Quantifying SV2A density and drug occupancy in the human brain using [11C]UCB-J PET imaging and subcortical white matter as reference tissue

• DOI: 10.1007/s00259-018-4119-8
• Publication Date: 2018-08-18T15:53:04Z
• Authors: Michel Koole; June van Aalst; Martijn Devrome; Nathalie Mertens; Kim Serdons; Brigitte Lacroix; Joel Mercier; David Sciberras; Paul Maguire; Koen Van Laere
• Keywords: Adult; Male; Positron emission tomography; Synaptic density; Pyridines; Nerve Tissue Proteins; Models, Biological; Multimodal Imaging; Young Adult; 03 medical and health sciences; Kinetic modelling; 0302 clinical medicine; Humans; Reference tissue; Membrane Glycoproteins; Middle Aged; Reference Standards; Magnetic Resonance Imaging; White Matter; [11C]UCB-J; Pyrrolidinones; Kinetics; Gene Expression Regulation; Positron-Emission Tomography; Female

A [11C]UCB-J blocking study was performed in healthy volunteers to validate simplified, non-invasive measures for quantifying presynaptic SV2A expression using subcortical white matter as reference tissue.Ninety minutes dynamic [11C]UCB-J PET scanning with arterial blood sampling was performed in 10 healthy volunteers (8 M/2F; age 27.6 ± 10.0 yrs), before and after administration of a novel chemical entity with selective affinity for SV2A. The centrum semi-ovale (SO) was validated as reference region by comparing baseline and post treatment distribution volume (VT). Using SO as reference tissue, Binding Potential (BPSO) using a Simplified Reference Tissue Model (SRTM, down to 60 min acquisition) and Standardized Uptake Value Ratios (60-90 min post injection - SUVRSO,60-90min) were compared with regional distribution volume ratios (DVR). Next, SV2A occupancy values based on SRTM BPSO and SUVRSO,60-90min were compared to occupancy estimates using regional VT values and a Lassen plot.After pretreatment, regional VT values were reduced significantly except for SO. Highly significant correlations were found between DVR, SRTM BPSO and SUVRSO,60-90min. Compared to DVR, baseline SRTM BPSO showed a small bias (≤ 6.1%) with lower precision for shorter acquisition times, while SUVRSO,60-90min showed 3.5% bias with similar precision. Differences between SV2A occupancy values based on SUVRSO,60-90min and occupancy estimates using VT and a Lassen plot were small but significant, while negligible bias was found for SRTM based occupancy estimates (at least 70 min acquisition).This [11C]UCB-J blocking study validated SO as a suitable reference region for non-invasive quantification of SV2A availability and drug occupancy in the human brain. Accurate quantification can be achieved by using either SUVRSO,60-90min with a 60-90 min PET acquisition or SRTM BPSOwith at least 70 min dynamic PET acquisition.