OF JOURNAL LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS J Label Compd Radiopharm 2006; 49: 1731.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.1016
Research Article
Automated radiosynthesis of [18F]SPA-RQ for imaging human brain NK1 receptors with PET
Frederick T. Chin*, Cheryl L. Morse, H. Umesha Shetty and Victor W. Pike
PET Radiopharmaceutical Sciences Section, Molecular Imaging Branch, National Institute of Mental Health, National Institutes of Health, Building 10, Room B3 C346, 10 Center Drive, Bethesda, MD 20892-1003, USA
Summary
[18F]SPA-RQ is an eective radioligand for imaging brain neurokinin type-1 (NK1) receptors in clinical research and drug discovery with positron emission tomography. For the automated regular production of [18F]SPA-RQ for clinical use in the USA under an IND we chose to use a modied commercial synthesis module (TRACERlab FXF-N; GE Medical Systems) with an auxiliary custom-made robotic coolingheating reactor, after evaluating several alternative radiosynthesis conditions. The automated radiosynthesis and its quality control are described here. [18F]SPA-RQ was regularly obtained within 150 min from the start of radiosynthesis in high radiochemical purity (>99%) and chemical purity and with an overall decay-corrected radiochemical yield of 15 2% (mean S.D.; n 10) from cyclotron-produced [18F]uoride ion. The specic radioactivity of [18F]SPA-RQ at the end of synthesis ranged from 644 to 2140 mCi/mmol (23.879.2 GBq/mmol). Copyright # 2005 John Wiley & Sons, Ltd. Key Words: [18F]SPA-RQ; Fluorine-18; automation; radioligand; neurokinin type-1; receptor
Introduction
The neurokinin type-1 (NK1) receptor is acted on by substance P, which has been implicated in several neuropsychiatric disorders such as depression,1,2 schizophrenia,3,4 Parkinsons disease5,6 and Alzheimers disease.5,7 Recently, [18F]SPA-RQ (3; Figure 1) has been developed for imaging brain NK1 receptors in vivo with positron emission tomography (PET) and validated for imaging in human subjects in Europe.810 However, this radioligand had not been used in human subjects in the USA preceding this work. In 2002 we
*Correspondence to: Frederick T. Chin, Molecular Imaging Program, Stanford University School of Medicine, Lucas MRS Center, MC 5484, Stanford, CA 94305-5484, USA. E-mail: [e-mail] Contract/grant sponsor: National Institute of Mental Health
Copyright # 2005 John Wiley & Sons, Ltd.
Received 5 August 2005 Accepted 21 September 2005
18
F. T. CHIN ET AL.
Figure 1. Synthesis of [18F]SPA-RQ aimed to adapt the then briey described11,y radiosynthesis of [18F]SPA-RQ from cyclotron-produced [18F]uoride ion (Figure 1) to allow regular production of [18F]SPA-RQ for clinical use under an INDz in the USA. Five main issues needed to be addressed, namely: (i) optimization of the yield of the labeling agent, [18F]uorobromomethane, from cyclotron-produced [18F]uoride ion, (ii) optimization of the yield of the subsequent 18F-uoromethylation of the phenolic precursor (1), (iii) optimization of deprotection of the intermediate (2), (iv) automation of the radiosynthesis to avoid excessive radiation burden to operating personnel, and (v) ecient formulation of the nal dose. Here we describe a satisfactory automated method with respect to radiochemistry, purication, formulation, and quality control for producing [18F]SPA-RQ for human studies under a United States IND.
Results and discussion
Radiochemistry preparation of labeling agent [18F]SPA-RQ is prepared via alkylation of the protected phenolic precursor 1 with [18F]uorobromomethane.8,10 In the reported procedure8 the labeling
y z
A full account of this radiosynthesis was published later (2004) by Solin et al.8 Investigational New Drug Application to the Food and Drug Administration. J Label Compd Radiopharm 2006; 49: 1731
Copyright # 2005 John Wiley & Sons, Ltd.
