In Vitro and In Vivo Evaluations of a Hydrophilic
Conjugate for Hypoxia Imaging
Simon R. Bayly1,2, Robert C. King3, Davina J. Honess3, Peter J. Barnard2, Helen M. Betts2, Jason P. Holland2,
Rebekka Hueting2,3, Paul D. Bonnitcha2, Jonathan R. Dilworth1,2, Franklin I. Aigbirhio4, and Martin Christlieb3
1Siemens Oxford Molecular Imaging Laboratory, Inorganic Chemistry Laboratory, University of Oxford, Oxford, United Kingdom;
2Chemistry ResearchLaboratory, Department of Chemistry, University of Oxford, Oxford, United Kingdom;3CRUK/MRC, Gray Institute
for Radiation Oncology and Biology, University of Oxford, Headington, Oxford, United Kingdom; and4The Wolfson Brain Imaging
Centre, Addenbrooke’s Hospital, Cambridge, United Kingdom
A water-soluble glucose conjugate of the hypoxia tracer64Cu-
thesized and radiolabeled (64Cu-ATSE/A-G). Here we report
our initial biological experiments with64Cu-ATSE/A-G and com-
pare the results with those obtained for
18F-FDG. Methods: The uptake of64Cu-ATSE/A-G and64Cu-
ATSM into HeLa cells in vitro was investigated at a range of dis-
solved oxygen concentrations representing normoxia, hypoxia,
and anoxia. Small-animal PET with64Cu-ATSE/A-G was per-
formed in male BDIX rats implanted with P22 syngeneic carcino-
sarcomas. Images of64Cu-ATSM and18F-FDG were obtained
in the same model for comparison. Results:64CuATSE/A-G
showed oxygen concentration–dependent uptake in vitro and,
under anoxic conditions, showed slightly lower levels of cellular
uptake than64Cu-ATSM; uptake levels under hypoxic conditions
were also lower. Whereas the normoxic uptake of64Cu-ATSM in-
creased linearly over time,64Cu-ATSE/A-G uptake remained at
low levels over the entire time course. In the PET study,64CuA-
TSE/A-G showed good tumor uptake and a biodistribution pat-
tern substantially different from that of each of the controls. In
marked contrast to the findings for64Cu-ATSM, renal clearance
and accumulation in the bladder were observed.64Cu-ATSE/A-G
did not display the characteristic brain and heart uptake of
18F-FDG. Conclusion: The in vitro cell uptake studies demon-
strated that64Cu-ATSE/A-G retained hypoxia selectivity and
had improved characteristics when compared with
tion pathways, with a shift from primarily hepatointestinal for
64Cu-ATSM to partially renal with64Cu-ATSE/A-G. This finding
is consistent with the hydrophilic nature of the glucose conju-
gate. A comparison with18F-FDG PET results revealed that
64Cu-ATSE/A-G was not a surrogate for glucose metabolism.
We have demonstrated that our method for the modification of
Cu-bis(thiosemicarbazonato) complexes allows their biodistri-
bution to be modified without negating their hypoxia selectivity
or tumor uptake properties.
Key Words: PET; molecular imaging;
J Nucl Med 2008; 49:1862–1868
Tumor hypoxia is an important feature of solid tumors. It
is closely correlated with resistance to radiotherapy and
chemotherapy (1–3) as well as the initiation of angiogenesis
and metastasis (4,5), which lead to malignant progression.
Because hypoxia has a profound impact on patient survival
(6), methods for the detection and quantification of hypoxia
are currently the subject of intense investigation. Several
radiolabeled markers have been developed for PET imag-
ing of hypoxia—notably,
and Cu-diacetyl-bis(N4-methylthiosemicarbazone) (Cu-ATSM,
where Cu is either60Cu,62Cu, or64Cu) (10,11).60Cu-ATSM
has been shown to be predictive of radiotherapy treatment
outcome in small-scale clinical studies, and64Cu-ATSM
recently entered clinical trials in the United States (12–14).
The blood clearance and excretion kinetics of Cu-ATSM are
not ideal, and high levels of liver and kidney uptake have
been observed. For improved hypoxia imaging, it is therefore
desirable to modify the biodistribution properties of Cu-
ATSM, in particular, its excretion pathway, without diminish-
ing its ability to discriminate tumor hypoxia.
