ARTICLE IN PRESS

Applied Radiation and Isotopes 65 (2007) 12441248

www.elsevier.com/locate/apradiso

Technical note

Synthesis procedure for routine production of

2-[ F]uoro-3-(2(S)-azetidinylmethoxy)pyridine (2-[18F]F-A-85380)
18

Andreas Schildan, Marianne Patt, Osama Sabri

Department of Nuclear Medicine, University of Leipzig, 04103 Leipzig, Germany
Received 6 October 2006; received in revised form 8 February 2007; accepted 19 February 2007

Abstract
2-[18F]Fluoro-3-(2(S)-azetidinylmethoxy)pyridine (2-[18F]F-A-85380) was among the rst subtype selective radioligands to visualise
the in vivo distribution of a4b2-containing neuronal nicotinic acetylcholine receptors (nAChRs) in human brain. We developed a one-pot
synthesis for the preparation of 2-[18F]F-A-85380 in a commercially available TRACERlab FXFN synthesis module. The synthesis
comprises a nucleophilic substitution followed by hydrolysis of a t-butyloxycarbonyl (BOC)-protected intermediate. After formulation
for intravenous application up to 20 GBq 2-[18F]F-A-85380 were produced from a starting activity of 100 GBq [18F]uoride in 60 min
with a specic activity of about 4  105 GBq/mmol and a mean radiochemical purity of more than 99%.
r 2007 Elsevier Ltd. All rights reserved.
Keywords: Nicotinic acetylcholine receptor; A-85380; 2-[18F]uoro-3-(2(S)-azetidinylmethoxy)pyridine; GMP synthesis

1. Introduction
nAChRs belong to the gene superfamily of ligand-gated
ion channels, consisting of a pentameric structure composed of 12 different subunits (a2a10 and b2b4), which
differ with regard to pharmacology and anatomical
distribution (for review cf. Gotti and Clementi, 2004).
nAChRs mediate neurotransmission throughout the central and peripheral nervous system. The two most
abundant nAChRs in brain, heteromeric a4b2-and homomeric a7-containing receptors, contribute to a variety of
brain functions such as cognition, learning and memory
and are affected in various pathophysiological conditions,
e. g. Alzheimers disease, Lewy body dementia, Parkinsons
disease and vascular dementia (Pimlott et al., 2004; Quik
and Kulak, 2002).
3-(2(S)-azetidinylmethoxy)pyridine (A-85380) (Abreo et
al., 1996) binds selectively to b2 subunit containing
nAChRs (Xiao et al., 2004) and its uorinated derivative
2-[18F]uoro-3-(2(S)-azetidinylmethoxy)pyridine (2-[18F]FCorresponding author. Tel.: +49 341 9718071; fax: +49 341 9718089.

E-mail address: andreas.schildan@medizin.uni-leipzig.de

(A. Schildan).
0969-8043/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apradiso.2007.02.009

A-85380) (Dolle et al., 1998, 1999) has, therefore, been used

to visualise a4b2-containing nAChRs in the human brain
in vivo (Gallezot et al., 2005; Kendziorra et al., 2005;
Meyer et al., 2005) and in post-mortem investigations in
swine preparations (Deuther-Conrad et al., 2006). Therefore, 2-[18F]F-A-85380-PET appears to be a valuable tool
to investigate the state of this subtype of nAChRs in
different types of neurological diseases such as Alzheimers
disease, Lewy body dementia, Parkinsons disease and
vascular dementia.
Previous automated radiosyntheses of no carrier added
(n.c.a.) 2-[18F]F-A-85380 use a two-reactor synthesis
module with one reactor for the conversion of the labelling
precursor 2-nitro-3-[2(S)-N-(tert-butoxycarbonyl)-2-azetidinylmethoxy]pyridine or (3-[2(S)-N-(tert-butoxycarbonyl)-2-azetidinylmethoxy]pyridin-2-yl)trimethylammonium
triuoromethanesulfonate with [18F]uorine to a tertbutyloxycarbonyl (BOC)-protected uorinated intermediate and with a second reactor for subsequent acidolysis of
the intermediate to 2-[18F]F-A-85380, respectively (Dolle et
al., 1998, 1999; Schmaljohann et al., 2004, 2005). This
restricts the use of commercially available synthesis
modules with usually one reactor as for example a
TRACERlab FXFN module. Therefore, we developed a