AUTOMATED RADIOSYNTHESIS OF [18F]SPA-RQ
19
agent is prepared by treating dibromomethane with cyclotron-produced [18F]uoride ion in acetonitrile12,13 and isolated by distillation11 or gas chromatography14 for entrapment in a second vessel for reaction with 1 in N,N-dimethylformamide (DMF). The decay-corrected radiochemical yield (RCY) of isolated labeling agent is an important factor aecting the overall eciency with which [18F]SPA-RQ may be produced. The highest RCY reported for this labeling agent is 62% by Coenen et al.,12 who used an extended reaction time (65 min), though Bergman et al.14 have regularly achieved an isolated RCY of 25 5% n 17 from an unspecied shorter reaction time. We made various attempts to improve on this yield, using a nitrogen purge to transfer volatile labeling agent into a second vessel containing precursor 1 in DMF. The use of neat dibromomethane, instead of a solution of dibromomethane in acetonitrile, gave [18F]uorobromomethane in high isolated RCY (37%) (Table 1). However, dibromomethane transferred with the labeling agent and undesirably reacts with the precursor 1 by forming a methylene-bridged dimer. Alternative volatile 18F-uoromethylating agents were considered, such as [18F]uoroiodomethane1517 or [18F]uorochloromethane. Attempts to prepare these labeling agents by treating the appropriate dihalomethane precursor with [18F]uoride ion gave inferior RCYs (Table 1). The RCY of [18F]uoroiodomethane was especially low (7.8%) in accord with the experience of Bergman et al.14 (RCY: 5.7 5.5%, n 30), although notably Zhang et al.17 (RCY: 1431%, n 13) and Zheng and Berridge15,16 (RCY: 40 8%) have reported higher yields. Hence, [18F]uorobromomethane, prepared from a solution of dibromomethane (10% v/v) in acetonitrile, remained the labeling agent of choice. In order to reduce the transfer of chemical impurities with the labeling agent to the entrapment vessel, a series of four disposable silica gel Sep-Pak Plus cartridges were inserted into the transfer line, as described by Iwata et al.18 The use of Sep-Paks was considered simpler and more convenient than the alternative of gas chromatographic separation described by Solin et al.8 Under
Table 1. Isolated decay-corrected radiochemical yields (RCYs) of 18F-uoromethylating agents
Labeling agent [18F]Fluorobromomethane [18F]Fluorochloromethane [18F]Fluorobromomethane [18F]Fluoroiodomethane Precursor (ml) CH2Br2 (500) CH2Cl2 (100) CH2Br2 (100) CH2I2 (100) Solvent None MeCN MeCN MeCN Isolated RCYa (%) 37 n 2 22 n 1 27.5 4.5 n 57 7.8 n 1
a All reactions were carried out in a sealed reaction vessel in the presence of K 2.2.2 (5 mg; 13.2 mmol) and potassium carbonate (0.5 mg; 3.6 mmol) at 1108C for 10 min. Volatile radioactive product was transferred by nitrogen stream into a second vessel over the course of 430 min until maximal radioactivity had been collected.
