Development and In Vivo Preclinical Imaging of Fluorine-18-Labeled Synaptic Vesicle Protein 2A (SV2A) PET Tracers.

Purpose: Synaptic vesicle protein 2A (SV2A) serves as a biomarker of synaptic density and positron emission tomography (PET) imaging of SV2A could provide a tool to assess progression of neurodegenerative diseases. Two tracers have primarily been reported and characterized in vivo: [C]UCB-J and [F]UCB-H. In early human studies, [C]UCB-J showed promising results, while its F-18-labeled analogue [F]UCB-H showed suboptimal specific signal in comparison to [C]UCB-J. Considering the limited use of [C]UCB-J to facilities with a cyclotron, having a F-18 variant would facilitate large, multicenter imaging trials. We have screened several F-18 derivatives of UCB-J in non-human primates and identified a promising F-18 PET candidate, [F]MNI-1126, with additional investigations of the racemate [F]MNI-1038, affording a signal comparable to [C]UCB-J.

Procedures: F-18 derivatives of UCB-J and UCB-H were synthesized and administered to non-human primates for microPET imaging. Following screenings, [F]MNI-1038 (racemate) and [F]MNI-1126 (R-enantiomer) were identified with the highest signal and favorable kinetics and were selected for further imaging. Kinetic modeling with one- and two-tissue compartmental models, and linear methods were applied to PET data using metabolite-corrected arterial input function. Pre-block scans with levetiracetam (LEV, 10, 30 mg/kg, iv) were performed to determine the tracers' in vivo specificity for SV2A. Two whole-body PET studies were performed with [F]MNI-1038 in one male and one female rhesus, and radiation absorbed dose estimates and effective dose (ED, ICRP-103) were estimated with OLINDA/EXM 2.0.

Results: All compounds screened displayed very good brain penetration, with a plasma-free fraction of ~ 40 %. [F]MNI-1126 and [F]MNI-1038 showed uptake and distribution the most consistent with UCB-J, while the other derivatives showed suboptimal results, with similar or lower uptake than [F]UCB-H. V of [F]MNI-1126 and [F]MNI-1038 was high in all gray matter regions (within animal averages ~ 30 ml/cm) and highly correlated with [C]UCB-J (r > 0.99). Pre-blocking of [F]MNI-1126 or [F]MNI-1038 with LEV showed robust occupancy across all gray matter regions, similar to that reported with [C]UCB-J (~ 85 % at 30 mg/kg, ~ 65 % at 10 mg/kg). Using the centrum semiovale as a reference region, BP of [F]MNI-1126 reached values of up to ~ 30 to 40 % higher than those reported for [C]UCB-J. From whole-body imaging average ED of [F]MNI-1038 was estimated to be 22.3 μSv/MBq, with tracer being eliminated via both urinary and hepatobiliary pathways.

Conclusions: We have identified a F-18-labeled tracer ([F]MNI-1126) that exhibits comparable in vivo characteristics and specificity for SV2A to [C]UCB-J in non-human primates, which makes [F]MNI-1126 a promising PET radiotracer for imaging SV2A in human trials.