Evaluation of the agonist PET radioligand [11C]GR103545 to image kappa opioid receptor in humans: Kinetic model selection, test–retest reproducibility and receptor occupancy by the antagonist PF-04455242

Abstract

IntroductionKappa opioid receptors (KOR) are implicated in several brain disorders. In this report, a first-in-human positron emission tomography (PET) study was conducted with the potent and selective KOR agonist tracer, [11C]GR103545, to determine an appropriate kinetic model for analysis of PET imaging data and assess the test–retest reproducibility of model-derived binding parameters. The non-displaceable distribution volume (VND) was estimated from a blocking study with naltrexone. In addition, KOR occupancy of PF-04455242, a selective KOR antagonist that is active in preclinical models of depression, was also investigated. MethodsFor determination of a kinetic model and evaluation of test–retest reproducibility, 11 subjects were scanned twice with [11C]GR103545. Seven subjects were scanned before and 75min after oral administration of naltrexone (150mg). For the KOR occupancy study, six subjects were scanned at baseline and 1.5h and 8h after an oral dose of PF-04455242 (15mg, n=1 and 30mg, n=5). Metabolite-corrected arterial input functions were measured and all scans were 150min in duration. Regional time-activity curves (TACs) were analyzed with 1- and 2-tissue compartment models (1TC and 2TC) and the multilinear analysis (MA1) method to derive regional volume of distribution (VT). Relative test–retest variability (TRV), absolute test–retest variability (aTRV) and intra-class coefficient (ICC) were calculated to assess test–retest reproducibility of regional VT. Occupancy plots were computed for blocking studies to estimate occupancy and VND. The half maximal inhibitory concentration (IC50) of PF-04455242 was determined from occupancies and drug concentrations in plasma. [11C]GR103545 in vivo KD was also estimated. ResultsRegional TACs were well described by the 2TC model and MA1. However, 2TC VT was sometimes estimated with high standard error. Thus MA1 was the model of choice. Test–retest variability was ~15%, depending on the outcome measure. The blocking studies with naltrexone and PF-04455242 showed that VT was reduced in all regions; thus no suitable reference region is available for the radiotracer. VND was estimated reliably from the occupancy plot of naltrexone blocking (VND=3.4±0.9mL/cm3). The IC50 of PF-04455242 was calculated as 55ng/mL. [11C]GR103545 in vivo KD value was estimated as 0.069nmol/L. Conclusions[11C]GR103545 PET can be used to image and quantify KOR in humans, although it has slow kinetics and variability of model-derived kinetic parameters is higher than desirable. This tracer should be suitable for use in receptor occupancy studies, particularly those that target high occupancy.