577. **Rafaelsen Award Presentation** Association between cortical serotonin release and cognitive function in schizophrenia: an in vivo [11C]Cimbi-36 PET study with a d-Amphetamine challenge

Abstract

Abstract Background Cognitive symptoms of schizophrenia are highly disabling. There are no licensed treatments and the underlying neurobiology is unclear. Indirect evidence implicates abnormalities in cortical serotonergic neurotransmission, but no prior study has tested if cortical serotonin release is associated with cognition in vivo in schizophrenia. Aims & Objectives We hypothesised a priori that neocortical serotonin release capacity would be positively correlated with general intellectual functioning in schizophrenia. Method This case-control neuroimaging study included stable outpatients with DSM-5 schizophrenia (antipsychotic-free, or taking antipsychotics with negligible 5-HT2a affinity), and healthy controls matched for age/sex/BMI. No subjects had used serotonergic medications in the past month. Subjects underwent a battery of validated cognitive tests; the a priori primary outcome was the Wechsler Adult Intelligence Scale (WAIS)-IQ. [11C]Cimbi-36 is a 5-HT2a receptor agonist radioligand sensitive to increases in extracellular serotonin induced by acute d-amphetamine administration. Subjects had T1-weighted MRI scans, and dynamic 90-minute [11C]Cimbi-36 PET scans at baseline and 3 hours after oral d-amphetamine (0.5mg/kg). Serotonin release capacity was prespecified as the percentage change in nondisplaceable binding potential (BPND) between scans, delta BPND (%). The entire neocortex was the primary region of interest. Neocortical BPND was calculated with the cerebellum as the reference region, from volumes of distribution derived from 2TCM modelling with metabolite-corrected arterial plasma input function. Between-group comparisons used two-tailed t-tests (cognition), two-sided ANCOVA (Delta BPND; adjustment for injected mass and MRI/PET scanner) and Fisher’s r-to-z (correlation coefficients). Results 61 subjects (33 schizophrenia [SZ], 28 healthy controls [HC]) completed cognitive testing and 54 (26 SZ, 28 HC) completed neuroimaging. Schizophrenia subjects (mean age 33 ± SD 9, 61% male, PANSS 58 ± 16) were well matched to controls (age 32 ± 9, 60% male). 85% with schizophrenia were antipsychotic-free, 15% were treated with haloperidol or amisulpride (negligible 5-HT2a affinity). Neocortical delta BPND was significantly greater in schizophrenia than in controls (n=54, mean difference=11.3%, p=0.040, d=0.62). Overall cognitive function was impaired in schizophrenia compared to controls (WAIS-IQ: n=61, HC 107 ± 16 vs SZ 92 ± 29, p=0.015, d=0.64). Greater neocortical serotonin release was positively correlated with WAIS-IQ in schizophrenia (n=26, r=0.49, p=0.011). This relationship was significantly different in schizophrenia in comparison to controls (z=2.2, p=0.031). Subgroup analysis of 21 antipsychotic-free schizophrenia subjects replicated increased neocortical delta BPND (p=0.036, d=0.68), and its correlation with WAIS-IQ (r=0.49, p=0.025). Exploratory analyses showed deficits in processing speed (d=-0.61), verbal learning (d=1.12) and verbal memory (d=0.68) in schizophrenia. Only processing speed was correlated with neocortical serotonin release (rho=-0.41, p=0.037). Discussion & Conclusions This is the first direct assessment of the relationship between serotonin release and cognition in schizophrenia. We show that neocortical serotonin release is directly associated with general intellectual function in schizophrenia. We also found greater neocortical serotonin release in schizophrenia compared to matched controls, and demonstrated a differential relationship between neocortical serotonin release and general intellectual functioning in schizophrenia and controls. These findings were replicated in antipsychotic-free patients. The results of this study indicate serotonergic dysfunction linked to cognitive function in schizophrenia, and identify the regulation of serotonin release and 5-HT2a receptors as potential targets to treat cognitive deficits.