The mechanism of hypoxia selectivity of Cu-ATSM
and related complexes (Fig. 1) is thought to involve an
initial intracellular proton-coupled reduction to produce
a copper(I) species, followed by reoxidation and efflux
from cells with high oxygen tension or by ligand dissoci-
ation and trapping of the copper in cells with low oxygen
tension (15). Dearling et al. conducted an extensive structure–
activity survey of Cu-bis(thiosemicarbazonato) complexes
and found that hypoxia selectivity was correlated with
Received May 2, 2008; revision accepted Jul. 17, 2008.
For correspondence or reprints contact: Martin Christlieb, Gray Institute for
Radiation Oncology and Biology, University of Oxford, Old Road Campus
Research Bldg., Old Road Campus, Headington, Oxford OX3 7DQ, United
COPYRIGHT ª 2008 by the Society of Nuclear Medicine, Inc.
1862THE JOURNAL OF NUCLEAR MEDICINE • Vol. 49 • No. 11 • November 2008
the copper(II/I) redox potential as dictated by the nature of
the ligand backbone (Fig. 1, R1and R2) (16,17). We have
developed methods for the synthesis of Cu-ATSM nonsym-
metrically substituted at R3and R4and have shown that
variations in the exocyclic amine substituents have a rela-
tively minor influence on the redox potential of the cop-
per(II) center (18,19) and should not affect hypoxia
selectivity. When R3or R4is a reactive amino group (Fig.
2A), the bis(thiosemicarbazonato) core can be conjugated
to a biologically active molecule, such as glucose (19).
Glucose is of particular interest because it is hydrophilic and
is avidly metabolized by many types of tumors. A Cu-
ATSM–glucose derivative may also possess these properties
and therefore permit improved hypoxic tumor imaging.
We previously described the synthesis of Zn-ATSE/A-G
used to prepare the64Cu-radiolabeled compound64Cu-ATSE/
A-G (Fig. 2C) (19). In this article, we describe the oxygen-
dependent uptake of64Cu-ATSE/A-G into HeLa cells in vitro
rats with implanted P22 syngeneic carcinosarcomas.
MATERIALS AND METHODS
Synthesis of Labeling Precursors
The proligand H2-ATSM, Zn-ATSM, and Zn-ATSE/A-G were
prepared according to published methods (17–20).
Preparation of64Cu-ATSE/A-G for Cell Uptake Studies
and Small-Animal PET
64Cu-ATSE/A-G was prepared by transmetallation of Zn-ATSE/
A-G. The precursor complex (1.0 mg) was dissolved in deionized
water (1.0 mL). A 50-mL sample of this solution was diluted
with deionized water (200 mL), and aqueous
[64Cu(OAc)2] (250 mL, 20–50 MBq) was added. After 20 min of
stirring at room temperature, high-performance liquid chromatog-
raphy (HPLC) analysis revealed 71.7% conversion of64Cu(OAc)2
(retention time, 2.62 min) to64Cu-ATSE/A-G (retention time,
9.33 min). The nonreacted64Cu(OAc)2was removed by use of a
Sep-Pak (C18) cartridge (Waters). The cartridge was conditioned
with ethanol (5.0 mL) and deionized water (3.0 mL) before the
reaction mixture was loaded. Deionized water (2.0 mL) was
passed through to remove nonreacted64Cu(OAc)2.64Cu-ATSE/
A-G was eluted with ethanol (2 · 0.1 mL) and diluted with
saline solution (0.9%, 1.8 mL) for use in PET experiments. The
radiochemical purity of the administered64Cu-ATSE/A-G was
greater than 95%, with specific activity in the range of 10–40
MBq/mL (or 0.2–0.8 MBq/mg of Zn-ATSE/A-G).
