Development of Customized [18F]Fluoride Elution Techniques for the Enhancement of Copper-Mediated Late-Stage Radiofluorination

In a relatively short period of time, transition metal-mediated radiofluorination reactions have changed the PET radiochemistry landscape. These reactions have enabled the radiofluorination of a wide range of substrates, facilitating access to radiopharmaceuticals that were challenging to synthesize using traditional fluorine-18 radiochemistry. However, the process of adapting these new reactions for automated radiopharmaceutical production has revealed limitations in fitting them into the confines of traditional radiochemistry systems. In particular, the presence of bases (e.g. K2CO3) and/or phase transfer catalysts (PTC) (e.g. kryptofix 2.2.2) associated with fluorine-18 preparation has been found to be detrimental to reaction yields. We hypothesized that these limitations could be addressed through the development of alternate techniques for preparing [18F]fluoride. This approach also opens the possibility that an eluent can be individually tailored to meet the specific needs of a metal-catalyzed reaction of interest. In this communication, we demonstrate that various solutions of copper salts, bases, and ancillary ligands can be utilized to elute [18F]fluoride from ion exchange cartridges. The new procedures are effective for fluorine-18 radiochemistry and, as proof of concept, have been used to optimize an otherwise base-sensitive copper-mediated radiofluorination reaction.

Additional anhydrous DMF was added (as required) to bring the total solution volume to 1000 µL. The vial was then heated in an aluminum block (Chemglass Part# CG-1991-04) without stirring at 110 ºC for 20 min. After 20 min, the reaction was allowed to cool to room temperature.

Radiolabeling Optimization Screens
4-acetylphenylboronic acid, pinacol ester (APBpin) was used as the [ 18 F]-fluorination substrate for all chemistry optimization screens. The reaction scheme, as well as accompanying tables in each subsection describes the reaction conditions employed; with bold typeface in the reaction scheme denoting the variable tested in each case. All reactant values are expressed in µmol quantities for brevity and simplicity. Red typeface denotes the [ 18 F]fluoride source used. S13

S3.3.1 Reagent Loading Screen
Stock solutions of 4-acetylphenylboronic acid, pinacol ester (APBpin, 100 µmol/mL), Cu(OTf) 2 (200 µmol/mL) and DMAP (500 µmol/mL) in anhydrous DMF were prepared 15-30 min prior to the start of the reaction(s). Appropriate volumes of the reagent solutions were added to colorless borosilicate 4 mL scintillation vials via disposable pipette to obtain the desired reagent quantity/ratios (see Table 3.3.1 below for exact reagent quantities), and additional anhydrous DMF was added to bring the total volume in the vials to 900 µL. The reaction vials were sealed under ambient atmosphere with a PTFE/Silicone septum cap, and 100 µL of [ 18 F]KF/KOTf DMF stock (approx. 500 µCi) was added to the reaction vial through the septum via syringe. The vial was then heated in an aluminum block (Chemglass Part# CG-1991-04) without stirring at 110 ºC for 20 min. The reaction was then allowed to cool, and once the reaction mixture was sufficiently cool to handle, Radio-TLC analysis was conducted to determine radiochemical conversion (RCC in %). Crude reaction mixture was spotted onto standard silica coated glass plates and developed with 1:1 hexane/ethyl acetate in a glass TLC  After this second round of heating, the reaction was allowed to cool, and once the reaction mixture was sufficiently cool to handle, Radio-TLC analysis was conducted to determine radiochemical conversion (RCC in %). Crude reaction mixture was spotted onto standard silica coated glass plates and developed with 1:1 hexane/ethyl acetate in a glass TLC chamber. The RCC was determined by dividing the integrated area under the fluorinated product spot by the total integrated area of all peaks on the TLC plate.

