Chemicals and Reagents
All chemicals and reagents were obtained from commercial vendors and used without further purification.
Automated Purification and Formulation of [13N]Ammonia
[13N]Ammonia is produced on a modified GE Tracerlab FXFDG synthesis module which was replumbed for the purification and formulation of [13N]ammonia. The graphic user interface is shown in Figure 1 and the purification and formulation steps are further detailed below. The QMA chloride and CM cartridges are prepared day of use by washing with 10 mL of sterile water for injection, USP and then capping the cartridges with sterile male/female luer caps. Prior to the first sub-batch of the day, the synthesis box vacuum is tested and the system fluid paths are washed with sterile water for injection, USP. The cyclotron [13N]ammonia target prior to the day’s use is primed with fresh 5 mM ethanol solution.
- [13N]Ammonia is produced on-site with a GE PETrace 860 cyclotron using the Wieland et al method3 of irradiation of 5 mM ethanol in water in-target synthesis method using a GE niobium body target with HAVAR/Niobium double target window or GE silver body target with HAVAR window (GE Medical Systems, Uppsala, Sweden). The target solution is pressurized to approximately 0.62 MPa (90 psi) for the silver body target or to 1.31 MPa (190psi) for the niobium body target with helium gas. The target is bombarded with 16.5 MeV protons using a beam current ranging from 15 to 50 µA using bombardment times between 5 and 30 minutes. For the [13N]ammonia batches cited in Table 1, 50 µA, 30 minute bombardments were used.
- [13N]Ammonia is transferred from the cyclotron target via 0.45 MPa (65 psi) helium overpressure to the automated synthesis unit and through an in-line anion exchange column (Waters, QMA Chloride) to remove any anionic impurities, such as [18F]fluoride.
- Using vacuum, the [13N]ammonia in water is trapped on a cation exchange column (Waters, Accell CM) to quantitatively trap [13N]ammonia.
- The [13N]ammonia is released from the cation exchange column using 8mL of 0.9% sodium chloride for injection, USP.
- The formulated [13N]ammonia in 0.9% sodium chloride for injection is then transferred through a 1/16” PFA line via 0.1MPa (14.5 psi) nitrogen overpressure to a ISO Class 5 isolator for sterile filtration through a vented 0.22µ polyethersulfone (PES) membrane filter (B Braun) into a vented 30mL sterile empty vial (ALK OKC Allergy Labs, Hollister Stier or Huayi Isotopes).
The total time of the purification, formulation and sterile filtration of the [13N]ammonia takes approximately 5 minutes.
Post-formulation in between sub-batches, the fluid pathways in the system unit and the transfer line are washed with sterile water for injection, USP and blown to dryness with nitrogen gas.
Quality Control of [13N]Ammonia
The quality control of [13N]ammonia was performed to ensure the PET drug product met the specifications in Table 1 to satisfy FDA and USP regulatory requirements. Due to the production of [13N]ammonia via in-target 5 mM ethanol solution, the EP/USP test for residual aluminum was not required. The TLC method was validated against the major potential radiochemical impurities, [13N]NOx and [18F]fluoride as detailed in Table 2 and cross-validated against the compendial HPLC method.
A thin layer chromatographic system was developed that uses a diethylaminoethyl cellulose (DEAE-C) stationary phase. The DEAE-C stationary phase of the chromatography system was chosen due to its ability to attract anionic species. The DEAE-C strip (J.T. Baker) is 1.5 cm × 8 cm and the mobile phase is composed of methanol: water 75:25. The Rf of [13N]ammonia is 0.7 – 0.9 and the major impurities of [13N]NOx and [18F]fluoride are retained at the origin (Rf = 0). A 0.5µL spot of [13N]ammonia is applied to the left hand side of the origin of the TLC strip via pipette and a 0.5µL spot of 100mg/mL ammonium chloride reference standard, USP is applied to the right hand side of the origin of the TLC strip. Both spots were allowed to dry prior to the TLC strip development in the mobile phase of methanol: water 75:25. The time for TLC strip development is approximately 8-10 minutes in a tightly sealed development chamber. Following development of the TLC strip, the strip was allowed to dry and was then counted using an AR-2000 radio-TLC plate reader (Eckert and Ziegler). Originally, the radiochemical identity was confirmed with 100mg/mL ammonium chloride USP reference standard which was visualized with a combination of spray the TLC strip with iodoplatinate reagent, which was allowed to develop for 10 min followed by placement in an iodine chamber for approximately 10 min to help highlight the ammonium chloride spot. The ammonium chloride spot appears as an orange brown spot against a light maroon background (Figure 3a).
