[18F]Fluoropyruvate: Radiosynthesis and initial biological evaluation

Article abstract of DOI:10.1002/jlcr.3183

The radiosynthesis of [¹⁸F]fluoropyruvate was investigated using numerous precursors were synthesized from ethyl 2,2-diethoxy-3- hydroxypropanoate (5) containing different leaving groups: mesylate, tosylate, triflate, and nonaflate. These precursors were evaluated for [ ¹⁸F]fluoride incorporation with triflate being superior. The subsequent hydrolysis step was investigated, and an acidic hydrolysis was optimized. After establishing suitable purification and formulation methods, the [¹⁸F]fluoropyruvate could be isolated in ca. 50% d.c. yield. The [¹⁸F]fluoropyruvate was evaluated in vitro for its uptake into tumor cells using adenocarcinomic human alveolar basal epithelial cells (A549) and unfortunately showed an uptake of approximately 0.1% of the applied dose per 100,000 cells after 30 min. Initial pharmacokinetic properties were assessed in vivo using nude mice showed a high degree of bone uptake from defluorination, which will limit its potential as an imaging agent for metabolic processes. Copyright

Full text of DOI:10.1002/jlcr.3183

Research Article
Received 13 October 2013,
Revised 10 December 2013,
Accepted 12 December 2013
Published online 23 January 2014 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3183
18
[ F]Fluoropyruvate: radiosynthesis and initial
biological evaluation
a
a,b
Keith Graham, Andre Müller, Lutz Lehmann,a Norman Koglin,a,b
*
Ludger Dinkelborg,a,b and Holger Siebeneichera
The radiosynthesis of [18F]fluoropyruvate was investigated using numerous precursors were synthesized from ethyl 2,
2-diethoxy-3-hydroxypropanoate (5) containing different leaving groups: mesylate, tosylate, triflate, and nonaflate. These
precursors were evaluated for [18F]fluoride incorporation with triflate being superior. The subsequent hydrolysis step was
investigated, and an acidic hydrolysis was optimized. After establishing suitable purification and formulation methods,
the [18F]fluoropyruvate could be isolated in ca. 50% d.c. yield. The [18F]fluoropyruvate was evaluated in vitro for its uptake
into tumor cells using adenocarcinomic human alveolar basal epithelial cells (A549) and unfortunately showed an uptake of
approximately 0.1% of the applied dose per 100,000 cells after 30 min. Initial pharmacokinetic properties were assessed
in vivo using nude mice showed a high degree of bone uptake from defluorination, which will limit its potential as an
imaging agent for metabolic processes.
Keywords: fluorine-18 radiolabeling; PET imaging; fluoropyruvate; pyruvate kinase; pyruvate dehydrogenase
Lactate and pyruvate are both described as transporter
substrates for MCT1 and MCT4. For MCT1, Km values of 0.7 and
3–5 mM were described for pyruvate and lactate, respectively.
Introduction
Pyruvic acid plays a pivotal role in numerous biochemical
pathways, both in the mammalian and bacterial systems. Pyruvate
is a substrate for a number of key enzyme systems and different
transport mechanisms including the Krebs cycle, monocarboxylic
For MCT4, Km values for L-lactate and pyruvate are 28 and
150 mM, respectively.8
A halogenated pyruvate derivative (3-bromopyruvate) is
acid transporter (MCT), pyruvate dehydrogenase, lactate
described in literature as potent agent for tumor therapy by
dehydrogenase (LDH), and pyruvate kinase.1 Pyruvate plays a
inhibiting the glycolytic enzyme hexokinase and causing ATP
depleting.9 It is hypothesized that 3-bromopyruvate enters the
major role in the metabolism of the heart and in addition is
tumor cells via MCT4 that release lactate.9,10 In addition, to
believed to play also a key role in cancer cell metabolism as it is
one of the metabolism products derived from the cancer cells’
support the hypothesis of using pyruvate to image tumors,
hyperpolarized [13C]-pyruvate has been recently described for
enhanced glycolytic activity (also known as the ‘Warbung
Effect’).2–4 The pyruvate produced during glycolysis acts as a key
tumor imaging in tumor-bearing rodents and is currently studied
in cancer patients.11 After injection, its conversion to either
intermediate between the glycolysis and the Krebs cycle (Figure 1).
Part of it can be transported from the cell cytoplasm into the
lactate in tumors or alanine in healthy tissues can be monitored
mitochondrial matrix, where it is converted to acetyl coenzyme A
by magnetic resonance spectroscopy.
The only radiolabeled derivatives of pyruvic acid are either [11C]-
which then enters the Krebs cycle. However, in cancer tissues,
labeled or [14C]-labeled, of which only the [11C]-derivative would
the majority of pyruvate remains in the cytosol and gets rapidly
reduced to lactate by LDH. Lactate is released from the tumor cells
be suitable for non-invasive imaging (Figure 2). The synthesis of
[11C]pyruvic acid has been reported either using enzymatic
by monocarboxylate transporters (MCT4). The reduction step
methods12 or chemical methods.13,14 This radiopharmaceutical
regenerates NAD+ from NADH in the cytosol which is highly
needed for a continued conversion of glucose in the glycolysis.
