In vivo imaging of the cannabinoid CB1 receptor using positron emission tomography

Sammanfattning: The actions of marijuana (cannabis) are mediated by receptors (primarily the cannabinoid CB1 receptor) that have unusually high density and wide distribution in brain. In addition to mediating the effects of exogenous drugs like cannabis, these receptors also receive signals from endogenous cannabinoids (endocannabinoids), which modulate the release of several other neurotransmitters. Since abnormalities of cannabinoid receptors and endocannabinoid transmission have been hypothesized to underlie disorders of the brain (e.g., memory impairment, schizophrenia, and seizures), this neurotransmitter system has been an active target for development of drug therapies and of biomarkers to measure its in vivo function. Positron emission tomography (PET) can image the distribution of receptors in the body, is a powerful tool for drug development, and can quantify the receptor as a biomarker to assess pathophysiology. The purpose of this thesis was to evaluate several candidate PET radioligands for their relative ability to quantify CB1 receptors in the living brain of animals and humans. We first assessed [11C]MePPEP as a PET radioligand through studies in rodents. Wildtype and genetically modified mice were used to determine that [11C]MePPEP is not a substrate for the P-glycoprotein efflux transporter and that the majority (about two-thirds) of its binding in brain is specific to the CB1 receptor. Pharmacologically active doses of CB1 agonists had no effect on [11C]MePPEP in rats, which suggests a large CB1 receptor reserve. Pharmacokinetic modeling of CB1 receptors using brain radioactivity and measurements of radioligand in arterial plasma yielded stable measures after 70 minutes of scanning. The results suggest that [11C]MePPEP might be successful in humans, although competition studies with endocannabinoids would not be possible, and radiometabolites might cause consistent overestimation of CB1 receptor density. Second, we examined [11C]MePPEP in healthy human subjects using the gold standard of compartmental modeling to quantify receptor density in brain. [11C]MePPEP had high uptake in brain, could be imaged for 210 minutes, and could quantify CB1 receptors within about 60 minutes of scanning. However, the accuracy and precision of the pharmacokinetic modeling hinged upon the accuracy of radioligand measurements in arterial plasma. A radioligand with a longer radioactive half-life, such as from 18F, would be expected to provide superior measurements in arterial plasma. We then evaluated several 18F-radiolabeled analogues of MePPEP in monkey. [18F]FMPEP-d2 was selected for study in human due to its superior uptake in brain compared to [18F]FEPEP, and reduced uptake of radioactivity in bone compared to [18F]FMPEP. In humans, [18F]FMPEP-d2 could image and quantify CB1 receptors with better accuracy and precision compared to that of [11C]MePPEP. As suspected, the accuracy in measuring radioligand in arterial plasma was the critical improvement needed in the pharmacokinetic modeling. Finally, we examined the biodistribution and radiation dosimetry estimates in human for [11C]MePPEP and [18F]FMPEP-d2. Both radioligands had high uptake in brain, liver, and lungs, and both had significant uptake of radioactivity in bone marrow, but not in bone. Regardless, both radioligands have an effective dose similar to that of other clinically used PET radioligands. In conclusion, we have shown that both [11C]MePPEP and [18F]FMPEP-d2 can quantify CB1 receptors in brain. However, [18F]FMPEP-d2 is superior to [11C]MePPEP because it has greater precision and accuracy. Thus, [18F]FMPEP-d2 is a promising PET radioligand to measure CB1 receptors in vivo, and can now be used to explore the role of this receptor in human health and disease.