Bolus followed by constant infusion for tracer administration (B/I) has been used for ligand-receptor dynamic PET study to measure changes of tracer binding due to neurological or pharmacological stimulation. Concentration ratio (CR) method is commonly used to quantify the stimulus-induced binding changes. CR method is based on the assumption that all tissue concentrations are at steady state within a given time period. The objective of this study is to develop and evaluate a full modeling approach to measure the amphetamine-induced binding potential (BP) changes in B/I dynamic PET study.
Methods
A 4-parameter generalized reference tissue model (GRTM), (R1, k2, BP0, BP1) is derived from a 3-parameter (R1, k2, BP) simplified reference tissue model (SRTM) by assuming R1, and k2 are constants, and BP is a step function that varies from BP0 in pre-phase to BP1 at post-amphetamine phase, where R1 is the ratio of transport rate from blood to brain tissue, and k2 is the efflux rate constant from free and nonspecific binding compartment to blood. A 4-parameter GRTM with only 2-parameter (k2 and BP1) SRTM for post phase is also derived from the assumption that reference tissue tracer concentration is constant in post phase (GRTM2). GRTM and GRTM2 were applied to 11 monkey [11C]raclopride (RAC). At each PET study, about 25 mCi high specific activity RAC was delivered by B/I with a 75:1 bolus to infusion ratio. Amphetamine was injected at 40 min post tracer bolus injection with a dose level of 2 mg/kg. Each Dynamic PET scan was performed on a GE Advance scanner with acquisition protocol of total 90 min. Images were reconstructed using filtered back projection with a ramp filter which resulted in a spatial resolution of about 4.5 mm FWHM at the center of the field of view. Irregular ROIs of cerebellum (reference tissue) and striatum were defined on the first 45 min integrated PET images. For comparison, two CR methods to estimated BP0 and BP1 are calculated as: CR1: BP0=Cstr(30–40)/Cref(30–40)-1, BP1= Cstr(60–90)/Cref(60–90)-1; CR2: BP0=Cstr_SRTM(60–90)/Cref(60–90)-1, BP1= Cstr(60–90)/Cref(60–90)-1, where Cstr_SRTM is extrapolated by SRTM with parameters determined by model fitting at pre-phase only. The amphetamine-induced dopamine release (DAR) is estimated by BP as DAR=(BP0-BP1)/BP0. GRTM and CR methods were also evaluated by computer simulation. A cerebellum time activity curve from a typical study and wide range of parameters were used to simulate striatum of different kinetics and DAR levels.
Results
In monkey study, the mean and standard deviation (mean(STD)) of estimates are listed in the following table 1. GRTM and GRTM2 provide almost same estimates. The DAR estimates from CR methods are significantly lower than those obtained by GRTM method (p<0.0001). Computer simulation shows that the DAR calculated by CR1 and CR2 are significantly underestimated in 71% and 30% from true values, respectively. A paired T-Test demonstrate that there is no significant difference in AIC values between GRTM and GRTM2 model fitting. In conclusion GRTM is a robust and reliable kinetic modeling approach to quantify the B/I dynamic PET with pharmacological challenge.
Mean and Standard (n=11) Deviation (mean(STD)) of Estimates