ARTICLE IN PRESS
A. Schildan et al. / Applied Radiation and Isotopes 65 (2007) 12441248

one-pot synthesis procedure for the routine preparation of

2-[18F]F-A-85380 for human use with optimised reaction
parameters for application in such an unmodied commercially available standard synthesis module under GMP
conditions.
2. Materials and methods
2.1. General
Acetonitrile for azeotropic distillation and as solvent for
the radiolabelling reaction was obtained from Merck
(Darmstadt, Germany; quality: for DNA synthesis). Acetonitrile for HPLC was obtained from Riedel-de Haen
(Seelze, Germany; Reag. European Pharmacopoeia (Ph.
Eur.)). Kryptoxs 222 and triuoroacetic acid were also
purchased from Merck as special quality (for synthesis) and
as Uvasols quality (for chromatography), respectively. The
labelling precursor 2-nitro-3-[2(S)-N-(tert-butoxycarbonyl)2-azetidinylmethoxy]pyridine (1; cf. Fig. 1) and the reference
standard 2-uoro-3-(2(S)-azetidinylmethoxy)pyridine tartrate was obtained from ABX (Radeberg, Germany).
Isotonic saline and Sodium Phosphate Braun (a ready-foruse additive for infusion solutions) were received from B.
Braun (Melsungen, Germany), water for injection from
DeltaSelect (Pfullingen, Germany). All other chemicals and
solvents were obtained from Merck, Riedel-de Haen or
Fluka (Buchs, Switzerland) as Ph. Eur. or Reag. Ph. Eur.
quality and used without further purication.
2.2. Production of [18F]fluoride
[18F]Fluoride was produced with a PETtrace cyclotron
(GE Medical Systems, Uppsala, Sweden) by irradiation of
enriched [18O]water (Rotem, Beer Sheva, Israel; 97%
enrichment) via the 18O(p,n)18F nuclear reaction. For
study purposes, irradiations have been carried out for
5090 min with a beam current of 30 mA resulting in
70100 GBq of [18F]uoride.
2.3. Automated production of 2-[18F]F-A-85380
Preparation of 2-[18F]F-A-85380 (3; cf. Fig. 1) was
performed according to a protocol published by Dolle et al.
(Dolle et al., 1998) with a number of modications.
Syntheses were carried out in an unmodied TRACERlab

1245

FXFN module (GE Medical Systems/Nuclear Interface,

Muenster, Germany; cf. Fig. 2).
The aqueous [18F]uoride solution was passed through
an anion exchange cartridge 18F Separation Cartridge 30PS-HCO3 (Macherey-Nagel, Dueren, Germany) preconditioned with 1 mL ethanol and 1 mL water for injection in
order to remove the enriched [18O]water. The separated
[18F]uoride was eluted from the cartridge using a solution
of 15 mg (40 mmol) Kryptoxs 222 and 2.75 mg (20 mmol)
potassium carbonate in a mixture of 800 mL acetonitrile
and 200 mL water for injection. Finally, the [18F]uoride
was evaporated to dryness at 95 1C within 6 min under
reduced pressure.
Fluorination of the labelling precursor was conducted by
adding a solution of the nitro compound (1; 3 mg or 10 mmol)
in 1 mL acetonitrile to the dried [KC2.2.2.]+/18F complex
and subsequent heating at 85 1C for 15 min. After evaporation of the solvent within 5 min at 95 1C under reduced
pressure, the BOC protection group was quantitatively
removed by acidolysis with a 1/1 mixture of dichloromethane
and triuoroacetic acid (1 mL) at 40 1C for 3 min. The
reaction mixture was again evaporated to dryness within
8 min at 95 1C under reduced pressure. The residue was taken
up with 3 mL water for injection, puried by passing over two
Sep-Paks Silica Plus cartridges (Waters, Milford, MA)
preconditioned with 5 mL ethanol and 20 mL water for
injection, resulting in the extraction of brownish-coloured
polar uorination and deprotection by-products. The elution
of the product fraction from the cartridges (still including
some impurities) was effected by 4 mL 0.015 M disodium
hydrogen phosphate. This crude product was puried by
subsequent semipreparative HPLC (column Supelcosil LCABZ+PLUS 12 mm, 250  10 mm [Sigma-Aldrich, St. Louis,
MO], eluent 0.2% triuoroacetic acid solution in water for
injection/acetonitrile, 97/3 [v/v], ow 14 mL min1). 2-[18F]FA-85380 had a retention time of 4.370.4 min (k0 3.170.5),
determined by both UV (l 280 nm) and radioactivity
detectors. The product fraction eluting from the HPLC was
cut into a separate dilution vessel (cf. Fig. 1) and after
basication with 2 mL 0.5 M sodium hydroxide the nal
formulation of the desired product was achieved by xation
on a solid-phase extraction cartridge Chromax 200-PS-RP
(Macherey-Nagel) preconditioned with 10 mL ethanol and
20 mL water for injection and removal of the HPLC eluent
via helium drying and washing with 10 mL water for
injection. The product was eluted with ethanol (1 mL) and
diluted with 9 mL saline. For pH adjustment of the nal