Order of Addition screen (Manual)
Procedure: step 1) Added X; stirred @ 110C for 20 min; cooled; step 2) added O; stirred @ 110C for 20 min; cooled; radio TLC   (Table 3.3.3, entry 6). Conducting this reaction using the optimized reagent ratio (see Figure 2b of the main manuscript) further increased the RCC to 38% and 45%, when the reaction was carried out at 140 °C and 110 °C, respectively (   Step 2 = 20 min). S19

Substrate Scope
Radiofluorination of several Bpin substrates was conducted as described in Section 3.2, however overall reagent loading was halved (see figure below). RCC was determined using radioTLC (in 1:1 hexanes: ethyl acetate eluent) and identity was determined using HPLC (see Section 6 for raw data).

HPLC analysis
Radio-HPLC analyses were conducted using a Shimadzu LC-2010A HT system equipped with a Bioscan B-FC-1000 radiation detector. To prepare samples for HPLC analysis, approx. 100 µL crude reaction mixture was added directly into HPLC vials without further dilution. To confirm identity, crude reaction mixture was spiked with 1 mg/mL 4-fluoroacetophenone in acetonitrile (typically 50 µL standard solution was added to 100 µL crude reaction mixture and briefly agitated). Eluent systems and columns used for HPLC analysis are described below.

Synthesis of [ 18 F]-4-fluoroacetophenone ([ 18 F]FAP) with DMAP elution
Waters QMA-light cartridges was sequentially rinsed with 10 mL ethanol, 10 mL 0.5M preconditioning salt, and 10 mL water prior to use. Dry [ 18 F]DMAP•HF was produced by trapping 18 F from target water on the QMA cartridge followed by elution with 0.5 mL aqueous DMAP (50 µmol, 1 equiv, 100 mM) (vial 1) and azeotropic drying with 1 mL acetonitrile ( Table 5A. Activities at various timepoints during synthesis as well as radiochemical conversions are detailed in Table 5A. Spectroscopic data (radio-TLC and HPLC spectra) and HPLC conditions (column, flow rate, and eluent used) are located in section 5.1.1-5.1.10.

Synthesis of [ 18 F]-4-Fluorophenacyl Bromide ([ 18 F]FPB) with DMAP Elution
Waters QMA-light cartridge was sequentially rinsed with 10 mL ethanol, 10 mL 0.5M preconditioning salt, and 10 mL water prior to use. containing methanesulfonic acid (1000 µmol, 20 equiv) and N-bromosuccinimide (150 µmol) in 1000 µL anhydrous acetonitrile was added to the reactor using Ar push gas, heated to 110 °C for 20 min and subsequently cooled to 50 °C with compressed air. 5 mL of DMF was added to the reactor, the mixture was stirred for ~2 min and was then transferred to an 8 mL sterile product vial with Argon push gas. Total activity of recovered material was measured using a  Table 5C.
Spectroscopic data (radio-TLC and HPLC spectra) and HPLC conditions (column, flow rate, and eluent used) are located in section 5.2.1. S24

