Positron emission tomography (PET) receptor occupancy imaging plays an increasingly important role in the development of central nervous system (CNS) drugs, providing critical information on drug brain penetration, target engagement and dosing. The conventional approach to measure occupancy of a CNS drug is to scan a subject twice, at baseline and after administration of the drug, independently apply image reconstruction and kinetic modeling to the data of each scan, and compute occupancy by measuring fractional reductions in specific ligand binding between the scans. The drawback of this approach, however, is the low precision of the estimated occupancy values. We propose to develop a novel parametric reconstruction approach that jointly reconstructs and analyzes the dynamic projections measured in the baseline and post-drug scans, leading to direct, quantitative estimation of receptor occupancy maps with a drastically higher signal-to-noise ratio. We expect our approach to significantly improve the precision and accuracy of occupancy quantification, allowing more robust characterization of dose-occupancy relationships and thereby greatly improving the quality of the information extracted from PET occupancy studies. The proposed methodology will be evaluated in an animal model.
In order to drastically improve the quality of information that can be extracted from PET receptor occupancy studies, we propose a novel parametric reconstruction approach that estimates occupancy maps directly and jointly from baseline and post-drug dynamic projection data sets.