Synthesis and evaluation of a hydrophilic ¹⁸F-labeled trifluoroborate derivative of 2-nitroimidazole for imaging tumor hypoxia with positron emission tomography

Objectives: The gold standard 2-nitroimidazole-based hypoxia tracer ¹⁸F-FMISO is normally imaged at late time points (several hours post-injection) due to its slow clearance from background tissues. To facilitate clearance from the background and to be able to image at an early time point, more hydrophilic 2-nitroimidazole-based tracers such as ¹⁸F-FAZA have been developed and evaluated in the clinic as well. In the past three years, our lab has successfully applied the ¹⁸F-¹⁹F isotope exchange reaction on a hydrophilic and zwitterionic ammoniomethyl-trifluoroborate (AmBF₃) motif for the design of PET tracers targeting several peptide receptors. In this study, we investigated if a hydrophilic AmBF₃ derivative of 2-nitroimidazole, ¹⁸F-AmBF₃-Bu-2NI, could have the potential to image tumor hypoxia at earlier time points.

Methods: AmBF₃-Bu-2NI was synthesized in four steps from commercially available 2-nitroimidazole: coupling of 2-nitroimidazole with 1,4-dibromobutane, formation of tertiary amine with dimethylamine, formation of quaternary ammonium iodide salt with iodomethylboronic acid pinacol ester, and the final conversion to trifluoroborate. ¹⁸F labeling was conducted via ¹⁸F-¹⁹F isotope exchange reaction directly in aqueous media (pH 2) at 80 °C, and purified by HPLC. LogP was measured by the shake flask

Methods: Stability of ¹⁸F-AmBF₃-Bu-2NI was determined by incubating the tracer with mouse plasma at 37 °C, and monitored by HPLC. PET/CT imaging and biodistribution studies were conducted in mice bearing HT-29 human colon cancer xenografts, and the results were compared with those obtained using ¹⁸F-FMISO in the same tumor model. Cell uptake assays were conducted at 27 °C using HT-29 cells for up to 3 hours, and acid wash was used to remove membrane-bound radioactivity.

Results: AmBF₃-Bu-2NI was obtained in 19% yield over 4 steps. ¹⁸F-AmBF₃-Bu-2NI was obtained in 14.8 ± 0.4% (n = 3) decay-corrected radiochemical yield with 24.5 ± 5.2 GBq/µmol specific activity and > 99% radiochemical purity. ¹⁸F-AmBF₃-Bu-2NI was relatively hydrophilic with LogP value of − 1.52 ± 0.02 (n = 3). For comparison, the reported LogP value for ¹⁸F-FMISO was − 0.40. ¹⁸F-AmBF₃-Bu-2NI was very stable in mouse plasma and no significant metabolites were observed after a 1-h incubation period at 37 °C. Imaging and biodistribution studies showed that ¹⁸F-AmBF₃-Bu-2NI cleared quickly from blood and was excreted via both hepatobiliary and renal pathways. However, the tumor was not visualized in PET images until 3 h post-injection. The tumor uptake values of ¹⁸F-AmBF₃-Bu-2NI at 1 h and 3 h post-injection were 0.54 ± 0.13 and 0.19 ± 0.04%ID/g, respectively. The tumor-to-muscle and tumor-to-blood ratios of ¹⁸F-AmBF₃-Bu-2NI were 0.51 ± 0.25 and 0.99 ± 0.32, respectively at 1 h post-injection, and 0.92 ± 0.08 and 2.62 ± 1.02, respectively at 3 h post-injection. For comparison, the tumor uptake of ¹⁸F-FMISO at 3 h post-injection using the same tumor model was 1.84 ± 0.52 %ID/g. The tumor-to-muscle and tumor-to-blood ratios for ¹⁸F-FMISO at 3 h post-injection were 4.52 ± 1.36 and 5.05 ± 0.50, respectively. The cell uptake studies showed no detectable radioactivity in acid wash and cell pellets at all tested time points (5, 15, 30, 60, 120, and 180 min), indicating that ¹⁸F-AmBF₃-Bu-2NI did not bind to cell membrane or get internalized into HT-29 cells. Conclusion: The low tumor uptake of ¹⁸F-AmBF₃-Bu-2NI is likely due to the highly hydrophilic and zwitterionic motif of AmBF₃ that prevents free diffusion of ¹⁸F-AmBF₃-Bu-2NI across the cell membrane. Our results suggest that highly hydrophilic and zwitterionic ¹⁸F-labeled AmBF₃ derivatives might not be suitable for imaging intracellular targets including nitroreductase, a common tumor hypoxia imaging target.