Automated production at the curie level of no-carrier-added 6-[18F]fluoro- l -dopa and 2-[18F]fluoro- l -tyrosine on a FASTlab synthesizer

Article abstract of DOI:10.1002/jlcr.3291

An efficient, fully automated, enantioselective multi-step synthesis of no-carrier-added (nca) 6-[¹⁸F]fluoro-L-dopa ([¹⁸F]FDOPA) and 2-[¹⁸F]fluoro-L-tyrosine ([¹⁸F]FTYR) on a GE FASTlab synthesizer in conjunctio

Full text of DOI:10.1002/jlcr.3291

Research Article
Received 15 December 2014,
Revised 6 March 2015,
Accepted 1 April 2015
Published online 26 May 2015 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3291
Automated production at the curie level of
18
no-carrier-added 6-[ F]fluoro-L-dopa and 2-
18
[ F]fluoro-L-tyrosine on a FASTlab synthesizer
a
a
b
a
b
a
*
C. Lemaire, L. Libert, X. Franci, J.-L. Genon, S. Kuci, F. Giacomelli,
and A. Luxena
An efficient, fully automated, enantioselective multi-step synthesis of no-carrier-added (nca) 6-[18F]fluoro-L-dopa ([18F]
FDOPA) and 2-[18F]fluoro-L-tyrosine ([18F]FTYR) on a GE FASTlab synthesizer in conjunction with an additional high-
performance liquid chromatography (HPLC) purification has been developed. A PTC (phase-transfer catalyst) strategy was
used to synthesize these two important radiopharmaceuticals. According to recent chemistry improvements, automation
of the whole process was implemented in a commercially available GE FASTlab module, with slight hardware modification
using single use cassettes and stand-alone HPLC. [18F]FDOPA and [18F]FTYR were produced in 36.3 3.0 % (n = 8) and
50.5 2.7 % (n = 10) FASTlab radiochemical yield (decay corrected). The automated radiosynthesis on the FASTlab module
requires about 52 min. Total synthesis time including HPLC purification and formulation was about 62 min. Enantiomeric
excesses for these two aromatic amino acids were always >95 %, and the specific activity of was >740 GBq/μmol. This
automated synthesis provides high amount of [18F]FDOPA and [18F]FTYR (>37 GBq end of synthesis (EOS)). The process, fully
adaptable for reliable production across multiple PET sites, could be readily implemented into a clinical good manufacturing
process (GMP) environment.
Keywords: automation; F-18, [18F]FDOPA; 2-[18F]fluoro-L-tyrosine; GMP; solid-phase extraction
positron emission tomography centers using the same
synthesis approach, and
• give the opportunity to synthesize from a single preparation
multiple doses for multi-sites distribution.
Introduction
[18F]Labeled fluoro aromatic amino acids (18FAA) such as 6-[18F]
fluoro-L-dopa and 2-[18F]fluoro-L-tyrosine are potential
radiolabeled probes for molecular imaging with positron
Among the commercially available synthesizers, the FASTlab
multi-tracer platform from GE healthcare and the use of
disposable components (single use cassette with syringes,
reactor, and reagents) seem well adapted to GMP human
production and large-scale synthesis. As the cassette mounted
on the module at the beginning of the synthesis is easily
eliminated at the end of the process, the risk of cross
contamination between two preparations is reduced.
Nevertheless, implementation of all the steps of the synthesis
(labeling, reduction, halogenation, alkylation, hydrolysis, and
purification; Figure 1) on the commercial FASTlab synthesizer is
challenging because the cassette contains only 25 positions with
some of them occupied by syringes and reagent barrels.
Through a collaboration with GE, we had the possibility to obtain
a FASTlab Developer Sequence (open software) and access to
each cassette component separately (Figure 2). The standard
emission tomography. Several clinical applications with these
two radiopharmaceuticals are described in the literature. [18F]
FTYR is used for oncology studies1,2 and [18F]FDOPA to image
3,4
the presynaptic dopaminergic function
as well as a variety
of neuroendocrine tumors.5,6
Recently, a phase-transfer catalyst strategy taking into account
the advantages of the no-carrier-added (nca) [18F]fluoride
production has been proposed in our laboratory for the
synthesis of these compounds (Figure 1).7
However, despite recent chemistry improvements
7,8
that
have allowed to simplify this very complex synthesis, preparation
of these 18FAA via this nca approach stays challenging for
untrained chemists. For wider use of these radiopharmaceuticals,
automation of the synthesis is then a prerequisite that should
• allow the standardization of the preparation,
• simplify the task of the different operators leading to less
failure in the synthesis,
• avoid radioprotection problems by handling multiple GBq of
aCyclotron Research Center, Université de Liège, Liège, Belgium
bGE Healthcare MDx Chemistry Systems, Ans, Belgium
[18F]fluoride,
• ease good manufacturing practice (GMP) synthesis and their
controls,
*Correspondence to: C. Lemaire, Cyclotron Research Center, Université de Liège,
Belgium.
