We have developed [(11)C]mirtazapine as a ligand for PET studies of antidepressant binding in living brain. However, previous studies have determined neither optimal methods for quantification of [(11)C]mirtazapine binding nor the pharmacological identity of this binding. To obtain that information, we have now mapped the distribution volume (V(d)) of [(11)C]mirtazapine relative to the arterial input in the brain of three pigs, in a baseline condition and after pretreatment with excess cold mirtazapine (3 mg/kg). Baseline V(d) ranged from 6 ml/ml in cerebellum to 18 ml/ml in frontal cortex, with some evidence for a small self-displaceable binding component in the cerebellum. Regional binding potentials (pBs) obtained by a constrained two-compartment model, using the V(d) observation in cerebellum, were consistently higher than pBs obtained by other arterial input or reference tissue methods. We found that adequate quantification of pB was obtained using the simplified reference tissue method. Concomitant PET studies with [(15)O]-water indicated that mirtazapine challenge increased CBF uniformly in cerebellum and other brain regions, supporting the use of this reference tissue for calculation of [(11)C]mirtazapine pB. Displacement by mirtazapine was complete in the cerebral cortex, but only 50% in diencephalon, suggesting the presence of multiple binding sites of differing affinities in that tissue. Competition studies with yohimbine and RX 821002 showed decreases in [(11)C]mirtazapine pB throughout the forebrain; use of the multireceptor version of the Michaelis-Menten equation indicated that 42% of [(11)C]mirtazapine binding in cortical regions is displaceable by yohimbine. Thus, PET studies confirm that [(11)C]mirtazapine affects alpha(2)-adrenoceptor binding sites in living brain.
(c) 2006 Wiley-Liss, Inc.