In Vivo Availability of Cannabinoid 1 Receptor Levels in Patients With First-Episode Psychosis

Importance Experimental and epidemiological studies implicate the cannabinoid 1 receptor (CB1R) in the pathophysiology of psychosis. However, whether CB1R levels are altered in the early stages of psychosis and whether they are linked to cognitive function or symptom severity remain unknown. Objective To investigate CB1R availability in first-episode psychosis (FEP) without the confounds of illness chronicity or the use of illicit substances or antipsychotics. Design, Setting, and Participants This cross-sectional, case-control study of 2 independent samples included participants receiving psychiatric early intervention services at 2 independent centers in Turku, Finland (study 1) and London, United Kingdom (study 2). Study 1 consisted of 18 volunteers, including 7 patients with affective or nonaffective psychoses taking antipsychotic medication and 11 matched controls; study 2, 40 volunteers, including 20 antipsychotic-naive or antipsychotic-free patients with schizophrenia or schizoaffective disorder and 20 matched controls. Data were collected from January 5, 2015, through September 26, 2018, and analyzed from June 20, 2016, through February 12, 2019. Main Outcomes and Measures The availability of CB1R was indexed using the distribution volume (VT, in milliliters per cubic centimeter) of 2 CB1R-selective positron emission tomography radiotracers: fluoride 18–labeled FMPEP-d2 (study 1) and carbon 11–labeled MePPEP (study 2). Cognitive function was measured using the Wechsler Digit Symbol Coding Test. Symptom severity was measured using the Brief Psychiatric Rating Scale for study 1 and the Positive and Negative Syndrome Scale for study 2. Results A total of 58 male individuals were included in the analyses (mean [SD] age of controls, 27.16 [5.93] years; mean [SD] age of patients, 26.96 [4.55] years). In study 1, 7 male patients with FEP (mean [SD] age, 26.80 [5.40] years) were compared with 11 matched controls (mean [SD] age, 27.18 [5.86] years); in study 2, 20 male patients with FEP (mean [SD] age, 27.00 [5.06] years) were compared with 20 matched controls (mean [SD] age, 27.15 [6.12] years). In study 1, a significant main effect of group on [18F]FMPEP-d2 VT was found in the anterior cingulate cortex (ACC) (t16 = −4.48; P < .001; Hedges g = 1.2), hippocampus (t16 = −2.98; P = .006; Hedges g = 1.4), striatum (t16 = −4.08; P = .001; Hedges g = 1.9), and thalamus (t16 = −4.67; P < .001; Hedges g = 1.4). In study 2, a significant main effect of group on [11C]MePPEP VT was found in the ACC (Hedges g = 0.8), hippocampus (Hedges g = 0.5), striatum (Hedges g = 0.4), and thalamus (Hedges g = 0.7). In patients, [11C]MePPEP VT in the ACC was positively associated with cognitive functioning (R = 0.60; P = .01), and [11C]MePPEP VT in the hippocampus was inversely associated with Positive and Negative Syndrome Scale total symptom severity (R = −0.50; P = .02). Conclusions and Relevance The availability of CB1R was lower in antipsychotic-treated and untreated cohorts relative to matched controls. Exploratory analyses indicated that greater reductions in CB1R levels were associated with greater symptom severity and poorer cognitive functioning in male patients. These findings suggest that CB1R may be a potential target for the treatment of psychotic disorders.


43
Exclusion criteria for all volunteers were as follows: 1) ages <18 or >60; 2) a history of a head injury leading to loss of consciousness, 3) personal or family 44 history of neurological, neurodevelopmental, endocrine or cardiovascular health problems, 4) contraindications to MRI safety, 5) current or lifetime history of 45 substance use or dependence as determined by the Structured Clinical Interview for DSM-IV-TR (SCID-I/P), 6) current or recent (within last month) 46 recreational use of illicit substances, 7) screened positive on a THC urine toxicology test that can detect THC metabolite THCCOOH for up to 30 days 47 (InstAlert; 50 ng/ml cut off), or 8) screened positive on a multi-panel urine drug screen (InstAlert) detecting the following substances: amphetamine (500ng/ml 48 cut-off), buprenorfine (5ng/ml cut-off), cocaine (300ng/ml cut-off), methadone (300ng/ml cut-off), opiates (300ng/ml cut-off).

50
Exclusion criteria for all volunteers were as follows: 1) ages <18 or >60; 2) history of a head injury leading to loss of consciousness; 3) personal or family 51 history of neurological, neurodevelopmental, endocrine or cardiovascular health problems; 4) contraindications to MRI including the presence of metal plates, 52 pins, bridges or dentures and pregnancy; 5) current or lifetime history of substance use or dependence as determined by the Structured Clinical Interview for 53 DSM-IV-TR (SCID-I/P); 6) current or recent (within last month) recreational use of illicit substances; 7) screened positive on a THC urine toxicology test that 54 can detect THC metabolite THCCOOH for up to 30 days (SureScreen, Diagnostics; 50 ng/ml cut off); 8) screened positive on a multi-panel drug screen 55 detecting the following substances: amphetamine (300 ng/ml cut off), cocaine (150 ng/ml cut off), ketamine (1000 ng/ml cut off), marijuana (50 ng/ml cut off), 56 methamphetamine (300 ng/ml cut off), opiates (2000 ng/ml cut off) (SureScreen Diagnostics).

60
Alcohol use was recorded using the Audit questionnaire 1 , tobacco use was recorded using a questionnaire adapted from the World Health Organization 2 and 61 cannabis use was measured using an illicit substance use questionnaire described previously 3 .