E-mail: christian.lemaire @ulg.ac.be
• provide reliable and robust synthesis of 6-[18F]fluoro-L-dopa
and 2-[18F]fluoro-L-tyrosine with reproducible yields in all
J. Label Compd. Radiopharm 2015, 58 281–290
Copyright © 2015 John Wiley & Sons, Ltd.

C. Lemaire et al.
Figure 1. General strategy for [18F]FDOPA and [18F]FTYR syntheses.
[18F]FDG cassette was then modified to meet the requirements detector. The flow rate was 5 mL/min, and the mobile phase
of the 18FAA synthesis presented earlier.
was aqueous acetic acid (50 mM, adjusted to pH 4 with pellets
In this paper, we describe the automated synthesis of both of sodium hydroxide, solvent A) with ascorbic acid (0.57 mM)
[18F]FDOPA and [18F]FTYR under GMP-oriented conditions on and Na2EDTA.3H2O (1 mM). For [18F]FTYR separation, ethanol
this commercially available FASTlab module.
(1.5%) was added in solvent A.
Radioactivity was measured in a dose calibrator, and all
radiochemical yields were decay corrected end of bombardment
(EOB).
Materials and methods
General
Preparation of reagents and cassette
Except for the trifluoromethanesulfonate salts 1 and 2 and the The main components of the disposable cassette are presented
chiral catalyst 9 (Figure 1), which were synthesized as previously in Figure 2. The skeleton of the cassette consists of 25 rotatable
reported,7,9 all reagents, QMA, tC18 were purchased from stopcock valves (three ways/three positions) numbered 1 to 25
commercial source. Powder of potassium carbonate (1.7 g) was with fluid pathways. Three syringes are used to drive the
used to fill cartridges (2 mL) from Supelco. Sterifix® Paed 0.2-μ different reagents (e.g., water and organic solvent) through the
m sterile filters were supplied from B. Braun (Melsungen, different components of the cassette. Position 3 has 1-mL
Germany). The nca [18F]fluoride was produced via the 18O(p,n) syringe (S1) and in positions 11 and 24, two syringes of 6 mL
nuclear reaction. See Libert et al. 7 for additional information.
are present (S2 and S3, respectively). Positions 2 and 12–16 of
Automated radiochemical synthesis of [18F]FDOPA and [18F] the manifold include barrels and spikes to put and pierce the
FTYR was carried out using a slightly modified FASTlab synthesis different vials containing the reagents placed either in the 11-
module. GE has implemented a second reactor, capable for rapid mm (2 mL) or 13-mm (4 mL) sealed vial. Solid-phase extraction
heating/cooling. Moreover, one additional two-way pinch valve (SPE) cartridges (QMA, tC18, and K2CO3) were also placed on this
(PV) was added for the alkylation/evaporation/hydrolysis steps skeleton. The cyclic olefin copolymer (COC) reactor is dedicated
and the loading of the HPLC loop, respectively.
to the [18F]fluoride drying and labeling steps. The Pyrex reactor
High-performance liquid chromatography purification was vial (35 mL), which contains the Schiff base 7 and the chiral
performed on an HPLC semi-preparative column connected to catalyst 9, was closed with a screw cap; and the septum was
a motorized rotary Rheodyne valve equipped with a loop of pierced with three sterile Teflon catheters (BD Angiocath, Becton
8 mL, directly connected to the FASTlab synthesis module via Dickinson, Sandy, Utah). The 13.3 cm one (i.e., 1.7 mm, gray)
the two-way PV. Purification occurred on an Xbridge® C18 allows either to mix or to aspirate the vial content and those of
column (10 μm, 250 × 10 mm, Waters, Milford, MA, USA) with a 2.1 cm (i.e., 1.7 mm, orange) for nitrogen flush (Figure 2). Silicone
Waters 600 pump and a 996 Waters PDA detector (281 nm) tubings of different lengths (14, 18, and 42 cm) with luer male
controlled by the Empower software. The radioactive elution connectors were used for the connection between the manifold
profile was monitored with a custom-made GM radioactivity and the different components. A water bag of 100 mL (B. Braun)
www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 281–290

C. Lemaire et al.
Figure 2. Main components of the cassette.
and its spike was placed in position 15 of the manifold. The
location of the different components, the length of the main
tubing, and amount of reagents on the pre-assembled cassette
are summarized in Table 1 and Figure 3.
and drying). The connection with the stand-alone HPLC system
was made via a silicone tube placed at position 10 of the
manifold and connected by a polypropylene Y connector to a
two-way solenoid pinch valve (PV, 24V DC). The flow directions
A and B of the PV are either passing or non-passing. At rest
position, the normal position of the valve, only pathway A is
open, and the venting of the system during the alkylation step
is permitted (pathway A connected to a gas balloon, Figure 3).
When PV is actuated, A is closed, and the valve allows liquid to
pass through B, which is lined to the inlet loop of the HPLC
system where a Sterifix Paed (0.2 μm) vented filter was
connected.
