Abstract
322
Objectives [11C]DPA713 has been used to image translocator protein (TSPO) in humans and has four-fold difference in the affinity, measured in vitro, between high- (HABs) and low-affinity binders (LABs). To assess the utility of [11C]DPA713 in different genotypes, we performed quantification in both regional and voxel data and estimated specific signals in healthy humans.
Methods Nine subjects genotyped for the polymorphism rs6971 (5 HABs, 2 mixed-affinity binders (MABs), and 2 LABs) had [11C]DPA713 PET scans with arterial sampling. Distribution volume (VT) was determined by 2-tissue compartment (2TC) for regional data and by Logan (LA) and Ichise’s bilinear (MA1) methods for voxel data. To estimate proportion of specific signal in VT, two methods were applied to fit the nondisplaceable distribution volume (VND); 1) by assuming that VND is the same among the three genotypes and 2) by the occupancy plot to HABs and MABs.
Results Average VT across regions was 4.5, 2.4, and 1.4 for HABs, MABs, and LABs, respectively. The two methods to estimate VND gave similar values of 1.7 and 1.9 indicating that these estimates are accurate and that most of VT of LABs is nonspecific binding. In the HABs and MABs, voxel VT determined by MA1 correlated well with VT by 2TC for regional data (r2 > 0.8) with 7% underestimation while voxel-based LA underestimated VT by 14%. On the other hand, in LABs, VT by both LA and MA1 for voxel data poorly correlated with VT from regional data and 2TC (r2 < 0.4) possibly due to barely detectable levels of specific binding.
Conclusions [11C]DPA713 has detectable levels of specific to nondisplaceable ratio, in HABs (1.5) and MABs (0.3), which allows quantification based on both regional and voxel levels. However, our preliminary results suggest that LABs have almost undetectable amounts of specific binding because these affinity ratios are different between in-vivo and in-vitro or affinity to LABs is too low to show the specific binding in PET.