Investigations on the radionuclide supply of fluorine-18 for the synthesis of ${}^{18}$F-labelled PET tracers using automated fluid synthesis technology

Positron emission tomography (PET) is used for the diagnosis of metabolic diseases with the application of specific ${}^{18}$F radiopharmaceuticals. In addition to established radiopharmaceuticals such as [${}^{18}$F]FDG, many novel radiopharmaceuticals with peptides (e.g. bombesin) as tracer molecules are also being developed. With regard to synthesis, various requirements are placed on radiochemistry: The synthesis time is limited due to the half-life of the radionuclide. The radiopharmaceuticals should have a high molecular activity, i.e. the amount of precursor used should be as small as possible. The dose-on-demand principle (DOD) requires the increasing use of automated synthesis modules. This work deals with the radionuclide preparation of fluorine-18 for nucleophilic syntheses without the time-consuming azeotropic drying step. The elaborated method for the preparation of [${}^{18}$F]fluoride was applied to the radiosyntheses of 2-[${}^{18}$F]FFluorodeoxyglucose and 2- [${}^{18}$F]fluoroethylcholine and compared with the azeotropic drying method. Furthermore, the syntheses were carried out using an automated fluid flow reactor and compared to syntheses in conventional reaction vials with respect to the amount of precursor required for high yields. In addition, a potential new ${}^{18}$F-PET tracer was synthesised using bombesin as the tracer molecule and labelled with fluorine-18 using a silicon building block (SIFA method). The method for providing [${}^{18}$F]fluoride without azeotropic drying could be successfully applied to the radiosyntheses of [${}^{18}$F]FDG and [${}^{18}$F]FEC. The advantage of azeotropic drying occurred only at precursor amounts below 0.5 mg. It could be shown that the yields are higher in the fluid flow reactor than in the reaction vial for the same amounts of precursor. In addition, the amount of precursor could be minimised significantly further in the syntheses with the fluid flow reactor than in the reaction vial. The one-step synthesis of [${}^{18}$F]FDG and the two-step synthesis of [${}^{18}$F]FEC including preparation were successfully transferred to the fluid flow reactor. The fully automated synthesis of [${}^{18}$F]FDG yielded 84 +/- 9% (radiochemical yield 59 +/- 8%). The fully automated synthesis of [${}^{18}$F]FEC required a synthesis time of 1 hour and yielded 58 +/- 2% (radiochemical yield 48 +/- 3%). The synthesis of a novel ${}^{18}$F-PET tracer could also be realised. In 6 synthesis steps, a derivative of bombesin and a silicon building block were coupled to the starting compound L-cysteic acid. The final product was labelled with fluorine-18 in the post-labelling variant by isotope exchange at the silicon building block. Radiolabelling in the fluid flow reactor yielded constant yields of 70% up to 0.02 mg precursor, while the yield in the reaction vial decreased continuously and fell below 50% at 0.02 mg.