The test sequence file was uploaded from the PC database into
the programmable logic controller (PLC) of the FASTlab module.
These preliminary synthesis tests as well as all the processes are
visualized on the screen of the computer. This test allowed to
check the hardware, electronics, and all the connections of the
manifold including cartridges, reactors, vacuum, and pressure.
The synthesis sequence was then uploaded into the internal
memory of the synthesizer. The punchers were then used to
Test cassette and reagents placement
The process was started, and the FASTlab hardware tested (e.g.,
vacuum, radioactivity detectors, reactor heater, syringes, and
nitrogen supply). If no hardware problem was detected, the
cassette was connected on to the synthesis device, and the
stopcock valves engaged into the corresponding pneumatic
rotating actuators of the module. The cassette exhaust located
on the right side of the manifold was connected to a source of
vacuum via the waste bottle (250 mL) used to collect all waste
fluids. The right male connector of the cassette was connected
via a filter (0.2 μm) to the nitrogen inlet of the module. A
nitrogen supply of at least 6 bars was used to activate all the
pneumatic components (e.g., rotating actuators, water bag, and
vials punchers). Another nitrogen supply of 1 bar was used for
the chemical synthesis (e.g., evaporation, transfer of reagents,
J. Label Compd. Radiopharm 2015, 58 281–290
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org

C. Lemaire et al.
Table 1. Location of reagents and materials on the manifold for 18FAA synthesis
Manifold
position
Reagents or materials
Silicone tubing to H128O recovery vial
Details
1
2
3
4
5
6
7
8
9
14 cm
K2CO3/K222/water/ACN
Syringe of 1 mL
11-mm vial (750 μL)
Part of the manifold
140 mg, 14 cm
14 cm
5 mL
14 cm
QMA cartridge with silicone tubing at position 5
Silicone tubing to QMA cartridge at position 4
H128O/18F inlet reservoir (part of manifold)
Silicone tubing to reactor vessel (left-hand side)
Silicone tubing to reactor vessel (central port)
Silicone tubing to alkylation reactor/hydrolysis reactor (gray catheter)
14 cm
28 cm
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Outlet to HPLC loop via silicone tubing to front additional two-way pinch valve 18 cm
Syringe of 6 mL
Part of the manifold
Ammonium precursor in DMF
Acetonitrile
NaBH4 (dry power)
Water bag spike
11 mm (30 mg/1.1 mL)
13 mm (4 mL)
13 mm (20 mg)
100 mL water bag
13 mm (4 mL)
14 cm
400 mg
1.7 g, 14 cm
14 cm
13-mm (3 mL)
18 cm
HI (57%)
Silicone tubing to tC18 cartridge at position 18
tC18 cartridge with silicone tubing at position 17
K2CO3 cartridge with silicone tubing at position 20
Silicone tubing to K2CO3 cartridge at position 19
Toluene
Silicone tubing to alkylation reactor/hydrolysis (orange catheter)
Silicone tubing to alkylation reactor/hydrolysis reactor (orange catheter)
18 cm (25 mg 7, 2 mg of chiral
catalyst 9, 330 μL of toluene)
Part of the manifold
42 cm
24
25
Syringe of 6 mL
Silicone tubing to reactor vessel (right-hand side)
ACN, acetonitrile; HPLC, high-performance liquid chromatography.
push down the different vials present in the barrels and to pierce via valves V3, V4, V5, and V8. The activity recovered into the
the septum of the vials containing the eluent, precursor, COC reactor was then dried by water evaporation at 120 °C for
acetonitrile (ACN), NaBH4, hydroiodic acid (HI), and the water 10 min by combination of vacuum (À200 to À1000 mbar) and
bag (position 15). To fill the line of the water bag and to insure nitrogen flow (200 mbar). The ammonium salt precursor in N,N-
reproducible water volume transfer, water was driven several Dimethylformamide (DMF) was then transferred with syringe
times. During these preliminary steps, sodium borohydride was S2 from position 12 to the COC reactor containing the dry
solubilized by 4 mL of water and all other 11-mm and 13-mm kryptofix [18F]fluoride/complex. Transfer was realized via the
vials slightly pressurized with nitrogen. Before the activity central tubing connection of the COC reactor. Radiolabeling
transfer, the C18 cartridge was also conditioned with 2 mL of was performed at 140 °C for 2.5 min with all the valves closed.
ACN and 4 mL of water. The toluene vial was then placed on The crude reaction mixture was then diluted with about 6 mL
the spike mounted at position 21 and the septum pierced of water aspirated from the water bag via S2 and stopcocks 11,
manually.