was investigated in tumor-bearing rabbits and enabled clear
Release of lactate from the cells is accompanied with protons,
visualization of the tumor and progressed to clinical trials for brain
tumors.15 These interesting results lead us to investigate whether a
which cause acidification of the tumor microenvironment a
process promoting local tumor growth and invasion. Initially, the
release of large amounts of lactate was considered as carbon
waste, and lactate production is solely needed for the regeneration
of NAD. However, more recent studies showed that the
[18F]-radiolabeled pyruvate derivative would be a suitable tool for
released lactate can be taken up and consumed by adjacent
tumor cells via MCT1.5 Another portion of the released lactate
enters the blood stream and can be used in the liver to
resynthesize glucose in an adopted Cori cycle. A high production
rate of lactate can cause lactic acidosis, which is a prominent
feature of the tumor microenvironment.6,7
aBayer Healthcare, Global Drug Discovery, Müllerstrasse 178, 13353 Berlin, Germany
bPiramal Imaging, Tegeler Strasse 6-7, 13353 Berlin, Germany
*Correspondence to: Keith Graham, Bayer Healthcare, Global Drug Discovery,
Müllerstrasse 178, 13353 Berlin, Germany.
E-mail: keith.graham@bayer.com
J. Label Compd. Radiopharm 2014, 57 164–171
Copyright © 2014 John Wiley & Sons, Ltd.

K. Graham et al.
Figure 1. Tumor-specific adaptations of the intermediary metabolism for assuring energy production, growth, ROS protection, and survival. Tumors are characterized by
adaptation of several catabolic and anabolic pathways as well as transport systems to meet their demands for energy, growth, and detoxification. In particular, glycolytic
and glutaminolytic pathways along with a truncated tricarboxylic acid cycle are pronounced. Pyruvate is a key metabolic intermediate and central hub in intermediary
tumor metabolism. [18F]-labeled pyruvate derivatives are supposed to feed into this system via uptake through monocarboxylate transporters. This figure is available in
color online at wileyonlinelibrary.com/journal/jlcr
O
O
O
the Bruker AVANCE-300 and 400 MHz and the Bruker DRX-600
NMR spectrometer. Chemical shifts (δ) are expressed in parts
per million relative to tetramethylsilane. Coupling constants (J)
are expressed in Hertz (Hz), and the following abbreviations
are used to describe the appearance of peak patterns: s (singlet),
11C
18F
OH
OH
OH
O
O
1
O
2
3
Figure 2. Structures of pyruvic acid (1), [11C]labeled pyruvic acid (2), and [18F]
d
(doublet), t (triplet), q (quartet), quint (quintet), and
combinations thereof (e.g. dd, doublet of doublet); furthermore,
m (multiplet), mc (centered multiplet), and br (broad signal).
Purity of all compounds was judged to be in the 95% range
according to 1H-NMR, with no chemical impurity detectable.
Radioanalytical HPLC runs were performed on an Agilent 1100,
and preparative HPLC runs were performed on a Knauer
Smartline 1000, both fitted with a GABI Star radioactivity
detector (Raytest). Analytical HPLC to check the radiochemical
purity and co-injections was performed using a Synergy Hydro
RP column (250 × 4.6 mm, 4 μ, Agilent): solvent A, water + 0.1%
TFA (v/v); solvent B, acetonitrile + 0.1% trifluoroacetic acid (v/v);
gradient, isocratic 0% B for 2 min then 0% B to 30% B in
10 min; flow, 1 mL/min. Semi-preparative HPLC purifications
were performed using an ACE RP C18 column (10 × 250 mm;
5 μm; Advanced Chromatographic Technologie): solvent A,
water; solvent B, acetonitrile; gradient, isocratic 5% B for 3 min
then 5% B to 100% B in 20 min; flow, 3 mL/min.
fluoropyruvate (3).
imaging metabolism in vivo (Figure 2). Surprisingly, despite
fluoropyruvate being commercially available, we could not find a
radiosynthesis for [18F]fluoropyruvate described in the literature;
the only reference to this tracer was found in a patent application
without any experimental details to its synthesis.16
Herein, we report on the synthesis of various precursors and
their ability to incorporate [18F]fluoride using non-carrier added
(n.c.a.) nucleophilic fluorination methods. The subsequent
deprotection of the radiolabeled intermediate was evaluated to
allow the radiofluorinated derivative [18F]-3 to be synthesized
in good yield and high radiochemical purity. This tracer was
evaluated in vitro with A549 tumor cells and in vivo in nude mice.
Materials and methods
The tumor cell lines were obtained from ATTC (USA) and
maintained according to protocols provided by the supplier. All
General
Chemicals and solvents were used as received from commercial chemicals used for biological studies were obtained from Sigma
vendors, Sigma Aldrich (Taufkirchen, Germany), Acros Organics Aldrich (Taufkirchen, Germany).
(part of Thermo Fisher-Scientific, Nidderau, Germany). [18F]
Fluoride was produced via an 18O(p,n)18F nuclear reaction and
was supplied by EuroPET (Berlin, Germany). Chemical reactions
were monitored by thin layer chromatography (TLC) and by
UPLC/MS analysis of reaction mixtures (Waters Aqcuity with
Synthesis of the precursors 6-9
Ethyl 2,2-diethoxy-3-hydroxypropanoate (5)
single-quadrupole ESI-MS detector). Crude products were, unless n-Butyllithium (4.00 mL, 10 mmol, 2.5 M in hexanes) was added to a
stated otherwise, purified on silica gel (KP-SIL cartridges, Biotage) solution of diisopropylamine (1.68 mL, 12 mmol) in dry THF (30 mL)
on automated flash chromatography devices (SP-4 or Isolera under nitrogen at À78°C. After 30 min at 0°C, the mixture was
Four by Biotage). Nuclear magnetic resonance (NMR) spectra cooled to À78°C, and methyl dimethoxyacetate 4 (1.78 mL, 10 mmol)
1
were recorded on H, and19F-NMR spectra were recorded with was added slowly. After 50 min at À78°C, paraformaldehyde
J. Label Compd. Radiopharm 2014, 57 164–171
Copyright © 2014 John Wiley & Sons, Ltd.
