Synaptic Density Imaging with SV2A using Positron Emission Tomography: Optimization of Reference Region Analysis and Applications in the Human Spinal Cord and the Developing Nonhuman Primate Brain

Date of Award

Fall 10-1-2021

Document Type


Degree Name

Doctor of Philosophy (PhD)


Biomedical Engineering (ENAS)

First Advisor

Carson, Richard


Positron emission tomography (PET) is a functional imaging modality in which radioactive ligands are used to target physiological functions, including metabolism or receptor densities, in vivo. Once administered into a subject, the PET radioligand distributes throughout the body and undergoes radioactive decay, emitting two gamma rays 180 degree from each other. These gamma rays are detected by the PET camera as coincidence events, and the PET raw data are reconstructed into 4-dimenstional images (x, y, z, t). Dynamic PET images can be quantified by applying various kinetic models to radioactivity concentrations in a volume or region of interest (ROI) over time to estimate outcome measures such as the binding potential BP, which is related to the in vitro measure of Bmax/KD. Gold standard quantification techniques typically require the acquisition of arterial blood samples to measure the arterial input function (AIF), which measures the radioactivity due to un-metabolized radiotracer in the arterial blood throughout the scan. However, reference region methods allow for quantification without blood sampling. A recent development in the field of PET imaging has been the motivation to target the synaptic vesicle glycoprotein 2A (SV2A) as a means to measure synaptic density in vivo. PET radioligands including [11C]UCB-J and [18F]SynVesT-1 have been developed at the Yale PET Center and show promising implications to image synaptic density by SV2A in clinical populations. Firstly, this dissertation reviews our work in assessing a reference region for PET imaging with [11C]UCB-J. Ex vivo studies have suggested that the white matter region of the centrum semiovale (CS) may be a suitable reference region for [11C]UCB-J, due to the region’s negligible concentration of SV2A. However, there was evidence of specific binding in the CS during displacement scans of human subjects, where the SV2A-specific drug levetiracetam were administered. We hypothesized that this observation was caused by a combination of partial volume effects and lack of convergence during image reconstruction. The CS was assessed as a reference region after optimizing the ROI to avoid spill-in from gray matter regions with high radioactivity concentrations and investigating convergence of the CS ROI values using the ordered subset expectation maximization (OS-EM) reconstruction algorithm. Both of these corrections led to the reduction, but not total elimination, of observed specific binding in the CS. The baseline total volume of distribution VT in the CS was also compared to the nondisplaceable volume of distribution VND in the gray matter, to evaluate the validity of using radiotracer uptake in the CS as a reference for brain [11C]UCB-J quantification. CS VT overestimated VND, but the measures are highly correlated, suggesting that the CS may be a useful, though somewhat biased, estimate of nondisplaceable uptake to allow for noninvasive quantification of SV2A PET. Next, we explored potential avenues for clinical and preclinical applications of [11C]UCB-J PET, starting with exploring the feasibility of SV2A PET imaging with [11C]UCB-J in the human spinal cord (SC). A simulation study was performed based on dimensions and expected SV2A distribution in the cervical SC (cSC) to determine potential baseline and blocking VT values, as well as VND and VS values, from [11C]UCB-J images on the HRRT. In addition, human baseline and blocking [11C]UCB-J HRRT images were used to estimate these same values. Finally, we used whole-body images to explore the feasibility of imaging SV2A in the full spinal cord. Distribution volume ratios (DVR) were estimated using automated SC ROIs for the full SC, cSC, and thoracic SC (tSC) with the brain gray matter as a reference region. Overall, the results from this study suggest that SV2A can be measured in the SC, though due to the very low density of SV2A in the SC compared to the brain, radiotracers with higher affinity or improved signal to noise ratios should be utilized. Finally, we used both [11C]UCB-J and [18F]SynVesT-1, another SV2A radioligand, in a longitudinal imaging study to evaluate SV2A density changes in the developing nonhuman primate brain. In collaboration with the California National Primate Research Center at University of California, Davis, a total of 8 gravid rhesus macaques were scanned twice during the third trimester (~120- and 145-days gestation age, with 165-day gestation period). Post-mortem samples were collected near term in a subset of subjects and regional SV2A densities were quantified by Western blot. Subcortical regions, including the amygdala and putamen, had the highest SV2A concentrations in the fetal brain as measured with in vivo PET and ex vivo Western blot. Though there are differences in synaptic density measures of fetal-to-adult ratios by SV2A PET and fetal-to-adult ratios by electron microscopy, the results suggest that SV2A PET methodology allows for the monitoring of regional SV2A changes in the fetal brain during the third trimester. Additionally, this dissertation summarizes ongoing work regarding longitudinal imaging in the neonate NHP (2 weeks old – 6 months old). This dissertation work utilizes and analyzes [11C]UCB-J PET radioligand to image SV2A in a number of settings, firstly by assessing the use of the CS as a reference region for noninvasive quantification for SV2A PET imaging with [11C]UCB-J. Further, we explore biological clinical and preclinical applications, starting with evaluating the feasibility of imaging SV2A in the human spinal cord. Finally, we utilized SV2A PET technology to longitudinally monitor synaptogenesis in the developing NHP brain by evaluating regional SV2A changes within subject throughout gestation and early postnatal life.

This document is currently not available here.