Development of α2δ PET ligand trans-[18F]4-fluorogabapentin and the characterization in non-human primate (NHP)

1619 Introduction: Neuropathic pain affects 7-10% of the world’s population. Gabapentinoids, which bind to the α2δ subunits of voltage-gated calcium channels, are the first line non-opioid treatments for neuropathic pain1. Immunohistochemistry indicates that the expression of α2σ-1 subunit of voltage-dependent calcium channels (VDCCs) increases 2-10 fold in the spinal cord and dorsal roots during neuropathic pain2. Based on these results, radiolabeled gabapentin is proposed as a potential radiotracer for imaging the neuropathic pain mechanism and development. Based on the structure-activity relationship (SAR) studies on gabapentin3, trans-4-fluorogabapentin was selected for radiolabeling. Methods: Similar to the method of synthesizing stereoisomer mixture 4-fluorogabapentin we described last year4, nonradioactive trans-4-fluorogabapentin (6a) was prepared in 4 steps. The trans- and cis-(3a) isomers of the hydroxyl intermediate (3) were isolated by semi-preparative HPLC using C18 semiprep-column. The deoxyfluorination and subsequent hydrolysis/deprotection of 3a using DAST/Alkylfluor and NaOH/HCl giving trans-4-fluorogabapentin (6a). The radioactive trans-[18F]4-fluorogabapentin (6b) was similarly synthesized, except that the cis-methylsulfonyl precursor (4) was used for nucleophilic radiofluorination. The binding affinity (EC50) of trans-4-fluorogabapentin (6a) was measured by competitive radioligand binding assay to rat spinal cord sections via quantitative autoradiography. Dynamic PET imaging and arterial blood sampling was performed on a rhesus monkey with and without preinjection of gabapentin (5 mg/kg) (1 scan per condition). The in vivo specific binding was tested by the imaging with 5 mg/kg of preinjected gabapentin. Blood samples were analyzed to determine the radioactivity concentration in whole blood and plasma and radiometabolites. Results: The nonradioactive trans-4-fluorogabapentin (6a) was synthesized and characterized. trans-[18F]4-fluorogabapentin (6b) was synthesized in ~2% non-decay corrected radiochemical yield and >99% of radiochemical purity in ~120 min of synthesis and purification time. The identity and purity of 6b was confirmed by analytical HPLC. Autoradiography in rat spinal cord sections showed binding of 6b in dorsal horn region consistent with the immunohistochemistry and reported expression of α2δ-1 receptors5. The binding affinities (EC50) of trans-4-fluorogabapentin (6a) and gabapentin were 104.26 ± 17.47 nM and 118.61 ± 19.81 nM, respectively. Blood analysis showed that 6b is not metabolized in vivo (>95% parent fraction 90 min post-injection) and that the tracer is largely free in plasma (PFF > 90%). Dynamic PET imaging of trans-[18F]4-fluorogabapentin showed moderate CNS penetration. Although the brain SUV was low, it could be accurately quantified (SUV120-180min: brain = 0.28 ± 0.005; spinal cord = 0.225 ± 0.013; muscle = 0.89 ± 0.02; blood (from blood sampling) = 0.96 ± 0.05). Preliminary data indicate blocking by gabapentin quantified by kinetic modeling methods. Conclusion: α2δ PET ligand trans-[18F]4-fluorogabapentin was synthesized and characterized successfully. trans-[18F]4-fluorogabapentin shows slightly higher binding affinity than gabapentin to rat spinal cord sections. PET imaging in NHP showed that trans-[18F]4-fluorogabapentin can enter the CNS and can be partially blocked by gabapentin. Acknowledgements: This work is supported by NIH/NINDS (PB, R21NS120139), S10OD018035 (MDN), P41EB022544 (MDN), and MGH Tosteson & Fund for Medical Discovery (YPZ). We also want to thank the cyclotron team of GCMI for the [18F]KF production. References: [1] Nat. Rev. Neurosci. 13, 542-555 (2012); [2] J. Neurosci. 21, 1868-1875 (2001); [3] J. Med. Chem. 41, 1838-1845 (1998); [4] J. Nucl. Med. 61 (supplement 1), 68-68 (2020); [5] Neuroscience 155, 510-521 (2008)