Fig. 1. Radiosynthesis of n.c.a. 2-[18F]F-A-85380 by nucleophilic substitution and subsequent acidolysis.

ARTICLE IN PRESS
1246

A. Schildan et al. / Applied Radiation and Isotopes 65 (2007) 12441248

Fig. 2. Schematic diagram for automated synthesis module of n.c.a. 2-[18F]F-A-85380 (unmodied TRACERlab FXFN, GE Medical Systems/Nuclear
Interface, Muenster, Germany).

product, 50 mL of an electrolyte concentrate (Sodium

Phosphate Braun, consisting of 3.05 g [8.52 mmol] disodium
hydrogen phosphate dodecahydrate and 0.462 g [2.96 mmol]
sodium dihydrogen phosphate dihydrate in 20 mL water for
injection) was used.
According to this protocol, up to 20 GBq 2-[18F]F-A85380 were produced from a starting activity of 100 GBq
[18F]uoride in 60 min synthesis time. On average, the
specic activity was 4.071.9  105 GBq/mmol, the mean
radiochemical purity was 99.270.8%.
During synthesis development and optimisation of the
reaction procedure, the protocol described above was
altered as described below: for examination of the
nucleophilic substitution and acidolysis, the reaction was
quenched with 2 mL water after the respective reaction
step. For the nucleophilic substitution, the dependence of
the radiochemical yields with regard to precursor amount
(16 mg of 1), reaction time (2.520 min) and solvents
(acetonitrile at 85 1C, DMF and DMSO at 85 and 140 1C,
respectively) were determined. Acidolysis of the BOCprotected intermediate was examined with triuoroacetic
acid in dichloromethane at 40 1C or alternatively 6 M
hydrochloric acid at 30 1C. Without evaporating the
reaction mixture after the substitution step acidolysis was

carried out with a volume of 0.6 mL pure triuoroacetic

acid and 6 M hydrochloric acid, respectively.
2.4. Analytical procedure
The product solution was analysed for chemical and
radiochemical purity with the exception of the triuoro
acetate content by means of an HPLC system consisting of
a UV detector (UVD 170U, Dionex, Sunnyvale, CA) and a
NaI(Tl) detector (2  2 in Bicron 2M2/2P, Gabi Star
system, Raytest, Straubenhardt, Germany). A Supelcosil
LC-ABZ+PLUS column (250  4.6 mm, 5 mm) was used
with a ternary gradient system (isocratic conditions till
15 min [eluent A methanol (2%), eluent B 0.2% triuoroacetic acid (95%), eluent C acetonitrile (3%)], then
gradient in 15 min [eluent A constant 2%, eluent B changed
to 48% and eluent C to 50%]) at a ow rate of 1 mL min1.
2-[18F]F-A-85380 was eluted with a retention time of
8.370.2 min (k0 1.570.1) and the BOC-protected uorinated intermediate (2; cf. Fig. 1) with a retention time of
30.570.1 min (k0 8.370.3). The triuoro acetate
concentration was determined via an HPLC system
containing a conductivity detector (CD25A, Dionex, detection with suppressed conductivity by Dionex suppressor