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2006; 49: 1731
20
F. T. CHIN ET AL.
these conditions the RCY of isolated labeling agent was 27.5 4.5% (range: 16.542.5%; n 57). Radiochemistry labeling and deprotection The RCY of the intermediate 2 from the alkylation of 1 with [18F]uorobromomethane (Figure 1) and the eciency of the deprotection of 2 are also major factors in the overall eciency of [18F]SPA-RQ production. In the reported procedure,8 cesium carbonate was used as base in DMF as solvent for the alkylation reaction. We explored other bases, (e.g. NaOH, t-butylammonium hydroxide (TBAOH) or potassium carbonate-18-crown-619) in attempts to improve the RCY of this reaction. In these experiments the RCY of the intermediate 2 was not measured directly but 2 was deprotected with triuoroacetic acid (TFA) to give [18F]SPA-RQ (Figure 1) and its overall RCY from [18F]uoride ion measured usually before or exceptionally after formulation (Table 2). In general, such deprotection reactions are very rapid.20 Solin et al.8 implemented deprotection of the intermediate 2 with TFA in 2 min
Table 2. RCYs of [18F]SPA-RQ from [18F]uoride ion under various conditions for alkylation of 1 and deprotection of 2
Precursor 1 amount (mg; mmol) 0.2; 0.2; 0.2; 0.3; 0.3; 0.3; 0.2; 0.3; 0.3; 0.4 0.4 0.4 0.6 0.6 0.6 0.4 0.6 0.6 Base; amount (mmol) Cs2CO3; 4 Cs2CO3; 4 Cs2CO3; 4 Cs2CO3; 4 Cs2CO3; 31 Cs2CO3; 31 TBAOH; 4 NaOH; 10 K2CO3-18-cr.6d; 15-15 K2CO3-18-cr.6; 15-15 K2CO3-18-cr.6; 15-19 K2CO3-18-cr.6; 15-19 K2CO3-18-cr.6; 15-19 Alkylation timea (min) 15 20 30 30 15 15 15 15 15 15 20 20 20 Deprotection timeb (min) 0.5 0.5 0.5 0.5 0.5 1.0 0.5 0.5 0.5 5 5 5 5 RCY of [18F]SPA-RQc (%) 3.7 5.7 3.6 5.4 6.6 6.1 4.2 5.5 3.9 10.6 17.0 8.4 3.6 n 35e,f 15 2.0 n 10e,g
0.3; 0.6 0.3; 0.6 0.3; 0.6 0.3; 0.6
a
In DMF (0.75 ml) heated at 1108C. At 1108C. c Overall for HPLC separated product from [18F]uoride ion, unless otherwise indicated, preceding formulation unless otherwise indicated. d 18-crown-6. e After formulation for intravenous injection. f Semi-automated preparations. g Automated preparations.
b
Copyright # 2005 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2006; 49: 1731
AUTOMATED RADIOSYNTHESIS OF [18F]SPA-RQ
21
at ambient temperature. However, we found that heating of the intermediate 2 with TFA for 5 min at 1108C was necessary to ensure complete deprotection (Table 2). These became the regular deprotection conditions. Among the bases used for the alkylation reaction (Table 2), potassium carbonate (2 mg; 15 mmol)-18-crown-6 (5 mg; 19 mmol) (heated with 1 (0.6 mmol) in DMF (0.70 ml) for 20 min at 1108C) gave consistently superior RCYs of [18F]SPARQ and became the base of choice. Shorter reaction time gave lower RCY (Table 2). It should be noted that once the alkylation reaction was complete, it was necessary to remove most of the DMF before the deprotection step. DMF was almost completely removed over a 79 min period by heating the vessel while purging it with nitrogen. Evaporation to dryness was routinely avoided since this tended to produce a byproduct that was dicult to separate from [18F]SPA-RQ. Automation of radiosynthesis No automated apparatus has been commercially available for the radiosynthesis of the labeling agent, [18F]uorobromomethane, or for its application in preparing radioligands, such as [18F]SPA-RQ. For the radiosynthesis of [18F]uorobromomethane, we opted to use the TRACERlab FXF-N apparatus oered by GE Medical Systems, since this apparatus is specically designed to perform automated reactions with cyclotron-produced [18F]uoride ion. The apparatus had to be recongured to allow transfer of the labeling agent into a second vessel for the labeling reaction (Figure 2). Initially, [18F]uorobromomethane synthesis and entrapment were demonstrated with the TRACERlab FXF-N module connected to an ice-cooled vial containing the precursor 1 in DMF (RCY: 26 7%; n 55). Deprotection of the generated intermediate 2, HPLC separation and formulation provided [18F]SPA-RQ ready for administration to human subjects in acceptable but rather variable overall RCY (8 4%; n 35) (Table 2). In order to achieve consistent production runs with greater radiation safety, a robotic Peltier coolingheating device (Figure 3) was designed, constructed and implemented as an auxiliary to the TRACERlab FXF-N module. This device was externally controlled to perform the alkylation reaction in a second reactor (reactor #2). Coupling of this device with the TRACERlab FXF-N maintained almost the same RCY for isolated [18F]uorobromomethane (26 5%; n 4). The same reactor was also used for the deprotection reaction and the preparation of the [18F]SPA-RQ for HPLC purication. The use of two reverse phase columns in series was eective in separating [18F]SPARQ from a multitude of low level chemical impurities (Figure 4). Following HPLC, the separated [18F]SPA-RQ was automatically and eciently formulated by adsorption on a C-18 Sep-Pak, elution with acidied
