Abstract:
The in vivo tracer kinetics of 3-(2'-[18F]fluoroethyl)spiperone
(FESP) in the caudate/striatum and cerebellar regions of the human and
monkey brain were studied with positron emission tomography (PET). The
minimal model configuration that can describe the kinetics was
determined statistically. Three two-compartment model configurations
were found to be suitable for describing the kinetics in
caudate/striatum and cerebellum: (1) a nonlinear model (five parameters)
applicable to studies using nontracer (partially saturating) quantities
of FESP in monkey striatum, (2) a linear four-parameter model applicable
to the caudate/striatal and cerebellar kinetics in human and monkey
studies with tracer quantities of FESP, and (3) a linear three-parameter
model derived from the four-parameter model by assuming irreversible
binding applicable to tracer studies of the human caudate. In the human
studies, when the caudate kinetics (n 3D 4) were fit by model 2 (with
four parameters), the value of the in vivo ligand dissociation constant
kd was found to be 0.0015 +/- 0.0032/min. The three-parameter model
(model 3) was found to fit the data equally well: this model is
equivalent to model 2 with kd set to zero. In the monkey studies, it was
found that for short (90 min) studies using tracer quantities of FESP,
model 2 fit the striatal kinetics better than model 3. The parameters
estimated using model 2 (four parameters) were in better agreement with
those estimated by the nonlinear model (model 1) than those estimated
using model 3 (three parameters). The use of a graphical approach gives
estimates of the plasma-tissue fractional transport rate constant K1 and
the net uptake constant K3 comparable to estimates using model 3 for
both human and monkey studies