In vivo imaging evaluation of a novel ¹⁸F-labeled SV2A PET tracer in nonhuman primates

Objectives: Synapse loss is one of the earliest and most robust biomarkers for Alzheimer’s disease (AD). In vivo PET imaging of synapses can provide crucial information on the pathophysiological dynamics of early and developing AD brain, unattainable through other approaches. For the first time, synaptic vesicle glycoprotein 2A (SV2A) PET imaging with ¹¹C-UCB-J allowed in vivo imaging of synapses [1, 2]. In order to enable multicenter clinical trials to validate the utility of SV2A PET imaging in AD and other neurologic disorders, we set out to develop ¹⁸F-labeled tracers with comparable imaging characteristics as ¹¹C-UCB-J. In this regard, we herein report our progress in that direction, with the discovery of a new SV2A PET imaging probe, and assess its unique in vivo binding characteristics and translational potential.

Methods: The racemic SV2A ligand SDM-7 and its fluorination precursor were synthesized from commercially available starting materials. Enantiomers of SDM-7 were separated on a chiral HPLC column. Binding affinity was measured through competition binding assay using rat brain homogenates. The iodonium ylide precursor was labeled with ¹⁸F-fluoride at 150 °C for 10 min. Both the radiolabeled racemic and enantiopure tracers were evaluated in nonhuman primates, with the enantiopure product separated on a chiral HPLC column. Brain PET imaging was performed on a FOCUS 220 scanner for 120 or 180 min. Levetiracetam (30 mg/kg, i.v.) was infused at 60 or 90 min post tracer injection to determine the binding specificity of the tracer. Arterial blood was taken for metabolite analysis and measurement of arterial input functions (AIF). Regional brain time-activity curves (TACs) were generated and analyzed by one-tissue (1T), two-tissue (2T) compartment models, and multilinear analysis-1 (MA1) to obtain regional volumes of distribution (V_T) and binding potential (BP_ND). Results obtained with the novel SV2A tracer were compared with other SV2A tracers. Results: SDM-7 was synthesized in 30% overall yield in 4 steps. Binding affinities of the two enantiomers were both in the nanomolar range with K_i of 2.1 nM and 9.4 nM for the (R)- and (S)-enantiomer, respectively, lower than UCB-J (0.27 nM). The iodonium ylide precursor was synthesized in 19% overall yield in 5 steps. (R)-¹⁸F-SDM-7 was prepared in high molar activity (A_m: 61 MBq/nmol) after chiral HPLC separation. The enantiopurity was 100% based on chiral HPLC analysis. In nonhuman primates, both racemic and (R)-¹⁸F-SDM-7 showed similar rates and patterns of metabolism, with radio-metabolites more polar than the parent tracers. Plasma free fraction was 41%. ¹⁸F-SDM-7 displayed high brain uptake (peak SUV of 6). Kinetics was faster than ¹¹C/¹⁸F-UCB-J, with brain activity-to-AIF ratios plateauing after 20 min for ¹⁸F-SDM-7, vs. 50 min for ¹¹C/¹⁸F-UCB-J. Both 1T and MA1 produced reliable V_T values with good fits. Regional BP_ND values were estimated using the white matter as reference region, and ranged from 0.66 in amygdala to 2.6 in cingulate cortex, which are comparable to ¹¹C/¹⁸F-UCB-J and much higher than ¹⁸F-UCB-H. Displacement with levetiracetam reduced the regional uptake of ¹⁸F-SDM-7 to background levels, confirming their binding specificity to SV2A.

Conclusions: We have successfully synthesized and evaluated a number of novel enantiopure SV2A ligands in nonhuman primates. Among these, (R)-¹⁸F-SDM-7 was found to possess attractive imaging properties: high brain uptake, fast tissue kinetics, and high specific-to-nonspecific binding ratios in brain. Comprehensive characterization of the SV2A PET imaging probes in nonhuman primates are underway to prioritize tracer(s) for advancing to first-in-human studies.