15, and 8 and thereafter transferred into reactor 1 to form a
diluted water/DMF solution, which was then aspirated into S2
and pushed down again into the reactor. A third of this solution
Automated radiosynthesis
At the end of the irradiation (EOB, t = 0), the nca [18F]fluoride (about 2 mL) was aspirated from reactor 1 and diluted again in
(185 GBq) in enriched oxygen-18 water (3.5 mL) was transferred S2 with an additional 4 mL of water from the water bag. This
by He pressure (2 bars) through a Teflon tubing line from the crude diluted DMF/water solution was then transferred from S2
cyclotron to the hot cell containing the FASTlab module. The through the tC18 cartridge, and the eluent recovered in the
aqueous solution containing the [18F]fluoride was directly waste bottle. The trapping step of 18F-3 was repeated twice.
recovered in the [18F]fluoride inlet conical reservoir (position 6). Thereafter, NaBH4 (4 mL) was aspirated into S2 and slowly
The water was then transferred using vacuum through the pushed down through the C18 cartridge via valves 11 and 17–
QMA-Light Sep Pak, and the enriched oxygen-18 water 18. Two of the 4 mL were passed through the SPE, and valve
recovered in the water recovery vial. The [18F]fluoride trapped 18 was closed. After 60 s, this valve was reopened, and an
on the SPE support was then eluted into the COC reactor with additional volume of 2 mL passed on the solid support. The
the eluent from vial 1 ((K222/ACN)/(K2CO3/water) (50/50)). This cartridge was then flushed for 15 s with a flow of nitrogen. For
pressurized solution (800 mbar) was firstly aspirated into S1 via the halogenation step, an aqueous HI solution (0.8 mL; 57%)
valves V2 and V3 and then pushed through the QMA cartridge was slowly passed through the solid support, and the SPE stayed
www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 281–290

C. Lemaire et al.
Figure 3. Diagram of the cassette for 18FAA synthesis on the GE FASTlab synthesis module.
wet for 90 s. During this step, S2 was rinsed with water (5 mL). At The line was flushed with nitrogen for 10 s, and the reactor
the end of the halogenation step, the main part of the acid on was rinsed twice with 2 mL of water. The motorized Rheodyne
the cartridge was removed with a stream of N2 for 30 s. SPE injection valve was then activated by the operator, and the
elution was then conducted with toluene (3 mL), and the excess HPLC purification was realized on an Xbridge C18 column
of acid removed by passing the eluate through a homemade (5 μm, 10 × 250 mm) at a flow rate of 5 mL/min and the HPLC
potassium carbonate cartridge (Supelco, 2 mL empty SPE tube).
Empower software.
[18F]Fluorobenzyl iodide 5 or 6 was directly recovered in the
The fraction containing [18F]FDOPA or [18F]FTYR was collected,
Pyrex reactor containing the Schiff base (25 mg), the catalyst after 7.5 and 9.3 min, respectively, in a clean product collection
(2 mg), toluene (330 μL), and KOH 12 N (200 μL). This vial located vial prefilled with 0.9% saline (30 mL), saline 10% (500 μL), and
in reactor heater 2 was connected to positions 9, 22, and 23 of ascorbic acid (500 μg/mL). [18F]FDOPA and [18F]FTYR were
the manifold via the three sterile Teflon catheters. The resulting obtained with high enantiomeric purity (ee > 95%) and high
solution was then mixed by vigorous nitrogen bubbling for specific activity (>740 GBq/μmol). For clinical use, the end
5 min. At the end of the alkylation step, hydroiodic acid (about formulated product was transferred to a pharmaceutical class A
1 mL) was taken via S2 and transferred through valve 22 to hot cell, where an aseptic dispensing in sterile vials through
reactor 2. Toluene evaporation was then conducted under two sterilizing 0.22-μm filters (B. Braun) was realized.
vacuum (À400 mbar) and a slight nitrogen flow (700 mbar) at
140 °C for 4 min. During this evaporation, the top of the reactor
Results and discussion
was cooled with nitrogen via the upper polyether ether ketone
(PEEK) ring. Once evaporation was completed, valves 22 and 23 The nca synthesis of [18F]FTYR and [18F]FDOPA was realized by a
were rotated, and the closed reaction vial was heated at either phase-transfer catalyst method according to a process previously
160 °C for 13 min ([18F]FDOPA) or at 140 °C for 9 min ([18F]FTYR), described in our laboratory.10 Recent chemistry improvements
7,8
respectively. During this hydrolysis step, nitrogen was flushed
through the upper designed PEEK ring.
and single use cassette have allowed us to automate this
reliable synthesis (Figure 1), under GMP conditions, in a FASTlab
After hydrolysis, the vial was cooled at approximately 100 °C module.
by means of a nitrogen flow (6 bars) applied via the lower PEEK
The first part of this multi-step synthesis consists of the pre-
ring located on the oven. The crude solution was diluted with paration of a [18F]fluoro-labeled benzyl iodide (5 and 6, Figure 1),
3 mL of water, aspirated via S2, and then the HPLC injection loop and the second consists of the enantioselective formation of a
(8 mL) located in the adjacent hot cell was loaded. Transfer was new carbon–carbon bond followed by acid hydrolysis and HPLC
realized with S2 via valves 11 and 10, the pinch two-way valves purification. In summary, after labeling, the [18F]aromatic
(pathway B), and the hydrophilic/hydrophobic filter (0.22 μm) aldehyde 3 was trapped on a tC18 cartridge, and the aldehyde
placed at the entrance of the loop of the Rheodyne injector. was converted to the corresponding [18F]fluorobenzyl iodide 5,
J. Label Compd. Radiopharm 2015, 58 281–290
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org

C. Lemaire et al.
6 in a two-step process occurring on the solid support (Figure 1). rather than 120 °C, this temperature parameter was modified.
