Initial biological evaluations of [¹⁸F]KS1, novel ascorbate derivative to image oxidative stress in cancer

Objectives: Excessive production of reactive oxygen species (ROS) and/or reactive nitrogen species (RNS) in the tumor microenvironment, through either endogenous or exogenous insults, results in oxidative stress. Oxidative stress has been implicated in many disorders including cancer, neurodegenerative, and cardiovascular diseases. Any clinical intervention aiming to reduce ROS will benefit from a noninvasive technique to measure real-time ROS in vivo. Several PET imaging agents were reported to primarily track ROS for neurological applications. However, all have limited translational potential due to their susceptibility to radiolysis, ability to intercalate DNA, complicated radiolabeling chemistry, poor target binding, and their highly unstable oxidation profiles. Therefore, a novel probe that is PET compatible and binds to ROS with high specificity and ideally with existing clinical safety records could overcome the aforementioned limitations. We have identified (E)-5-(2-chloroethylidene)-3-((4-(2-fluoroethoxy)benzyl)oxy)-4-hydroxyfuran-2(5H)-one (KS1), a vitamin C derivative that detects ROS in solid tumors with high affinity and selectivity. Herein, we describe the synthesis, radiosynthesis, initial in vitro and in vivo evaluation of [¹⁸F]KS1 as a PET radiotracer for oxidative stress imaging in cancer.

Methods: Both the reference standard KS1 and the corresponding F-18 tosyl precursor were synthesized following the published methods, starting from 2-benzyloxy L-ascorbic acid. Radiosynthesis of [¹⁸F]KS1 was achieved by standard nucleophilic substitution of the tosyl precursor with [¹⁸F]^¯ in TRASIS-AIO module. In vitro cell uptake assay was performed in (a) two differentially ROS expressing head and neck cancer (HNC) and (b) prostate cancer (PCa) cell lines, with and without different ROS blockers to evaluate the specificity of [¹⁸F]KS1. MicroPET imaging and standard biodistribution studies were performed in PCa bearing mice (n=3) for 20 min after 45 min post-injection of [¹⁸F]KS1. Blocking mPET experiments were performed with KS1 (5.0 mg/kg), 20 min prior to radiotracer administration.

Results: All the final analogs were characterized using NMR spectra. KS1 and corresponding tosyl precursor were synthesized in 31% and 27% yields respectively. [¹⁸F]KS1 was synthesized in 18±3% radiochemical yield and specific activity of 1.8-2.0 Ci/mmol at EOS. In vitro radioactive uptake in SCC-61 cell line (high oxidative stress HNC cells) was ~2-fold higher compared to rSCC-61 (low oxidative stress HNC cells) uptake. Additionally, the uptake was ~60%, and ~80% blocked by superoxide dismutase (SOD) and KS1 respectively, and ~18% increased by ROS inducer doxorubicin (DOX). [¹⁸F]KS1 uptake in PCa cells under hypoxic conditions was ~2.5-fold higher than normoxic, consistent with the known associated increase in ROS. The uptake was ~55%, and ~60% blocked by SOD and KS1 respectively, and 24% increased by DOX. From mPET ROI analysis, we observed high tumor uptake and significantly blockade uptake (~3-fold lower than baseline), demonstrating selectivity and specificity. Biodistribution from 30 min to 60 min post [¹⁸F]KS1 injection displayed ~1.7-fold increased tumor uptake: %ID/g 4.23 ± 0.531 (30 min) to 6.81 ± 0.33 (60 min). Additionally, there was a >50% increase in the tumor: muscle ratio from 30 min to 60 min. mPET imaging results corroborated well with the biodistribution studies, by demonstrating excellent tumor uptake and blocking.

Conclusions: [¹⁸F]KS1 was synthesized and radiolabeled with high radiochemical purity and specific activity. This would be the first PET tracer, based on a natural antioxidant to measure oxidative stress levels in several solid tumor cells in vivo. Based on the promising data, we hypothesize that our ascorbate-based PET ligand strategy will expand the ascorbate scaffold for potential imaging agent(s) to measure oxidative stress in vivo.