www.jlcr.org

K. Graham et al.
(1.2 g, 40 mmol) was added and allowed to warm to RT. After 2 h flash chromatography with hexane/EtOAc (5:1) to give 9 as a
1
at RT, saturated NaHCO3(aq) was added. The reaction was diluted colorless oil (194 mg, 0.4mmol, 48%). H-NMR (400 MHz, CDCl3):
with water extracted with EtOAc. The organic layers were dried δ = 1.26 (t, J = 7.06 Hz, 6H) 1.34 (t, J = 7.16 Hz, 3H) 3.50–3.73 (m,
(MgSO4) and filtrated. After evaporation, the crude mixture was 4H) 4.31 (q, J = 7.03 Hz, 2H) 4.68 (s, 2H) ppm. 19F-NMR (376 MHz,
purified via flash chromatography with hexane/EtOAc (6:1) to CDCl3) δ = À126.09 to À125.67 (m, 2F) À121.51 to À121.14 (m,
give 5 as a colorless oil (835 mg, 40%). 1H-NMR (400 MHz, CDCl3): 2F) À110.76 to À110.39 (m, 2F) À80.64 (t, J = 9.75 Hz, 3F) ppm.
δ = 1.26 (t, J = 7.06 Hz, 6H) 1.33 (t, J = 7.06 Hz, 3H) 1.92 (br s, 1H) 13C-NMR (100 MHz, CDCl3): δ = 14.04, 15.00, 58.91, 62.49, 72.52,
3.51–3.73 (m, 4H) 3.82 (br s, 2H) 4.29 (q, J = 7.16 Hz, 2H) ppm. 97.76, 166.11 ppm. MS (ESI).
13C-NMR (100 MHz, CDCl3): δ = 14.19, 15.24, 58.40, 61.81, 63.23,
100.57, 168.65 ppm.
Radiosynthesis of [18F]fluoropyruvate (3)
[18F]Fluoride was produced by an 18O (p, n) 18F nuclear reaction
Ethyl 2,2-diethoxy-3-[(methylsulfonyl)oxy]propanoate (6)
by bombardment of a 98% 18O-enriched water target with an
5 (206 mg, 1.00 mmol) and triethylamine (205 μL, 1.49 mmol) were
dissolved in CH2Cl2 (7mL) and cooled to 0°C. A solution of
methanesulfonyl chloride (77 μL, 0.99 mmol) in CH2Cl2 (3mL) was
added slowly. The reaction was gradually warmed to room
temperature and stirred for additional 5 h. The reaction mixture
was diluted with water and extracted with methylene chloride.
The organic layers were dried (MgSO4) and filtrated, concentrated
and purified by column chromatography (SiO2:hexanes/EtOAc 5:1)
to give 6 as a colorless oil (277mg, 97%). 1H-NMR (400 MHz, CDCl3):
δ = 1.26 (t, J = 7.07 Hz, 6H), 1.34 (t, J = 7.20 Hz, 3H), 3.07 (s, 3H), 3.51–
3.59 (m, 2H), 3.63–3.70 (m, 2H), 4.30 (q, J = 7.16 Hz, 2H), 4.44 (s, 2H)
ppm. 13C-NMR (100MHz, CDCl3): δ = 14.22, 15.08, 38.16, 58.69,
62.24, 67.17, 98.66, 167.02 ppm. MS (ESI): [M+ Na]+ = 307.
11 MeV proton beam at the RDS111 cyclotron. The aqueous
[18F]fluoride solution was trapped in a small anion exchange
Sep-PakTM Plus QMA cartridge (Waters) (preconditioned with
5 mL 0.5 M K2CO3 solution and 10 mL water). The radioactivity
was eluted with a solution mixture (1.0 mg K2CO3 in 0.5 mL water
and 5.27 mg K222 in 1.5 mL MeCN) from the QMA cartridge into a
5 mL conic Wheaton vial. The solvent was evaporated under a
stream of nitrogen at 110°C. Azeotropic drying was repeated
three times with 1.0 mL portions of acetonitrile. Then, the
precursor ethyl 2,2-diethoxy-3-{[(trifluoromethyl)sulfonyl]oxy}
propanoate (8) (4.5 mg, 13.3 μmol) in dry MeCN (300 μL) was
added to the dried [18F]KF–K222. After heating at 90°C for
15 min in a sealed vial, the reaction mixture was diluted with
1.7 mL water and given on a semi-prep HPLC (ACE, RP C18,
(10 × 250 mm), 5 μ, 3 mL/min (Agilent), solvent A: H2O, solvent
B: MeCN, gradient: isocratic 5% B for 3 min then 5% B to 100%
Ethyl 2,2-diethoxy-3-{[(4-methylphenyl)sulfonyl]oxy}propanoate (7)
5 (206 mg, 1.00 mmol) was dissolved in a mixture of CH2Cl2 (3 mL),
Et3N (150μL, 1.08 mmol), and DMAP (25 mg, 0.21 mmol) and cooled
to 0°C. A solution of para-toluenesulfonyl chloride (191 mg,
1.00 mmol) in CH2Cl2 (2 mL) was added dropwise. After stirring
at 0°C for 5 h, the mixture was washed with water, saturated
NaCl(aq), dried over MgSO4, and filtrated. After evaporation, the
crude mixture was purified via flash chromatography with
hexane/EtOAc (7:1) to give 7 as a colorless oil (212 mg, 59 %).