This electrophilic agent was then eluted from the SPE cartridge Radiochemical labeling yields of 60% and 75% were obtained
with toluene and then used to generate the new chiral center with precursor of [18F]FDOPA and [18F]FTYR, respectively. On
by enantioselective alkylation of a glycine Schiff base 7 in the completion of the labeling reaction, the [18F]fluorobenzaldehyde
presence of a chiral phase-transfer catalyst 9. Finally, removal 3 was trapped on a tC18 SPE cartridge. This purification step
of the protective groups of the [18F]fluoro alkylated compound removes most of the unreacted [18F]fluoride, excess precursor,
8 followed by HPLC purification afforded 18FAA at the nca level potassium carbonate, kryptofix, and DMF.
with high chemical, radiochemical, and enantiomeric purities.
For the two next steps of the radiosynthesis (reduction and
The complexity of this synthesis (e.g., number of steps and halogenation), the aldehyde was kept on the tC18 solid support.
reagents) and the limited number of positions available on the When the SPE cartridge is wetted with NaBH4, the reduction step
manifold complicate the automation.
proceeds quickly at room temperature even if the sodium
borohydride reagent is solubilized several hours before the start
of the synthesis. However, the amount of NaBH4 was limited to
5 mg/mL to avoid hydrogen overpressure that could cause a
Cassette
FASTlab synthesizer belongs to a new generation of automated leaky septum and lead to reagent loss. Under these reaction
synthesizer using disposable cassettes, which decrease the risk conditions, reduction is quantitative at room temperature in less
of cross contamination between several productions. This than 2 min.
synthesizer, initially dedicated for the production of [18F]FDG,
As the halogenation proceeds also at room temperature and is
seems also well adapted for the development of other [18F]fluo- nearly quantitative after 2 min, automation of this step was easy.
rinated radiopharmaceuticals.
A
18FAA cassette assembled However, for the subsequent alkylation step, traces of residual HI
with the disposable, single use [18F]FDG components, should meet need to be removed by passing the tC18 toluene eluent through
the basic GMP rules required for radiopharmaceutical preparation. a small potassium carbonate cartridge. The neutralized eluent
The cassette manifold includes 25 stopcocks numbered 1 to was directly recovered into reactor 2.
25 (Figures 2 and 3). To achieve the [18F]fluoride recovery from
the cyclotron, its drying, and the labeling of the ammonium
salt, 12 of the 25 manifold positions are necessary for the
Radiosynthesis: part 2
corresponding [18F]FTYR and [18F]FDOPA radiochemistry. The The most crucial point in the process was the alkylation step.
conversion of the aldehyde to the corresponding benzyl halide According to our recent publication,7 nearly quantitative
required 10 other positions, but because as all the dedicated alkylation is obtained in toluene with KOH 9 N and vigorous
barrel vial positions are used, the toluene vial was placed on a stirring in the presence of a Schiff base and a chiral catalyst.
normal spike. So, for the last part of the process (alkylation, We have also previously demonstrated that high enantio-
hydrolysis, and HPLC injection) only three of the 25 positions selectivities (>96%) were obtained at room temperature with a
are free. One of them was dedicated to the HPLC transfer. catalyst derived from cinchona or one having a conformationally
Obviously, some syringes and reagents are common to the fixed biphenyl core.7 For this part of the process, only results
different parts of the synthesis. For example, the same syringe obtained with the chiral catalyst 9 (Figure 1) and toluene are
S2 is used during all the processes for transferring the different presented.
solutions in selected quantities through the manifold (e.g., the
For the alkylation step, reactors of different sizes and volumes
precursor, the [18F]labeled aldehyde after dilution of the crude were evaluated. Firstly, the alkylation reaction was carried out in
reaction mixture with water, NaBH4, HI, toluene, and water) and sterile test tubes (10 mL; outer diameter (od) 10 × 100 mm)
for the filling of the HPLC loop.
equipped with a rubber septum connected to the cassette via
The same reagent, HI, is used both for the synthesis of 5 and 6 silicone tubing and three metallic needles pierced through the
and the hydrolysis step. ACN is used for the [18F]fluoride drying, septum. To allow the recovery of the product at the end of the
the C18 cartridge conditioning, and the washing of the cassette synthesis, one of the needles was pushed down to the bottom
at the end of the process. Water is used for the C18 cartridge of the reactor. As previously notified, the alkylation reaction
conditioning to dilute the crude-labeled mixture, to solubilize requires vigorous stirring. So, to avoid additional equipment
NaBH4, and to wash the cassette. Nevertheless, even some such as a magnetic stirrer and many changes to the module, this
components are common to the different parts of the process; needle was also used to pass nitrogen from the module through
the placement of all the reagents on the manifold was not the glass reactor containing the reagents for the alkylation step.
