18F-Chemistry in HPLC vials - a microliter scale radiofluorination approach

Abstract:

Objectives: Finding optimal 18F-fluorination conditions represents a critical and often time-consuming step in radiotracer development. Microfluidic or lab-on-a-chip devices are modern tools for that purpose but not accessible everywhere. Inspired by 18F-chemistry in low volumes based on a MAX/MCX-trapping technique for one complete 18F-batch,[1] we herein present an 18F-labeling approach using low volume aliquots (<100 µL) of defined QMA-eluates in HPLC vials as reaction vessels for optimizations as well as the preparation of radiotracers for preclinical studies.

Methods: For optimization experiments, 1-80 MBq [18F]fluoride was eluted from QMA cartridges with defined mixtures of phase transfer catalyst (Kryptofix®222), base (such as K2CO3, KHCO3, KH2PO4), and 2-3% water in acetonitrile[2]. 25 or 50 µL aliquots of the eluates were pipetted into 1.5 mL HPLC vials and the sealed vials were consecutively dried once at 90°C for 3 min with a helium stream. After processing up to 15 vials in one batch, 25 or 50 µL precursor stock solution (1-20 mg/mL in MeCN, DMF, or DMSO) was added and the sealed vials were heated at defined temperatures (60-140°C) and times (5-15 min) by using three heaters equipped with aluminum blocks. Samples were 4-fold diluted with acetonitrile/water 50/50 for analysis by radio-TLC/radio-HPLC. 18F-trapping/elution at higher activity levels (~20 GBq [18F]fluoride) was performed by using a radiosynthesizer (Tracerlab FXFN) followed by transfer into a lead container having three small bores as inlet, vent needle and withdrawal port. A self-made pipet tip-to-cannula adapter allowed withdrawal of defined 25-100 µL aliquots at an activity level of ~13 MBq/µL for further processing under optimized conditions as described above.

Results: The use of small reaction volumes and HPLC vials enabled an efficient 18F-optimization workflow to examine rapidly a variety of reaction parameters (>50 reactions/day) using minimal amounts of precursor. Utilizing mixtures of 2-3% water in acetonitrile for 18F-elution allowed for radiolabeling even without azeotropic drying in certain cases, which can further accelerate the entire labeling process. For more demanding radiolabeling reactions, drying of only 3 min with helium furnished highly reactive [18F]fluoride with comparable reactivity even several hours after the elution step. This optimization workflow was used successfully to identify high yielding reaction conditions for a variety of precursor molecules like aliphatic tosylates or mesylates as well as aromatic sydnones, boronic acids or trimethylammonium salts. As examples, optimization results for 18F-labeling of the FDG mannose triflate precursor and the tosylate precursor of celecoxib-derivative [18F]A[3] are shown in Figure 1. Transfer into isolated RCY at activity levels suitable for preclinical evaluations was achieved (e.g. for [18F]A: 25-30% after 18F-labeling in acetonitrile at 90°C and purification) and turned out to be comparable to results from conventional automated radiosyntheses.

Conclusion: This microliter scale optimization and labeling procedure can easily be adapted and transferred to the synthesis of known and new radiotracers. The straightforward approach represents a valuable radiochemistry tool to enter rapidly into subsequent preclinical studies.

References: [1] Iwata (2018), J.Labelled.Comp.Radiopharm., 61, 540; [2] Kniess (2017), Appl.Radiat.Isot 127, 260; [3] Laube (2020), RSC Advances, 10, 38601.

Figure 1. Setup of HPLC vial experiments and representative results of optimization.