Kinetic analysis of the cannabinoid-1 receptor PET tracer [18F]MK-9470 in human brain

Quantitative imaging of the type 1 cannabinoid receptor (CB1R) opens perspectives for many neurological and psychiatric disorders. We characterized the kinetics and reproducibility of the CB1R tracer [F-18]MK-9470 in human brain. [F-18]MK-9470 data were analysed using reversible models and the distribution volume V (T) and V (ND) k (3) (V (ND) k (3) = K (1) k (2)) were estimated. Tracer binding was also evaluated using irreversible kinetics and the irreversible uptake constant K (i) and fractional uptake rate (FUR) were estimated. The effect of blood flow on these parameters was evaluated. Additionally, the possibility of determining the tracer plasma kinetics using a reduced number of blood samples was also examined. A reversible two-tissue compartment model using a global k (4) value was necessary to describe brain kinetics. Both V (T) and V (ND) k (3) were estimated satisfactorily and their test-retest variability was between 10% and 30%. Irreversible methods adequately described brain kinetics and FUR values were equivalent to K (i). The linear relationship between K (i) and V (ND) k (3) demonstrated that K (i) or FUR and thus the simple measure of tracer brain uptake provide CB1R availability information. The test-retest variability of K (i) and FUR was < 10% and estimates were independent of blood flow. Brain uptake can be used as a receptor availability index, albeit at the expense of potential bias due to between-subject differences in tracer plasma kinetics. [F-18]MK-9470 specific binding can be accurately determined using FUR values requiring a short scan 90 to 120 min after tracer administration. Our results suggest that [F-18]MK-9470 plasma kinetics can be assessed using a few venous samples.