feasible. Therefore, the alkylation reagents (e.g., Schiff base, The resulting bubbles, created in the liquid lead to an efficient
KOH, and chiral catalyst) have to be introduced before the mixing of the two phases. The over pressure was immediately
beginning of the synthesis in the second glass reactor. The final removed via the waste bottle and one of the shorter needles
layout of the cassette is summarized in Figure 3.
of the reactor. One of the first optimization consisted to
determine the optimal pressure and flow rate to obtain a good
mixing without loss of activity. Unfortunately, each time that
the vacuum pump connected to the waste module starts, the
Radiosynthesis: part 1
The classical kryptofix/[18F]fluoride approach was used for the bubble emulsion phase (similar to soap foam) formed above
labeling of the ammonium salt, which was solubilized in DMF the liquid was directly aspirated to the waste. As it was not
just before its location on the cassette. As the [18F]fluoride possible to switch off the vacuum pump of the FASTlab module,
recovery, evaporation, and labeling steps were already well the exhaust of nitrogen from the reactor was then carried out by
implemented in the FDG sequence, automation of this part of using the HPLC way (position 10) and the two-way PV (path A,
the process was conducted with the FDG sequence. However, Figure 3). Furthermore, for the following experiments,
a
as the labeling reaction proceeds in DMF at 140 °C for 2.5 min commercially available Pyrex reactor vial of larger diameter with
www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 281–290

C. Lemaire et al.
a screw cap and a septum in Teflon was selected (25 mm control of this additional unit is realized by the FASTlab software.
(OD) × 100 mm, 35 mL).
The mantle of this copper oven (Figure 5) is heated with a
In comparison with the experiments conducted manually and resistor of 60 W (12 V). This new device includes also a
despite vigorous stirring, the alkylation reaction realized with the radioactive detector for the measurement of the activity during
module was not quantitative. At least 10 min was required to the alkylation, evaporation, and hydrolysis steps. To ensure a
obtain the full conversion of the benzyl iodide into the alkylated reflux in the reaction vessel, the top of the glass vial can be
product. Nevertheless, with a more concentrated solution of cooled with a flow of nitrogen (4 bars) passing through the PEEK
potassium hydroxide (KOH 12 M),9 all the [18F]fluoro ring placed above the reactor heater. The bottom of the heating
methoxybenzyl halides 5 and 6 are consumed in less than block can also be cooled with nitrogen via the second PEEK ring.
5 min. The HPLC chromatogram of the crude reaction mixture Two additional solenoid valves, controlled by the module,
(Figure 4) shows in addition to the expected alkylated provide the flow of nitrogen necessary for the cooling. With
product (Tr = 10.1 min) other [18F]fluorinated by-products. some minor hardware modifications to the synthesizer, the
Obviously, they are present in very small quantities or are not second reactor heater and the two cooling devices are
detectables.
controlled in a fully automated manner.
From several other tests, it appeared that at least 1 mg
Toluene evaporation was realized under vacuum at 140 °C for
(1.2 μmol) of the diphenic catalyst 9 is required to achieve the 4 min. The foam that appears sometimes at the bottom of the
reaction in 5 min. reactor and that is aspirated to the waste bottle can be broken
At the end of the alkylation, HI was directly transferred into if the outside of the glass reactor is cooled with a flush of
reactor 2. This addition that proceeds before the evaporation nitrogen applied via the upper PEEK ring. Under these
of toluene avoids potential racemization if the alkylation reaction conditions, evaporation of toluene proceeds without significant
was not completed in the allotted time.
loss of activity.
For the subsequent toluene evaporation and hydrolysis steps,
Three sterile Teflon catheters were used to resist to the harsh
the module was equipped with an additional reactor heater that conditions of the hydrolysis (HI, 180 °C). They were more
can be rapidly cooled down. This device was designed by GE to appropriate than the metallic one, which might afford some
receive the Pyrex reactor vial (35 mL) described earlier. The full metallic contamination.
Figure 4. Typical radioactive HPLC profile of the alkylation step (X terra RP 18 (3.5 μm, 4.6 × 150 mm); CH3CN/H2O (70/30); 1 mL/min).
Figure 5. Modified FASTlab and the second reactor heater.
J. Label Compd. Radiopharm 2015, 58 281–290
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org

C. Lemaire et al.
In the literature, hydrolysis of protected [18F]FDOPA is realized in the final vial. After dilution with isotonic solution of sodium
at 180 °C for 20 min. A shorter reaction time being preferable, the chloride and subsequent filtration is performed, the sterilized
efficiency of this hydrolysis step has been verified both at lower formulated product is now ready to be tested for quality control
temperature and for shorter duration. According to these (QC) before it is approved for direct human administration. The
experiments, hydrolysis of [18F]FDOPA and [18F]FTYR can now flow chart of the synthesis for [18F]FDOPA is summarized in
be conducted at 160 °C for 13 min or 140 °C for 9 min, Table 2.
respectively.