ARTICLE IN PRESS
A. Schildan et al. / Applied Radiation and Isotopes 65 (2007) 12441248

ASRSs-ULTRA 2 mm). A Dionex IonPacs AS15 column

(250  2 mm) was used with a potassium hydroxide gradient
generated by a Dionex EG 40 eluent generator (isocratic
conditions till 6 min [eluent 10 mM potassium hydroxide],
then gradient in 10 min [eluent changed to 100 mM
potassium hydroxide]) at a ow rate of 0.45 mL min1.
Triuoro acetate was eluted with a retention time of
21.870.3 min (k0 13.670.3).
Tests on clarity, colourlessness, Kryptoxs 222, residual
solvents, pH value, bacterial endotoxines, sterility, radionuclidic purity and radionuclide identity were performed
according to the monographs on radiopharmaceutical
preparations and udeoxyglucose (18F) injection prescribed by the Ph. Eur.
3. Results and discussion
Based on previously published results (Dolle et al., 1998,
1999) we developed a one-pot synthesis of 2-[18F]F-A85380, which allows the application of an unmodied
widely used and commercially available synthesis module
TRACERlab FXFN with one reactor for the GMP
conform routine preparation for human use. Starting by
introducing [18F]uorine to the labelling precursor 2-nitro3-[2(S)-N-(tert-butoxycarbonyl)-2-azetidinylmethoxy]pyridine via standard nucleophilic substitution conditions, the
product was prepared by acidolysis of the protective BOC
group by triuoroacetic acid (cf. Fig. 1).
The dependence of the labelling yields on precursor
amount, reaction time and solvents were optimised rst
aligned to the preconditions of the automated synthesis
module. A radiochemical yield of 7580% in acetonitrile
and 8590% in DMF as well as DMSO could be achieved
with 3 mg of 1 after a reaction time of 2.5 min at a reaction
temperature of 85 1C (radiochemical yields determined in
solution after nucleophilic substitution by HPLC). In
contrast, at a reaction temperature of 140 1C, the conversion into the BOC-protected intermediate 2 was nearly
quantitative in both solvents, DMF and DMSO. Nevertheless, acetonitrile was chosen as solvent, because of the
better results for the deprotection reaction. As the optimal
substitution conditions 3 mg of 1, a reaction time of 15 min
and a reaction temperature of 85 1C were determined with a
radiochemical yield for 2 of 9095% in solution.
The decision for acetonitrile was based on examinations of
the acidolysis of the BOC-protected intermediate. Cleavage
of the protecting group in DMSO with pure triuoroacetic
acid or 6 M hydrochloric acid under various conditions was
not successful. Besides the insufcient radiochemical yield of
maximal 20% of 3 (due to incomplete deprotection of 2;
radiochemical yield determined in solution after hydrolysis
by HPLC) in all these cases there was obviously a side
reaction between DMSO and triuoroacetic acid, probably
as a Pummerer rearrangement-like reaction with formation
of an activated sulfoxonium intermediate and subsequent
decomposition into a sulphide compound (De Lucchi et al.,
1991). The acidolysis in DMF with triuoroacetic acid yields

