Increased Hippocampal Blood Flow in People at Clinical High Risk for Psychosis and Effects of Cannabidiol

BACKGROUND
Hippocampal hyperperfusion has been observed in people at Clinical High Risk for Psychosis (CHR), is associated with adverse longitudinal outcomes and represents a potential treatment target for novel pharmacotherapies. Whether cannabidiol (CBD) has ameliorative effects on hippocampal blood flow (rCBF) in CHR patients remains unknown.


METHODS
Using a double-blind, parallel-group design, 33 CHR patients were randomized to a single oral 600 mg dose of CBD or placebo; 19 healthy controls did not receive any drug. Hippocampal rCBF was measured using Arterial Spin Labeling. We examined differences relating to CHR status (controls v. placebo), effects of CBD in CHR (placebo v. CBD) and linear between-group relationships, such that placebo > CBD > controls or controls > CBD > placebo, using a combination of hypothesis-driven and exploratory wholebrain analyses.


RESULTS
Placebo-treated patients had significantly higher hippocampal rCBF bilaterally (all pFWE<0.01) compared to healthy controls. There were no suprathreshold effects in the CBD v. placebo contrast. However, we found a significant linear relationship in the right hippocampus (pFWE = 0.035) such that rCBF was highest in the placebo group, lowest in controls and intermediate in the CBD group. Exploratory wholebrain results replicated previous findings of hyperperfusion in the hippocampus, striatum and midbrain in CHR patients, and provided novel evidence of increased rCBF in inferior-temporal and lateral-occipital regions in patients under CBD compared to placebo.


CONCLUSIONS
These findings suggest that hippocampal blood flow is elevated in the CHR state and may be partially normalized by a single dose of CBD. CBD therefore merits further investigation as a potential novel treatment for this population.


FIGURE S1. Schematic of proposed neural circuit mechanisms of hippocampal
dysfunction in the pathophysiology underlying psychosis onset.In (1), low glutamate signal/input from hypofunctioning NMDARs (akin to faulty homeostatic sensors) prompts GABAergic interneurons to homeostatically increase excitation by reducing inhibition (disinhibition) of glutamatergic pyramidal cells.However, by disinhibiting pyramidal cells (and thus increasing glutamate signalling) in this dysfunctional neural environment, the potential homeostatic adaptation becomes allostatic, with enhanced excitatory drive inducing (2) hypermetabolism and hyperperfusion (elevated blood flow to meet increased metabolic demand), and (3) an overdrive in the responsivity of midbrain dopamine neurons, which project to the associative striatum.Note that the connection between hippocampal pyramidal cells and midbrain dopamine neurons is presented as monosynaptic but is in fact polysynaptic via the ventral striatum and ventral pallidum.Completing the (simplified) circuit, local glutamatergic tone is increased in (4) but is not detected as such by hypofunctioning NMDARs on GABAergic interneurons.Figure reproduced and adapted with permission (CCBY 4.0) from (Davies et al., 2019).For original diagrams and discussion of evidence for this proposed circuit, see (Lisman et al., 2008;Krystal and Anticevic, 2015;Modinos et al., 2015;Krystal et al., 2017;Lieberman et al., 2018).Abbreviations: Glu, glutamate; NMDAR, N-methyl-Daspartate receptor; CA1, Cornu Ammonis 1.
Resting Cerebral Blood Flow (CBF) was measured using 3D pseudo-Continuous Arterial Spin Labelling (CASL) scans acquired with a 3D Fast Spin Echo (FSE) spiral multi-shot readout, following a post-labelling delay of 1.5s.The spiral acquisition used a short (10ms) TE, and 8 spiral arms (interleaves) with 512 points in each arm.FSE TE= 32.26ms, TR = 5500ms.64 slices of 3mm thickness were obtained and the in-plane FoV was 240×240mm.Three pairs of tagged-untagged images were collected.The whole ASL pulse sequence, including the acquisition of calibration images, was performed in 6:08min.

Image Processing
Data were preprocessed using FMRIB Software Library (FSL) 6.0.2 using the following procedure: (1) T1 and T2 images were skull-stripped and corresponding brain-only binary masks created; (2) original CBF images were coregistered to the T2 images and (3) multiplied by the binary T2 mask to create a skull-stripped CBF image in T2 space; (4) skull-stripped T2 was coregistered to skull-stripped T1; (5) skull-stripped T1 was first linearly coregistered to the MNI152 T1 2mm brain template, before non-linear registration (FNIRT) of the original T1 to MNI space; (6) original T2 images were registered to the MNI template (via T1 space) in a single concatenated step, using the T2-to-T1 transformation matrix (from step 4) and T1-to-MNI warp (from step 5); (7) skull-stripped CBF images (already in T2 space) were registered to the MNI template using the concatenated procedure in step 6; (8) normalised CBF images were spatially smoothed with a 6mm Gaussian kernel.The final voxel size was 2 x 2 x 2 mm.All images were visually inspected for preprocessing errors.

