Abstract
579
Objectives: PET scanning during gestation can allow for the evaluation of changes in the nonhuman primate (NHP) brain during development, which in the rhesus macaque lasts ~ 165 days. One challenge of fetal PET imaging is the accuracy of quantification, as arterial blood samples might need to be avoided to minimize risk to the mother and fetus. Additionally, due to the placental barrier between maternal and fetal circulation, the true input function to the fetal brain may be different from the mother’s input function. Previously, we have reported the uptake of 11C-UCB-J, a synaptic vesicle protein 2A (SV2A) tracer, at equilibrium in the fetal NHP brain as a standardized uptake value ratio (SUVr) by normalizing the fetal brain SUV to the maternal brain SUV. Other methods to noninvasively yet accurately represent radiotracer distribution include reference tissue methods, in which the uptake in a region of interest is analyzed with a reference region. Specifically, the Simplified Reference Tissue Model (SRTM) can estimate the distribution volume ratio (DVR) between the target and reference regions as a relationship between the estimated relative influx kinetic parameter (R1) and efflux parameters of the target and reference regions (k2, k2’). The purpose of this work is to investigate the use of SRTM in fetal brain PET quantification by using the maternal brain as a reference tissue, and to compare this value to the previously reported SUVr measures, in order to validate these quantification measures.
Methods: Two pregnant rhesus macaques underwent PET scanning with 11C-UCB-J at approximately 80, 110, and 140 post-conception days (pcd). The radiotracer was administered using a bolus plus constant infusion paradigm (Kbol=150) for a total scan duration of 75 minutes on the Siemens Biograph mCT. A CT image was taken prior to PET scanning for attenuation correction and anatomical delineation. Both the fetal and maternal brain volumes were defined on the CT images. No blood samples were taken during the scan. Both equilibrium analysis (EA) and SRTM were implemented for two scan durations, using TMAX = 60 min or TMAX = 75 min. For each scan duration, equilibrium uptake in the fetal and maternal brain were averaged during the last 15 min (i.e., 45-60 and 60-75 minutes, respectively) and were quantized as fetal/maternal SUVr.
Results: Fetal/maternal brain DVR from SRTM and SUVr from EA increased in both subjects during the second half of gestation. Estimates from SRTM and EA were consistent at each scan (Table 1). The average SRTM R1 estimate was 0.10 (± 0.04) using 60 minutes and 0.11 (± 0.03) using 75 minutes of data (Table 1). On average, the % difference of SUVr with respect to DVR was 3.3 (± 5.1)% using TMAX = 60 min minutes and 1.8 (± 7.5)% using TMAX = 75 minutes. Within each scan, there was an average percent difference of 4.8 (± 4.8)% between DVR estimates at 60 and 75 minutes. Similarly, an average percent difference of 7.0 (± 5.2)% was observed within scan between 45-60 and 60-75min SUVr.
Conclusions: Using the maternal brain as a reference region for quantification of the fetal brain with SRTM provides similar outcome measurements as the equilibrium analysis of SUVr between the fetal and maternal brains. These preliminary results suggest that using SUVr to quantify fetal brain uptake of 11C-UCB-J relative to the maternal brain is a fair alternative to kinetic modeling with SRTM. It is important to note that the reference region is not representative of the nondisplaceable uptake, as there is specific binding of 11C-UCB-J to SV2A in the maternal brain. Thus, these values must be interpreted as a measure of fetal to maternal total brain uptake. Current and ongoing work includes investigating comparisons between kinetic modeling of fetal brain images and neonate brain images with 11C-UCB-J. Quantification using kinetic modeling might facilitate comparisons within subject between these two developmental periods.