Abstract (eng)
Aim
This thesis investigated the use of novel elution solutions for separation of cyclotron-produced [18F]fluoride from water. The aim was to find an elution solution that avoids or minimizes the azeotropic drying step following the elution to activate the [18F]fluoride for nucleophilic radiofluorination. A special focus was on the applicability into a microfluidic reaction system as conventional solutions cause occasional clogging of the reactors.
Methods
Elution solutions A through F were prepared with Kryptofix® 222 (58.4 µmol/ml) and potassium carbonate (32.6 µmol/ml) with acetonitrile and a water content from 0 % to 20 %. Elution solution WA contained potassium hydroxide (167 µmol/ml) and Kryptofix® 222 (250 µmol/ml) in pure acetonitrile. Tetrabutylammonium hydroxide (56.3 µmol/ml) and tetraethylammonium bicarbonate (78.4 µmol/ml) were used in 80/20 (v/v) acetonitrile and water.
Elution profiles were generated for four different columns: 30-PS-HCO3- (Macherey-Nagel), Sep-Pak AccelPlus QMA Plus Light (Waters) and Fluoride Trap & Release (25 mg and 9 mg sorbent; ORTG Inc.).
[18F]FE@SUPPY and [18F]FE@SUPPY:2 radiosyntheses were used for indicating the reactivity of eluted [18F]fluoride. Dried [18F]fluoride was heated with an acetonitrile precursor solution (15 mg/ml Tos@SUPPY or 20 mg/ml Tos@SUPPY:2) at 75 °C for 20 minutes and the reaction was quenched with water. Microfluidic syntheses of [18F]FE@SUPPY:2 were conducted in the automated Advion NanoTek® system with 5 mg/ml Tos@SUPPY:2 in acetonitrile at temperatures from 150 – 170 °C and an overall pump rate during the reaction of 60 µl/min. Purity was determined via radio-HPLC and the radiochemical incorporation yield was determined via radio-TLC.
Results
Reducing the water content in conventional elution solutions from 20 % to 1 % only slightly impaired the elution efficiency (93.7 ± 0.0 % compared to 98.0 ± 1.5 %; column: PS-HCO3-; volume: 1 ml). In the same setup, solution WA eluted 98.4 ± 0.8 %, TBAH solution eluted 97.2 ± 0.6 % and TEAB solution eluted 94.4 ± 3.8 % of [18F]fluoride.
Manual [18F]FE@SUPPY:2 radiosynthesis with conventional conditions showed high standard deviations with 250 µl reaction volume (15.7 ± 13.5 % for azeotropic dried eluate from solution A; n=9), while at 500 µl reaction volume, only TBAH solution without azeotropic drying proved to be more efficient than standard conditions (32.8 ± 6.5 % improvement). No improved method could be found for [18F]FE@SUPPY radiosyntheses.
While microfluidic radiosynthesis of [18F]FE@SUPPY:2 achieved 82.2 ± 23.4 % incorporation rate under conventional conditions (solution A, 160 °C, 60 µl/min overall pump rate, azeotropic drying), use of TBAH solution increased the conversion to 97.5 ± 1.2 %.
Conclusion
Several of the investigated elution solutions sufficiently eluted [18F]fluoride. In manual vessel-based radiosyntheses, no reproducible significant increase in reactivity could be achieved. In microfluidic radiosyntheses, TBAH solution increased the incorporation yield and azeotropic drying proved to be essential in all reactions.