Preparation of64Cu-ATSM for Cell Uptake Studies and
Samples for small-animal PET were prepared by 2 different
methods: by labeling of the bis(thiosemicarbazone) proligand H2-
ATSM (15) and by transmetallation of Zn-ATSM. A dimethyl
sulfoxide solution of Zn-ATSM or H2-ATSM (50 mL, 1.0 mg/mL)
was diluted with deionized water (200 mL), and aqueous64Cu(OAc)2
(250 mL, 20–50 MBq) was added. After 10 min of stirring at room
temperature, the samples were purified by use of a Sep-Pak cartridge
as described earlier. The radiopharmaceutical was formulated in a
solution of 10% ethanol and saline (1.0 mL). Radio-HPLC results
were identical for samples produced by either method. Mixtures
spiked with isolated cold Cu-ATSM confirmed the identity of the
radio-HPLC peak as64Cu-ATSM. The radiochemical purity of the
administered radiopharmaceutical was greater than 98%, with spe-
cific activity in the range of 10–40 MBq/mL (or 0.2–0.8 MBq/mg of
A 1.0 mM solution of cold Cu-ATSE/A-G (200 mL) was
prepared as previously described (19). This solution was added
to rat serum (500 mL; Sigma-Aldrich) maintained at 37?C in a
water bath. Immediately after mixing of the solution, a 200-mL
sample was withdrawn and added to water (1.5 mL) in a UV/
visible-light cuvette. The absorption spectrum of the solution was
measured with a Perkin-Elmer Lamda 19 UV/visible-light spec-
trometer. A cuvette containing the same concentration of serum in
water (also kept at 37?C) was used as a background reference.
Further samples were withdrawn at 30 and 60 min, and their
spectra were measured in the same way. The absorptivity of the
solution at 460 nm was assumed to be proportional to the
concentration of intact Cu-ATSE/A-G in the solution.
Cell Uptake Studies
For radiotracer uptake experiments, suspensions of HeLa cells
(5 mL of 106cells/mL) were incubated in glass vessels in a humidified
atmosphere under anoxic, hypoxic, or normoxic conditions with 5%
CO2and the balance N2(with a constant flow of gas) at 37?C. The
medium used was minimum essential medium with the Eagle–
Spinner modification (Earle’s salts and sodium bicarbonate), without
calcium chloride and L-glutamine, and supplemented with L-glutamine
(0.292 g/L), penicillin–streptomycin, and nonessential amino acids.
After 1 h, the oxygen conditions had reached equilibrium (probed with
an Oxford Optronics Oxylab pO2tissue oxygenation monitor), and
complexes (Cu-ATSM R15
R25 R35 R45 methyl).
derivatives. (A) Zn-ATSE/A. (B) Zn-ATSE/
HYDROPHILIC64CU-ATSM DERIVATIVE • Bayly et al.1863
the N2-purged radiotracer was added. Samples (1 mL) were taken by
use of a long needle syringe at 5, 15, 30, 45, and 60 min, and three
300-mL portions of each sample were dispensed into Eppendorf tubes.
The tubes were spun to pellet the cells, and the supernatant liquid
was removed. The activities of the cell pellet and of the supernatant
liquid were measured with a Perkin–Elmer Wizard 1470 automatic
g-counter. The uptake of64Cu as a percentage of the activity injected
was calculated and plotted. A control experiment was performed
with the radiotracer in minimum essential medium under normoxic
conditions to assess the amount of64Cu adhering to the plastic
Eppendorf tubes; this value was measured and subtracted from
each data point.
Early-generation transplants of P22 syngeneic carcinosarcomas
were used for all experiments (21). The tumors were grown
were used in the imaging experiments at approximately 14–21 d
after implantation, when the tumors were between 1.5 and 2.0 cm
in maximum diameter. The animals were treated according to
protocols approved by the U.K. Home Office and the local ethics
committee, in accordance with the U.K. Animals (Scientific Proce-
dures) Act of 1986.64Cu tracers were produced as described earlier
with64CuCl2provided by The Wolfson Brain Imaging Centre.
18F-FDG was purchased from PETNET Solutions Ltd.
The rats were anesthetized with halothane–oxygen vapor and
then injected intraperitoneally with fentanyl–fluanisone and mid-
azolam. A tail vein and a tail artery were cannulated to permit the
intravenous administration of the tracer and to monitor mean
arterial blood pressure, respectively. Rats were kept warm in the
prone position on a custom-made cradle with an electrically
warmed mat fitted with feedback control from a rectal probe.
Images were acquired (Concorde microPET 220) (22) beginning
with the injection of the tracer. The scanner field of view is
approximately 80 mm long; therefore, multiple bed positions were
required to image an entire animal, excluding the tail. Images were
acquired at 3 different bed positions corresponding to the head (at
0–20 min and 60–80 min), the middle section (at 20–40 min and
80–100 min), and the rear section (at 40–60 min and 100–120 min).
