Non-steady-state measurement of in vivo radioligand binding with positron emission tomography: specificity analysis and comparison with in vitro binding.

We previously have developed a non-steady-state method for in vivo measurement of radioligand-receptor binding in brain using positron emission tomography (PET) and 18F-spiperone (18F-SP). This method has proven to be highly sensitive to the detection of decreases in the apparent number of available specific binding sites. The purposes of this investigation are to demonstrate the specificity of this PET assay and compare findings to in vitro binding assays. Three to six studies were performed in each of five male baboons. Each animal was pretreated with either ketanserin [serotonergic (S2)], eticlopride [dopaminergic (D2)], or unlabeled SP to compete with 18F-SP for specific binding sites. Sequential PET scans and arterial-blood samples were collected for 3 hr after intravenous injection of 18F-SP. Data were analyzed with a three-compartment model that considered the accumulation of radiolabeled metabolites in arterial blood. Five baboons were killed, and radioligand-receptor binding in vitro was measured by homogenate techniques. There was no detectable in vitro or in vivo specific binding of SP in cerebellum. The specific binding of SP in striatal tissue in vitro was approximately 74% to D2 sites and 26% to S2 sites, whereas ketanserin displaced all specific binding in frontal cortex. In close agreement, specific binding measured in vivo with PET revealed that 68% of apparent striatal binding could be blocked by pretreatment with eticlopride, and 34% by ketanserin. The small apparent difference between receptor binding in vitro and in vivo may result from the relatively poor resolution of PET.(ABSTRACT TRUNCATED AT 250 WORDS)