1247

up to 10% of 3, the cleavage of the BOC group in acetonitrile

more than 90% of 3. Quantitative removal of the protecting
group could only be achieved by evaporating the solvent,
especially acetonitrile, and subsequent acidolysis with a
deprotecting agent consisting of at least 50% triuoroacetic
acid in dichloromethane at moderate temperatures. Smaller
amounts of triuoroacetic acid result in an incomplete
deprotection. During large scale productions by means of the
above-mentioned synthesis module, a signicant loss of the
starting activity was observed after addition of the deprotecting agent to the reaction mixture, which is most likely to be
explained by loss of [18F]hydrogen uoride formed from
unreacted [18F]uoride into the gas phase.
As mentioned above, product 3 was puried before
semipreparative HPLC with Silica Plus cartridges from
polar uorination and deprotection by-products, which
would not entirely be removed under the applied HPLC
conditions. During the evaporation of triuoroacetic acid
and dichloromethane no loss of activity was detected,
nevertheless, the solid phase extraction procedure results in
an overall loss of activity of approximately 25% of the
initial activity, most likely due to an incomplete recovery of
the desired product from the cartridges, since no degradation of the product was observed during the evaporation
step as conrmed by HPLC analysis before and after
evaporation. The elution of the product fraction (still
including some impurities) from the two combined silica
cartridges was effected by disodium hydrogen phosphate.
For the nal purication step after semipreparative
HPLC and the formulation for intravenous application, a
reversed-phase cartridge on a polystyrenedivinylbenzene
copolymer was used. Silica-based cartridges together with
the basied reaction mixture and saline form precipitating
sodium silicate. The low level of crosslink of the PS-RP
cartridges and therefore the minor retardation of the target
compound also allows the facile elution of product 3 with a
small and physiological compatible amount of ethanol. By
this procedure up to 60% of the puried product could be
recovered. The stability of the nal product was determined over 8 h by HPLC. The radiochemical purity then
still exceeds 95%.
The syntheses described in the literature up to now use 1
or alternatively (3-[2(S)-N-(tert-butoxycarbonyl)-2-azetidinylmethoxy]pyridin-2-yl)trimethylammonium
triuoromethane-sulfonate as labelling precursors. These are
likewise transformed with [18F]uorine into the BOCprotected uorinated intermediate with subsequent acidolysis of this intermediate into 3 (Dolle et al., 1998, 1999;
Schmaljohann et al., 2004, 2005). In comparison with these
synthetic approaches using precursor 1 the overall radiochemical yield presented by our method, is somewhat lower
(cf. Dolle et al., 1998: precursor 1, 4.06.0 mg, synthesis
time 105110 min, uncorrected yields 2533%; Schmaljohann et al., 2004: precursor 1, 1.5 mg, synthesis time
70 min, uncorrected yields 48%, based on [18F]uoride
starting activity) due to compromises because of realisation
of the reaction as a one-pot synthesis (e.g. choice of

ARTICLE IN PRESS
1248

A. Schildan et al. / Applied Radiation and Isotopes 65 (2007) 12441248

solvent, three evaporation processes), whereas analytical

results are quite comparable or even better especially with
regard to radiochemical purity and specic activity.
However, the main advantages of the here-presented onepot synthesis compared to the previously published
methods are that a fully automated synthesis by means of
a standard radiochemistry module is realised (compared to
Schmaljohann et al., 2004), that there is no need for a
second reaction vessel (compared to Dolle et al., 1998;
Schmaljohann et al., 2005) and that the synthesis is realised
by the application of only one HPLC purication instead
of two, what so far to our knowledge demands the
application of a robotic system, that is, while very exible,
usually demanding hot cells of large dimensions, which are
not available in most PET centres (compared to Dolle
et al., 1998).
In summary, a one-pot synthesis procedure for the
production of 2-[18F]F-A-85380 with a corrected yield of
24.976.2% was established on an unmodied, commercially
available synthesis module. The achieved activity quantities
are entirely sufcient for clinical routine. The purity of the
product was suitable for human use with a mean specic
activity of 4.071.9  105 GBq/mmol, a mean radiochemical
purity of 99.270.8% and no remarkable macroscopic
impurities. The nal product was stable over 8 h.
References
Abreo, M.A., Lin, N.-H., Garvey, D.S., Gunn, D.E., Hettinger, A.-M.,
Wasicak, J.T., Pavlik, P.A., Martin, Y.C., Donelly-Roberts, D.L.,
Anderson, D.J., Sullivan, J.P., Williams, M., Arneric, S.P., Holladay,
M.W., 1996. Novel 3-pyridyl ethers with subnanomolar afnity for
central neuronal nicotinic acetylcholine receptors. J. Med. Chem. 39,
817825.
De Lucchi, O., Miotti, U., Modena, G., 1991. The Pummerer reaction of
sulnyl compounds. Org. React. 40, 157405.
Deuther-Conrad, W., Wevers, A., Becker, G., Schildan, A., Patt, M.,
Sabri, O., Steinbach, J., Brust, P., 2006. Autoradiography of 2-[18F]FA-85380 on nicotinic acetylcholine receptors in the porcine brain in
vitro. Synapse 59 (4), 201210.
Dolle, F., Valette, H., Bottlaender, M., Hinnen, F., Vaufrey, F., Guenther,
I., Crouzel, C., 1998. Synthesis of 2-[18F]Fluoro-3-[2(S)-2-azetidinylmethoxy]pyridine, a highly potent radioligand for in vivo imaging

central nicotinic acetylcholine receptors. J. Labelled Compd. Radiopharm. 41, 451463.