1H-NMR (400 MHz, CDCl3): δ = 1.17–1.27 (m, 9H), 2.45 (s, 3H),
3.40–3.59 (m, 4H), 4.18 (q, J = 7.16 Hz, 2H), 4.21 (s, 2H), 7.35 (d,
J = 8.10 Hz, 2H), 7.78 (d, J = 8.48 Hz, 2H) ppm. 13C-NMR (100 MHz,
CDCl3): δ = 14.11, 15.03, 21.71, 58.42, 62.08, 67.09, 98.33, 128.14,
129.85, 132.47, 145.10, 166.76 ppm. MS (ESI): [M+ Na]+ = 383.
B
in 20 min, retention time of ethyl 2,2-diethoxy-3-[18F]
fluoropropanoate: 20.9–21.1 min). The collected fraction (1.0–
1.2 mL) was given into a 5 mL conic Wheaton vial, and 4 N
H2SO4 (600 μL) was added. The vial was sealed and the reaction
mixture heated for 20 min at 140°C. Afterwards, the reaction
mixture was neutralized with 6 N NaOH, and under gentle stream
of nitrogen, the reaction mixture was heated in the open
Wheaton vial at 100°C to dryness (20–30 min). The residue was
taken up in Ampuwa® water (5000 μL) to get 23–325 MBq [18F]
fluoropyruvate (radiochemical yield, 25 11.4% d.c; synthesis
time, 145 13.4 min; n = 4). The radiochemical purity was
>95%. The [18F]-labeled product was confirmed by co-injection
with the F-19 cold standard on an analytical HPLC (column:
Synergy Hydro RP, 250 × 4.6 mm, 4 μ, 1 mL/min (Agilent); solvent
A, H2O + 0.1% TFA; solvent B, MeCN + 0.1% TFA; gradient,
isocratic 0% B for 2 min then 0% B to 30% B in 10 min). This
co-injection was also spiked with [18F]fluoride, and the baseline
separation of the [18F]fluoride peak from the product peak
showed a resolvable product retardation from the void on the
used Synergy Hydro column. In addition, silica TLC, using the
solvent system 1-butanol:water:acetic acid:ethanol (12:5:3:3),
was carried out with the [18F]-labeled product, fluoropyruvate,
and [18F]-fluoride; the fluoride remained on the baseline, and
both the radiolabeled [18F]fluoropyruvate and fluoropyruvate
gave the same Rf value of 0.49.
Ethyl 2,2-diethoxy-3-{[(trifluoromethyl)sulfonyl]oxy}propanoate (8)
To a solution of 5 (309 mg, 1.5 mmol) in dry CH2Cl2 (10 mL) was
added pyridine (157 μL, 1.95 mmol). After cooling to À78°C,
trifluoromethylsulfonyl anhydride (328 μL, 1.95 mmol) was
added dropwise. The mixture was stirred at 0°C for 1 h. The
reaction was quenched with saturated NH4Cl(aq) and extracted
with CH2Cl2 three times. The combined organic layers were dried
over Na2SO4 and concentrated. The crude residue was purified
by silica gel flash chromatography with EtOAc/hexane (1:10) to
give 8 as a colorless oil (430 mg, 85%). 1H-NMR (400 MHz, CDCl3):
δ = 1.22–1.30 (m, 6H) 1.34 (t, J = 7.16 Hz, 3H) 3.49–3.75 (m, 4H)
4.32 (q, J = 7.16 Hz, 2H) 4.64 (s, 2H) ppm. 19F-NMR (376 MHz,
CDCl3): δ = À74.32 ppm.
Cell lines
Ethyl 2,2-diethoxy-3-{[(nonafluorobutyl)sulfonyl]oxy}propanoate (9)
To a stirred solution of 5 (172 mg, 0.83 mmol) in MeCN (7.0mL) was For tracer uptake and competition studies A549 tumor cells were
added perfluoro-1-butanesulfonyl fluoride (180 μL, 1.00 mmol), used. Twenty-five thousand cells/well were seeded in 48 well
i-Pr2NEt(HF)3 (270μL, 1.08 mmol), and i-Pr2NEt (543 μL, 3.17 mmol) plates 2 days prior to the cell uptake study and grown under
at 0°C. After stirring at 25°C for 5 h, the mixture was quenched with standard conditions (37°C, 5% CO2) until sub-confluency. The cell
water, extracted with EtOAc, washed with saturated NaCl(aq), dried number at the day of the uptake assay was determined by
over MgSO4, and concentrated. The crude mixture was purified via detaching cells in three representative wells and cell counting
www.jlcr.org
Copyright © 2014 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2014, 57 164–171

K. Graham et al.
in a Neubauer cell chamber. Uptake data were normalized to precursors, that is, different solvents (MeCN, DMSO, and DMF),
100,000 cells. Prior to the radioactive uptake assay, the cell different temperatures (90–160°C), and different times (5–20 min)
culture medium was removed, and the cells were washed twice are illustrated in Table 1, where it is clear to see that the triflate
with 37°C warm PBS buffer. The radiotracer was added to the precursor 8 was superior with excellent incorporation under
assay buffer (PBS/0.1% w/v BSA) using a concentration of relatively mild conditions. The radiolabeled intermediate 5
250 kBq/well. For competition experiments, the cells were co- was purified by semi-preparative HPLC methods before the
incubated with pyruvate as competitor used in excess of 1 mM. hydrolysis step as we believed this would simplify the process
Tracer uptake was stopped by removal of the assay buffer at as a purification after the hydrolysis step could prove to be
the time points indicated. Cells were quickly washed twice with non-trivial as the polar [18F]fluoropyruvate would have to be
PBS and lysed by the addition of 1 N NaOH. The cell lysate was separated from other polar impurities, both radioactive and
removed from the plates, and radioactivity of 18F-samples was non-radioactive (Scheme 2).
determined in a gamma counter (Wizard 3, Perkin Elmer, USA).