Synthesis of [ 18 F]-4-Fluorophenacyl Bromide ([ 18 F]FPB) with DMAP Elution: in-box purification and reformulation.
Waters QMA-light cartridge was sequentially rinsed with 10 mL ethanol, 10 mL 0.5M preconditioning salt, and 10 mL water prior to use. Dry [ 18 F]DMAP•HF was produced by trapping 18 F from target water on the QMA cartridge followed by elution with 0.5 mL aqueous DMAP (50 µmol, 1 equiv, 100 mM) (vial 1) and azeotropic drying with 1 mL acetonitrile (vial 2) at 100 °C. The reactor housing the dry [ 18 F]fluoride was cooled to 50 °C with compressed air, and then a solution (vial 3) containing 4-acetylphenylboronic acid, pinacol ester BPin-1 (50 µmol, 1 equiv) in 500 µL anhydrous DMF was added to the reactor using Ar push gas. The reactor was heated to 110 °C for 5 min and subsequently cooled to 50 °C with compressed air.
A second solution (vial 4) containing Cu(OTf) 2 (8 µmol, 0.16 equiv) and DMAP (50 µmol, 1 equiv) in 500 µL anhydrous DMF was added to the reactor using Ar push gas, heated to 110 °C for 20 min and subsequently cooled to 50 °C with compressed air. Finally, a third solution (vial 5) containing methanesulfonic acid (1000 µmol, 20 equiv) and N-bromosuccinimide (150 µmol, 3 equiv) in 1000 µL anhydrous acetonitrile was added to the reactor using Ar push gas, heated to 110 °C for 20 min and subsequently cooled to 50 °C with compressed air. A solution (vial 6) containing KOH (750 µmol, 15 equiv) in 2 mL of water was added to the reactor to neutralize excess acid, and the mixture was loaded onto a semi-preparative column (HPLC condition C). The [ 18 F]FPB product peak was collected (22.1-24.0 min) into a dilution flask containing 50 mL of water. The diluted product was then passed through a Waters C18 1cc vac cartridge, and the cartridge was rinsed with 10 mL of water (vial 9). Product trapped on the cartridge was eluted with 0.5 mL ethanol (vial 8), followed by 4.5 mL sterile saline buffer into a collection flask. The formulated product was then transferred to an 8 mL sterile product vial with Argon push gas. Total activity of recovered material was 13 mCi as measured using a Capintec dose calibrator. The identity (matched with co-injected FPB standard), specific activity (8,097 Ci/mmol) and radiochemical purity (99%) were determined using HPLC (condition D). HPLC data and HPLC conditions (column, flow rate, and eluent used) are located in section 5.2.2.

Synthesis of [ 18 F]-4-Fluoroacetophenone ([ 18 F]FAP) with Cu(OTf) 2 Elution
Waters QMA-light cartridge was sequentially rinsed with 10 mL ethanol, 10 mL 0.5M potassium trifluoromethanesulfonate, and 10 mL water. Dry [ 18 F]Cu(OTf) 2 •Fwas produced by trapping 18 F from target water on the QMA cartridge followed by elution with 500 µL of aqueous S25 (5.3.1) or methanolic (5.3.2) Cu(OTf) 2 (50 µmol, 6.25 equiv, 100 mM, vial 1), and azeotropic drying with 1 mL acetonitrile (vial 2) at 100 °C. The reactor housing the dry [ 18 F]fluoride was cooled to 50 ⁰C with compressed air and a solution containing 4-acetylphenylboronic acid (8 µmol, 1 equiv, 4 mM), and pyridine (1000 µmol, 125 equiv, 500 mM) in 2 mL anhydrous DMF (vial 3) was added to the reactor by applying Ar push gas through the vial containing the reagent solution. All open valves leading out of the reactor were then closed and the mixture was stirred for 20 min at 140 ºC. The mixture was cooled to 50 ºC with compressed air cooling and 5 mL of DMF was added to the reactor. The mixture was stirred for ~2 min and was transferred to an 8 mL sterile product vial with Ar push gas. The product vial was transferred out of the synthesis module in a lead pig. Total activity of recovered material was measured using a Capintec dose calibrator. Radiochemical conversion (RCC) was determined with radio-TLC (Eluent = 1:1 hexanes: ethyl acetate). Identity of the product was confirmed with HPLC. The activities at various timepoints during synthesis and radiochemical conversions are detailed in Table 5B. Spectroscopic data (radio-TLC and HPLC spectra) and HPLC conditions (column, flow rate, and eluent used) are located in section 5.3.1 and 5.3.2.                    [Note: this compound was very poorly visible in the 200-300 nm range normally utilized for HPLC UV detection and can be seen as an increase in the size of the UV peak immediately prior to the RAD product peak].

Regression Analysis
To develop a predictive model for [ 18 F]fluoride recovery vs. pKa, regression analysis was conducted on the data collected in this experiment using GraphPad software (90% confidence level,