Dolle, F., Dolci, L., Valette, H., Hinnen, F., Vaufrey, F., Guenther, I.,
Fuseau, C., Coulon, C., Bottlaender, M., Crouzel, C., 1999. Synthesis
and nicotinic acetylcholine receptor in vivo binding properties of 2uoro-3-[2(S)-azetidinylmethoxy]pyridine: a new positron emission
tomography ligand for nicotinic receptors. J. Med. Chem. 42,
22512259.
Gallezot, J.-D., Bottlaender, M., Gregoire, M.-C., Roumenov, D.,
Deverre, J.-R., Coulon, C., Ottaviani, M., Dolle, F., Syrota, A.,
Valette, H., 2005. In vivo imaging of human cerebral nicotinic
acetylcholine receptors with 2-18F-uoro-A-85380 and PET. J. Nucl.
Med. 46, 240247.
Gotti, C., Clementi, F., 2004. Neuronal nicotinic receptors: from structure
to pathology. Prog. Neurobiol. 74, 363396.
Kendziorra, K., Meyer, P., Barthel, H., Becker, G., Sorger, D., Schildan,
A., Hesse, S., Patt, M., Seese, A., Richter, K., Hammerstein, E.,
Wagenknecht, G., Zimmer, C., Gertz, H.J., Sabri, O., 2005. Cerebral
nicotinic acetylcholine receptors (nAChRs) in patients with Alzheimers disease (AD) or vascular dementia (VaD) evaluation with 2[18F]F-A85380 and positron emission tomography (PET). J. Nucl.
Med. 46 (Suppl. 2), 11P.
Meyer, P., Kendziorra, K., Barthel, H., Hesse, S., Becker, G., Sorger, D.,
Schildan, A., Patt, M., Seese, A., Lincke, T., Wagenknecht, G.,
Lobsien, D., Wegner, F., Brust, P., Steinbach, J., Zimmer, C.,
Schwarz, J., Sabri, O., 2005. Quantitative assessment of the
cerebral a4b2 nicotinic acetylcholine receptors in Parkinsons disease:
a PET study using 2-[18F]F-A85380. J. Nucl. Med. 46 (Suppl. 2),
65P.
Pimlott, S.L., Piggott, M., Owens, J., Greally, E., Court, J.A., Jaros, E.,
Perry, R.H., Perry, E.K., Wyper, D., 2004. Nicotinic acetylcholine
receptor distribution in Alzheimers disease, dementia with Lewy
bodies, Parkinsons disease, and vascular dementia: in vitro binding
study using 5-[125I]-A-85380. Neuropsychopharmacology 29,
108116.
Quik, M., Kulak, J.M., 2002. Nicotine and nicotinic receptors; relevance
to Parkinsons disease. Neurotoxicology 23, 581594.
Schmaljohann, J., Minnerop, M., Karwath, P., Gundisch, D., Falkai, P.,
Guhlke, S., Wullner, U., 2004. Imaging of central nAChReceptors
with 2-[18F]F-A85380: optimized synthesis and in vitro evaluation in
Alzheimers disease. Appl. Radiat. Isotopes 61, 12351240.
Schmaljohann, J., Gundisch, D., Minnerop, M., Joe, A., Bucerius, J.,
Dittmar, C., Jessen, F., Guhlke, S., Wullner, U., 2005. A simple and
fast method for the preparation of n.c.a. 2-[18F]F-A85380 for human
use. Appl. Radiat. Isotopes 63, 433435.
Xiao, Y., Baydyuk, M., Wang, H.P., Davis, H.E., Kellar, K.J., 2004.
Pharmacology of the agonist binding sites of rat neuronal nicotinic
receptor subtypes expressed in HEK 293 cells. Bioorg. Med. Chem.
Lett. 14, 18451848.