Deprotection of the radiolabeled intermediate 10
Animals
The hydrolysis of the purified [18F]-labeled intermediate [18F]-5
The animal positron emission tomography (PET) imaging study
was performed in compliance with the current version of the
German law concerning animal protection and welfare. NMRI
nude mice (Taconic, Ry, Denmark) were used for this study. A
PET imaging study was performed using the Inveon small animal
PET/computed tomography (CT) scanner (Siemens, Knoxville,
TN). Ten megabecquerel of 18F-fluoropyruvate was injected
intravenously in conscious animals via the tail vain. After
50 min post-injection (p.i.), the animals were anesthetized with
isoflurane (Abbot), and the PET data were acquired from 60 to
70 min p.i.
was tested under both acidic and basic conditions. Basic
hydrolysis was tested with 2 M NaOH at 25°C and was complete
after 10 min with no defluorination observed; a second acidic
step was tested under 6 M HCl at 120°C for 15 min. To simplify
the synthesis, the acidic conditions were tested as this would
enable a one-step deprotection. Therefore, the intermediate
[18F]-5 was subjected to different strongly acidic deprotection
conditions (2–6 M HCl and 2–4 M H2SO4) using varied
temperatures and times, and the results are summarized in
Table 2. The hydrolysis conditions were monitored by HPLC
and TLC (n-BuOH:water:AcOH:EtOH, 12:5:3:3), and 4 M H2SO4 at
100°C for 20 min were found to be the optimal conditions. The
resulting strongly acidic reaction mixture was neutralized,
evaporated to dryness, and taken up in 5 mL Ampuwa® water
to give the desired product [18F]fluoropyruvate 3 as a formulated
solution ready for biological testing. The identity of [18F]
Results
Synthesis of the precursors 6-9
As outlined in Scheme 1, the key alcohol intermediate 5 was
synthesized in moderate yield via deprotonation of dimethoxy
acetate 4 with lithium diisopropylamide generated in situ and
subsequent use of paraformaldehyde as the elecrophile. This
key intermediate was reacted with methanesulfonyl chloride,
para-toluenesulfonyl chloride, trifluoromethylsulfonyl anhydride,
or perfluoro-1-butanesulfonyl fluoride using standard chemistry
to yield the mesylate (6), tosylate (7), triflate (8), or nonafluoro
butylsulfonate (9), respectively.
fluoropyruvate
3 was confirmed by co-injection of the
corresponding cold reference fluoropyruvate (Figure 3A and B)
and also injected with free [18F]fluoride (Figure 3C) to confirm
that the polar product peak was not free [18F]fluoride.
Biological characterization
The uptake of 18F-fluoropyruvate was studied in A549 tumor
cells. After 10 min of incubation, a tracer uptake/binding of
1752 504 cpm per 100,000 cells was measured, which
corresponds to 0.08% of the applied dose. In a 10 min blocking
Radiolabeling of the precursors 6-9
To establish a suitable radiosynthesis method of [18F]fluoropyruvate, experiment using excess of pyruvate (1 mM), the uptake was slightly
the n.c.a. nucleophilic [18F]fluorination was evaluated using the reduced to 1216 107 cpm per 100,000 cells corresponding to
mesylate (6), tosylate (7), triflate (8), and nonafluorobutylsulfonate 0.05% of the applied dose. No increasing tracer uptake was
(9) precursors. Numerous conditions were tested with the four observed for the following time points. The 20 min uptake value
Scheme 1. Synthesis of radiolabeling precursors 6-9.
J. Label Compd. Radiopharm 2014, 57 164–171
Copyright © 2014 John Wiley & Sons, Ltd.
www.jlcr.org

K. Graham et al.
Table 1. Evaluation of the precursors 6-9 under different radiolabeling conditions
Precursor
Solvent
Temperature (°C)
Time (min)
Incorporation (%)
Mesylate 6
MeCN
DMF
DMSO
MeCN
DMF
DMSO
MeCN
MeCN
MeCN
DMF
90
120
160
90
120
160
90
90
90
120
160
120
10
10
10
10
10
10
5
10
20
10
10
10
0
0
23
0
Tosylate 7
Triflate 8
0
11
77
82
97
83
49
81
DMSO
DMF
Nonaflate 9
Standard conditions of 5 mg K222, 1 mg K2CO3, and 4 mg precursor in 300 μL solvent were investigated.
Scheme 2. Radiosynthesis of [18F]fluoropyruvate.