The bed positions were chosen to ensure a small amount of overlap
between images. The images were connected (incorporating a
decay correction) by use of Asipro software (Siemens). Transmis-
sion scans were obtained with a sealed57Co point source to correct
for attenuation of the 511-keV photons. At the completion of image
acquisition, the rats were euthanized by intravenous injection of
pentobarbital sodium (Euthatal; Merial Animal Health Ltd.).
Calculation of Standardized Uptake Values (SUVs) and
The PET projection data (list-mode data) were binned to 23
time frames: 5 · 10 s, 7 · 15 s, 10 · 30 s, and 1 · 300 s. Regions
of interest were defined by use of Inveon Research Workplace
software (Siemens) to paint the entire tissue, except for intestinal
tissue, in which only the main regions of activity were character-
ized, and muscle tissue, which was sampled by painting a region
within the right thigh muscle. The SUV was measured as the
decay-corrected radiotracer concentration normalized to the in-
jected dose and animal weight. SUVs were obtained at 18 min for
the brain; at 32 min for the liver, intestine, left and right kidneys,
and heart; and at 42 min for the tumor, bladder, and muscle.
Different time points were required for each set of organs because
of the acquisition time for the relevant portion of the animal.
When the activity observed was too low for an organ to be
delineated, the SUV was recorded as 0. For comparison of the
differences in cell uptake and SUVs, the Student t test was
performed. Differences at the 95% confidence level (P , 0.05)
were considered significant.
In Vitro Uptake into HeLa Cells
HeLa cells were equilibrated at 21%, 0.5%, and 0%
oxygen concentrations, and the uptake of64Cu-ATSE/A-G
was measured in triplicate at 5, 15, 30, 45, and 60 min after
addition (Fig. 3A). After 5 min, the percentages of uptake
were similar (;8.5%) for all oxygen concentrations. Under
and remained at the same level for the remaining 30 min.
Under hypoxic (0.5% O2) conditions, uptake increased lin-
early over the entire time course to reach 22.5% at 60 min.
Cell uptake under anoxic (0% O2) conditions also increased,
dependent on the O2concentration, and the ratio of hypoxic
uptake to normoxic uptake at 60 min was 2.3.
64Cu-ATSM (B) into HeLa cells over time at various oxygen
concentrations. Errors, if not indicated, are within symbols.
Percentages of uptake of64Cu-ATSE/A-G (A) and
1864THE JOURNAL OF NUCLEAR MEDICINE • Vol. 49 • No. 11 • November 2008
For comparison, the uptake of64Cu-ATSM (the parent
compound) was measured in the same cell line (Fig. 3B).
O2concentration–dependent uptake was observed, as ex-
pected (15,16), but was higher than that of64Cu-ATSE/A-G
under all conditions. The percentages of uptake at 60 min
with 21%, 0.2% and 0% O2concentrations were 34.1%,
26.4%, and 19.3%, respectively. The differences in the
percentages of uptake were found to be significant for both
21% and 0% O2(P 5 0.018). However, under normoxic
conditions,64Cu-ATSM uptake did not plateau after 30 min
but continued to increase throughout the time course. The
ratio of hypoxic uptake to normoxic uptake at 60 min was
1.4, considerably lower than that for
(especially considering that in the64Cu-ATSM experiment,
a lower partial pressure of oxygen was used for the hypoxic
samples). The hypoxic uptake values obtained for64Cu-
ATSM were lower than those reported for the more widely
used EMT6 (murine) cell line (16). Lewis et al. reported
that uptake in this type of experiment is cell line dependent,
and the lower level of uptake found in our study was thus
ascribed to differences in the HeLa cell phenotype (23).
Cold Cu-ATSE/A-G was incubated in rat serum, and the
amount of the complex remaining in the solution was
assessed with UV/visible-light spectroscopy at 30 and 60 min.
After 30 min, 92% of the complex remained in the solution.
At 60 min, the amount had decreased to 82%. With this
method it was not possible to determine whether the tracer
underwent decomposition or whether it became bound to
serum proteins. The results are in accordance with those of
a recent study of closely related lipophilic64Cu-ATSM
derivatives in mouse serum, which showed that 20%–25%
of the activity was protein bound after 60 min (24).