62
Study 2 63 Current and previous use of alcohol, nicotine and illicit substances were recorded. If volunteers reported current use, the quantity (cigarettes/alcohol units 64 etc.), duration (years) and frequency (number of uses per week) of use was recorded. Prior cannabis use was recorded using the Cannabis Experiences after a bolus injection of 314 ± 34.4 MBq of [ 11 C]MePPEP, synthesized using methods reported elsewhere 6,7 . CT scans were acquired prior to each PET scan 80 for correction for attenuation and scatter. Continuous arterial blood sampling took place for the first 15 minutes of the scan which was followed by discrete 81 blood sampling at 2, 5, 10, 15, 20, 25, 35, 40, 50, 60, 70, 80 and 90 minutes after the radioligand injection. Images were reconstructed with filtered back 82 projection including corrections for attenuation and scatter.

98
Mathworks Inc., Sherborn, Massachusetts). Motion correction was applied for all PET scans as follows. Attenuation and scatter corrected images were realigned 99 to a single reference frame (the 12 th frame, which contained the highest uptake average). Realigned frames were then summated to create single-subject 100 motion-corrected maps which were then used for MRI and PET co-registration, prior to PET data quantification. T1-weighted structural images were co-101 registered to the PET image using rigid body transformations. Normalization parameters were obtained by warping the co-registered structural MRI to MNI 102 space (International Consortium for Brain Mapping ICBM/MNI) using probabilistic tissue classification with bias correction. The inverse of these parameters was used to fit a neuroanatomical atlas to each individual PET scan using the Hammersmith atlas (Hammers et. al 2003).

104
Decay corrected whole blood tissue activity curves (TAC) derived from automatic blood pump sampling were converted to plasma activity using hematocrit and 105 a population derived tracer specific distribution function. Automated and manual sample TACs were then combined and PET count rate curves were used to 106 reference peak tissue activity to correct for temporal delay between blood sample measurement and target the tissue data. Plasma and whole blood TAC values 107 were extrapolated from 100 to 120 minutes using a biexponential function fit starting at two times the peak activity location. The time delay of the peak metabolites, was used as parent input for modelling.

114
Data pre-processing was performed using a combination of Statistical Parametric Mapping 12 (http://www.fil.ion.ucl.ac.uk/spm) and FSL

116
Non-attenuated corrected frames were realigned to a single "reference" frame (corresponding to that with the highest number of counts) by employing a 117 mutual information algorithm. The transformation parameters were then applied to the corresponding attenuated-corrected dynamic images, creating a 118 movement-corrected dynamic image which was used for the analysis. Realigned frames were then summated to create single-subject motion-corrected maps 119 which were then used for MRI and PET co-registration, prior to PET data quantification. T1-weighted structural images were co-registered to the PET image 120 using rigid body transformation. Normalization parameters were obtained by warping the co-registered structural MRI to MNI space (International Consortium While the use of Logan has been previously validated for FMPEP, returning test-retest data comparable to two-tissue compartmental modelling 9 we validated 130 the use of Logan by comparing VT estimates with coefficient variation <10% derived from Logan and 2TCM modelling. In this context, we demonstrated that 131 the mean relative difference between Logan and 2TCM was low (whole brain mrd: -4%+/-11%) and that VT estimates derived using these models were highly

& 2 138
For studies 1-2, cumulative scanner movement was defined as the sum of total frame-to-frame movement during imaging acquisition. For studies 1-2, motion 139 spikes were defined as frame-to-frame scanner movement exceeding 5mm since the resolution of PET images were 5mm. A voxel-based morphometry (VBM) analysis was conducted using SPM12 in order to determine if there were volumetric differences between patients and 144 controls. Structural T1-weighted structural scans were segmented, warped to a template created using the DARTEL algorithm which improves the accuracy of the anterior cingulate, thalamus, hippocampus and striatum. The height threshold was set to p=0.001 and peak-level family-wise error corrected thresholds 149 (p<0.05) were used. To determine if CB1R availability was lower in patients, a repeated measures ANOVA using a 2 (group: control vs. patient) x 13 (region: amygdala, caudate, 154 putamen, insula, cerebellum, posterior cingulate, temporal lobe, parietal lobe, occipital lobe, frontal lobe) design was used for each dataset. These regions of 155 interest were also defined using the Hammersmith atlas 10 . The main effect of group tested whether the VT of the respective tracer was different between 156 patients and controls, and the group x region interaction tested whether mean VT across ROIs were different between groups, where a null result indicates a To identify if there were volumetric differences between patients and controls, tissue volumes were compared between patients and controls in whole-brain 202 and region of interest analyses of the anterior cingulate, thalamus, hippocampus and striatum.

205
There were no significant group differences in tissue volumes between patients and controls in whole-brain and region of interest analyses of the anterior 206 cingulate, thalamus, hippocampus and striatum. No statistics are reported because there were no suprathreshold clusters.

213
To determine if differences in gray matter volumes in ROIs influenced our results, the analyses were repeated using a gray matter mask applied to each ROI 214 to restrict the image analysis to gray matter. The results for each study are as follows:

229
To determine if tobacco or lifetime cannabis use influenced our results, the analyses were repeated including data on these variables as covariates.

260
In a whole-brain voxel-wise analysis, patients relative to controls showed significantly lower VT of [ 18 F]FMPEP-d2 in a cluster encompassing the anterior and 261 middle cingulate, superior and middle frontal gyrus, inferior orbital gyrus, middle and inferior temporal gyrus, inferior opercular gyrus, inferior frontal pars 262 triangularis (see supplementary figure 1). There were no differences in the control<patient contrast.