Table 2. Evaluation of the hydrolysis of 10 to [18F]
fluoropyruvate 3
Discussion
Our efforts have been focused on investigating and developing
new [18F]labeled metabolic tracers that could potentially
address the downfalls of [18F]FDG (the PET imaging ‘workhorse’),
that is, differentiation between tumor and inflammation.17 These
new tracers could offer complimentary imaging tools to answer
key clinical questions and help offer patients the best and
appropriate treatment. We have reported on a number of
glutamate tracers, 4-[18F]fluoroglutamate (BAY 858050)18 and
FSPG (BAY 949392)19 which were successful in preclinical studies
for tumor imaging and showed no uptake in inflammation. Both
these tracers progressed to the clinical setting where FSPG
proved superior and is being investigated further.20,21 Multiple
mechanisms and many metabolic pathways are adapted in the
metabolism of tumor cells. They are connected and converge
to support rapid energy generation, to increase biosynthesis of
macromolecules, to maintain appropriate cellular redox status,
and detoxification potential.22 These include the aerobic
glycolytic, glutaminolytic, and lipogenesis pathway that provide
key intermediates for anabolic or anaplerotic reactions.
Molarity
(M)
Temperature
(°C)
Time
(min)
Hydrolysis
(%)
Acid
HCl
2
2
6
2
2
4
4
25
100
160
140
160
140
140
10
5
10
5
5
10
20
0
15
23
49
91
H2SO4
>95
>95
was higher (3009 2457 cpm per 100,000 cells corresponding to
0.13% of the applied dose) but associated with a higher
standard deviation and thus statistically not significant. The
30 min time point was again in the range of 10 min point
(1647 421 cpm per 100,000 cells, corresponding to 0.08% of
the applied dose) (Figure 4).
Pyruvate is a particularly interesting molecule as it is involved
in numerous metabolism pathways and, that is, glycolysis, Krebs
cycle, MCT, pyruvate dehydrogenase, LDH, and pyruvate kinase.1
Surprisingly, despite being an attractive molecule for the PET
imaging of various disease progresses either in oncology or
Micro-positron emission tomography/computed
tomography imaging
A PET/CT imaging study was conducted to study the behavior of cardiology, there are very few radiolabeled derivatives of
the new radiotracer in vivo. PET images were acquired at 60 min pyruvate and only one PET imaging probe synthesized ([11C]
p.i. Axial, coronal and sagittal sections, and a whole body pyruvate) either by enzymatic12 or chemical methods.13,14
projection are shown (Figure 5).
Interestingly, [11C]pyruvate has been investigated pre-clinically
www.jlcr.org
Copyright © 2014 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2014, 57 164–171

K. Graham et al.
Figure 4. Cell assay to study the uptake of [18F]fluoropyruvate in A549 tumor
cells. Cell uptake/binding was measured at 10, 20, and 30 min incubation. A
blocking was performed by co-incubation with excess pyruvate (1 mM). Twenty-
three thousand counts per minute corresponds to 0.1% of the applied dose. This
figure is available in color online at wileyonlinelibrary.com/journal/jlcr
different sulfonates as these are the leaving groups of choice
for aliphatic radiofluorinations.30 In order to synthesize these
precursors, the commercially available methyl dimethoxyacetate
4 was treated with lithium diisopropylamide and the electrophile
paraformaldehyde to give the key intermediate alcohol, ethyl
2,2-diethoxy-3-hydroxypropanoate 5, in moderate yield (40%)
as illustrated in Scheme 1. Simple standard sulfonation methods
on the alcohol 5 gave the desired precursors with the mesylate 6
(97%), tosylate 7 (59%), and triflate 8 (85%) leaving groups.
The precursor with the nonaflate 9 leaving group was isolated
from unsuccessfully attempts to make the cold fluorine-19
reference compound of 10 using the Vorbrüggen fluorination
methodology.31 The radiofluorination was explored using
standard radiofluorination methods using Kryptofix 2.2.2 (K222
,
5 mg) as the phase transfer catalyst and potassium carbonate
(1 mg) as the base. Initial radiolabelings with the standard
solvents and temperatures (acetonitrile at 100°C, DMF at 120°C,
and DSMO at 160°C) showed that the triflate was the superior
precursor with >90% incorporation of fluoride (90°C in
acetonitrile for 20 min). The radiolabeled intermediate 10 was
purified using semi-preparative HPLC methods as we envisioned
Figure 3. Analytical HPLC chromatograms of the final formulated
[
18F]
fluoropyruvate (3) (Figure 3A gamma detector); the co-injection of the [18F]
fluoropyruvate (3) with fluoropyruvate (Figure 3B, UV detector at 210 nm); and
co-injection with [18F]fluoride (Figure 3C gamma detector). This figure is available
in color online at wileyonlinelibrary.com/journal/jlcr
in tumor-bearing rabbits, in one patient with a brain tumor15 which difficulties in purifying the final radiolabeling product away from
looked promising and in patients with cerebral ischemic hypoxia similar polar by-products which could interfere with any
or infarction.23 Because of the short half-life of carbon-11, we subsequent biological testing. The deprotection step on the
decided to look into a fluorine-18 derivative as fluoropyruvic acid HPLC purified intermediate 10 was investigated, first under basic
is a known substrate for pyruvate carboxylase,24 LDH25 and condition using 2 M NaOH at RT, and the hydrolysis was found to
inhibitor of the pyruvate dehydrogenase complex,26,27 and is be quantitative after 10 min with no free [18F]fluoride by TLC
considerably less toxic than fluoroacetate.28 Surprisingly, in the analytics. The ketal protecting group was removed by treating
literature, the only reference to [18F]fluoropyruvic acid was found basic solution with 6 M HCl and heating at 120°C for 15min
in a patent application where no radiosynthesis was described.16
which gave the desired [18F]fluoropyruvic acid 3. The
In designing a suitable precursor for the synthesis of [18F] deprotection step was investigated further to see whether a
fluoropyruvic acid 3, it seemed that pursing a simple precursor such simple acidic deprotection step would be sufficient instead of
as simple alkyl esters of bromopyruvic acid, which are commercially the two step hydrolysis. The acidic hydrolysis conditions were
available, might not be advantageous as it is known that simple tested using different molarities (2–6 M) of either HCl or H2SO4
esters of pyruvic acid undergo spontaneous hydration,29 which at elevated temperatures as shown in Table 2. This single acidic
we believed would hinder the radiofluorination.