Statistical Thresholds in SPM
Statistical thresholds for exploratory wholebrain analyses (cluster-forming threshold: p<.005; cluster reported as significant at p<.05 using FWE cluster correction in SPM) were determined a priori based on previous work at our Institute investigating the effects of potential novel pharmacotherapies on rCBF in humans (Paloyelis et al., 2016;Martins et al., 2020bMartins et al., , 2022)), including in our previous work in CHR patients (Davies et al., 2019), and are standardly applied in ASL studies measuring rCBF (Joe et al., 2006;Takeuchi et al., 2011;Loggia et al., 2013;Mutsaerts et al., 2019;Martins et al., 2020a;Nwokolo et al., 2020).
Further details required for adherence to CONSORT (including recruitment periods, power calculations, randomisation and further blinding details, etc) can be found in the Supplementary Material of our previous publication in the same sample, where the study protocol is also appended (Bhattacharyya et al., 2018a).

Supplementary Wholebrain Analyses: SPM vs FSL Randomise
To test the robustness of the wholebrain findings, we re-ran our pairwise wholebrain analyses using two independent t-tests (controls vs placebo; placebo vs CBD) using randomise in FSL/6.0.1.De-meaned age, sex, years of education, smoking status and mean grey matter CBF per subject were included as covariates in the design matrix.The analysis used 5000 permutations and was restricted using a grey matter mask (thresholded at >.50).Clusterbased thresholding (threshold=2.3due to modest sample size) was used, corrected for multiple comparisons by using the null distribution of the max (across the image) cluster size.

Healthy Control vs CHR Placebo
We found significantly higher CBF in a single (large, k=3660) cluster in the CHR placebo group vs healthy controls (cluster pFWE=.025;see randomise output table below).In terms of anatomical location, the significant cluster found here using

Placebo vs CBD
For the CHR placebo vs CBD contrast, we did not observe any significant clusters using FSL's randomise.However, the cluster found using SPM (where CBD > placebo) was present at a relaxed statistical threshold (cluster pFWE=.15;see randomise output table below).In terms of anatomical location, the (non-significant) cluster found using FSL's randomise (shown in red in the figure below) was almost identical to the significant cluster in our original analyses using SPM (shown in yellow in the figure below, superimposed on the FSL cluster [red]).In terms of spatial extent, the FSL-derived cluster included slightly more voxels than the SPM results, but note that this cluster was not significant in the FSL analysis.The reasons for the differential findings for this contrast in SPM vs FSL are unclear, but it is possible that the magnitude of the difference (SPM results: T(21)=4.51,pFWE=.014)was not sufficiently large-combined with the modest sample size-to be significant with FSL's non-parametric statistical tests.

Exploratory Correlations between CAARMS and rCBF
As we collected CAARMS data prior to drug administration and scanning, only baseline values from the CHR-placebo group could have been used in analyses with the imaging data, and given the modest sample size, we did not plan any a priori hypotheses along these lines.
However, to explore whether hippocampal rCBF may be related to CAARMS scores, we correlated CAARMS positive symptoms (sum of the product of severity x frequency for each of the 4 positive symptoms) and total symptoms with rCBF values in the hippocampus (from the significant cluster from the linear trend/relationship analyses) in the placebo group.We found no significant correlation between rCBF and either CAARMS positive symptoms (r=.065, p=.83, n=14) or total symptoms (r=-.073,p=.80, n=14).This is consistent with two previous studies conducted at our institute, which found no significant association between attenuated positive symptom scores and elevated hippocampal rCBF in CHR patients (Allen et al., 2016(Allen et al., , 2018)).
hippocampal endocannabinoid tone, it is possible that CBD modulates CBF through these direct receptor/circuit mechanisms.Further proposed mechanisms include inhibition of anandamide hydrolysis (Bisogno et al., 2001) and actions on 5-HT1A (Russo et al., 2005), vanilloid type 1 (Bisogno et al., 2001) and GPR55 receptors (Ryberg et al., 2007;Pertwee, 2008).Recent work has also implicated effects on the glutamate system (Gomes et al., 2015;Linge et al., 2016), which is of particular relevance to psychosis pathophysiology (Lodge and Grace, 2011;Howes et al., 2015;Bossong et al., 2019).On the human neuroimaging level, CBD modulates hippocampal glutamate in patients with early psychosis while concomitantly reducing psychotic symptoms (O'Neill et al., 2021), and may partially ameliorate glutamatergic dysfunction in those at CHR (Davies et al., 2023).CBD also alters glutamate and GABA in ASD and neurotypical individuals (Pretzsch et al., 2019).The mechanisms underlying the effects of CBD on these neuroimaging parameters as well as on symptoms therefore remains an important avenue for future research.
FSL's randomise (shown in red in the figure below) was almost identical to the significant clusters in our original analyses using SPM (shown in yellow in the figure below, superimposed on the FSL results [red]), with the addition of some further left cerebellar coverage with the FSL results.In terms of spatial extent, the FSL-derived clusters included slightly more voxels than the SPM results.