Up to 185 GBq of [18F]fluoride was used to perform the
Because the FASTlab does not include an HPLC system, the radiochemical synthesis with a FASTlab module. Radiochemical
purification was conducted with the HPLC located inside the yield, synthesis duration, and enantiomeric excess are presented
next hot cell. The HPLC loop was filled up with S2 via path B of in Table 3.
the PV (Figure 3). The hydrophilic/hydrophobic filter (0.22 μm)
The automated radiosynthesis on the FASTlab module
placed at the entrance of the Rheodyne loop makes that the requires about 52 min (n = 18). Total synthesis time including
injection does not depend on the length of the tubing. HPLC purification and formulation was around 1 h. (Table 3).
Moreover, HPLC lines can be rinsed and flushed with nitrogen Starting from such a high level of activity, more than 45 GBq of
without having to empty the loop. This method of injection with [18F]FDOPA and 64 GBq of TYR was available at the end of the
vented sterile filter is reliable, and any experience of filter clogs synthesis.
was observed. Despite injected volume close to 7 mL, peaks were
A complete QC has been developed in accordance with the
not broader, and good separation was still achieved. The guidelines of the European pharmacopeia. The enantiomeric
radioactive peak eluted from the column was directly collected and radiochemical purities of [18F]FTYR and [18F]FDOPA were
Table 2. Flow chart of the automated synthesis for [18F]FDOPA
Step duration
(min)
Synthesis
time (min:s)
Act [18F]FDOPA
(GBq, d.c.)
Step
T (°C)
RT
[18F]Fluoride
recovery
and aldehyde
labeling
0
SOS: end of 18F-fluoride transfer from the
cyclotron to the conical vial of the FASTlab
synthesizer (Position 6)
0:00
185
1.7
10.8
End of [18F]fluoride trapping on QMA
End of [18F]fluoride drying and water
evaporation
1:40
12:15
RT
120
(15 min)
2.5
End of labeling
14:50
17:20
140
RT
[18F]
End of the third transfer of the diluted crude
of reaction through the tC18
End of the reduction step
End of the halogenation step
End of the tC18 toluene elution through the
tC18
fluorobenzyliodide:
synthesis and
purification
(8.5 min)
20:10
21:40
23:20
RT
RT
RT
and K2CO3 cartridges and collection of the
benzyl
iodide derivative in reactor 2
End of the alkylation into reactor 2
End of toluene evaporation
End of the hydrolysis step
Alkylation
Evaporation
Hydrolysis
HPLC preparation:
cooling, rinsing,
and injection
HPLC purification
5
6
13
5
28:20
33:55
46:55
51:50
RT
140
160
RT
End of loading of the HPLC loop
10
End of HPLC purification and collection of the
18FAA into the vial containing 30 mL of NaCl
0.9%, 0.5 mL of NaCl 10%, and ascorbic acid.
62:00
RT
67.1
d.c., decay corrected; SOS, start of synthesis; RT, room temperature.
Table 3. Radiochemical yield and enantiomeric excess
Synthesis duration (min)
Radiochemical Yield (EOB, d.c., %)
Enantiomeric excess (%)
n
[18F]FTYR
58
62
50.5 2.7
36.3 3.0
97.5 0.7
98.2 0.6
10
8
[18F]FDOPA
www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 281–290

C. Lemaire et al.
Figure 6. HPLC chromatograms of the final [18F]TYR (radiochemical [γ, A] and ultraviolet [UV, B] 270 nm).Crownpack CR(+) (5 μm, 150 × 4 mm), HClO4 (aqueous (pH 2),
0.8 mL/min).
Table 4. Overview of quality control results of [18F]FDOPA and [18F]TYR at release
[18F]FDOPA
release
Test
Method
Acceptance criteria
[18F]TYR release
Conform
Appearance
Visual observation
Colorless and free
Conform
from particle matter
Radionucleidic identity
Radionucleidic purity
Measure of radioactive decay 105–115 min
Conform
Conform
Conform
Conform
Conform
Conform
Gamma spectroscopy
Gamma spectroscopy after
decay
511 KeV
>99.9%
Radioactive content
at EOS
Radiochemical purity
Dose calibrator
≈1.1 GBq/mL
≈1.8 GBq/mL
UPLC analysis
[18F]FAA (D + L) >
95%
Conform
Conform
Enantiomeric purity
pH
Chiral HPLC
pH Meter
>90% (L)
3–5
99.1%
3.70
98.7%
3.70
Osmolarity
Osmometer
280 ≤ (mOsm/kg) ≤
340
300
320
Specific activity
UPLC
>37 GBq/μmol
740–1850 GBq/μmol 740–1850 GBq/μmol
Residual solvents
Gas chromatography with FID ACN < 410 ppm
<LOQ
<LOQ
<LOQ
<LOQ
NA
<LOQ
<LOQ
<LOQ
<LOQ
0.4%
detector (limit test)
Toluene < 890 ppm
DMF < 880 ppm
HOAc < 5000 ppm
EtOH < 5000 ppm
<17.5 IU/mL
(if maximum dose is
equal to 10 ml)
Sterile
Bacterial endotoxin
testing
LAL: methods A and D, Eur.
