Potentials and pitfalls using high affinity radioligands in PET and SPET determinations on regional drug induced D2 receptor occupancy--a simulation study based on experimental data

The D2 dopamine receptor density ranges from 0.2 to 40 nM among human brain regions. For high density regions radioligands like [(11)C]raclopride provide accurate and reliable estimates of the receptor density. In research on neuropsychiatric disorders there is, however, a growing need for quantitative approaches that accurately measure D2 dopamine receptor occupancy induced by drugs or endogenous dopamine in regions with low receptor density. The new high affinity radioligands [(11)C]FLB 457 and [(123)I]epidepride have been shown to provide a signal for extrasriatal D2 dopamine receptor populations in the human brain in vivo. Initial observations indicate, however, that the time required to reach equilibrium is dependent on receptor density. Ratio analyses may thus not be readily used for comparisons among different brain regions. The aim of the present simulation study was to examine commonly used approaches for calculation of drug induced D2 dopamine receptor occupancy among regions with widely different receptor density. The input functions and the rate constants of [(11)C]FLB 457 and the reference ligand [(11)C]raclopride were first used in a simulation estimating the effect of receptor density on equilibrium time. In a second step we examined how errors produced by inaccurate determination of the binding potential parameter propagate to calculations of drug induced receptor occupancy. The simulations showed a marked effect of receptor density on equilibrium time for [(11)C]FLB 457, but not for [(11)C]raclopride. For [(11)C]FLB 457, a receptor density above about 7 nM caused the time of equilibrium to fall beyond time of data acquisition (1 h). The use of preequilibrium data caused the peak equilibrium and the end time ratio approaches but not the simplified reference tissue model (SRTM) approach to underestimate the binding potential and thus also the drug occupancy calculated for high-density regions. The study supports the use of ratio and SRTM analyses in extrastriatal low-density receptor regions for which the high affinity ligand [(11)C]FLB 457 was developed. However, in high-density regions such as the human striatum simple ratio approaches cannot be validly applied, whereas the SRTM approach has higher potential to provide valid estimates. Interestingly, the results suggest that published data on a proposed extrastriatal selectivity for the antipsychotic drugs clozapine and olanzapine may be due to erroneous estimations of the binding potential when using ratio approaches.