Copyright # 2005 John Wiley & Sons, Ltd. J Label Compd Radiopharm 2006; 49: 1731
22
F. T. CHIN ET AL.
Figure 2. Automated apparatus for entrapment of [18F]uorobromomethane for the preparation of [18F]SPA-RQ
ethanol, neutralization with sodium bicarbonate in saline and sterile-ltration through a Millex-MP lter. Acidication of the ethanol with acetic acid ensured ecient elution of [18F]SPA-RQ from the Sep-Pak. The Millex-MP lter was superior to alternative lters in that it adsorbed only a very small percentage (54%) of the [18F]SPA-RQ. The fully automated process gave an improved and more consistent overall RCY of puried and formulated [18F]SPA-RQ from [18F]uoride ion (15 2%; n 10) compared to the partially automated procedure (Table 2). This arrangement also reduced the radiation exposure to the production chemist to an acceptable level of about 70 mrem when using a starting radioactivity of 500 mCi (18.5 GBq). The total time required for the radiosynthesis, purication and formulation was 150 min. Quality control of nal dose The radiochemical purity of [18F]SPA-RQ preparations exceeded 99% at the end of synthesis and this purity was maintained for at least 6 h (Figure 5). Preparations were also greater than 99% chemically pure, based on the area of the absorbance peak for carrier relative to those of impurities} in the HPLC
}
Carrier SPA-RQ and impurities are assumed to have the same extinction coecient at l 254 nm. J Label Compd Radiopharm 2006; 49: 1731
Copyright # 2005 John Wiley & Sons, Ltd.
AUTOMATED RADIOSYNTHESIS OF [18F]SPA-RQ
23
Figure 3. Conguration of TRACERlab FXF-N used to synthesize [18F]uorobromomethane analysis when earlier eluting low level compounds introduced with the formulation vehicle are discounted. Specic radioactivities were determined from the HPLC analysis, which was calibrated for mass of SPA-RQ. LCMSMS was used to verify carrier identity rapidly (53 min) preceding
Copyright # 2005 John Wiley & Sons, Ltd. J Label Compd Radiopharm 2006; 49: 1731
24
F. T. CHIN ET AL.
Figure 4. Chromatogram from the HPLC purication of [18F]SPA-RQ
Figure 5. Chromatogram from the HPLC quality control of [18F]SPA-RQ release of the radioactive product for use. Preparations were free of any signicant level of contaminating acetonitrile (40.04% w/v), had acceptable pH (within range 4.57.5) and were sterile and apyrogenic.
Copyright # 2005 John Wiley & Sons, Ltd. J Label Compd Radiopharm 2006; 49: 1731
AUTOMATED RADIOSYNTHESIS OF [18F]SPA-RQ
25
Conclusion
An automated method for the regular and consistent production of [18F]SPARQ for safe intravenous injection into human subjects was accomplished by coupling a commercial automated apparatus with a custom-made device.
