Computer simulations of the analysis of equilibrium, steady-state, correction for nonspecific binding, and receptor and delivery sensitivity for the in vivo quantification of a neuroreceptor-binding radioligand

The quantitative potential of PET or SPECT for imaging neuroreceptors is degraded by inaccuracies in kinetic modeling. The authors have simulated the pharmacokinetics for the in vivo binding of carfentanil to the mu opiate receptor in both normal and epileptic human brain. Brain regions within transverse slices through hippo/amyg/cereb and basal ganglia/occ ctx in the mathematical Hoffman 3D brain model were assigned realistic mu concentrations, with 20% elevation in left post temp ctx and hippo for the epileptic brain. Radioligand time-courses, simulated using published plasma curves and rate constants, were used to determine: (1) when equilibrium or state-state exist so that modeling and data analysis might be simplified; (2) when the radioactivity in a control region can be subtracted from the radioactivity in a region of interest as a measure of nonspecific binding; and (3) whether the receptor (delivery) sensitivity is too low (high) to permit accurate estimation of changes in pharmacokinetic parameters in a disease state. These results indicate the caution required for interpretation of clinical data