity, task-based responses) to clarify relationships between regional receptor occupancy, acute subjective effects and enduring therapeutic responses. Overall, the study team concludes that the data demonstrate cortical 5-HT 2A engagement by psilocybin and motivate further work to link regional occupancy to individual differences in psychedelic effects and clinical outcomes.