In Vivo PET with64Cu-ATSE/A-G
Two tumor-bearing rats were injected with 6 and 18 MBq
tracer was completely water soluble, ethanol (10%) was
added for consistency with the64Cu-ATSM experiment. In
experiment. The bladder and kidneys showed high uptake,
suggesting that the primary excretion pathway for the water-
was observed, as was uptake in the liver and areas of the
intestines. SUVs in the major organs were estimated for
comparisons with18F-FDG and64Cu-ATSM (Fig. 5).
To test the possibility that64Cu-ATSE/A-G acts as a
marker for glucose metabolism, we performed a control
experiment with18F-FDG in the BDIX/P22 tumor model
and with the same image acquisition and reconstruc-
tion protocols (Fig. 4). A strong correlation between the
observed in vivo distributions of64Cu-ATSE/A-G and
18F-FDG would indicate that the glucose subunit of the
novel tracer controls its biodistribution pathway. Differ-
ent patterns of uptake were immediately apparent on
obtained with18F-FDG (A),64Cu-ATSE/A-G (B), and64Cu-ATSM
(C) injected into tail vein of P22 tumor–bearing rats. Slices
shown in same panel were taken from one animal (superior to
Coronal slices showing small-animal PET images
HYDROPHILIC64CU-ATSM DERIVATIVE • Bayly et al. 1865
comparison of the64Cu-ATSE/A-G and18F-FDG im-
ages. In particular, the localization of18F-FDG in the
brain and the heart was clearly visible. The SUVs in
these organs were considerably higher for18F-FDG (5.1
and 12.6, respectively) than for64Cu-ATSE/A-G (0.9
and 2.5, respectively). Statistical analysis showed the
differences to be significant (P 5 0.040 for the brain and
P 5 0.018 for the heart). Liver uptake was also ob-
served, but the SUV was significantly lower for18F-FDG
than for64Cu-ATSE/A-G (3.5 and 8.6, respectively; P 5
0.027). The tumor and kidneys also retained18F-FDG,
but although the levels of uptake of18F-FDG appeared to
be lower, the SUVs were not significantly different from
those for64Cu-ATSE/A-G. The muscle uptake of18F-
FDG was significantly higher than that of64Cu-ATSE/A-
G (SUVs of 1.5 and 0.5, respectively; P , 0.001). The
observed tumor/muscle SUV ratios were 3.9 for18F-FDG
and 7.3 for64Cu-ATSE/A-G.
A key aim of this work was to demonstrate that the in
vivo distribution of
64Cu-ATSM could be modified by
derivatization. A control experiment with
was therefore performed with the BDIX/P22 tumor model
and with the same image acquisition and reconstruction
protocols (Fig. 4).64Cu-ATSM produced by the standard
method from H2-ATSM and
transmetallation of Zn-ATSM showed identical biodistri-
butions. Characteristic kidney, liver, and tumor retention
was observed.64Cu-ATSM uptake in the brain was also
discernible (SUV, 2.0), with very strong localization ob-
served in the olfactory bulbs (SUV, 12.4). In contrast,64Cu-
ATSE/A-G uptake in the brain was minimal (SUV, 0.9; in a
comparison with64Cu-ATSM, the P value was 0.042), and
the olfactory bulbs could not be delineated. Also, unlike
64Cu-ATSM produced by
64Cu-ATSE/A-G, activity in the bladder due to64Cu-ATSM
was too low for this organ to be delineated.
The results of the in vitro study demonstrated that64Cu-
ATSE/A-G possesses oxygen-dependent uptake properties
similar to those of64Cu-ATSM. Thus, conjugation of the
exocyclic nitrogen atom of the Cu-bis(thiosemicarbazonato)
complex does not negate its hypoxia selectivity. In fact,
although the hydrophilic glucose unit did appear to reduce
overall uptake at all oxygen concentrations (perhaps be-
cause of a reduced rate of diffusion through the cell
membrane), in HeLa cells the hypoxia selectivity of
64Cu-ATSE/A-G was improved. The lower level of nor-
moxic uptake shown by64Cu-ATSE/A-G led to a far higher
ratio of hypoxic uptake to normoxic uptake for this tracer
One problem limiting the clinical use of64Cu-ATSM is
its high level of liver uptake (25). The complex is highly
lipophilic (water-to-octanol partition coefficient [logP] 5
1.5) and hence has extremely low solubility in water.