The radio-TLC method has since been simplified with the use of resazurin (Millipore Sigma), a visible dye which is used as a marker of system suitability eliminating the need to use ammonium chloride and the complicated development process of iodoplatinate reagent and iodine chamber to visualize the standard. Additionally, the visible dye aids the operator by providing a visual pink-purple streak indicating proper TLC development. A 0.5µL spot of [13N]ammonia is applied to the left hand side of the origin of the TLC strip via pipette and a 0.5µL spot of resazurin dye (1mg/mL) is applied to the right hand side of the origin of the TLC strip. Both spots were allowed to dry prior to the TLC strip development in the mobile phase of methanol: water 75:25. The other components of the TLC assay remain the same as described above. The Rf of resazurin is 0.43 – 0.63 and an example TLC strip is shown in Figure 3b. Table 2 below contains the validation data of the [13N]ammonia TLC assay using resazurin as a system suitability marker. For system suitability purposes, we require the front of the resazurin peak to be 34 – 50 mm.
Complete quality control testing and the [13N]ammonia test specifications for five high activity stability batches is described in Table 1. For routine clinical production, a quality control (QC) sub-batch is performed prior to manufacturing any sub-batches for patient use. The routine clinical QC sub-batch QC process takes approximately 25 minutes. All of the QC tests are detailed below which are performed on the quality control sub-batch, except for the periodic quality indicating tests (PQIT), which are performed at their defined testing periods. For the patient sub-batches of [13N]ammonia, partial quality control testing is performed, which is composed of final product vial visual inspection, product assay, and sterile filter integrity test.
Sterile filter integrity is performed using a manual bubble point test using a variable pressure gas source with a calibrated pressure gauge according to filter manufacturing directions. Visual inspection of the final product vial is performed by a qualified operator observing the vial through the lead glass window of the dispensing hot cell to verify vial integrity as well as to ensure the solution is clear and particulate free. The pH testing was performed using two pH strips (0-6, 2-9, EMD Millipore) and comparing the result to pH strips spotted with the closest US NIST traceable pH reference standards. Radio-TLC is performed to determine radiochemical purity and identity as described above using the radio-TLC method. Sterility testing is performed within 30 hours of end of synthesis of the QC sub-batch using a validated modification of USP <71> direct inoculation method of 0.1 – 0.3mL of [13N]ammonia into TSB and FTM hungate 10mL sterility tubes. Bacterial Endotoxin testing is performed using the Endosafe Nexgen PTS system (Charles River). Residual solvent testing for ethanol content (ICH Class III solvent), is a periodic quality indicating test (PQIT) performed at least quarterly using a GC (Agilent 7890) with direct split injection (15:1) onto a USP G16 wax column (Agilent DB Wax ETR, 30m x 0.25mm x 0.5µm), using hydrogen as a carrier gas (1.3mL/min) and FID detector. GC oven at time of injection is 40℃, and then ramps 40℃/min to 110℃, where it holds for 1 minute (3.25min run time). Ethanol elutes at approximately 2.7 minutes. Radionuclidic identity is performed as a PQIT at least annually using a high purity germanium detector system with Genie software (Mirion) which automatically detects and identifies the 511keV, 1022keV, Compton scatter photopeaks as well as any unknown photopeaks. Radionuclidic purity is an annual PQIT which is performed using the high purity germanium detector system using a two-hour count to be able to quantitate 3.7Bq (100pCi). The Genie software automatically calculates the amount of the known radionuclidic impurities feasible from the target body and target windows as well as flags any identified unknown photopeaks for further analysis and identification.
Synthesis of [13N]Ammonia Major Radiochemical Impurities, [13N]NOx and [18F]Fluoride
[13N]NOx was produced by cyclotron bombardment of high purity water using identical irradiation conditions as described above for [13N]ammonia. [13N]NOx was used for the methods validation without further purification. [18F]Fluoride was produced by cyclotron bombardment via (p,n) reaction of ≥98% enriched [18O]water (Rotem or Taiyo Nippon Sanso) using the GE Niobium [18F]fluoride target with variable beam currents up to 65µA. [18F]Fluoride was used for the methods validation after allowing for decay of [13N]-species.