hydrolysis step was not so straightforward as envisioned, and
Our strategy was to investigate precursors with the ketone only 4 M H2SO4 with heating at 120°C for 10 min or longer was
moiety at the two-position protected as a ketal, which should successful. After formulation, the identity of [18F]fluoropyruvic
facilitate a quick and simple removal under acidic conditions. acid 3 was confirmed by co-injection of the corresponding cold
The leaving groups for the n.c.a. radiofluorinations would be reference (Figure 3A and B). [18F]Fluoropyruvic acid 3 was
J. Label Compd. Radiopharm 2014, 57 164–171
Copyright © 2014 John Wiley & Sons, Ltd.
www.jlcr.org

K. Graham et al.
Figure 5. Micro-positron emission tomography/computed tomography (PET/CT) study of [18F]fluoropyruvate in mice. PET data acquisition was started at 60 min after
injection of 10 MBq in conscious animals. Corresponding axial (A), coronal (B), and sagittal (C) sections covering the heart and vertebra are shown. Whole body PET/CT
is shown in (D). Low PET signals were observed from most tissues except the bone (especially vertebra and joints) and bladder. The increased bone signal is presumably
caused by defluorination and the bladder signal from renal excretion. This figure is available in color online at wileyonlinelibrary.com/journal/jlcr
analyzed by HPLC to confirm that no co-elution was observed metabolic degradation pathway in humans might be different
with free [18F]fluoride (Figure 3C).
than observed in mice.
By targeting MCTs with [18F]fluoropyruvate, we tried to
visualize a common metabolic adaptation in tumor cells. A549 Acknowledgements
cells are known to have a glycolytic phenotype by consuming
glucose and releasing lactate. Therefore, strong MCT activity The authors would like to acknowledge the excellent technical
can be assumed. However, the degree of uptake of [18F]
assistance of Jörg Pioch, Melanie Senftleben, Jörg Jannsen, and
Eva-Maria Bickel.
fluoropyruvate in A549 tumor cells was low, and no time
dependent uptake was observed, which is in contrast to other
18F-labeled compounds studied in this assay (Figure 4).18,19 This References
may suggest that [18F]fluoropyruvate is either not recognized
[1] A. L. Lehninger, D. L. Nelson, M. M. Cox, Principles of Biochemistry,
by the transporter or binds with a low affinity that is not
sufficient for transport into the cells. This was unexpected in light
of the uptake and biological activity of 3-bromopyruvate in
cancer cells and the uptake and intracellular conversion of
hyperpolarized-[13C]pyruvate. As the fluoropyruvate is known
to form mainly the hydrated isoform in aqueous solutions32
whereas bromopyruvate is mainly present in the ketoform, this
slight difference could be a reason for the different uptake
behavior of these homologuous compounds.
Positron emission tomography-imaging study in nude mice
demonstrated that 18F-fluoropyruvate is quickly excreted via
the kidneys into the bladder (Figure 5). Neither the heart nor
other organs showed accumulation and retention of the tracer
except the bones. The observed uptake of radioactivity can be
most likely attributed to accumulation of released free 18F-
fluoride after metabolism in the blood and liver as reported for
other non-stable 18F-compounds.18
2nd ed. Worth, New York, 1993.
[2] M. G. Vander Heiden, L. C. Cantley, C. B. Thompson, Science 2009,
324, 1029.
[3] H. R. Christofk, M. G. Vander Heiden, M. H. Harris, A. Ramanathan,
R. E. Gerszten, R. Wie, M. D. Fleming, S. L. Schreiber, L. C. Cantley,
Nature 2008, 452, 230.
[4] M. G. Vander Heiden, J. W. Locasale, K. D. Swanson, H. Sharfi,
G. J. Heffron, D. Amador-Noguez, H. R. Christofk, G. Wagner,
J. D. Rabinowitz, J. M. Asara, L. C. Cantley, Science 2010, 329, 1492.
[5] P. Sonveaux, F. Végran, T. Schroeder, M. C. Wergin, J. Verrax,
Z. N. Rabbani, C. J. De Saedeleer, K. M. Kennedy, C. Diepart,
B. F. Jordan, M. J. Kelley, B. Gallez, M. L. Wahl, O. Feron, M.
W. Dewhirst, J. Clin. Invest. 2008, 118, 3930.
[6] N. C. Denko, Nature Rev. Cancer 2008, 8, 705.
[7] J. L. Chen, J. E. Lucas, T. Schroeder, S. Mori, J. Wu, J. Nevins,
M. Dewhirst, M. West, J.-T. Chi, PLoS Genet. 2008, 4, e1000293.
[8] A. P. Halestrap, D. Meredith, Pflugers Arch. 2004, 447, 619.