Previous work suggested that64Cu-ATSM does not circu-
late as a molecular species but is more likely to be bound to
lipophilic sites in proteins or membranes. The related tracer
64Cu-ATSE showed 20% protein-associated radioactivity
when tested after dilution in serum (26). These properties
also explain why the primary excretion pathway for64Cu-
ATSM is hepatointestinal, as observed in our PET study.
Small lipophilic molecules are also able to cross the blood–
brain barrier (27), leading to the observed brain uptake of
the complex. The production of
metallation of Zn-ATSM rather than by the standard
method resulted in no discernible difference in the observed
in vivo distribution.
A serum stability test with cold Cu-ATSE/A-G produced
results similar to those observed previously for
ATSM, with 18% of the tracer being lost from the solution
after 1 h. This finding was somewhat surprising, as Cu-
ATSE/A-G is highly water soluble (logP 5 0.5) because of
the hydrophilic glucose unit and was expected to have a
low affinity for serum proteins (19). It is possible that Cu-
ATSE/A-G acts as an amphiphile and that the lipophilic Cu-
bis(thiosemicarbazonato) unit binds to serum proteins with
an affinity similar to Cu-ATSE. Alternatively, the complex
could undergo hydrolysis of the linkage between the Cu-
bis(thiosemicarbazonato) and glucose units, resulting in
decomposition and release of the copper(II) ion.
The high levels of kidney and bladder accumulation
observed in the in vivo PET studies with64Cu-ATSE/A-G
indicated that the renal pathway functioned as a route of
excretion for this tracer (64Cu-ATSM accumulated in the
kidneys but was not excreted into the bladder). The brain
uptake of64Cu-ATSE/A-G was also very low, consistent
with the low logP of the complex. These results imply that
the glucose subunit exerts some control over the biodis-
64Cu-ATSM by trans-
64Cu-ATSM,64Cu-ATSE/A-G, and18F-FDG in BDIX rats with
P22 carcinosarcomas (average of 2 imaging experiments).
Results obtained at 18 min (A), 32 min (B), and 42 min (C) are
shown. Error bars show SDs. Because of large variations
between experiments, bladder SUV for18F-FDG was obtained
from single PET image (error was estimated).
SUVs estimated from PET images obtained with
1866THE JOURNAL OF NUCLEAR MEDICINE • Vol. 49 • No. 11 • November 2008
tribution of the64Cu complex, because it is directed to
target organs for excretion of hydrophilic species and is
excluded from compartments accessible only to lipophilic
(or actively transported) species. Our PET study was not
designed to test the hypoxia selectivity of64Cu-ATSE/A-G.
It was intended as an initial qualitative study of biodis-
tribution, the tumor model used was not thought to be
hypoxic, and no attempt was made to measure tumor
18F-FDG, a marker for glucose metabolism, is by far the
most studied PET tracer, is in routine clinical use, and is
produced commercially (28). Recently, much attention has
been directed toward producing analogs of18F-FDG that
incorporate a metallic radionuclide (particularly99mTc for
SPECT) instead of18F (29–31). None of these metallo-
glucose conjugates has yet shown the ability to map
glucose metabolism. This result is perhaps not surprising,
because the addition of a bulky metal-bearing group to a
small molecule such as glucose will lead to gross changes
in its biochemical properties. In our PET study,
ATSE/A-G and18F-FDG showed obviously divergent bio-
distributions (particularly in the brain and the heart); this
result seems to confirm the notion that64Cu-ATSE/A-G does
not participate in glucose-specific transport or metabolism.
We have demonstrated that conjugation of the copper-
bis(thiosemicarbazonato) complex at the exocyclic nitrogen
with glucose leads to a modified biodistribution profile,
with the excretion pathway being switched from primarily
hepatointestinal (for64Cu-ATSM) to partially renal (for
64Cu-ATSE/A-G). We have also confirmed that
ATSE/A-G retains hypoxia selectivity in vitro, with an
improved ratio of hypoxic uptake to normoxic uptake,
when compared with
64Cu-ATSM. A quantitative PET
study with a hypoxic tumor model is required to determine
whether these characteristics are replicated in vivo. Thus,
we have obtained proof of the concept that the biodistri-
bution of64Cu-ATSM can be modified by derivatization
without negating its favorable tumor-targeting properties
and hypoxia selectivity.
