PET kinetics of radiolabeled antidepressant, [N-methyl-11C]mirtazapine, in the human brain

BACKGROUND We compared six kinetic models with and without the requirement of arterial cannulation for estimating the binding potential of [N-methyl-11C]mirtazapine in the living human brain. METHODS Distribution volumes of [N-methyl-11C]mirtazapine in brain regions were estimated using single- and two-tissue compartment models as well as a graphical plasma input model. The two-tissue compartment model provided a direct estimate of the binding potentials of [N-methyl-11C]mirtazapine in brain regions, while binding potentials of the single-tissue compartment model and the graphical plasma input model were estimated indirectly from ratios of distribution volumes in brain regions. We obtained also direct estimates of binding potentials using a graphical reference tissue model and two nonlinear reference tissue models. RESULTS The two-tissue compartment model required several fits with different initial guesses for avoiding negative values of parameters. Despite the extra fits, estimates of distribution volumes and binding potentials of [N-methyl-11C]mirtazapine obtained by the two-tissue compartment model were far more variable than those produced by the other methods. The graphical plasma input method and the graphical reference tissue method provided estimates of the binding potential that correlated closely, but differed in magnitude. The single-tissue compartment model provided relatively low estimates of binding potentials with curves that failed to fit the data as well as the three other methods that used the entire series of positron emission tomography data. The reference tissue method and the simplified reference tissue method provided similar, consistent estimates of binding potentials. However, certain assumptions of the simplified reference tissue method may not be fulfilled by the radioligand. CONCLUSION The reference tissue method is appropriate for estimating the binding potential of [N-methyl-11C]mirtazapine in regions of the human brain so that the binding potential of [N-methyl-11C]mirtazapine can be estimated without arterial cannulation.