Experimental
Materials (2S,3S)-1-t-Butoxycarbonyl-2-phenyl-3-[2-hydroxy-5-(50 -triuoromethyl-tetrazol-1-yl)phenylmethyleneamino]piperidine (1) and authentic SPA-RQ ((2uoromethoxy-5-(5-triuoromethyl-tetrazol-1-yl)-benzyl)((2S,3S)2-phenylpiperidin-3-yl)-amine) (3) were supplied by Merck Research Laboratories and were of >99% chemical purity. The following were obtained commercially, as indicated: sodium chloride injection (0.9% USP; APP Inc.; Schaumburg, IL), sodium bicarbonate injection (8.4% USP; Abbott Laboratories, North Chicago, IL), ethanol (dehydrated USP; Abbott Laboratories), Millex-MP and Millex-GV lters (Millipore; Bedford, MA), silica gel and C-18 Sep-Pak Plus cartridges (Waters Corporation; Milford, MA), Luna C-18 columns (Phenomenex; Torrance, CA), LAL test kit vials and control standard endotoxin (Cape Cod Associates; Falmouth, MA) and ColorpHast1 indicator strips (EM Science; Gibbstown, NJ). All other reagents and solvents were obtained from Aldrich (St. Louis, MO).
General methods No-carrier-added [18F]uoride ion was prepared by the 18O(p,n)18F nuclear reaction on a PETtrace cyclotron (GE Medical Systems; Milwaukee, WI). 18 O-Enriched water (1.8 ml, >95% isotopic enrichment) was added to a titanium target equipped with titanium foils. The [18O]water was bombarded with 18 MeV protons (typically 20 mA for 120 min). Portions of the irradiated water (0.020.5 ml), containing up to 500 mCi (18.5 GBq) of [18F]uoride ion were used for individual experiments. Radioactivity measurements were carried out with an AtomlabTM 300 dose calibrator (Biodex Medical Systems; Shirley, NY), which was calibrated daily using Cs-137 and Co-57 sources. Each radiochemical yield (RCY) is and decaycorrected reported as a mean S.D. HPLC separations and analysis were performed on Gold HPLC modules (System Gold 126 gradient solvent module coupled with a variable wavelength 166 UV absorbance detector; Beckman Coulter; Fullerton, CA). Radioactivity in HPLC eluates was detected with a pin-diode radioactivity detector (FlowCount; Bioscan; Washington, DC).
Copyright # 2005 John Wiley & Sons, Ltd. J Label Compd Radiopharm 2006; 49: 1731
26
F. T. CHIN ET AL.
Automated radiosynthesis apparatus description The automated apparatus for radiosynthesis was housed in a lead-shielded hot-cell. The apparatus comprised a modied commercial device (TRACERlab FXF-N; GE Medical Systems) (Figure 2), which is designed for performing single-step reactions with cyclotron-produced [18F]uoride ion, linked to a custom-made robotic Peltier coolingheating device (Figure 3) and to HPLC. The TRACERlab FXF-N, as supplied, has a glassy carbon reaction vessel (reactor #1; Figure 2) and features automated facilities for drying aqueous [18F]uoride ion by cycles of addition and evaporation of acetonitrile, for forming a complex of this ion with K+-Kryptox1 2.2.2 (4,7,13,16,21,24hexaoxa-1,10-diazabicyclo-[8,8,8]hexacosane; K 2.2.2) and for conducting single-pot radiochemistry. The apparatus also has facility for heating the glassy carbon vessel up to 1358C for reactions and solvent evaporation. A PC, loaded with TracerLab software, controls and monitors the function of the apparatus. The TRACERlab FXF-N apparatus was re-congured as shown in Figure 2, mainly by switching valve functions from their original purpose. The major changes and their purposes are further explained in the following section on operation of the apparatus. The robotic Peltier coolingheating device, incorporates a second reaction vessel (reactor #2; Figure 3) that may be heated or cooled and penetrated through its top septum with a variety of needles. The device is mainly constructed from stainless steel and aluminum parts on an aluminum platform. Thus, an aluminum wheel (diameter 14 cm), equipped with six aluminum clamps each holding two needles (length: 0.6752.00 in; diameter 2518 gauge), was mounted on a stainless steel driveshaft. The driveshaft rides on seal ball bearings under power from an electric-servo motor (HITEC, 6.0 V, 0.19 s/608 speed, 11.5 kg cm torque). The wheel was indexed with a stainless steel 6-position Geneva-gear (WM Berg; East Rockaway, NY). Micro-limit switches were used to control the home position and index location of the wheel. The wheel assembly has a Lexan needle guard and moves up/down vertically on a stainless steel shaft under pneumatic power; micro-limit switches regulate the motion of the pneumatic cylinders. Electrically-activated air solenoids (Norgren; Norgren.com), pneumatic actuators (Clippard Instrument Labs Inc., OH) on aluminum brackets and quick-disconnect air-line ttings (Clippard Instrument Labs Inc.) direct air ow through the pneumatic assembly under remote electronic control. A modied Peltier device (which may cool to 58C) and dual rod heaters (which may heat to 2008C) regulate the temperature of the Teon-insulated vial cavity, which is monitored with an attached thermocouple. Two radiation detectors (Bioscan), held by a uni-directional clamping device, externally monitor the radioactivity
Copyright # 2005 John Wiley & Sons, Ltd. J Label Compd Radiopharm 2006; 49: 1731
AUTOMATED RADIOSYNTHESIS OF [18F]SPA-RQ
27
in the reaction vial when it resides in the vial cavity. A stainless steel pneumatic arm with a nylon tip enables the reaction vial to be held in place during removal of wheel assembly needles from the vial via the septum. Nitrogen ow to the secondary reactor was controlled by a ow controller connected to the Local Set Point Module using a Top Connector Socket (Alicat Scientic Inc.; Tucson, AZ). Automated radiosynthesis apparatus general features of operation General features of operation of the automated radiosynthesis apparatus (Figures 2 and 3) are as follows. Valve 2 (Figure 2) allows [18F]uoride ion, K 2.2.2 and potassium carbonate in [18O]water, followed by acetonitrile, to be introduced into the glassy carbon reaction vessel (reactor #1). Then azeotropic drying is performed through two subsequent cycles of acetonitrile addition and evaporation. The PEEK tee between the outlet lines of valves 1 and 2 allows acetonitrile from the rst azeotropic drying step to be transferred through valve 1, under reduced pressure, to rinse out any residual activity from the preceding loading of activity into the reaction vessel. Acetonitrile is added through valve 5 for the second azeotropic drying step. After the [18F]uoride complex is dry, precursor (dihalomethane) in acetonitrile is added through valve 3 for the synthesis of the radioactive labeling agent. After this synthesis, the glassy carbon reaction vessel is washed twice with acetonitrile introduced via valves 4 and 6. A conical vial (10 ml) is present between valves 12 and 26 and used to store residual radioactivity in a shielded environment after the synthesis. A Teon tee introduced before valve 20 provides for a new nitrogen source which is regulated with an adjustable ow meter (150 mm Cole-Palmer) and switched on or o by valve 11. Valve 11 is connected to the normally closed side of three-way valve 14. Labeling agent is transferred out of the glassy carbon reaction vessel in a nitrogen gas stream (ow rate: 3035 ml/min), released by valve 11, and allowed to bubble through the reaction mixture and out through valve 13. Labeling agent is puried by passage through a series of four silica gel Sep-Paks and trapped in a cooled secondary reaction vessel (reactor #2; Figure 3) containing des-uoromethyl precursor (1), potassium carbonate and 18-crown-6 in DMF with the Peltier device in position #1 (Inset 1; Figure 2). Radioactivity transfer was monitored with detectors (Bioscan) at the terminus of the silica gel cartridges and at the secondary reactor (Figure 3). The rotating wheel assembly was put in position #5 (Inset 2, Figure 2), to reconnect the secondary reactor to the TRACERlab FXF-N synthesis module at the HPLC loop. The crude reaction mixture was loaded onto the loop by pressurizing the secondary reactor with an external syringe (20 ml).