We are grateful to Dr. Sally Hill and Dr. Katie Wood for
helpful discussions. We thank Peter Wardman for resources
at the Gray Cancer Institute in Northwood. We thank
Vivien Prise and Ian Wilson for implanting tumors; Anne
Clark, Ann Marriot, and Valerie Edwards for the welfare
and care of the animals; and Dr. Katrin Probst and Oksana
Golovko for their assistance with radio-HLPC. The work
was funded by CRUK (C5255/A8591), Siemens Molecular
Imaging Ltd., and the U.K. Department of Trade and
Industry. We thank GlaxoSmithKline for funding as well as
Merton College and the U.K. Engineering and Physical
Sciences Research Council for a studentship.
1. Hedley D, Pintilie M, Woo J, et al. Carbonic anhydrase IX expression, hypoxia,
and prognosis in patients with uterine cervical carcinomas. Clin Cancer Res.
2. Nordsmark M, Overgaard J. A confirmatory prognostic study on oxygenation
status and loco-regional control in advanced head and neck squamous cell
carcinoma treated by radiation therapy. Radiother Oncol. 2000;57:39–43.
3. Nordsmark M, Overgaard M, Overgaard J. Pretreatment oxygenation predicts
radiation response in advanced squamous cell carcinoma of the head and neck.
Radiother Oncol. 1996;41:31–39.
4. Pugh CW, Ratcliffe PJ. Regulation of angiogenesis by hypoxia: role of the HIF
system. Nat Med. 2003;9:677–684.
5. Dachs GU, Tozer GM. Hypoxia modulated gene expression: angiogenesis,
metastasis and therapeutic exploitation. Eur J Cancer. 2000;36:1649–1660.
6. Isa AY, Ward TH, West CML, Slevin NJ, Homer JJ. Hypoxia in head and neck
cancer. Br J Radiol. 2006;79:791–798.
7. Dubois L, Landuyt W, Haustermans K, et al. Evaluation of hypoxia in an
experimental rat tumor model by [F-18]fluoromisonidazole PET and immuno-
histochemistry. Br J Cancer. 2004;91:1947–1954.
8. Troost EGC, Bussink J, Laverman P, et al. Comparison of
autoradiography with immunohistochemistry after modification of hypoxia in
human gliomaand squamous
9. Troost EGC, Laverman P, Kaanders JH, et al. Imaging hypoxia after
oxygenation-modification: comparing [18F]FMISO autoradiography with pimo-
nidazole immunohistochemistry in human xenograft tumors. Radiother Oncol.
10. Fujibayashi Y, Taniuchi H, Yonekura Y, Ohtani H, Konishi J, Yokoyama A.
Copper-62-ATSM: a new hypoxia imaging agent with high membrane perme-
ability and low redox potential. J Nucl Med. 1997;38:1155–1160.
11. Lewis JS, McCarthy DW, McCarthy TJ, Fujibayashi Y, Welch MJ. Evaluation of
Cu-64-ATSM in vitro and in vivo in a hypoxic tumor model. J Nucl Med.
12. Dehdashti F, Grigsby PW, Mintun MA, Lewis JS, Siegel BA, Welch MJ.
Assessing tumor hypoxia in cervical cancer by positron emission tomography
with Cu-60-ATSM: relationship to therapeutic response—a preliminary report.
Int J Radiat Oncol Biol Phys. 2003;55:1233–1238.
13. Dehdashti F, Mintun MA, Lewis JS, et al. In vivo assessment of tumor hypoxia
in lung cancer with60Cu-ATSM. Eur J Nucl Med Mol Imaging. 2003;30:
14. Dehdashti F, Grigsby PW, Lewis JS, Laforest R, Siegel BA, Welch MJ.
Assessing tumor hypoxia in cervical cancer by PETwith Cu-60-labeled diacetyl-
bis(N-4-methylthiosemicarbazone). J Nucl Med. 2008;49:201–205.