Ph. 2.6.14
<1 IU/mL
<1 IU/mL
Sterility
Eur. Ph. 2.6.1
Conform
Conform
EOS, end of synthesis; UPLC, ultra performance liquid chromatography; ACN, acetonitrile; LOQ, limit of quantification; FID, flame
ionisation detector; NA, not applicable; LAL, limulus amebocyte lysate; Eur. Ph., European pharmacopeia.
determined after HPLC analysis on a CrownPack CR (Daïcel, QC tests and acceptance criteria for several batches of [18F]
Illkirch, France) (+) column from Daicel (Figure 6).
FDOPA and [18F]TYR are summarized in Table 4.
With the activity of 1800 MBq/mL and ascorbic acid, the
Controls performed at release correspond to analyses of the
stability studies performed at 6 h post-end of synthesis (EOS) final product at EOS. Specific activity was measured as previously
7
demonstrate that no degradation occurs, and that radiochemical described
purity is unchanged relative to QC results obtained at EOS. Other ascorbic acid.
but on synthesis realized without addition of
J. Label Compd. Radiopharm 2015, 58 281–290
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org

C. Lemaire et al.
Conclusion
References
Syntheses of [18F]FDOPA and [18F]FTYR have been fully automated
on a slightly modified FASTlab synthesizer from GE. The automated
multi-step procedure affords a reliable and convenient access
to [18F]FDOPA and [18F]FTYR in a reproducible decay-corrected
radiochemical yield of 36% and 50%, respectively. According to
QC, these 18FAA obtained in maximum 62 min (HPLC included)
exhibit high specific activity (>740 GBq/μmol) and high chemical,
radiochemical, and enantiomeric purities. This process, realized
with disposable single use cassette, is adapted to routine
GMP production of high amount of activity in 18FAA. This fully
automated process facilitates access for clinical application and
should contribute to the future development of these compounds.
[1] R. Hustinx, C. Lemaire, G. Jerusalem, P. Moreau, D. Cataldo, B. Duysinx,
J. Aerts, M. F. Fassotte, J. Foidart, A. Luxen, J. Nucl. Med. 2003; 44, 533.
[2] I. Rutten, J. E. Cabay, N. Withofs, C. Lemaire, J. Aerts, V. Baart, R.
Hustinx, J. Nucl. Med. 2007; 48, 720.
[3] J. Kalbitzer, L. Deserno, F. Schlagenhauf, A. Beck, T. Mell, G. Bahr,
H.-G. Buchholz, M. Plotkin, R. Buchert, Y. Kumakura, P. Cumming,
A. Heinz, M. A. Rapp, Eur. J. Nucl. Med. Mol. Imaging 2012; 39, 1462.
[4] P. Jokinen, M. Karrasch, A. Bruck, J. Johansson, J. Bergman, J. O. Rinne,
J. Neurol. Sci. 2013; 329, 23.
[5] S. Balogova, J.-N. Talbot, V. Nataf, L. Michaud, V. Huchet, K. Kerrou, F.
Montravers, Eur. J. Nucl. Med. Mol. Imaging 2013; 40, 943.
[6] M.-Y. Lu, Y.-L. Liu, H.-H. Chang, S.-T. Jou, Y.-L. Yang, K.-H. Lin, D.-T. Lin,
Y.-L. Lee, H. Lee, P.-Y. Wu, T.-Y. Luo, L.-H. Shen, S.-F. Huang, Y.-F. Liao,
W.-M. Hsu, K.-Y. Tzen, J. Nucl. Med. 2013; 54, 42.
[7] L. C. Libert, X. Franci, A. R. Plenevaux, T. Ooi, K. Maruoka, A. J. Luxen,
C. F. Lemaire, J. Nucl. Med. 2013; 54, 1154.
Acknowledgements
[8] C. Lemaire, L. Libert, A. Plenevaux, J. Aerts, X. Franci, A. Luxen,
J. Fluor. Chem. 2012; 138, 48.
[9] Y.-G. Wang, M. Ueda, X. Wang, Z. Han, K. Maruoka, Tetrahedron 2007;
63, 6042.
[10] C. Lemaire, S. Gillet, S. Guillouet, A. Plenevaux, J. Aerts, A. Luxen, Eur.
This work was supported by a scientific grant from GE. We thank
P. Hawotte for providing us with fluorine-18.
J. Org. Chem. 2004; 13, 2899.
Conflicts of Interest
The authors have no conflicts of interest to disclose.
www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2015, 58 281–290