[9] Y. H. Ko, B. L. Smith, Y. Wang, M.G. Pomper, D. A. Rini, M. S. Torbenson,
J. Hullihen, P. L. Pedersen, Biochem. Biophys. Res. Commun. 2004, 324, 269.
[10] J. P. Geschwind, C. S. Georgiades, Y. H. Ko, P. L. Pedersen, Expert Rev.
Anticancer Ther. 2004, 4, 449.
[11] K. Golman, R. I. Zandt, M. Lerche, R. Pehrson, J. H. Ardenkjaer-Larsen,
Cancer Res. 2006, 66, 10855.
[12] M. B. Cohen, L. Spolter, C. C. Chang, J. S. Cook, N. S. MacDonald, Int. J.
Appl. Radiat. Isot. 1980, 31, 45.
More detailed research would be needed to further
characterize this new 18F-labeled tracer, however, its low stability
in mice limits its use and further exploration in vivo.
[13] M. R. Kilbourn, M. J. Welch, Int. J. Appl. Radiat. Isot. 1982, 33, 359.
[14] J. R. Ropchan, J.R. Barrio, J. Nucl. Med. 1984, 25, 887.
[15] T. Hara, M. Iio, R. Izuchi, T. Tsaukiyama, F. Yokoi, Eur. J. Nucl. Med.
1985, 11, 275–278.
[16] J. Beliczey, U. Kragl, A. Liese, C. Wandrey, K. Hamacher, H.H. Coenen,
T. Tierling. US Patent US 6,355,453 B1, 2002.
Conclusions
The synthesis of four precursors (6-9) derived from ethyl 2,2-
diethoxy-3-hydroxypropanoate 5 bearing the mesylate, tosylate,
triflate, and nonaflate leaving groups was successfully [17] P. D. Shreve, Y. Anzai, R. L. Wahl, RadioGraphics 1999, 19, 61.
[18] R. N. Krasikova, O. F. Kuznetsova, O. S. Fedorova, Y. N. Belokon,
synthesized and evaluated for the their ability to incorporate
[18F]fluoride. The triflate precursor 8 was found to give the
V. I. Maleev, L. Mu, S. Ametamey, P. A. Schubiger, M. Friebe,
M. Berndt, N. Koglin, A. Mueller, K. Graham, L. Lehmann, L. M.
Dinkelborg, J. Med. Chem. 2011, 54, 406.
best results. The subsequent hydrolysis step was investigated
and optimized to give the desired [18F]fluoropyruvate in an
[19] N. Koglin, A. Mueller, M. Berndt, H. Schmitt-Willich, L. Toschi,
A. W. Stephens, V. Gekeler, M. Friebe, L. M. Dinkelborg, Clin.
Cancer Res. 2011, 17, 6000.
[20] S. Baek, C.-M. Choi, S. H. Ahn, J. W. Lee, G. Gong, J. S. Ryu, S. J. Oh,
C. Bacher-Stier, L. Fels, N. Koglin, C. Hultsch, C. A. Schatz,
L.M. Dinkelborg, E. S. Mittra, S. S. Gambhir, D. H. Moon, Clin. Cancer
Res. 2012, 18, 5427.
excellent 50% d.c. yield. This tracer was tested in vitro resulting
in low uptake into adenocarcinomic human alveolar basal
epithelial A549 cells. In vivo PET imaging of the compound
showed considerable defluorination that would limit the use
of this tracer as an imaging agent for cancer, although the
www.jlcr.org
Copyright © 2014 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2014, 57 164–171

K. Graham et al.
[21] S. Baek, A. Mueller, Y. S. Lim, H.C. Lee, Y. L. Lee, G. Gong, J. S. Kim,
J. S. Ryu, S. J. Oh, S. J. Lee, C. Bacher-Stier, L. Fels, N. Koglin,
[28] Y. Avi-Dor, J. Mager, Biochem. J. 1956, 63, 613.
[29] Y. Pocker, J. E. Meany, C. Zadorojny, J. Phys. Chem. 1971, 75, 792.
C. A. Schatz, L. M. Dinkelborg, D. H. Moon, J. Nucl. Med. 2013, 54, 1. [30] S. M. Ametamey, M. Honer, P. A. Schubiger, Chem. Rev. 2008,
[22] R. A. Cairns, I. S. Harris, T. W. Mak, Nat. Rev. Cancer 2011, 11, 85.
[23] T. Hara, F. Yokoi, M. Iio, Eur. J. Nucl. Med. 1986, 12, 21.
[24] Y. F. Cheung, C. Walsh, Biochemistry 1976, 15, 3749.
[25] D. R. Grassetti, M. E. Brokke, J. F. Murray, J. Med. Chem. 1966, 9, 149.
[26] D. S. Flournoy, P. A. Frey, Biochemistry 1989, 28, 9594.
108, 1501.
[31] For reviews see: (a) H. Vorbrüggen, Helv. Chim. Acta 2011, 94, 947; (b)
S. Takamatsu, S. Katayama, N. Hirose, E. De Cock, G. Schelkens, M.
Demillequan, J. Brepoels, K. Isawa, Nucleos. Nucleot. Nucleic Acids
2002, 21, 849.
[27] L. S. Leung, P.A. Frey, Biochem. Biophys. Res. Commun. 1978, 81, 274. [32] F. C. Kokesh. J. Org. Chem. 1976, 41, 3593.
J. Label Compd. Radiopharm 2014, 57 164–171
Copyright © 2014 John Wiley & Sons, Ltd.
www.jlcr.org