15. Vavere AL, Lewis JS. Cu-ATSM: a radiopharmaceutical for the PET imaging of
hypoxia. Dalton Trans. 2007;November:4893–4902.
16. Dearling JLJ, Lewis JS, McCarthy DW, Welch MJ, Blower PJ. Redox-active
metal complexes for imaging hypoxic tissues: structure-activity relationships in
copper(II) bis(thiosemicarbazone) complexes. Chem Commun. 1998;November:
17. Dearling JLJ, Lewis JS, Muller GED, Welch MJ, Blower PJ. Copper
bis(thiosemicarbazone) complexes as hypoxia imaging agents: structure-activity
relationships. J Biol Inorg Chem. 2002;7:249–259.
18. Christlieb M, Dilworth JR. Ligands for molecular imaging: the synthesis of
bis(thiosemicarbazone) ligands. Chemistry. 2006;12:6194–6206.
19. Holland JP, Aigbirhio FI, Betts HM, et al. Functionalized bis(thiosemicarbazonato)
complexes of zinc and copper: synthetic platforms toward site-specific radio-
pharmaceuticals. Inorg Chem. 2007;46:465–485.
20. Cowley AR, Davis J, Dilworth JR, et al. Fluorescence studies of the intra-cellular
distribution of zinc bis(thiosemicarbazone) complexes in human cancer cells.
Chem Commun. 2005;February:845–847.
21. Sensky PL, Prise VE, Tozer GM, Shaffi KM, Hirst DG. Resistance to flow
through tissue-isolated transplanted rat tumors located in two different sites.
Br J Cancer. 1993;67:1337–1341.
22. Tai YC, Ruangma A, Rowland D, et al. Performance evaluation of the microPET
Focus: a third-generation microPET scanner dedicated to animal imaging. J Nucl
23. Burgman P, O’Donoghue JA, Lewis JS, Welch MJ, Humm JL, Ling CC. Cell
line-dependent differences in uptake and retention of the hypoxia-selective
nuclear imaging agent Cu-ATSM. Nucl Med Biol. 2005;32:623–630.
24. Bonnitcha PD, Vavere AL, Lewis JS, Dilworth JR. In vitro and in vivo
evaluation of bifunctional bisthiosemicarbazone
positron emission tomography imaging of hypoxia. J Med Chem. 2008;51:
64Cu-complexes for the
HYDROPHILIC64CU-ATSM DERIVATIVE • Bayly et al. 1867
25. LaforestR,DehdashtiF,LewisJS,SchwarzSW.DosimetryofCu-60/61/62/64-ATSM: Download full-text
a hypoxia imaging agent for PET. Eur J Nucl Med Mol Imaging. 2005;32:764–770.
26. McQuade P, Martin KE, Castle TC, et al. Investigation into Cu-64-labeled
bis(selenosemicarbazone) and bis(thiosemicarbazone) complexes as hypoxia
imaging agents. Nucl Med Biol. 2005;32:147–156.
27. Pardridge WM. Transport of small molecules through the blood-brain-barrier:
biology and methodology. Adv Drug Deliv Rev. 1995;15:5–36.
28. Gambhir SS, Czernin J, Schwimmer J, Silverman DHS, Coleman RE, Phelps
ME. A tabulated summary of the FDG PET literature. J Nucl Med. 2001;42(5
29. Bayly SR, Fisher CL, Storr T, Adam MJ, Orvig C. Carbohydrate conjugates for
molecular imaging and radiotherapy: Tc-99m(I) and Re-186(I) tricarbonyl
complexes of N-(29-hydroxybenzyl)-2-amino-2-deoxy-D-glucose. Bioconjug
30. Schibli R, Dumas C, Petrig J, et al. Synthesis and in vitro characterization of
organometallic rhenium and technetium glucose complexes against Glut 1 and
hexokinase. Bioconjug Chem. 2005;16:105–112.
31. Ferreira CL, Bayly SR, Green DE, et al. Carbohydrate-appended 3-hydroxy-4-
pyridinone complexes of the [M(CO)3]1core (M 5 Re,
Bioconjug Chem. 2006;17:1321–1329.
1868THE JOURNAL OF NUCLEAR MEDICINE • Vol. 49 • No. 11 • November 2008