Copyright # 2005 John Wiley & Sons, Ltd. J Label Compd Radiopharm 2006; 49: 1731
28
F. T. CHIN ET AL.
Synthesis of [18F]uorohalomethanes standard automated procedure Cyclotron-produced [18F]uoride ion in [18O]water was delivered into a vial containing K 2.2.2 (5 mg; 13.2 mmol) and potassium carbonate (0.5 mg; 3.6 mmol) in acetonitrilewater (9:1 v/v; 100 ml). The [18F]uoride ion mixture was transferred to reactor #1 of the automated radiosynthesis module, chased by acetonitrile (1 ml). A further amount of acetonitrile was added (1 ml) and the mixture taken to dryness at 908C under reduced pressure with nitrogen ow. A further cycle of acetonitrile (2 ml) addition and evaporation was performed. Dihalomethane (CH2Cl2, CH2Br2 or CH2I2; 100 ml) in anhydrous acetonitrile (1.0 ml) was added to the dry [18F]uoride ion-K 2.2.2-potassium carbonate complex which was then heated to 1108C for 15 min. The reaction vessel was then cooled to 358C. Nitrogen gas was used to transfer the volatile [18F]uorohalomethane through a series of four silica gel cartridges (Sep-Pak Plus) and into a pre-cooled vessel, a V-vial (volume 1 ml) with a crimp-seal silicone-Teon septum cap. Synthesis of [18F]SPA-RQ standard automated procedure The glass reaction vessel for reactor #2 was charged with 1 (0.3 mg; 0.6 mmol), potassium carbonate (2 mg; 15 mmol) and 18-crown-6 (5 mg; 19 mmol) in DMF (0.8 ml) and cooled to 58C. [18F]Fluorobromomethane was transferred to this solution under computer control from the TRACERlab FXF-N module. Radioactivity transfer was monitored by two external Bioscan detectors (Figure 2) with Hotcell software and was stopped when radioactivity in the vessel reached a maximum. The vessel was then heated at 1108C for 20 min and the DMF removed almost to dryness with a nitrogen stream. Triuoroacetic acid (TFA; 0.1 ml) was added to the dry radioactive residue and heated at 1108C for 5 min. The acidic mixture was diluted with aqueous mobile phase A (waterMeCNTFA, 95:5:0.1 by vol.; 0.9 ml) and injected remotely onto HPLC for purication on two Luna C-18 columns (10 250 mm; 10 mL) connected in series and eluted over 40 min at 9 ml/min starting with mobile phase A changing linearly to AMeCN (78:22 v/v) over 20 min. Eluate was monitored for absorbance at 254 nm and radioactivity (Figure 4). Puried [18F]SPA-RQ was collected and diluted with sterile water (85100 ml). Formulation of [18F]SPA-RQ standard procedure A Sep-Pak Plus C-18 cartridge was eluted at about 10 ml/min with ethanol (10 ml) and then water (10 ml). Under TRACERlab FXF-N module control, the aqueous solution of puried [18F]SPA-RQ was passed through the prepared cartridge. The loaded cartridge was then washed with sterile water (5 ml). Finally, [18F]SPA-RQ was eluted from the cartridge with ethanol (USP; 0.95 ml) containing glacial acetic acid (2 ml). Sodium chloride injection (0.9%
Copyright # 2005 John Wiley & Sons, Ltd. J Label Compd Radiopharm 2006; 49: 1731
AUTOMATED RADIOSYNTHESIS OF [18F]SPA-RQ
29
USP; 9 ml) with sodium bicarbonate injection (8.4% USP; 40 ml) were pushed through the Sep-Pak C-18 cartridge and combined with the ethanolic solution of [18F]SPA-RQ. This solution was thoroughly mixed and pushed through a sterile Millex-MP lter (25 mm) into a sterile dose vial (10 or 30 ml). QC analysis of [18F]SPA-RQ An aliquot (1 ml) of the formulated [18F]SPA-RQ was removed from the dose vial with a pre-attached sterile syringe (1 ml size). A portion (0.8...